From e684af40b358aa1ba0180ccdba39fcb9289871b8 Mon Sep 17 00:00:00 2001 From: Lauren Chambers Date: Mon, 20 May 2019 15:28:57 -0400 Subject: [PATCH 01/14] Initial commit of photutils notebooks --- .../01_photutils_background_estimation.ipynb | 619 ++++++++++ .../01_background_estimation/requirements.txt | 4 + .../02_photutils_source_detection.ipynb | 1100 +++++++++++++++++ .../02_source_detection/requirements.txt | 4 + .../03_photutils_aperture_photometry.ipynb | 978 +++++++++++++++ .../requirements.txt | 4 + .../04_photutils_psf_photometry.ipynb | 253 ++++ .../04_psf_photometry/requirements.txt | 4 + .../photutils_notebook_style.mplstyle | 20 + 9 files changed, 2986 insertions(+) create mode 100644 notebooks/photutils/01_background_estimation/01_photutils_background_estimation.ipynb create mode 100644 notebooks/photutils/01_background_estimation/requirements.txt create mode 100644 notebooks/photutils/02_source_detection/02_photutils_source_detection.ipynb create mode 100644 notebooks/photutils/02_source_detection/requirements.txt create mode 100644 notebooks/photutils/03_photutils_aperture_photometry/03_photutils_aperture_photometry.ipynb create mode 100644 notebooks/photutils/03_photutils_aperture_photometry/requirements.txt create mode 100644 notebooks/photutils/04_psf_photometry/04_photutils_psf_photometry.ipynb create mode 100644 notebooks/photutils/04_psf_photometry/requirements.txt create mode 100644 notebooks/photutils/photutils_notebook_style.mplstyle diff --git a/notebooks/photutils/01_background_estimation/01_photutils_background_estimation.ipynb b/notebooks/photutils/01_background_estimation/01_photutils_background_estimation.ipynb new file mode 100644 index 00000000..46991a0e --- /dev/null +++ b/notebooks/photutils/01_background_estimation/01_photutils_background_estimation.ipynb @@ -0,0 +1,619 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
![photutils logo](photutils_banner.svg)
\n", + "\n", + "# Background Estimation with `photutils`\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "##### What is background estimation?\n", + "In order to most accurately do photometric analysis of celestial sources in image data, it is important to estimate and subtract the image background. Any astronomical image will have background noise, due to both detector effects and background emission from the night sky. This noise can be modeled as uniform, or as varying with position on the detector. \n", + "\n", + "The `photutils` package provides tools for estimating 2-dimensional background noise, which can then be subtracted from an image to ensure the most accurate photometry possible.\n", + "\n", + "##### What does this tutorial include?\n", + "This tutorial covers the basics of background estimation and subtraction, including the following methods:\n", + "- Scalar Background Estimation\n", + "- 2-D Background Estimation\n", + "\n", + "##### Which data are used in this tutorial?\n", + "We will be manipulating Hubble eXtreme Deep Field (XDF) data, which was collected using the Advanced Camera for Surveys (ACS) on Hubble between 2002 and 2012. The image we use here is the result of 1.8 million seconds (500 hours!) of exposure time, and includes some of the faintest and most distant galaxies that have ever been observed. \n", + "\n", + "Background subtraction is essential for accurate photometric analysis of astronomical data like the XDF.\n", + "\n", + "##### The methods demonstrated here are available in narrative form within the `photutils.background` [documentation](http://photutils.readthedocs.io/en/stable/background.html)." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
**Note:** This notebook focuses on global background estimation. Local background subtraction with annulus apertures is demonstrated in the [aperture photometry notebook](03_photutils_aperture_photometry.ipynb).
\n", + "\n", + "
**Important:** Before proceeding, please be sure to update your versions of `astropy`, `matplotlib`, and `photutils`, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the [AstroConda](https://astroconda.readthedocs.io) distribution.
\n", + "\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Import necessary packages" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "First, let's import packages that we will use to perform arithmetic functions and visualize data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "import numpy as np\n", + "\n", + "from astropy.io import fits\n", + "import astropy.units as u\n", + "from astropy.stats import sigma_clipped_stats, SigmaClip\n", + "from astropy.visualization import ImageNormalize, LogStretch\n", + "import matplotlib.pyplot as plt\n", + "from matplotlib.ticker import LogLocator\n", + "\n", + "# Show plots in the notebook\n", + "% matplotlib inline" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [shared style file](photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "plt.style.use('photutils_notebook_style.mplstyle')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Retrieve data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", + "\n", + "(Generally, the best package for web queries of astronomical data is `astroquery`; however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with `astroquery`.)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "url = 'https://archive.stsci.edu/pub/hlsp/xdf/hlsp_xdf_hst_acswfc-60mas_hudf_f435w_v1_sci.fits'\n", + "with fits.open(url) as hdulist:\n", + " hdulist.info()\n", + " data = hdulist[0].data\n", + " header = hdulist[0].header" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Modifying data\n", + "For the purposes of this notebook example, we're going to add a background effect to this data, but don't worry about this. (Pay no attention to that man behind the curtain!)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "mask = data == 0\n", + "n_data_pixels = len(data[~mask])\n", + "background = np.linspace(-1e-4, 5e-4, num=n_data_pixels)\n", + "\n", + "modified_data = np.copy(data)\n", + "modified_data[~mask] += background[:]" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Data representation\n", + "\n", + "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons (counts) per second." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from astropy.nddata import CCDData\n", + "unit = u.ct / u.s\n", + "xdf_image = CCDData(modified_data, unit=unit, meta=header)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's look at the data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", + "xdf_image_clipped = np.clip(xdf_image, 1e-4, None) # clip to plot with logarithmic stretch\n", + "fitsplot = ax1.imshow(xdf_image_clipped, norm=norm_image)\n", + "\n", + "# Define the colorbar and fix the labels\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "*Note: Double-click on any inline plot to zoom in.*" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Mask data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "You probably noticed that a large portion of the data is equal to zero. The data we are using is a reduced mosaic that combines many different exposures, and that has been rotated such that not all of the array holds data. \n", + "\n", + "We want to **mask** out the non-data, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Define the mask\n", + "xdf_image.mask = xdf_image.data == 0" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, [ax1, ax2] = plt.subplots(1, 2, figsize=(12, 6), sharey=True)\n", + "plt.tight_layout()\n", + "\n", + "# Plot the mask\n", + "ax1.imshow(xdf_image.mask, cmap='Greys')\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('Mask')\n", + "\n", + "# Plot the masked data\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", + "fitsplot = ax2.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", + "\n", + "# Define the colorbar and fix the labels\n", + "cbar_ax = fig.add_axes([1, 0.09, 0.03, 0.87])\n", + "cbar = fig.colorbar(fitsplot, cbar_ax, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "ax2.set_xlabel('X (pixels)')\n", + "ax2.set_title('Masked Data')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Perform scalar background estimation" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Now that the data are properly masked, we can calculate some basic statistical values to do a scalar estimation of the image background. \n", + "\n", + "By \"scalar estimation\", we mean the calculation of a single value (such as the mean or median) to represent the value of the background for our entire two-dimensional dataset. This is in contrast to a two-dimensional background, where the estimated background is represented as an array of values that can vary spatially with the dataset. We will calculate a 2D background in the upcoming section.\n", + "\n", + "Here we will calculate the mean, median, and mode using sigma clipping. With sigma clipping, the data is iteratively clipped to exclude data points outside of a certain sigma (standard deviation), thus removing some of the noise from the data before determining statistical values." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "mean, median, std = sigma_clipped_stats(xdf_image.data, sigma=3.0, iters=5, mask=xdf_image.mask)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "But what difference does this sigma clipping make? And how important is masking, anyway? Let's visualize these statistics to get an idea:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Calculate the data without masking\n", + "stats_nomask = sigma_clipped_stats(xdf_image.data, sigma=3.0, iters=5)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, [ax1, ax2] = plt.subplots(1, 2, figsize=(12, 4), sharey=True)\n", + "\n", + "# Plot histograms of the data\n", + "flux_range = (-.5e-3, 1.5e-3)\n", + "ax1.hist(xdf_image[~xdf_image.mask], bins=100, range=flux_range)\n", + "ax2.hist(xdf_image[~xdf_image.mask], bins=100, range=flux_range)\n", + "\n", + "# Plot lines for each kind of mean\n", + "ax1.axvline(mean, label='Masked \\& Clipped', c='C1', lw=3)\n", + "ax1.axvline(np.average(xdf_image[~xdf_image.mask]), label='Masked', c='C2', ls='--', lw=3)\n", + "ax1.axvline(stats_nomask[0], label='Clipped', c='C3', ls='-.', lw=3)\n", + "ax1.axvline(np.average(xdf_image), label='Neither', c='C6', ls=':', lw=3)\n", + "\n", + "ax1.set_xlim(flux_range)\n", + "ax1.set_xlabel(r'Flux Count Rate ($e^{-1}/s$)', fontsize=14)\n", + "ax1.set_ylabel('Frequency', fontsize=14)\n", + "ax1.set_title('Effect of Sigma-Clipping and Masking on Mean', fontsize=16)\n", + "ax1.legend(fontsize=11)\n", + "\n", + "\n", + "# Plot lines for each kind of median\n", + "# Note: use np.ma.median rather than np.median for masked arrays\n", + "ax2.axvline(median, label='Masked \\& Clipped', c='C1', lw=3)\n", + "ax2.axvline(np.ma.median(xdf_image[~xdf_image.mask]), label='Masked', c='C2', ls='--', lw=3)\n", + "ax2.axvline(stats_nomask[1], label='Clipped', c='C3', ls='-.', lw=3)\n", + "ax2.axvline(np.ma.median(xdf_image), label='Neither', c='C6', ls=':', lw=3)\n", + "\n", + "ax2.set_xlim(flux_range)\n", + "ax2.set_xlabel(r'Flux Count Rate ($e^{-1}/s$)', fontsize=14)\n", + "ax2.set_title('Effect of Sigma-Clipping and Masking on Median', fontsize=16)\n", + "ax2.legend(fontsize=11)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Just from simply looking at the distribution of the data, it is pretty easy to see how sigma-clipping and masking improve the calculation of the mean and median.\n", + "\n", + "But enough looking at numbers, let's actually remove the background from the data. By using the `subtract()` method of the `CCDData` class, we can subtract the mean background while maintaining the metadata and mask of our original CCDData object:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Calculate the scalar background subtraction, maintaining metadata, unit, and mask\n", + "xdf_scalar_bkgdsub = xdf_image.subtract(mean * u.ct / u.s)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, [ax1, ax2] = plt.subplots(1, 2, figsize=(12, 6), sharey=True)\n", + "plt.tight_layout()\n", + "\n", + "# Define the normalization\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", + "xdf_scalar_bkgdsub_clipped = np.clip(xdf_scalar_bkgdsub, 1e-4, None) # clip to plot with logarithmic stretch\n", + "\n", + "# Plot the original data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('Original Data')\n", + "\n", + "# Plot the subtracted data\n", + "fitsplot = ax2.imshow(np.ma.masked_where(xdf_scalar_bkgdsub.mask, xdf_scalar_bkgdsub_clipped), norm=norm_image)\n", + "ax2.set_xlabel('X (pixels)')\n", + "ax2.set_title('Scalar Background-Subtracted Data')\n", + "\n", + "# Define the colorbar and fix the labels\n", + "cbar_ax = fig.add_axes([1, 0.09, 0.03, 0.87])\n", + "cbar = fig.colorbar(fitsplot, cbar_ax, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Note that both plots above use the same normalization scheme, represented by the colorbar at right. That is to say, if two pixels have the same color in both arrays, they have the same value.\n", + "\n", + "That looks better! You can tell that the background is darker, especially in the top corner. However, the background still does not seem to be completely removed. In this case, the background varies spatially; it is two-dimensional. Thankfully, `photutils` includes functions to remove background like this." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + "

**Exercises:**


\n", + "Perform a median scalar background subtraction on our sigma-clipped data. Plot it and visually inspect it. How does it compare to the original data?\n", + "

\n", + "Compare the median background subtraction to the mean background subtraction. Which is better?\n", + "
" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Perform 2-D background estimation" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The `Background2D` class allows users to model 2-dimensional backgrounds, by evaluating the background signal in small boxes, and smoothing these boxes to reconstruct a continuous 2D background. The class includes the following arguments/attributes:\n", + "* **`box_size`** - the size of the boxes used to calculate the background. This should be larger than individual sources, yet still small enough to encompass changes in the background.\n", + "* **`filter_size`** - the size of the median filter used to smooth the final 2D background.\n", + "* **`filter_threshold`** - threshold below which the smoothing median filter will not be applied.\n", + "* **`sigma_clip`** - an ` astropy.stats.SigmaClip` object that is used to specify the sigma and number of iterations used to sigma-clip the data before background calculations are performed.\n", + "* **`bkg_estimator`** - the method used to perform the background calculation in each box (mean, median, SExtractor algorithm, etc.).\n", + "\n", + "For this example, we will use the `MeanBackground` estimator." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils.background import Background2D, MeanBackground" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "sigma_clip = SigmaClip(sigma=3., iters=5)\n", + "bkg_estimator = MeanBackground()\n", + "bkg = Background2D(xdf_image, box_size=200, filter_size=(10, 10), mask=xdf_image.mask,\n", + " sigma_clip=sigma_clip, bkg_estimator=bkg_estimator)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "So, what does this 2D background look like? Where were the boxes placed?" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", + "background_clipped = np.clip(bkg.background, 1e-4, None) # clip to plot with logarithmic stretch\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, background_clipped), norm=norm_image)\n", + "\n", + "# Plot the meshes\n", + "bkg.plot_meshes(outlines=True, color='lightgrey')\n", + "\n", + "# Define the colorbar\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('2D Estimated Background')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "And how does the data look if we use this background subtraction method (again maintaining the attributes of the CCDData object)?" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Calculate the 2D background subtraction, maintaining metadata, unit, and mask\n", + "xdf_2d_bkgdsub = xdf_image.subtract(bkg.background * u.ct / u.s)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, [ax1, ax2] = plt.subplots(1, 2, figsize=(12, 6), sharey=True)\n", + "plt.tight_layout()\n", + "\n", + "# Define the normalization\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", + "xdf_2d_bkgdsub_clipped = np.clip(xdf_2d_bkgdsub, 1e-4, None) # clip to plot with logarithmic stretch\n", + "\n", + "# Plot the scalar-subtracted data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_scalar_bkgdsub.mask, xdf_scalar_bkgdsub_clipped), norm=norm_image)\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_title('Scalar Background-Subtracted Data')\n", + "\n", + "# Plot the 2D-subtracted data\n", + "fitsplot = ax2.imshow(np.ma.masked_where(xdf_2d_bkgdsub.mask, xdf_2d_bkgdsub_clipped), norm=norm_image)\n", + "ax2.set_xlabel('X (pixels)')\n", + "ax2.set_title('2D Background-Subtracted Data')\n", + "\n", + "# Plot the colorbar\n", + "cbar_ax = fig.add_axes([1, 0.09, 0.03, 0.87])\n", + "cbar = fig.colorbar(fitsplot, cbar_ax, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Note how much more even the 2D background-subtracted image looks; especially the difference between these two images in the bottom corner and top corner. This makes sense, as the background that `Background2D` identified was a gradient from the top corner down to the bottom!" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + "

**Exercises:**


\n", + "Calculate the standard deviation (with sigma-clipping and masking!) for the original data, the scalar background-subtracted data, and the 2D background-subtracted data. How do the values compare? Which has the smallest standard deviation?

\n", + "\n", + "Notice that the difference between each dataset's standard deviation is small - why might this be?\n", + "
" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Conclusions\n", + "\n", + "The `photutils` package provides a powerful tool in the `Background2D` class, allowing users to easily estimate and subtract spatially variant background signals from their data.\n", + "\n", + "**To continue with this `photutils` tutorial, go on to the [source detection notebook](02_photutils_source_detection.ipynb).**" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "September 2018\n", + "\n", + "Author: Lauren Chambers (lchambers@stsci.edu), Erik Tollerud (etollerud@stsci.edu)\n", + "\n", + "For more examples and details, please visit the [photutils](http://photutils.readthedocs.io/en/stable/index.html) documentation." + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.5.5" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notebooks/photutils/01_background_estimation/requirements.txt b/notebooks/photutils/01_background_estimation/requirements.txt new file mode 100644 index 00000000..a0d3505e --- /dev/null +++ b/notebooks/photutils/01_background_estimation/requirements.txt @@ -0,0 +1,4 @@ +astropy>=3.1.2 +matplotlib>2.2.2 +numpy>=1.13.3 +photutils>=0.4 diff --git a/notebooks/photutils/02_source_detection/02_photutils_source_detection.ipynb b/notebooks/photutils/02_source_detection/02_photutils_source_detection.ipynb new file mode 100644 index 00000000..ead20219 --- /dev/null +++ b/notebooks/photutils/02_source_detection/02_photutils_source_detection.ipynb @@ -0,0 +1,1100 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "\n", + "\n", + "[\n", + "\n", + "](http://photutils.readthedocs.io/en/stable/index.html)\n", + "\n", + "\n", + "# Source Detection with `photutils`\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "##### What is source detection?\n", + "In order to do photometry on astronomical image data, one must first determine the locations of the sources in the image. Source detection methods find sources algorithmically, by looking for regions of an image where the signal from a source is statistically higher than the signal from background noise. Some algorithms search for sources whose profiles match specific data models, such as a 2-D Gaussian, while others simply look for local maxima.\n", + "\n", + "The `photutils` package provides a variety of tools that use different detection algorithms to locate sources in an image.\n", + "\n", + "##### What does this tutorial include?\n", + "This tutorial covers different tools for source detection with `photutils`, including the following methods:\n", + "* Source Detection with `DAOStarFinder`\n", + "* Source Detection with `IRAFStarFinder`\n", + "* Source Detection with `find_peaks`\n", + "* Image Segmentation\n", + "\n", + "##### Which data are used in this tutorial?\n", + "We will be manipulating Hubble eXtreme Deep Field (XDF) data, which was collected using the Advanced Camera for Surveys (ACS) on Hubble between 2002 and 2012. The image we use here is the result of 1.8 million seconds (500 hours!) of exposure time, and includes some of the faintest and most distant galaxies that have ever been observed. \n", + "\n", + "##### The methods demonstrated here are available in narrative form within the `photutils.detection` [documentation](http://photutils.readthedocs.io/en/stable/detection.html) and `photutils.segmentation` [documentation](http://photutils.readthedocs.io/en/stable/segmentation.html)." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + "\n", + "**Important:** Before proceeding, please be sure to update your versions of `astropy`, `matplotlib`, and `photutils`, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the [AstroConda](https://astroconda.readthedocs.io) distribution.\n", + "\n", + "
\n", + "\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Import necessary packages" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "First, let's import packages that we will use to perform arithmetic functions and visualize data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from astropy.io import fits\n", + "import astropy.units as u\n", + "from astropy.nddata import CCDData\n", + "from astropy.stats import sigma_clipped_stats, SigmaClip\n", + "from astropy.visualization import ImageNormalize, LogStretch\n", + "import matplotlib.pyplot as plt\n", + "from matplotlib.ticker import LogLocator\n", + "import numpy as np\n", + "from photutils.background import Background2D, MeanBackground\n", + "\n", + "# Show plots in the notebook\n", + "%matplotlib inline" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [style file shared with the other photutils tutorials](photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "plt.style.use('photutils_notebook_style.mplstyle')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Retrieve data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", + "\n", + "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "url = 'https://archive.stsci.edu/pub/hlsp/xdf/hlsp_xdf_hst_acswfc-60mas_hudf_f435w_v1_sci.fits'\n", + "with fits.open(url) as hdulist:\n", + " hdulist.info()\n", + " data = hdulist[0].data\n", + " header = hdulist[0].header" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "As explained in the [previous notebook](01_photutils_background_estimation.ipynb) on background estimation, it is important to **mask** these data, as a large portion of the values are equal to zero. We will mask out the non-data portions of the image array, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Define the mask\n", + "mask = data == 0" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons (counts) per second." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "unit = u.electron/ u.s\n", + "xdf_image = CCDData(data, unit=unit, meta=header, mask=mask)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's look at the data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Set up the normalization and colormap\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch(), clip=False)\n", + "cmap = plt.get_cmap('viridis')\n", + "cmap.set_over(cmap.colors[-1])\n", + "cmap.set_under(cmap.colors[0])\n", + "cmap.set_bad('white') # Show masked data as white\n", + "xdf_image_clipped = np.clip(xdf_image, 1e-4, None) # clip to plot with logarithmic stretch\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", + "\n", + "# Define the colorbar and fix the labels\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "*Tip: Double-click on any inline plot to zoom in.*" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## Source Detection with `DAOStarFinder`" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "With the `DAOStarFinder` [class](http://photutils.readthedocs.io/en/stable/api/photutils.DAOStarFinder.html), `photutils` provides users with an easy application of the popular [DAOFIND](http://stsdas.stsci.edu/cgi-bin/gethelp.cgi?daofind) algorithm ([Stetson 1987, PASP 99, 191](http://adsabs.harvard.edu/abs/1987PASP...99..191S)), originally developed at the Dominion Astrophysical Observatory. \n", + "\n", + "This algorithm detects sources by:\n", + "* Searching for local maxima\n", + "* Selecting only sources with peak amplitude above a defined threshold\n", + "* Selecting sources with sizes and shapes that match a 2-D Gaussian kernel (circular or elliptical)\n", + "\n", + "It returns:\n", + "* Location of the source centroid\n", + "* Parameters reflecting the source's sharpness and roundness\n", + "\n", + "Generally, the threshold that source detection algorithms use is defined as a multiple of the standard deviation. So first, we need to calculate statistics for the data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "mean, median, std = sigma_clipped_stats(xdf_image.data, sigma=3.0, maxiters=5, mask=xdf_image.mask)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Now, let's run the `DAOStarFinder` algorithm on our data and see what it finds. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import DAOStarFinder" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "daofind = DAOStarFinder(fwhm=5.0, threshold=20.*std)\n", + "sources_dao = daofind(xdf_image * ~xdf_image.mask) \n", + "print(sources_dao)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", + "ax1.scatter(sources_dao['xcentroid'], sources_dao['ycentroid'], s=30, marker='o', \n", + " lw=1, alpha=0.7, facecolor='None', edgecolor='r')\n", + "\n", + "# Define the colorbar and fix the labels\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('DAOFind Sources')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's randomly pull out some of these sources to get a closer look at them." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "fig, axs = plt.subplots(3,3, figsize=(3, 3))\n", + "plt.subplots_adjust(wspace=0.1, hspace=0.1)\n", + "\n", + "cutout_size = 20\n", + "\n", + "srcs = np.random.permutation(sources_dao)[:axs.size]\n", + "for ax, src in zip(axs.ravel(), srcs):\n", + " slc = (slice(int(src['ycentroid'] - cutout_size), int(src['ycentroid'] + cutout_size)),\n", + " slice(int(src['xcentroid'] - cutout_size), int(src['xcentroid'] + cutout_size)))\n", + " ax.imshow(xdf_image_clipped[slc], norm=norm_image)\n", + " ax.text(2, 2, str(src['id']), color='w', va='top')\n", + " ax.set_xticks([])\n", + " ax.set_yticks([])" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + "\n", + "

Exercises:


\n", + "\n", + "Re-run the `DAOStarFinder` algorithm with a smaller threshold (like 5σ), and plot the sources that it finds. Do the same, but with a larger threshold (like 100σ). How did changing the threshold affect the results?\n", + "\n", + "
" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Source Detection with `IRAFStarFinder`" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Similarly to `DAOStarFinder`, `IRAFStarFinder` is a class that implements a pre-existing algorithm that is widely used within the astronomical community. This class uses the `starfind` [algorithm](http://stsdas.stsci.edu/cgi-bin/gethelp.cgi?starfind) that was originally part of IRAF.\n", + "\n", + "`IRAFStarFinder` is fundamentally similar to `DAOStarFinder` in that it detects sources by finding local maxima above a certain threshold that match a Gaussian kernel. However, `IRAFStarFinder` differs in the following ways:\n", + "* Does not allow users to specify an elliptical Gaussian kernel\n", + "* Uses image moments to calculate the centroids, roundness, and sharpness of objects\n", + "\n", + "Let's run the `IRAFStarFinder` algorithm on our data, with the same FWHM and threshold, and see how its results differ from `DAOStarFinder`:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import IRAFStarFinder" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "iraffind = IRAFStarFinder(fwhm=5.0, threshold=20.*std)\n", + "sources_iraf = iraffind(xdf_image * ~xdf_image.mask) \n", + "print(sources_iraf)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", + "ax1.scatter(sources_iraf['xcentroid'], sources_iraf['ycentroid'], s=30, marker='o', \n", + " lw=1, alpha=0.7, facecolor='None', edgecolor='r')\n", + "\n", + "# Define the colorbar and fix the labels\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('IRAFFind Sources')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Again, let's randomly select some sources for a closer look:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "fig, axs = plt.subplots(3,3, figsize=(3, 3))\n", + "plt.subplots_adjust(wspace=0.1, hspace=0.1)\n", + "\n", + "cutout_size = 20\n", + "\n", + "srcs = np.random.permutation(sources_iraf)[:axs.size]\n", + "for ax, src in zip(axs.ravel(), srcs):\n", + " slc = (slice(int(src['ycentroid'] - cutout_size), int(src['ycentroid'] + cutout_size)),\n", + " slice(int(src['xcentroid'] - cutout_size), int(src['xcentroid'] + cutout_size)))\n", + " ax.imshow(xdf_image_clipped[slc], norm=norm_image)\n", + " ax.text(2, 2, str(src['id']), color='w', va='top')\n", + " ax.set_xticks([])\n", + " ax.set_yticks([])" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + "\n", + "

Exercises:


\n", + "\n", + "Re-run the `IRAFStarFinder` algorithm with a smaller full-width-half-max (FWHM) – try 3 pixels – and plot the sources that it finds. Do the same, but with a larger FWHM (like 10 pixels). How did changing the FWHM affect the results? What astronomical objects might be better captures by smaller FWHM? Larger?\n", + "\n", + "
" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Note: Comparing `DAOStarFinder` and `IRAFStarFinder`" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "You might have noticed that the `IRAFStarFinder` algorithm only found 211 sources in our data - 14% of what `DAOStarFinder` found. Why is this?\n", + "\n", + "The answer comes down to the default settings for the two algorithms: (1) there are differences in the upper and lower bounds on the requirements for source roundness and sharpness, and (2) `IRAFStarFinder` includes a minimum separation between sources that `DAOStarFinder` does not have:\n", + "\n", + "| | `IRAFStarFinder` | `DAOStarFinder` |\n", + "|------|------|------|\n", + "| sharplo | 0.5 | 0.2 |\n", + "| sharphi | 2.0 | 1.0 |\n", + "| roundlo | 0.0 | -1.0 |\n", + "| roundhi | 0.2 | 1.0 |\n", + "| minsep_fwhm | 1.5 * FWHM | N/A |\n", + "\n", + "Thinking about this, *it then makes sense* that `IRAFStarFinder` would find fewer sources. It has stricter restrictions on source roundness, meaning that it eliminates more elliptical galactic sources (this is the eXtreme Deep Field, after all!), and the minimum separation requirement further rules out sources that are too close to one another.\n", + "\n", + "If we set all these parameters to be equivalent, though, we should find much better agreement between the two methods:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "iraffind_match = IRAFStarFinder(fwhm=5.0, threshold=20.*std, \n", + " sharplo=0.2, sharphi=1.0, \n", + " roundlo=-1.0, roundhi=1.0,\n", + " minsep_fwhm=0.0)\n", + "sources_iraf_match = iraffind_match(xdf_image * ~xdf_image.mask) \n", + "print(sources_iraf_match)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The number of detected sources are in much better agreement now - 1415 versus 1470 - but the improved agreement can also be seen by plotting the location of these sources:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, [ax1, ax2] = plt.subplots(1, 2, figsize=(12, 6))\n", + "plt.tight_layout()\n", + "\n", + "# Plot the DAOStarFinder data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", + "ax1.scatter(sources_dao['xcentroid'], sources_dao['ycentroid'], s=30, marker='o', \n", + " lw=1, alpha=0.7, facecolor='None', edgecolor='r')\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('DAOStarFinder Sources')\n", + "\n", + "# Plot the IRAFStarFinder data\n", + "fitsplot = ax2.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", + "ax2.scatter(sources_iraf_match['xcentroid'], sources_iraf_match['ycentroid'], \n", + " s=30, marker='o', lw=1, alpha=0.7, facecolor='None', edgecolor='r')\n", + "ax2.set_xlabel('X (pixels)')\n", + "ax2.set_title('IRAFStarFinder Sources')\n", + "\n", + "# Define the colorbar\n", + "cbar_ax = fig.add_axes([1, 0.09, 0.03, 0.87])\n", + "cbar = plt.colorbar(fitsplot, cbar_ax, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Take this example as reminder to be mindful when selecting a source detection algorithm, and when defining algorithm parameters! Don't be afraid to play around with the parameters and investigate how that affects your results." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Source Detection with `find_peaks`" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "For more general source detection cases that do not require comparison with models, `photutils` offers the `find_peaks` function. \n", + "\n", + "This function simply finds sources by identifying local maxima above a given threshold and separated by a given distance, rather than trying to fit data to a given model. Unlike the previous detection algorithms, `find_peaks` does not necessarily calculate objects' centroids. Unless the `subpixel` argument is set to `True`, `find_peaks` will return just the integer value of the peak pixel for each source.\n", + "\n", + "This algorithm is particularly useful for identifying non-stellar sources or heavily distorted sources in image data.\n", + "\n", + "Let's see how it does:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import find_peaks" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "sources_findpeaks = find_peaks(xdf_image.data, mask=xdf_image.mask, \n", + " threshold=20.*std, box_size=30, subpixel=True) \n", + "print(sources_findpeaks)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", + "ax1.scatter(sources_findpeaks['x_peak'], sources_findpeaks['y_peak'], s=30, marker='o', \n", + " lw=1, alpha=0.7, facecolor='None', edgecolor='r')\n", + "\n", + "# Define the colorbar\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('find\\_peaks Sources')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "And a closer look:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "fig, axs = plt.subplots(3,3, figsize=(3, 3))\n", + "plt.subplots_adjust(wspace=0.1, hspace=0.1)\n", + "\n", + "cutout_size = 20\n", + "\n", + "srcs = np.random.permutation(sources_findpeaks)[:axs.size]\n", + "for ax, src in zip(axs.ravel(), srcs):\n", + " slc = (slice(int(src['y_peak'] - cutout_size), int(src['y_peak'] + cutout_size)),\n", + " slice(int(src['x_peak'] - cutout_size), int(src['x_peak'] + cutout_size)))\n", + " ax.imshow(xdf_image_clipped[slc], norm=norm_image)\n", + " src_id = np.where((sources_findpeaks['x_peak'] == src['x_peak']) & \n", + " (sources_findpeaks['y_peak'] == src['y_peak']))[0][0]\n", + " ax.text(2, 2, str(src_id), color='w', va='top')\n", + " ax.set_xticks([])\n", + " ax.set_yticks([])" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Comparing Detection Methods\n", + "\n", + "Let's compare how each of these different strategies did.\n", + "\n", + "First, how many sources did each method find?" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "print('''DAOStarFinder: {} sources\n", + "IRAFStarFinder: {} sources\n", + "find_peaks: {} sources'''.format(len(sources_dao), len(sources_iraf), len(sources_findpeaks)))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Next, how many of these sources match? We can answer this question by using [sets](https://docs.python.org/3/library/stdtypes.html#set-types-set-frozenset) to compare the centroids of the different sources (rounding to the first decimal place)." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": true + }, + "outputs": [], + "source": [ + "# Make lists of centroid coordinates\n", + "centroids_dao = [(x, y) for x, y in sources_dao['xcentroid', 'ycentroid']]\n", + "centroids_iraf = [(x, y) for x, y in sources_iraf['xcentroid', 'ycentroid']]\n", + "centroids_findpeaks = [(x, y) for x, y in sources_findpeaks['x_centroid', 'y_centroid']]\n", + "\n", + "# Round those coordinates to the first decimal place and convert them to be sets\n", + "rounded_centroids_dao = set([(round(x, 1), round(y, 1)) for x, y in centroids_dao])\n", + "rounded_centroids_iraf = set([(round(x, 1), round(y, 1)) for x, y in centroids_iraf])\n", + "rounded_centroids_findpeaks = set([(round(x, 1), round(y, 1)) for x, y in centroids_findpeaks])\n", + "\n", + "# Examine the intersections of different sets to determine which sources are shared\n", + "all_match = rounded_centroids_dao.intersection(rounded_centroids_iraf).intersection(rounded_centroids_findpeaks)\n", + "dao_iraf_match = rounded_centroids_dao.intersection(rounded_centroids_iraf)\n", + "dao_findpeaks_match = rounded_centroids_dao.intersection(rounded_centroids_findpeaks)\n", + "iraf_findpeaks_match = rounded_centroids_iraf.intersection(rounded_centroids_findpeaks)\n", + "\n", + "print('''Matching sources found by:\n", + " All methods: {}\n", + " DAOStarFinder & IRAFStarFinder: {}\n", + " DAOStarFinder & find_peaks: {}\n", + " IRAFStarFinder & find_peaks: {}'''\n", + " .format(len(all_match), len(dao_iraf_match), \n", + " len(dao_findpeaks_match), len(iraf_findpeaks_match)))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "And just for fun, let's plot these matching sources. (The colors chosen to represent different sets are from [Paul Tol's guide for accessible color schemes](https://personal.sron.nl/~pault/).)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", + "ax1.scatter([x for x, y in list(dao_findpeaks_match)], [y for x, y in list(dao_findpeaks_match)],\n", + " s=30, marker='s', lw=1.5, facecolor='None', edgecolor='#EE7733',\n", + " label='Found by DAO \\& find\\_peaks')\n", + "ax1.scatter([x for x, y in list(dao_iraf_match)], [y for x, y in list(dao_iraf_match)],\n", + " s=30, marker='D', lw=1.5, facecolor='None', edgecolor='#EE3377',\n", + " label='Found by DAO \\& IRAF')\n", + "ax1.scatter([x for x, y in list(iraf_findpeaks_match)], [y for x, y in list(iraf_findpeaks_match)],\n", + " s=30, marker='o', lw=1.5, facecolor='None', edgecolor='#0077BB',\n", + " label='Found by IRAF \\& find\\_peaks')\n", + "ax1.scatter([x for x, y in list(all_match)], [y for x, y in list(all_match)],\n", + " s=30, marker='o', lw=1.2, linestyle=':',facecolor='None', edgecolor='#BBBBBB',\n", + " label='Found by all methods')\n", + "ax1.legend()\n", + "\n", + "# Define the colorbar\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('Sources Found by Different Methods')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "*Remember that you can double-click on the plot to zoom in and look around!*" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Custom Detection Algorithms\n", + "\n", + "If none of the algorithms we've reviewed above do exactly what you need, `photutils` also provides infrastructure for you to generate and use your own source detection algorithm: the `StarFinderBase` object can be inherited and used to develop new star-finding classes. Take a look at the [documentation](https://photutils.readthedocs.io/en/latest/api/photutils.detection.StarFinderBase.html#photutils.detection.StarFinderBase) for more information.\n", + "\n", + "If you do go that route, remember that `photutils` is open-developed; you would be very welcome to [open a pull request](https://github.com/astropy/photutils/blob/master/CONTRIBUTING.rst) and incorporate your new star finder into the `photutils` source code - for everyone to use!" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "\n", + "## Image Segmentation" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Beyond traditional source detection methods, an additional option for identifying sources in image data is a process called **image segmentation**. This method identifies and labels contiguous (connected) objects within an image. \n", + "\n", + "You might have noticed that, in the previous source detection algorithms, large and extended sources are often incorrectly identified as more than one source. Segmentation would label all the pixels within a large galaxy as belonging to the same object, and would allow the user to then measure the photometry, centroid, and morphology of the entire object at once.\n", + "\n", + "#### Creating a `SegmentationImage`\n", + "\n", + "In `photutils`, image segmentation maps are created using the threshold method in the `detect_sources()` function. This method identifies all of the objects in the data that have signals above a determined **`threshold`** (usually defined as a multiple of the standard deviation) and that have more than a defined number of adjoining pixels, **`npixels`**. The data can also optionally be smoothed using a kernel, **`filter_kernel`**, before applying the threshold cut.\n", + "\n", + "The `detect_sources()` function returns a `SegmentationImage` object: an array in which each object is labeled with an integer. As a simple example, a segmentation map containing two distinct sources might look like this:\n", + "\n", + "```\n", + "0 0 0 0 0 0 0 0 0 0\n", + "0 1 1 0 0 0 0 0 0 0\n", + "1 1 1 1 1 0 0 0 2 0\n", + "1 1 1 1 0 0 0 2 2 2\n", + "1 1 1 0 0 0 2 2 2 2\n", + "1 1 1 1 0 0 0 2 2 0\n", + "1 1 0 0 0 0 2 2 0 0\n", + "0 1 0 0 0 0 2 0 0 0\n", + "0 0 0 0 0 0 0 0 0 0\n", + "```\n", + "where all of the pixels labeled `1` belong to the first source, all those labeled `2` belong to the second, and all null pixels are designated to be background.\n", + "\n", + "Let's see what the segmentation map for our XDF data will look like." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import detect_sources\n", + "from photutils.utils import random_cmap" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Define threshold and minimum object size\n", + "threshold = 5. * std\n", + "npixels = 15\n", + "\n", + "# Create a segmentation image\n", + "segm = detect_sources(xdf_image.data, threshold, npixels)\n", + "\n", + "print('Found {} sources'.format(segm.max_label))" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, [ax1, ax2] = plt.subplots(1, 2, figsize=(12, 6))\n", + "plt.tight_layout()\n", + "\n", + "# Plot the original data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('Original Data')\n", + "\n", + "# Plot the segmentation image\n", + "rand_cmap = random_cmap(random_state=12345)\n", + "rand_cmap.set_under('black')\n", + "segplot = ax2.imshow(np.ma.masked_where(xdf_image.mask, segm), vmin=1, cmap=rand_cmap)\n", + "ax2.set_xlabel('X (pixels)')\n", + "ax2.set_title('Segmentation Map')\n", + "\n", + "# Define the colorbar\n", + "cbar_ax = fig.add_axes([1, 0.09, 0.03, 0.87])\n", + "cbar = plt.colorbar(segplot, cbar_ax)\n", + "cbar.set_label('Object Label', rotation=270, labelpad=30)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Compare the sources in original data to those in the segmentation image. Each color in the segmentation map denotes a separate source.\n", + "\n", + "You can easily see that larger galaxies are shown in the segmentation map as contiguous objects of the same color - for example, the two yellow and pink galaxies near (1200, 2500). Each pixel containing light from the same galaxy has been labeled as belonging to the same object." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "For a closer look, let's see what the sources we found with `find_peaks` look like in this segmentation map:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "fig, axs = plt.subplots(3,3, figsize=(3, 3))\n", + "plt.subplots_adjust(wspace=0.1, hspace=0.1)\n", + "\n", + "cutout_size = 20\n", + "\n", + "srcs = np.random.permutation(sources_findpeaks)[:axs.size]\n", + "for ax, src in zip(axs.ravel(), srcs):\n", + " slc = (slice(int(src['y_peak'] - cutout_size), int(src['y_peak'] + cutout_size)),\n", + " slice(int(src['x_peak'] - cutout_size), int(src['x_peak'] + cutout_size)))\n", + " ax.imshow(segm.data[slc], cmap=rand_cmap, vmin=1, vmax=len(sources_findpeaks))\n", + " src_id = np.where((sources_findpeaks['x_peak'] == src['x_peak']) & \n", + " (sources_findpeaks['y_peak'] == src['y_peak']))[0][0]\n", + " ax.text(2, 2, str(src_id), color='w', va='top')\n", + " ax.set_xticks([])\n", + " ax.set_yticks([])" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + "\n", + "

Exercises:


\n", + "\n", + "Recompute the `SegmentationImage`, but alter the threshold and the minimum number of pixels in a source. How does changing the threshold affect the results? What about changing the number of pixels?\n", + "\n", + "
" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Analyzing `source_properties`\n", + "\n", + "Once we have a `SegmentationImage` object, `photutils` provides many powerful tools to manipulate and analyze the identified objects. \n", + "\n", + "Individual objects within the segmentation map can be altered using methods such as `relabel` to change the labels of objects, `remove_labels` to remove objects, or `deblend_sources` to separating overlapping sources that were incorrectly labeled as one source.\n", + "\n", + "However, perhaps the most powerful aspect of the `SegmentationImage` is the ability to create a catalog using `source_properties` to measure the centroids, photometry, and morphology of the detected objects:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import source_properties" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "catalog = source_properties(xdf_image.data, segm)\n", + "table = catalog.to_table()\n", + "print(table)\n", + "print(table.colnames)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Creating apertures from segmentation data\n", + "\n", + "We can use this information to create isophotal ellipses for each identified source. These ellipses can also later be used as photometric apertures!" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import EllipticalAperture" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Define the approximate isophotal extent\n", + "r = 4.\n", + "\n", + "# Create the apertures\n", + "apertures = []\n", + "for obj in catalog:\n", + " position = (obj.xcentroid.value, obj.ycentroid.value)\n", + " a = obj.semimajor_axis_sigma.value * r\n", + " b = obj.semiminor_axis_sigma.value * r\n", + " theta = obj.orientation.value\n", + " apertures.append(EllipticalAperture(position, a, b, theta=theta))" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", + "\n", + "# Plot the apertures\n", + "for aperture in apertures:\n", + " aperture.plot(color='red', lw=1, alpha=0.7, ax=ax1)\n", + "\n", + "# Define the colorbar\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('Segmentation Image Apertures')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + "\n", + "

Exercises:


\n", + "\n", + "Play with the isophotal extent of the elliptical apertures (defined above as `r`). Observe how changing this value affects the apertures that are created.\n", + "\n", + "
" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "It is clear that using `photutils` for image segmentation can allow users to generate highly customized apertures - great for complex data that contain many different kinds of celestial sources." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## Conclusions\n", + "\n", + "The `photutils` package provides users with a variety of methods for detecting sources in their data, from familar algorithms such as `DAOFind` and `starfind`, to more complex and customizable image segmentation algorithms. These methods allow for easy creation of a diverse array of apertures that can be used for photometric analysis.\n", + "\n", + "**To continue with this `photutils` tutorial, go on to the [aperture photometry](03_photutils_aperture_photometry.ipynb) or [PSF photometry notebook](04_photutils_psf_photometry.ipynb).**" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## Additional Resources\n", + "For more examples and details, please visit the [photutils](http://photutils.readthedocs.io/en/stable/index.html) documentation." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## About this Notebook\n", + "**Authors:** Lauren Chambers (lchambers@stsci.edu), Erik Tollerud (etollerud@stsci.edu), Tom Wilson (towilson@stsci.edu)\n", + "
**Updated:** May 2019" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "[Top of Page](#title_ID)\n", + "\"STScI" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.6.8" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notebooks/photutils/02_source_detection/requirements.txt b/notebooks/photutils/02_source_detection/requirements.txt new file mode 100644 index 00000000..a0d3505e --- /dev/null +++ b/notebooks/photutils/02_source_detection/requirements.txt @@ -0,0 +1,4 @@ +astropy>=3.1.2 +matplotlib>2.2.2 +numpy>=1.13.3 +photutils>=0.4 diff --git a/notebooks/photutils/03_photutils_aperture_photometry/03_photutils_aperture_photometry.ipynb b/notebooks/photutils/03_photutils_aperture_photometry/03_photutils_aperture_photometry.ipynb new file mode 100644 index 00000000..0368dbe4 --- /dev/null +++ b/notebooks/photutils/03_photutils_aperture_photometry/03_photutils_aperture_photometry.ipynb @@ -0,0 +1,978 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "\n", + "\n", + "[\n", + "\n", + "](http://photutils.readthedocs.io/en/stable/index.html)\n", + "\n", + "# Aperture Photometry with `photutils`\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "##### What is aperture photometry?\n", + "The most common method to measure the flux of a celestial source is aperture photometry. This kind of photometry measures the amount of flux within a region of an astronomical image of defined shape and size (an aperture) surrounding a source. The ideal aperture would capture all of the flux emitted by a desired source, and none of the flux emitted by the surrounding sky or nearby sources. Especially when performing photometry on image data that includes a number of sources with varying size and shape, it is important to perform aperture corrections to account for imperfect apertures and better constrain photometric errors.\n", + "\n", + "The `photutils` package provides tools for performing photometry with apertures of various shapes.\n", + "\n", + "##### What does this tutorial include?\n", + "This tutorial covers how to perform aperture photometry with `photutils`, including the following methods:\n", + "* Creating Apertures\n", + " * Circular Apertures\n", + " * Elliptical Apertures\n", + " * Sky Apertures with WCS\n", + "* Performing Aperture Photometry\n", + "* Calculating Aperture Corrections with Local Background Subtraction\n", + "\n", + "##### Which data are used in this tutorial?\n", + "We will be manipulating Hubble eXtreme Deep Field (XDF) data, which was collected using the Advanced Camera for Surveys (ACS) on Hubble between 2002 and 2012. The image we use here is the result of 1.8 million seconds (500 hours!) of exposure time, and includes some of the faintest and most distant galaxies that have ever been observed. \n", + "\n", + "##### The methods demonstrated here are available in narrative form within the `photutils.aperture` [documentation](http://photutils.readthedocs.io/en/stable/aperture.html)." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + "\n", + "**Important:** Before proceeding, please be sure to install or update your [AstroConda](https://astroconda.readthedocs.io) distribution. This notebook may not work properly with older versions of AstroConda.\n", + "\n", + "
\n", + "\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Import necessary packages" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "First, let's import packages that we will use to perform arithmetic functions and visualize data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from astropy.io import fits\n", + "import astropy.units as u\n", + "from astropy.nddata import CCDData\n", + "from astropy.stats import sigma_clipped_stats, SigmaClip\n", + "from astropy.visualization import ImageNormalize, LogStretch\n", + "import matplotlib.pyplot as plt\n", + "from matplotlib.ticker import LogLocator\n", + "import numpy as np\n", + "from photutils.background import Background2D, MeanBackground\n", + "\n", + "# Show plots in the notebook\n", + "%matplotlib inline" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [shared style file](photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "plt.style.use('photutils_notebook_style.mplstyle')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Retrieve data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", + "\n", + "(Generally, the best package for web queries of astronomical data is `astroquery`; however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with `astroquery`.)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "url = 'https://archive.stsci.edu/pub/hlsp/xdf/hlsp_xdf_hst_acswfc-60mas_hudf_f435w_v1_sci.fits'\n", + "with fits.open(url) as hdulist:\n", + " hdulist.info()\n", + " data = hdulist[0].data\n", + " header = hdulist[0].header" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "As explained in the [previous notebook](01_photutils_background_estimation.ipynb) on background estimation, it is important to **mask** these data, as a large portion of the values are equal to zero. We will mask out the non-data, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Define the mask\n", + "mask = data == 0" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons (counts) per second." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "unit = u.ct / u.s\n", + "xdf_image = CCDData(data, unit=unit, meta=header, mask=mask)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's look at the data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", + "xdf_image_clipped = np.clip(xdf_image, 1e-4, None) # clip to plot with logarithmic stretch\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", + "\n", + "# Define the colorbar and fix the labels\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "*Tip: Double-click on any inline plot to zoom in.*" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## Creating Apertures" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "With `photutils`, users can create apertures with the following shapes:\n", + "\"Examples\n", + "\n", + "Each of these can be defined either in pixel coordinates or in celestial coordinates (using a WCS transformation).\n", + "\n", + "It is also possible for users to create custom aperture shapes.\n", + "\n", + "Any aperture object is created by defining its position and size (and, if applicable, orientation). Let's use the `find_peaks` method that we learned about in a [previous notebook](02_photutils_source_detection.ipynb) to get the positions of sources in our data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import find_peaks" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Calculate statistics\n", + "mean, median, std = sigma_clipped_stats(xdf_image.data, sigma=3.0, iters=5, mask=xdf_image.mask)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "sources_findpeaks = find_peaks(xdf_image.data, mask=xdf_image.mask, \n", + " threshold=20.*std, box_size=30, subpixel=True) \n", + "# Display the table\n", + "sources_findpeaks" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "And let's plot the centroids of each of these sources:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", + "ax1.scatter(sources_findpeaks['x_centroid'], sources_findpeaks['y_centroid'], s=10, marker='.', \n", + " lw=1, alpha=0.7, color='r')#facecolor='None', edgecolor='r')\n", + "\n", + "# Define the colorbar\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('find\\_peaks Sources')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "So thanks to `find_peaks`, we now we know the positions of all our sources. Next, we need to define apertures for each source. First, as the simplest example, let's try using circular apertures of a fixed size.\n", + "\n", + "### Circular Apertures" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import CircularAperture" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# define the aperture\n", + "position = (sources_findpeaks['x_centroid'], sources_findpeaks['y_centroid'])\n", + "radius = 10.\n", + "circular_aperture = CircularAperture(position, r=radius)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", + "\n", + "# Plot the apertures\n", + "circular_aperture.plot(color='red', alpha=0.7)\n", + "\n", + "# Define the colorbar\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('Circular Apertures')\n", + "\n", + "# Crop to show an inset of the data\n", + "ax1.set_xlim(2000, 3000)\n", + "ax1.set_ylim(2000, 1000)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "As you can see, these circular apertures don't fit our data very well. After all, this is the Hubble eXtreme Deep Field, so there aren't any nice, round, nearby Milky Way stars in this image! \n", + "\n", + "Let's use ellipses instead, to better match the morphology of the galactic blobs in our image." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Elliptical Apertures" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import (detect_sources, source_properties, \\\n", + " EllipticalAnnulus, EllipticalAperture)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "In a [previous notebook](02_photutils_source_detection.ipynb), we showed how you can use the `photutils.detect_sources` feature to generate segmentation maps, which identify and label contiguous (connected) objects within an image. Then, with `source_properties`, you can access descriptive properties for each unique object - not just their centroid positions, but also their pixel areas, eccentricities, orientations with respect to the coordinate frame of the image, and more.\n", + "\n", + "Here we'll use the centroid, semimajor axis, semiminor axis, and orientation values from `source_properties` to generate elliptical apertures for each of the sources in our image." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "# Define threshold and minimum object size\n", + "threshold = 5. * std\n", + "npixels = 15\n", + "\n", + "# Create a segmentation image\n", + "segm = detect_sources(xdf_image.data, threshold, npixels)\n", + "\n", + "# Create a catalog using source properties\n", + "catalog = source_properties(xdf_image.data, segm)\n", + "table = catalog.to_table()\n", + "\n", + "# Display the table\n", + "table" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "r = 3. # approximate isophotal extent of semimajor axis\n", + "\n", + "# Create the apertures\n", + "elliptical_apertures = []\n", + "for obj in catalog:\n", + " position = (obj.xcentroid.value, obj.ycentroid.value)\n", + " a = obj.semimajor_axis_sigma.value * r\n", + " b = obj.semiminor_axis_sigma.value * r\n", + " theta = obj.orientation.value\n", + " elliptical_apertures.append(EllipticalAperture(position, a, b, theta=theta))" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", + "\n", + "# Plot the apertures\n", + "for aperture in elliptical_apertures:\n", + " aperture.plot(color='red', alpha=0.7, ax=ax1)\n", + "\n", + "# Define the colorbar\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('Elliptical Apertures')\n", + "\n", + "# Crop to show an inset of the data\n", + "ax1.set_xlim(2000, 3000)\n", + "ax1.set_ylim(2000, 1000)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Clearly, these custom-made elliptical apertures fit our XDF galaxies much better than the one-size-fits-all circular apertures from before." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Sky Coordinates & Apertures\n", + "\n", + "At the moment, the positions of our apertures are in pixels, relative to our data array. However, if you need aperture positions in terms of celestial coordinates, `photutils` also includes aperture objects that can be integrated with Astropy's `SkyCoords`.\n", + "\n", + "Fortunately this is extremely easy when we use the [World Coordinate System (WCS)](http://docs.astropy.org/en/stable/wcs/) to produce a WCS object from the header of the FITS file containing our image, and then the `to_sky()` method to transform our `EllipticalAperture` objects into `SkyEllipticalAperture` objects." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from astropy.wcs import WCS" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "wcs = WCS(header)\n", + "sky_elliptical_apertures = [ap.to_sky(wcs) for ap in elliptical_apertures]" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "If we wanted to generate `SkyEllipticalAperture` objects from the get-go, we could have used that WCS object in the following way:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import SkyEllipticalAperture\n", + "from astropy.coordinates import SkyCoord\n", + "\n", + "r = 3. # approximate isophotal extent of semimajor axis\n", + "\n", + "# Create the apertures\n", + "sky_elliptical_apertures = []\n", + "for obj in catalog:\n", + " # Convert the centroids into RA/Dec using WCS\n", + " ra, dec = wcs.all_pix2world(obj.xcentroid.value, obj.ycentroid.value, 0)\n", + " # Convert the positions to an Astropy SkyCoord object, with units!\n", + " sky_position = SkyCoord(ra, dec, unit=u.deg)\n", + " \n", + " # Define the elliptical parameters, now with units\n", + " a = obj.semimajor_axis_sigma.value * r * u.pix\n", + " b = obj.semiminor_axis_sigma.value * r * u.pix\n", + " theta = obj.orientation.value * u.rad\n", + " \n", + " # Convert the theta from radians from X axis to the radians from North \n", + " x_to_north_angle = (90. + header['ORIENTAT']) * u.deg\n", + " x_to_north_angle_rad = x_to_north_angle.to_value(u.rad) * u.rad\n", + " theta -= x_to_north_angle_rad\n", + " \n", + " # Define the apertures\n", + " ap = SkyEllipticalAperture(sky_position, a, b, theta=theta)\n", + " sky_elliptical_apertures.append(ap)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Unfortunately, you can't plot SkyApertures. However, you can use them just as easily to perform aperture photometry!" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "----\n", + "## Performing Aperture Photometry" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Now that we have aperture objects that fit our data reasonably well, we can finally perform photometry with the `aperture_photometry` function. This function takes the following arguments:\n", + "\n", + "* **`data`** - the background-subtracted data array on which to perform photometry.\n", + "* **`apertures`** - an aperture object containing the aperture(s) to use for the photometry.\n", + "* **`error`** (optional) - an array of values that represent the pixel-wise Gaussian 1-sigma errors of the input data.\n", + "* **`mask`** (optional) - a mask for the `data` to exclude certain pixels from calculations.\n", + "* **`method`** (optional) - how to place the aperture(s) onto the pixel grid (see below).\n", + "* **`unit`** (optional) - unit of `data` and `error`.\n", + "* **`wcs`** (optional) - the WCS transformation to use if `apertures` is a `SkyAperture` object. \n", + "* **`subpixels`** (optional) - the factor by which pixels are resampled (see below).\n", + "\n", + "The following methods are the options for how to place apertures onto the data pixel grid:\n", + "\n", + "* **exact** (default) - calculate the exact fractional overlap of each aperture for each overlapping pixel. This method is the most accurate, but will also take the longest. \n", + "* **center** - a pixel is either entirely in or entirely out of the aperture, depending on whether the pixel center is inside or outside of the aperture.\n", + "* **subpixel** - a pixel is divided into `subpixels` x `subpixels` subpixels, each of which are considered to be entirely in or out of the aperture depending on whether its center is in or out of the aperture. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from photutils import aperture_photometry\n", + "from astropy.table import QTable" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's see what this looks like using the first aperture in our image:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# The CCDData mask will be automatically applied\n", + "phot_datum = aperture_photometry(xdf_image, elliptical_apertures[0]) \n", + "phot_datum" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The `aperture_sum` value is what reports the number of counts within the aperture: 3.47 ct/s.\n", + "\n", + "And, just as a check, to make sure our sky apertures give basically the same answer..." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# The CCDData mask will be automatically applied\n", + "sky_phot_datum = aperture_photometry(xdf_image, sky_elliptical_apertures[0], wcs=wcs)\n", + "sky_phot_datum" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Woohoo!\n", + "\n", + "Unfortunately for our purposes, the `aperture_photometry` function can be only used alone for one of the two cases:\n", + "* Identical apertures at distinct positions (e.g. circular apertures with `r = 3` for many sources)\n", + "* Distinct apertures at identical positions (e.g. two circular apertures with `r = 3` and `r = 5` for one source)\n", + "\n", + "Since our elliptical apertures are distinct apertures at distinct positions, we need to do a little more work to get a single table of photometric values.\n", + "\n", + "(This step will take a while, almost 5 minutes, so hang tight!)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# The CCDData mask will be automatically applied\n", + "phot_table = aperture_photometry(xdf_image, elliptical_apertures[0])\n", + "id = 1\n", + "for aperture in elliptical_apertures[1:]:\n", + " id += 1\n", + " phot_row = aperture_photometry(xdf_image, aperture)[0]\n", + " phot_row[0] = id\n", + " phot_table.add_row(phot_row)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's take a look at all these apertures we've made:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Display the table\n", + "phot_table" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "There's only so much you can learn from looking at a table of numbers, so let's explore alternate ways to examine these data." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "plt.figure(figsize=(8, 5))\n", + "\n", + "values = [phot.value for phot in phot_table['aperture_sum']]\n", + "logbins=bins = 10.**(np.linspace(-1, 2, 100))\n", + "plt.hist(values, bins=logbins)\n", + "\n", + "plt.yscale('log')\n", + "plt.xscale('log')\n", + "plt.title('Histogram of Source Photometry')\n", + "plt.xlabel('Count Rate [ct/s]')\n", + "plt.ylabel('Number of Sources')" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "plt.figure(figsize=(8, 5))\n", + "\n", + "plt.scatter(table['area'], values, alpha=0.5)\n", + "\n", + "plt.yscale('log')\n", + "plt.xscale('log')\n", + "plt.title('Count Rate v. Aperture Area')\n", + "plt.xlabel('Aperture Area [pixels$^2$]')\n", + "plt.ylabel('Count Rate [ct/s]')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + "\n", + "

Exercise:


\n", + "\n", + "Re-calculate the photometry for these elliptical apertures - or just a subset of them - using the `subpixel` aperture placement method instead of the default `exact` method. How does this affect the count sum calculated for those apertures?\n", + "\n", + "
" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "# Aperture Corrections" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "We've done photometry with some lovely apertures, but unfortunately even using elliptical apertures with unique sizes and orientations does not account for an important source of extraneous flux: the sky background.\n", + "\n", + "## Local Background Subtraction\n", + "\n", + "In the [background estimation notebook](01_photutils_background_estimation.ipynb), we explored how to perform global background subtraction of image data with `photutils`. However, you can also use `photutils` to perform local background estimations for aperture corrections.\n", + "\n", + "To estimate the local background for each aperture, measure the counts within annulus apertures around (but not including!) each source. In our example, we defined elliptical apertures with `r = 3` to measure the counts within each source. To calculate the background for each source, let's measure the counts elliptical annuli between `r = 3.5` and `r = 5`." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "r_in = 3.5 # approximate isophotal extent of inner semimajor axis\n", + "r_out = 5. # approximate isophotal extent of inner semimajor axis\n", + "\n", + "# Create the apertures\n", + "elliptical_annuli = []\n", + "for obj in catalog:\n", + " position = (obj.xcentroid.value, obj.ycentroid.value)\n", + " a_in = obj.semimajor_axis_sigma.value * r_in\n", + " a_out = obj.semimajor_axis_sigma.value * r_out\n", + " b_out = obj.semiminor_axis_sigma.value * r_out\n", + " theta = obj.orientation.value\n", + " elliptical_annuli.append(EllipticalAnnulus(position, a_in, a_out, b_out, theta=theta))" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image)\n", + "\n", + "# Plot the apertures\n", + "for aperture in elliptical_annuli:\n", + " aperture.plot(color='red', alpha=0.4, ax=ax1, fill=True)\n", + "for aperture in elliptical_apertures:\n", + " aperture.plot(color='white', alpha=0.7, ax=ax1)\n", + "\n", + "# Define the colorbar\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')\n", + "ax1.set_title('Elliptical Annuli')\n", + "\n", + "# Crop to show an inset of the data\n", + "ax1.set_xlim(2000, 3000)\n", + "ax1.set_ylim(2000, 1000)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Calculating Aperture Corrections\n", + "\n", + "Now that our apertures have been defined, we can do photometry with them to estimate and account for the background. The aperture correction is calculated by:\n", + "- Calculating the count rate within each annulus using `aperture_photometry`\n", + "- Dividing each annulus' count rate by each annulus' area to get the mean background value for each annulus\n", + "- Taking the mean of those annulus means to get a mean background value for the entire image\n", + "- Multiplying the global background mean value times the area of each elliptical photometric aperture, to get the estimated background count rate within each aperture\n", + "- Subtracting the estimated background count rate from the photometric count rate for each aperture\n", + "\n", + "(Just like when we did photometry with the elliptical apertures above, the below step will take almost 5 minutes.)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# The CCDData mask will be automatically applied\n", + "bkg_phot_table = aperture_photometry(xdf_image, elliptical_annuli[0])\n", + "id = 1\n", + "for aperture in elliptical_annuli[1:]:\n", + " id += 1\n", + " phot_row = aperture_photometry(xdf_image, aperture)[0]\n", + " phot_row[0] = id\n", + " bkg_phot_table.add_row(phot_row)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Display table\n", + "bkg_phot_table" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "You might have noticed that these background count rates are *really* small. In this case, this is to be expected - since our example XDF data is a high-level science product (HLSP) that already has already been background-subtracted." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Calculate the mean background level (per pixel) in the annuli \n", + "bkg_area = [annulus.area() for annulus in elliptical_annuli]\n", + "bkg_mean_per_aperture = bkg_phot_table['aperture_sum'].value / bkg_area\n", + "bkg_mean = np.average(bkg_mean_per_aperture) * (u.ct / u.s)\n", + "print('Background mean:', bkg_mean)\n", + "\n", + "# Calculate the total background within each elliptical aperture\n", + "bkg_sum = bkg_mean * table['area'].value\n", + "\n", + "# Subtract the background from the original photometry\n", + "flux_bkgsub = phot_table['aperture_sum'] - bkg_sum\n", + "\n", + "# Add this as a column to the original photometry table\n", + "phot_table['aperture_sum_bkgsub'] = flux_bkgsub" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Finally, let's see the difference between our original count rates and our background-subtracted count rates (it should be small for us!):" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Display table\n", + "phot_table" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "plt.figure(figsize=(8, 5))\n", + "\n", + "values = [phot.value for phot in phot_table['aperture_sum']]\n", + "values_bkgsub = [phot.value for phot in phot_table['aperture_sum_bkgsub']]\n", + "logbins=bins = 10.**(np.linspace(-1, 2, 100))\n", + "plt.hist(values, bins=logbins, alpha=0.7, label='Original photometry')\n", + "plt.hist(values_bkgsub, bins=logbins, alpha=0.7, label='Background-subtracted')\n", + "\n", + "plt.yscale('log')\n", + "plt.xscale('log')\n", + "plt.title('Histogram of Source Photometry')\n", + "plt.xlabel('Count Rate [ct/s]')\n", + "plt.legend()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "# Conclusions\n", + "\n", + "The `photutils` package provides a comprehensive toolkit for astronomers to perform aperture photometry, including customizable aperture shapes that allow for more precise photometry and easy photometric correction.\n", + "\n", + "**To continue with this `photutils` tutorial, go on to the [PSF photometry notebook](04_photutils_psf_photometry.ipynb).**" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## Additional Resources\n", + "For more examples and details, please visit the [photutils](http://photutils.readthedocs.io/en/stable/index.html) documentation." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## About this Notebook\n", + "**Authors:** Lauren Chambers (lchambers@stsci.edu), Erik Tollerud (etollerud@stsci.edu), Clare Shanahan (cshanahan@stsci.edu)\n", + "
**Updated:** May 2019" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "[Top of Page](#title_ID)\n", + "\"STScI" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.5.5" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notebooks/photutils/03_photutils_aperture_photometry/requirements.txt b/notebooks/photutils/03_photutils_aperture_photometry/requirements.txt new file mode 100644 index 00000000..a0d3505e --- /dev/null +++ b/notebooks/photutils/03_photutils_aperture_photometry/requirements.txt @@ -0,0 +1,4 @@ +astropy>=3.1.2 +matplotlib>2.2.2 +numpy>=1.13.3 +photutils>=0.4 diff --git a/notebooks/photutils/04_psf_photometry/04_photutils_psf_photometry.ipynb b/notebooks/photutils/04_psf_photometry/04_photutils_psf_photometry.ipynb new file mode 100644 index 00000000..0a6aa615 --- /dev/null +++ b/notebooks/photutils/04_psf_photometry/04_photutils_psf_photometry.ipynb @@ -0,0 +1,253 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "[](http://photutils.readthedocs.io/en/stable/index.html)\n", + "\n", + "# PSF Photometry with `photutils`\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "##### What is PSF photometry?\n", + "A more specific form of photometry than aperture photometry, PSF photometry takes into account the shape of a source's point spread function (PSF). The PSF is a model that represents the distribution of light from a point source as it falls onto a detector. An example of a basic PSF is simply a 2-D Gaussian, while more complex PSFs can include distortion, diffraction, or interference effects associated with a particular telescope. For instance, the PSFs from the Hubble Space Telescope and the James Webb Space Telescope have been meticulously modeled, and can be simulated with the [Tiny Tim](http://www.stsci.edu/hst/observatory/focus/TinyTim) and [WebbPSF](https://github.com/mperrin/webbpsf) software packages, respectively. However, for datasets that do not have readily available PSF models, such models can be statistically generated by analyzing the image itself.\n", + "\n", + "The `photutils` package provides tools that combine background estimation, source detection, and model-fitting to perform PSF photometry on image data.\n", + "\n", + "##### What does this tutorial include?\n", + "This tutorial covers how to perform PSF photometry with `photutils`, including the following methods:\n", + "* Gaussian PSF Photometry\n", + "* Iterative Subtraction\n", + "* Point Response Function (PRF) Photometry\n", + "\n", + "The methods demonstrated here are available in narrative form within the `photutils.psf` [documentation](http://photutils.readthedocs.io/en/stable/psf.html)." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
**Warning:** The PSF photometry API is currently considered experimental and may change in the future. The photutils development team will aim to keep compatibility where practical, but will not finalize the API until sufficient user feedback has been accumulated.
**Important:** Before proceeding, please be sure to install or update your [AstroConda](https://astroconda.readthedocs.io) distribution. This notebook may not work properly with older versions of AstroConda.
\n", + "\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Import necessary packages" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "First, let's import packages that we will use to perform arithmetic functions and visualize data:" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [], + "source": [ + "import numpy as np\n", + "from astropy.io import fits\n", + "from astropy.stats import sigma_clipped_stats, SigmaClip\n", + "from astropy.visualization import ZScaleInterval, ImageNormalize\n", + "from photutils import make_source_mask\n", + "from photutils.background import Background2D, MedianBackground\n", + "import matplotlib\n", + "import matplotlib.pyplot as plt\n", + "from matplotlib.colors import LogNorm\n", + "% matplotlib inline" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's also define some `matplotlib` parameters, to make sure our plots look nice. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": {}, + "outputs": [], + "source": [ + "matplotlib.rc('font', family='serif', weight='light', size=12)\n", + "matplotlib.rc('mathtext', bf='serif:normal')\n", + "matplotlib.rc('axes', titlesize=18, titlepad=12, labelsize=16)\n", + "matplotlib.rc('xtick', labelsize=14)\n", + "matplotlib.rc('ytick', labelsize=14)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Retrieve data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "We have place the data for this tutorial in the github repository, for easy access. The data were originally retrieved from the STScI archive: https://archive.stsci.edu/prepds/udf/udf_hlsp.html." + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": {}, + "outputs": [], + "source": [ + "with fits.open('data/h_udf_wfc_v_drz_img.fits') as hdulist:\n", + " v_data = hdulist[0].data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's look at the data:" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkQAAAHACAYAAABDKXcJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3Xd4VFX++PH3vdOTSSa9F9JDDz2EIj0UAVFRLGCXYgOs69p23bWsi3WlqhQLCgoIAgYU6YTQQguQXie9l+lzf3+g+/hz97uLihvR83qe+wRmzhw+c+4wfDj3c86VFEVBEARBEATh90zu7AAEQRAEQRA6m0iIBEEQBEH43RMJkSAIgiAIv3siIRIEQRAE4XdPJESCIAiCIPzuiYRIEARBEITfvSs2IZIkaYokSUckSdorSdIBSZL6d3ZMgiAIgiBcmaQrcR8iSZL6AXuAgYqi5EiSdDWwGuiuKEpV50YnCIIgCMKV5kpNiD7lYuzXfe+xHOAzRVGe7rzIBEEQBEG4Eqk7O4CfaAzw8g8eOwKMBURCJAiCIFxx0kd6KvUNrsva57FTtgxFUcZf1k5/o664hEiSJD/ABFT+4KkqYML/PiJBEARB+PnqG1xkZURd1j5VoXkBl7XD37ArLiECPL/9afvB4zbA44eNJUm6F7gXQNJq+2mCgn7Z6ARBEITfHWdjA662dqmz4xB+uisxIWr/9qfuB4/rgI4fNlYUZTmwHEAXGamEPTz/l41OEARB+N0xL3r9Z/ehAG7cPz+Yn0CSpClcLDmxACrgIUVRjv6H9t7Am0D3b9vvBP6oKIrze21CgaVAEBf/jf5IUZS//5u+bgXeAhYoirLqB8+NAFYBxT942WxFUS78mPf431xxCZGiKA2SJDUBIT94KgQo6ISQBEEQBOEyUHAp//uE6NuV2x/x/6/czpAk6T+t3F4FtCuKMkCSJC2wG/gz8OS3fcrAFmC7oihPS5JkAo5LktTy7UQFkiRpgLVAHeDzH0JcpSjKcz/3ff43V+o+RF8BP9x3qP+3jwuCIAiCcOn+AGQoipIDoCjKF0A1cN+/ayxJUg9gGvC3b9vbgdeB+ZIkGb9tNhFIARZ926YZWAY8JUnSd5cWNcByRVHm/BJv6se6UhOil4B0SZK6AkiSNBEIBd7u1KgEQRAE4Se6eMlMuazHJRoD/PDy2Hcrt/+v9lbgzA/aG4Ch32tToChK0w/aRAJJAIqidCiKsuNSg/ylXXGXzAAURTkmSdItwBpJkr673pkuNmUUBEEQrmS/QA1RgCRJ3092ln93yQp+8srtWKBa+f83Mqz63nPf/fx3fX733PlLCx+AwZIkZXBxUVUj8A9FUTJ+xOsvyRWZEAEoirIZ2NzZcQjCb038wkzyX03t7DAEQbg86hRF+U+3tvpRK7e/95p/157vveZS2lyKZqAEeERRlBZJkkYB2yRJmqkoyvof0c9/daVeMhME4RJ4Ff24v+Ixm+1kmLN/oWgEQfhPFBRcyuU9LsGPWrn9vdf8u/Z87zWX0ua/UhTlhKIo9yqK0vLt73cB64E/Xmofl0okRILwG1UwYymtMZc+/R6/MBPp6VoAVFaJghlLf6nQBEH4lVAUpQH4sSu3C4Gg7xVHf9ee772m8P/o8/ttfqoCIP5n9vEvREIkCL9BOTe+9c9fKxqFF69e+x/bxy/MJMOczQC/EmI/m03CCvMvHaIgCP9GJxVV/9iV2zu5WEDd/QftLcCB77WJlyTJ5wdtyn7M/kGSJM2XJCnmBw+HA6WX2selEgmRIPzGFMxYyrQe47ijdBivTV7D+5OWcIOx+f9sH9WzkuZbL9YMvRB8CsXgov5tDRNHT+fw9EXEL8z8X4UuCL9rCuBCuazHJfqPK7clSfqLJElnJEnSAyiKchbYCDz67fMa4CHgdUVR2r7tczuQDSz4to03F+8a8ZcfOSwpwH3fzUZ9G+ONwOIf2c9/JRIiQfiNOWB1Y1nnzcqoffzp77cxc8s8AK65KutfkhvFz0HLp2G0Rcgk7p1F0ntzSbz7KA6XjKJREaDyFDVFgvAbpyjKMeC7ldt7uVif8/2V23ouFkJ//xLZ7QCSJB0BDgMHgWe+16cbmAL0kSTpELCHH6xw+/b1iyVJ2v3tb5+QJGm3JEnfv6HbUiAKOCBJ0j7gfeAxfoFtdiTl0oqufhPErTuE37rvLn0BjJ98C3JzB8/v/IQnYwaSYc7m2vyxtA+v/ecqsqCutdTUeVMweiWjZt3FrjXv/p99x338q9g7TRB+lcyLXsdWVvaz7mWW0lur7NweeLlCAiAo3Hzsv6wyE74lZogE4TfCs0Km9Jk0ZpUMJz0shZkffYkrv4h4jYsMczZfdujYEL+TDHM2ikahYMZSWvYGc2OPY4y94XbaFrSQHpYCQNK+Wf/SvyiyFoRflgKdscpM+JZIiAThNyB+YSaKCmzxVjKLY8h/LZU1SZEA3BAxmPSwFO7bfAcAw09PY/fURdxcNBJdk8KW4h7k36pF+iSAyoVppF87C/VJI9cXjCHmy7v/5c8RBEH4LRIJkSBc4QpmLEVO6YZHpULCrOPM6n6Y+AWZXHXKgmNcf1JOQIY5m4IZS/mw1Z9Pu33A1BN346m2c/zpJUyNOU23v9XgvzkH/ehaMjasQdsCJ47GI2tdAPQ8fDPpYSmYH0sTM0WC8AtyX+ZDuHQiIRKEK9zonCm4s3PQdLjJMGezr5eeppmDCdY0s2vVO2T3gfhv7iBm291015oZ8uEjeK4xYVDZ6XpgJnur41m/dx15i2PQqFzUudrxLnaibZYpGLWSu0qHEjYth5r703B8u6etz/mfVSohCILwqyMSIkG4gr07dTnqMaVkmLOpu7GDSQMnAdCUBHeZqv5ZE+T/pZ7Eu4/yeMwgvIrB87PDbMnqw7kh72NIL2LIC/Nx1emoqjPR6la4d9FnaJuh52vzeDdqPxnmbCyB4Ey4uMFswPJDvDllVSe9a0H4bVIu85L7H7HsXkAkRIJwxYpfmMldh27DPTSFpU3h6A56sTVrK/V3D6bnVXnEfHEPd1woYXPFERQV5L+WSv5rqTSl2Qg86IPk4eKO0mEULEqleZCVgGMy8aG1TH/+UV7NHU3Iawdp72nlQfMA2txWooeWola7eLK6F6XPpPFmfLK4fCYIl5MCrst8CJdOJESCcIWS+nTHeNRAc5yBOT4VnHxsMcnvzOXon5fwadxXaE02Xn1xBl2/no3dS0Lxt4MEqXFFHDibgOKW2L+nB/k3LaVw7HsEZhRS/lUU/qfaaKj1pvj5wSTMOs6Obf25LmoIBVWBvNRnIwONhVij7eS+e3Elryrskm9LJAiC8KslEiJBuAIZurTiMmoJX5tPW6REufPi5rDn714CQL/n5nJ/jz1kvbAEWe2muYeD6NB6Elc2c3RfMon3HGHhoJ30Tsv7Z5/Fd8QR8eJBiqYZeWDgLrpsaSPv7UHY/d1UbUhk29B/4CVb8JE76B1fRtGEd4hdP4fc4WsI3yvKNwXh51IQRdWdSSREgnCFuX5kJmdSP+SPq1az7cQOYscUEaE2MnHEdfTKugm4WOPzXv7F5fb5I1bhc1rD4IAizs/2Iv7danock1l2YSgNf4om9bE5JK+Yx1/vXENEphFHgJNgTTOqvHJ8zsjcPOQgoc/LnLaFMtrg4q+3zeJUUQQAioeLeRWptESqueaqrM4cFkH4DZBwXeZDuHQiIRKEK0jik9kcfH4QSftm8czCuxn0xFy+SNxO12XzGLfxGLazPhf3HMrLJSmghrYbUhn0xFzah7bxQvAporcoKOZqtm9IxVrgjb60Cb/tuTiNbkJUzRxZ3wvZ4OSZI1OwDohDZYPPNg6jcpiJKqeJ1MfmUDhbQnFLpIf3IXGFlRqrEZsfHKiOJbZ3RWcPkSAIwk8ibt0hCFeIpCU1bNuzAYD0sBReKMriyZiBAFx3roZ7TWZiv7oTWjUYw1swbPBBvqmG6ZEn+MfX4/DPlrB7Szi8wDi4FnmtPw3dJPT1En7nHGgersL8VSTRn9eRP8ufmCcOcdUpC9ufHUFTgor2bjYujF2ORlKRtG8W0gVPgo650DU50JibCf+gmtJB7eSuGIDcqurMoRKE/7nLceuOHr20ymdbAy5XSAAkR1WKW3dcIjFDJAhXiNphwaS8OI9r88cC8NS029HsDsW035/PugaRuHouV3c7jd9xmY4CE7JDwUPj4Iv5I9E2yQRuvkBrogt9nYJ7UwBWX5nQQy6cHlA6XqawMgCVHexBRiaPPUzDF4msXz6a1kgVD9/5KYYLOkafuZ70sBS63HiKoGMugh4ppHCqjvw/e7Ei8gAAyW+389m0NzpzqARBEH40kRAJwhWgYMZSvvnza7h0ULk0jgxzNqUTfSjdEkPz0HoatybgWSaRP7MLGouCV4HM5D98Q+mxcMz32nEldHDhjWgUnYvjzyxh09OvoLYqOPUSnuUKHhUqTAf0+KWbqbhKz66VqXT1ryLweDvtYQq3e9fgWalQnhtE1fw06rYk4pVVyoW6IApmLOXxlAwAHOP68/Km93j8pnvEknxB+AlEDVHnUXd2AIIg/GfxCzPpXTIPZWQjEctOU3ZfT9IWzuGGP+xmkGcBLbP1HGmTcN9VwM5xSUQ8VsuYDSd4f+l4pAgFRZFYNGAdbyck0nZDKgNDp+OhcTBu3gEOPJPKdc9m8PaOcdx/2xYONsdR2c2b/l0KKPhjMgF/K+bj8O2khw3EtkDCJ7qJ1g4/nOU+zPwqi9WLJ3Kt/1jaHTrWdS2n6iktj3ZJxbajhUlpU4js7qJstLh8JgiXQgGRxHQiMUMkCL9i382ynHx8MbYzPrhbW/lozqt8+PLf2fXUUF6L78oNxmYyPhzMhiP96RFYRcXEYNaX9cHz6ioS3i4jcJ2BhVk38kpxJpXDFaZHH6fU7M/W1UMpnermrQOjCTwGeZZgct7rjr1dS+mgdoqnqDlyPoYnYwbSft0gQl47yKy4w7g83XjnqnnrxEiG3H6M3Log8k9HYH40DWuwiwxzNrpxxZg+bMXxQD2m+MZOHkVBEIT/TiREgvArlh6WQuuMVGK/upMuTx1CGtCTXlo9942ZxZ5ly//Z5tTDi5FsMtnbu9IerlBV5E/HpmCKZ0ZRPUgm/rbT9NLqUftbWfv2OLxO6tA1KnSNM6P1seF3pJZtm1ORptYTvEtN1YI0rk47TsABDRWPp7HgxbUUfNiHVcsmErfeTmtvG+GBTez4qi8hb+qIn5+JJcVC4XXLAMhb3ZemewJpbPNgVa/VJHzU1pnDKAhXDLciXdZDuHQiIRKEX6nvZoeqxjt4KfUzMszZKEdOkx6WQuObEulhKai6J1H+ZBoj77yHbr1L6DH+Av6nwS9bRXOSgv9ZJ0qEhf5HbQDIuZ64tBLvP/QqWS8uoeBgNP4bPKgeGUTCqEJMi4zU95Jw6uHNsCM4PSU0aQ289NItSBV6XFqQ7G4Sbj9GnKmOGybsR7X7OKqEWBKfbQLgqtn3Ujj2PVznCzBu8WJB/g18ufkDUVMkCP/Fd5fMRA1R5xAJkSD8CkVvc5B+7mrKn0wj4fZjvJsYQ/9n5sLAnmSYs2nMCmZKTj32QE+umnackhluyjfEcDwzAU2Hm4b+TuRgK613tqDVOtlRkQyA3L0FlwGmfnU/vV+Zx81X78HqK+GccDGZqRqkx7sA3DoY9MRc7F5gO+qHd4mdsSNP4F3iQvpLPRPONnF6RQ8y3hhK7nv9eeurNbT1CCTp3bkoKohfO4fyxwfh0sPX3TaT8uI8ZpUMJ/7j9s4cVkEQhP+TSIgE4VcmfmEmskuB0eVEvnIUgJabU7FObMHloSE9LIXoZw/y2fx07l+xjrwFyST9vR3XyCY845vRzakkOKIRv+0GmktMWJr07Oy9hphN92Kp8UAzuAFttQaXFgo6Aki96wTaz30I1LWRMiWHpiQFRYKGbmCoVbAmWCm8QSbGUEv0/Fzyzoez+aHRvPf0ayx6egnRn0lsa+vOnxe9Q/T2DvYuXo5nXDMRLx7E67pKur81DyQI1rWSd5uHmCkShP+DgoQL+bIewqUToyUIvzL5r6ai+uY4qm/CyPtbX9QR4Xh/lEnoqxraIrQA5L01iPp57fzt6VupGeCB+8x53G6J1iYPyg+F05AdiEsHYXshYaWTB8rSCcxSIbkkrovJRkpow2FUOHiwG7s396VuuJ1dZ5M59mU3bhp1ALuvi+gvbbSMaSf54TK6bISM+68iMy8WdbPMrH9s4ZGbZmN2+LJ7xQq2DQjnnvWzQZJIeXEeLbVGLFMHwhuBKP1aOPL4W2zdOJjCaRdrjLZft6gzh1gQBOFfiIRIEH5FjLHNqMM7yH23P9VtRiYPP8qdu/ZxX14uzXEGMl56DSWtN0jgPuhLfQ8J5apGSp9J46+9Picxqgpdk4T/GYWmkVYqr7GjarfT5tDREgvaBhXvZg7D53NPHNE2jCUyvkOqKEp/F02Nhth3i1m7ewiK3k3BDWqCfNqomRKP7HDT8mgrilPC4evib2fGUTHSk9f+MoOJw6fx7Jl9LJ2+nOpHbUgKPD5kG1UzbOxesQJVlje9lj2AugN6LZpH76ybeCB6CIpYjS8I/0IUVXcekRAJwq+Erl7Gd5kR7wxPbuh7lGP91tHNw8zyxFimeHYQkFXHjXEjKJ3vRh/STtAxGx6VEm0tBrqPzWXhjlso3h/FqJuzaOgqER1cjzbXwPmHDDwRuQ1buAN9HahaVLg1EopNhXNEMwd6XbwdiCvSyoiMC3iYZUynNehr1LgUiYbebkomalB94I+kdaPqkPHI8MKjUqEmzYUrv4iHnr2fRaXpuA/6IqfXMcozlxC/FiYNmUrovnZiPq2jtYeNUw8vJuSac2SYs9G0SvTuX9DJoy4Ivx6iqLpziYRIEH4F4hdmMmjyaXTbj+D33iFO3tmdVS1B/OPCVaj8/QDIe9qThg1RhC/VYq32JOqvuQS9fRDcEkXvJ5A4/wTn717C5rO9sIc7aLdr0TdAUnQVj903jyFd82lJduHyd2D3kpCsMtfEnqLnq/Po88I8AjL0bH10FAOvPYV2fC2KrNC6J5jC65fh8nVSlyLRK7oC3BJWP4n6Pm4S52Yx4WwTSfPO4nwykMk37SfkQRuTDtzHs/GbyZ0bRvuzrbhycsGmIj0sBYDEvbNAgqrFcRiqxdeQIAidT3wTCUInK5ixFPfXkZhTW3Ff1YdpObWUPw0vfXI9rZVeFD1wcYVY9HKZwDkWHF4qCqctY8/h7uS/mgptaizBEnW396PbwVvR5+nBJlObGwBjG8g/FkX1AA2n13dD7WfF67SOQTNPELXNzdpdQ/AdW4nVD+onWGic3cbxNb1wbA0k+os2nH1bif9oDijgWS5RsDWOmL7lhB7ooG+fAiqeSKPG7s0joTuQDp3k8MMDqB4VBsDHdal4F8CBXhuwTRxAUmIFNfel0f5lLI/33kFIpou6FIkDDywifl1HZ54CQfiVkHAp8mU9hEsnRksQOlmPN+ZRVBWAY1x/CqfqmONTgf9KT6RurSAr2PxdTBp0NbXzLcRtrGbvkuXEfX0HhioZD7OMZ6kKU4GbhqE2nHleaFog4QM7KOA84Idb50bXCF6lLsIDmmhNdrAzpxtjXt5Hz35FlOcHITsg6HM9LZVeOMY0M+Ge/bhfbOL80PfxPSuhrtPQMaQN9ZAGyup9qJjvJDszgfYoJ5s+HcoTk26j5r40dFWtJN91DsoNrIg8gGFaNZOGTKX0RhfKqAqC3j7I8OB8lvx9Gi1RavS1EibZwJT3vmF++vbOPhWC0KkUwI18WQ/h0onREoROVDBjKeEvHyQlqozWCA2mPInHq1PYs2w5thoPiiavIPkvReQ8GYZKdnOn/37urxjEnD57kV3AkCZMhS7aQ2R0RXrmTdmO7AR1kwVViIWQ9DKit7poHWDBPFyi4YtwUCtITRo2FPemh8mMKUeFbnA9jckyRVOXY8vzZt3XaYR6NBO74y5WPfsq8R82EvuKi+tisol8S41pgxHCrCy8KgNrgo2i57S0xri5cK8fB04lggK3Fo+gLiuYrQc+R1Wpo/XGVOKO6Mla0J9PnnkF28gWvvsPrI+qg3VPjSd+YWanng9BEH6/REIkCJ0kfmEmX3boUPmYOPNNAnUDXLz32Os8EXCQKXnjSbjvMADzD3xDZAa80v1TXjJP4Ost/TjS1AWXFsL/ImGcV862+X/DrVF4I2s0Tb0cFE33JzqogeIjEdgXNKDNN6AOshB40oq6RkNQUi2xvvV8eGwQSMB2P7RN0P3QLSgqBXW7zNFtPZDUbh7Mv5HzDxkpekxFlc2E4S+V+N9bglSu550Vk4haJxMTUI/Lw41ffAN6fwtIcGJLN8ZPOgLA8BGnsfpK5DSGUHSNlmlvPoal1oPTCxYDcItXPYte/QcAUoi1U86HIPwaiKLqziMSIkHoJBnmbBauvIuSud05f/cSxvU7zcMF06lygXPyxZqa2J138vjZ66gaqOLjulTOfp6Mb66bwlWJyA4omubNhQvhDNsxH22zRPBXGsJ3yJgG1pCfG4rkhsY2D+6/cQuOVi1FU7U4Axz4zpco/CgBjYcD61WttAyz0NLTjnzYhClXQpEV5t+8idilCsGGVgL3aXDUGNh+vhvnDsdwNjeCuPWtTL59H0nPnaG6zYjGz8o9cftxFhpRd0j4XXARp6/lyw4dp+tDCVx6CM9ZFvxOSUjDGnltzEdMHHEde61wzGbnmaJryFvVj7ibs8VMkSAI/3MiIRKETlAwYymxn80m577FdMQ4SA9LYVnEIYYGFvDI0OkUPN6d3GUDuL/fbh5K2oVLr5D1cW8GXnsKl1aivp+LiPQS5KQ2UCv4H9YwfsYhGq62ELEwD5POSmRsLW4N5KR9wKbZY4hZ58bDLDOixwVynzViH9uCJCukRRYxOv4C/oc0tEe68LmhgtnTMthQ2YeCe2TyGgMxmh2k9buAV5YBp9HF6jEryJ1p5FBdDPs39qH1rD+GTCMvfT0ZIi24dArS7Bre3DqRR965i9oSX8r/kMY9+w5g95JoqTGy+PbrsS12oMXFkzEDUUZVUDjuXTLM2cgeHiR82NrZp0kQ/qcURRRVdyYxWoLwP1YwYykJu28n4YHDDLt/NucnLWZpyX6GzJ/DgfmDyLsvCpdewfuchk1/HMPaWemoQi3YfGBPQQKO6xuQDC6aV0QCoK1R09Dbzd7XU3ltwCcMNBWTWxJCWUkALm8nfZ+fS/7NGmr66/ApcLH7eFc8D3vg86ERd4knPb3K2Xm4F+2REopGweWWWehXSMWX0dyWkknjBT+KrpfI2tMVa6CCvlrNbbvuRtMi0dDuQdo1Jwk+7Cb15hMkrOnA1aDD76wCSwKRnCA7QPZ2EDSqggW7bsKrwkXYVypq+npgc6q5ee+9rCrdz6wLZYyfcivJ+2eyPf8grX+1kLBW3PtM+H1xI13WQ7h0IiEShP+hmM12hp+eRt6IVZT/IY19/1iGTtIw/S+PUtdLwu/5Ehz+TlwGNw5PcKslLM+3oTnlydMzPuHsiOUMCCnF45yOli4yun1eOKOtXJeWRU2qi2dfuoPFp4bj5deOrlJDz+Qy2oZ3IHk4MQytoyFJhUeZmrb+FiqHSsh2WPz5BMJ3QdSwUiSXxFXBeSTvn4l3sZtP1o3ArVHwOanBaXRjqJLQNsHUPtkYU+o5MeBjAJx6iXONITQ/Z0G2SjRO7qApTo2hWsJ7bBUarRP9NXUkLe/AfLUT8xgXvrl2WreH4HlWx7hj9/LC6hvJu8WINtOL9LAUjOMLUY6chkBbJ581QRB+D9SdHYAg/F58Vxez96NsJvYZh+utVnq8MY8zDy3GOL2S9xM/YuHVd+I3RI1LJ+Gb68Dqq6L+QChqFzx3bDLL/JupPBGC7K3gf0qhOU6mT3QZm3am4lMi0TTSQohfCxFeTbhH11D9chyuGx3oPBzUF/siB7hxebvwNXXQrnWh1Tppq/WkPF3Cq9WLpCXNrFENwStPTW2KQlDfKhTA7OlHwntOKgd70Bbn5POsvgRHNwDwdVYPQtwwKLCYLV+kErXXztWv72fFufF4VihUFAaQ8IGNmN0uxvl8zsvP3kpjsoS+upmQjKO8UpxJmdOH8JRmUnQ60sNSkDRaKtbFo97pg/dhCMpqJX+GZyeePUH45V3cqVrMU3QWMfKC8D9y9dlGMszZfNjqD04nf07ZTNTmWtLDUnC8E8KN2XdRcIsv1nEtpM08Ts09HVh9ZWxBLjp6WXBX6zHXm/ColDh1+5t0XXCG3pPOUbE4Hn1iM83dXcglepT3gjj9RTLHDyVSMcOBNt+AtUVH8uJGVFYJySZj0Dq4v8ce5L0+BO9R0e1PpbQXmii5xo+N6W+hSOD0VKjLCqayIJDEFXbyb9ShbVGYPOAEAZFNtH8TBMBVA3Kou8bC3tdTsQW6qO2j4639Y3jqpk9oSlYAaOjmQcb5rjyyaSZWPxlNq8T5Bzx4MP88L1RM5JUHZrLwnnlMTB5OhjmbyP0aTg/6COvIVoL+cRBFLYtCa+F3QNQQdSYxWoLwP1AwYylvfjERgHcfmMaWU1+xcvpEXOfyyDBnYx7h5rqYk3ww400slUb2re+LdpcJTYdCQEwDnscNuL2cuJq0+OY66P3OQ+w6k0xWcRdqJtpQ7fLB95TMA1O30dJFxubv5oYxB7ilRxZdNjURHVnHhbv90DZJqNplEn1qWVmQSmuMi4arLZTeGovKKuE7tIoHFj5Iey8rfqclfAfUgAsKH5TRV6uw+0hsye5NfZ0XlmA3D5oHcOKjniSHVePwhLBvZE4vWEy/7oU8nTmVxOV1mM6pUWQoGL2SpMVm2oe0097TStGEd3hi6Z2Y20zEPXsOu0nNvceOM2nIVI6u7k3Ki/NIiywiw5xNwbyLX1VuD3dnnkZBEH7DREIkCL+w+IWZDH1wNuG7nbxQl8SuVe8wMbwvxdN8yTBnE/vZbAwValZnp3Lr2gfxylfRHuHCYYSRDx3CU2vH+d3VIoMLh1HGu38tyBDg28qbg9ciuUEztRarosb/tIOwntVsKujF+3uHkneLCfOxUHzOS4ybkUnezCWldlyzAAAgAElEQVSsjNqH9bA/hdcvQynzQJFBm9xC9ZkgrHc14nFWT/1AJ7bNQUwfdhgf7w5889zYfBSQQFOmxWV088WBfrQOtGB7NJDYm/OovbaDhDVzKV6TgNFkofvaAmKn52H1l5g0ZCq5s8MI/UiHzuAg9rPZhP39IOpF/nxzuAfmEfDW7BvRre7gxFOLCV1xnHej9jPkodkUjFpJ+ZNpJM7JImlJTWeeTkH4xYidqjuXGC1B+AUdnr4IVVI8xpIOdr+7gicDLhD38RwKPkrh3L2LmTDxZmSrTP+rzxD9kYxPLoTubyHpidMEnHKwKbcXZTV++A2pAkVCW66loZuKNosO0wktze0GEjT1OIzQ1GZgW2UPZIebmsxQLLUe9O+Tj8oOHmaJ9jFtHKiO/WdsgyefIn7tHLRNEgxqZnB4MbeM3UfzKX+QQGuyYRvTwldLB3Ok7zque2YHAacUhnTNJ3RQJTcMykIf3sabqWupGOXFhbogQj/Q031wIQ1DbXisN7Hli1RaHXrOPrCYc4+E4I6wMvave4l+wc2paW/QvC2ekpkuCqcvJSyhFttjjQTq24jZdC/Fj/clPSwFr7nlPGgegLvPxWX4VYs0Ykm+IAiXnUiIBOEXUjBjKUPfe5TQ1VV8+fn7AIzOmULQEYgMbASgbLwPM8bu50BBHJaHmqhNdWH30RG2S6a2jwaT0YIsu6nMCWJYjws4om04PRRClutp7u7EuM3I1Qfn4TAq2Cs9+ab751Sm6bCF29k4/i281DZ8LgASGHQOhgQX/jO+fbt6MnPMXmQHjIrK5b6gXRy5PglXpJXkSbmQ54m1xIuOUImzdgurVo6naqIdnewEoMlhwF7kxYL1d/Dq3SvQZJhoSlBT+HkcNGvoCJGx+7spb/Ah5aV5yBYJ70wD7+29igv3evJFeyiZKZ/y4qANDLt/Nvt7bcDrES25zUEkzssibmQRCUd0lH4TzZthR4iafpoMczaSpJC4LBf/k2JJsfDb41Kky3oIl05SFKWzY/if0UVGKmEPz+/sMITfidjeFWhvc5Pzp1CkDhXqQAvrBq3g8ZhBuL+OJD8vFGO+GtkBAZPLqfkygo7eFoK/0FE5wk1kbC17e27k5foEPlw5FsfgVkybPGmNkrEGu4na7qRihAbZLuHdv5aWYwH49K+lutwXXBKBUY00Hw/A4ecmIEumfoyVNwevZZKHlUZXB8OO3IPDoSI97hxbjvZB06hCXythDVLQJTXju9qI+QY7SpUeY6lMR6iCFN2O5z4jASc7KB/lgboDbP3bMBwy4nfejt1bRdstLbTnm5AdEg5/J97nNLSHuzFUy7R1teN1RsvaBxcx9bMFeBXIWEa04io2ErnDTtUcG12etlE4I5CQwWY8NHbqV0bTHCcR/exBAKbk1HOfTxnpYSk8U3icWZvmdfKZFgQwL3odW1nZz8pA4np6Ki9tSr5cIQFwQ/zxY4qi9L+snf5GiRkiQfgFuE1OmtdEMHlHNrEfKUi+dr4YvITz9hBuOm+mutWIZJf46+xVWIMUqlu8kJ2gzzGgvrOa1enL+VvCpwDM8T3FqUcWY60z0Bot413sxj9bwjxcg8ug0HfMOZpPBKBpk5gQnkN4VD2BUY14LPahz6gLqFplQu4s4tZeWXxaN4DUx+bQd+t8PujzHo5qA19/NgDvkFaMpeDWQFBKNe0VXlRc5yBgqx6/MxIe1W50DRKmHZ4wroG2SD2BJ52kTj+J7rgRm79CyRQVDTd0YFhvwuXpZvzYo+h9rVgGtSE7JJAgNqqGtl427sqZiSKBywDaTC8SBpRQMkmD0WCjNdmPmA2N7O6xicblUfjcVoYjzkKGOZu3Sg5wn08Zs8sHE5bpxVlbRCefaUEQfitEQiQIl1nQUVAbnAQcqmGOTwU191soGLWSRI0nL79xE3ZFTWu1EXWglUc/vg1jr3osxV74TarAGuimosaH4Xq49dDdxG6YjUk2ABAdV4PToOC3t5TGZJAS2nB7O3ku4gtceoWgozZWnxhMRYk/bRYdAU8UMdBUTMwWC+Y1MXywaxiDTQU0dJNIfM/CLSsWMDb1FPNmbkH7uQ9OD4nQgxYqa00gwa29svC8w0xdfxd1fSRGTT9C3WAn9kw/rnnya5ruaOXo6t4wuAlbkAt9cDuy7Kb9+hY0vlZ2bh7As72+IDm0Bk2LRHuUi6KzYXw0fDk1NSb6DMhnwPRTzLv7c85diKB732Ja2vW4Z9fi9NGTHpaCUy9RuieKxL/bmDjiOh6IHkJ6WArFAy3kLO7B6menoI9upWDG0k4+64JwebgV+bIewqUToyUIl1H8wkx0TS7yRqzC5ePBxNHTCZuWw/jogYy77jYibyzkpZ2TCdmjQinxwB7ipLHei9BuNZSfCMOtdyNX6kncO4u4kFpuG7aP5P0zqXS2sbvHJnSNEvlzovGokrA16/HI1zJx/cMEHgO7jxptiQ7JqsKd40XOjkQ2Pj2W8EUFNCeA2+hioL4Ih68LdXUTHVFOdp7vyqKscQy+7yiWIAVNbgVIEP+hlc/fvYqyI+Go2lU4A+0cXNIfvyNq9Gl1LD0wEqtFi90ET3TLAElBOeMNR02YPvDCZ7snXTY38fSmGZw/EIOihvcnLsGUKzP/6ftQrCrejdlC5ue9eHnvJLS+VjYnfAnnvPC5y4rDU03zLakc+esS+o3PweGjp+0tN7lLB+J7wI8/Fmbjd7IJq6+EdpeJLJuDAYNyO/v0C8LP8t3GjJfzEC6dGC1BuIwyzNnoth3BpjgomG7k/Bx/1DHRqAL8aOjhgcWpIWaTE68SKxH9zATvUaHP01F3OARFAp8zapz+DvRZRh6OzuCjzVfh791O2lfz6f/sXJAgLNXM3Hs/x/ushrP3L8ZQLVM1zE1dDxUhWQ6itzlBkbBEOiif6ObANz2Q3BAc3si1Wx7EWKDGPCkS1G7iwmtR3BKnG8OI+7SNslnxdI+spGysJ+0RCkq0BbeHG+9sHV4VTlx6Cf0aX24alImzxoAtwM1Tu65jcr9sFFnB1aeV9lAVTUlQMMNE4muFOCNsKDLMWXI/1kBQZAlJ72J++Tg64u0kvmeBfE8mDb0Gl14hbEMzDfe2YQmQidl0L403e6N6uoZPu31A4pws2hw65i2fR2NPE/oGNx6Tq1hWPZJGqweJq5s7+yMgCMIVSiREgnCZxC/MJGbb3agSYrlm4izcejcJD2VSPCOcortjmbtwI/n5IRRN1dAca6DsZCg2bxnvITV4FStoWyTUFgVkhcEzTjA381amT96P9nV/ABpS3LRFuxgdfIH3Xp5CW38LAJYQN4agDqJHlFDXU4OkgMoGsocTjZcNryJwhNpp3xWEOsiC7AKXHvwPaSir98HzgpaqfeGUj/bCObCVnIoQPMsVfLvX4bXPQGpKLi1dnXg+UU57hMKBN5ax9thA8LUTeAxki8yByhh8+tci5XjR1MOJy+NiYXbuq6Ho8/T4XnDjO6YSa7gDTYebgMAWbgzIYnLvk+heqYX4dgyr24j7tI2vznQls/9qbENa0bSoiPu0EkaXM/bY3WSYs+nqXUXYvg4mPLqXlhgVFruG8tQ2Gj6IRHq9mS+ufbUzPwaC8JMpXN4VZmKV2Y8jEiJBuAy+u63EkfQ3cOUVsn37WvxOymSYs+mIdGINcvHS59PwLLy4KqwjWKLLVjvNA21Ul/jRHibhe97F1AXf4OFtZd+WPuhzDJjUHRRfr4BNRhNgIfwbePfQMK575CuMRw2kh6WQ3K+ESbFnaVgdhcNToex2JyorRHysQSn1xDKuFVWdlrAJpShumZOPLsalh6bRFjx2GQGwhjgJH1uKrdIDVZGBwHVnqK020R4GR/ck4xvWTM7ZKFSWi1+w+jItnqf01KVIuD1dAGiX+2GLsaJuVhHbs4JZCYfJG7EKt0aheiBUngwBScHhKVNr9uGdymGcbgyj6dUojF974qG20xrjSdRGmSer0/D16kAd38rOomRyFw8k/I9uur09j03ne1P9qI0bTEeJm1TA8f6fEHXYE7+VhzhXGMaz5ZNR/O2d80EQBOGKJRIiQfiZCmYspfH2wahDQ7jxwk04R/cjPSwFy4QW4nbdwUNX7QAJjOUSyVfnEtaripDDFmr66EFSUHk7sIa6sPrKrDw5mLiAehyeCs7ebRxp6gIuibh1F/f+8SxtQ+NtZ+3ysbT1s2DYE8y549FsOJtCczy4tRD+gQZrkEJNXzWBxxQ0B7xBViioDsDVoKPHG/MIHGHG1aSlJUFBk9aA5OHC/koo+moV2m7NWD7zxzegFVPfOmSHRFOpDyP7nWXcpCOMzplC+F4rpxcuRrYDKoXuAVV4PFjB62kf43MBkrxrWPL1WJL2zcIZb0HTKhN60EX3+Ar8j9YT84mb7NJIXotfh3mYTHMC1DwQRdVgaLmnhXNNIdScD2RCTA72Mk+CD8hceMIDTRvkjVhF6PMyD6dMQK92kLxiHmN9z5C7fACJdx2leWg9Cbcd79TPhCD8VGKn6s4jRksQLoOGsVZynovk626bUX99jAxzNqnhJSQ9XMGOa/oSHFdH4LF2ThyJp25vKL4vltIW40Kjd+Jq1dC3dwF3zv+CqJAGypp88D+tELjOwOkdSaib1Sh/rOONfp9QNM0bZ72eG+75muCtOtr/GIrKLtGnSxluDQT0qqE9RM209EMEnHZROc5JS4ITnxwJ0x4DhdcuQ9uiUJEdimyVcXm5MBmsqKq0+D5ZjCXcif2sCd24YqxZ/sT61GOLsIME+4pj2bq/HyXV/jQsaGfc9bfRdVghIeGNZJZ0waixYXb40pQMW4/3QjE66RNRTsSHGtxahdreahwjKin6kw6Pp8wkh1fxYO4MNG0Xv4Zil+Qza9RepC99qdgRxdMTNnDgtYHoGmRcNzXgbtcQ8sZBkvfPJH5pPp+e2Umj1YPIoWU8/84tJD1wigxzNhnmbOrvGixuBitccRQFcXPXTiRGSxB+Brf3xZmb/JErSZx9hPExgyj+pBezywdTafEmb0Ec5a/osG8MovABiYIZS0lOz+Nofhd0IR2MiMkjPqGS5icjWfz+ZCqOheHe7UfNYIWKsQq2WBvOQDvFp8J48NM78S4Ar3wVe3oZMJZZcetUDBx+juI1CbgjrHg9byT0tiLWn+pL5fV21o5chmdIO409FVRT6uj56jwUWUJfI+GdL6P3tVK7O4zwvpU0vNyF4C4NdBlcRvmTaegaILc+EEOxlqiEasJXahkwIBevQwaaSn0omCdjXhlLdX4Aft4dXB14ipf3TMIZaCcxvhJJpXBuXTKVaWp09RJ2Hzd5q/vi5WGlcm0X8vZ3oXVLKHvueoX8W5aw/URPsqYlAtAe48Tq1qCyKRgG1RHq1YLPGTVtX8ZybcJJvimNJ/XobQwLzKcoO5xH71zHozlHybbZ6J11E888sRr3VX24Y8xusSRfEIRLInaqFoSfKP6RI0j9ulH2hBvDdm9G33eIl4Ozidl0L8ZCNW3xTpJWtFN4nTeOACdqowPvPQZCviwHp4uIjY3sONaT+I8dVA42YPdWGDP2BAc+6osxvQrrxmA6xrThY+zA8VkQrWPbcdQZiN7ipmSyjGSTCDoKzKylqsIXzzwtlp4W3G0a1C0q/M5Ae6hElwlFDPPPZ9mJYQTu1FEzwoGkdiNr3KhUbvSHjLi14NKBLcmCXKFn6FVnOLKpJ765LqwmmZa4i5fjjEmNnBjwMQP/MBdLoMTEWw6y8XxvDIc9GTMzk6y/DgCgI1DGo9bN2tcXMen4Pchf+dKc6CJ6q4uqwVocRje6mFZCTK1Ut3jR0aJHV6TDGuFA1aLCs1zGNL6Sxp2huHQQOrKchk0RdIQokNCOvVFPt+dKuLAoDFWRHnWHhOwAtxpMRW5iHzzPB112E/P5vTw0fAf7G+I5cSS+cz8wwm/a5dipuksPL+WpDSmXKyQA7knaL3aqvkRihkgQfqK26wegHDmN8XNvHnxkPQf+Moj0sBQwuDCMrEXra0W9qIGRY7LR+1qRJIXGNBtl10fS+p6OrNV96N61jJ1rV9IR6kZ2wjfFCXhWuqlu8MY5oQnvzUZqG7ypT3WgO2wkNSWX5nktaJpkpBArzXEymqX+6EsvJkO6CwZki4x3PrROaQUJcqsCWZo5AnW5joB9FYzqfp6IDWpczVqMuzzpCFXwqFLQNoE210Bw72rM7SbSp2fSkKTCJ9+KR89Gumy2sCDpKwBmP7ER05gqIrSNIClox9QRrmtCX2tn1LP7aY2G8jEKEWojapWLyOmFYHLQ0FWLNdSBbJcIXqKnJiMCrdpJyHYNsgO6J5Sjr5UZdvMxmix6FBUYahVKj4ZjKnAQetDJ8C4FSC6Jcy9FoD9tQNejiWdv/5A7bvuSiBcPYjrfQtbersTuvJNPxr9Nh0vHIN+iTv60CILwaycSIkH4CeLWW2i9+eKeNx3TmpnlXYemzYXcIxlf/zbcGwN4td86+vmWcujjPuh3e3FD1+P4HtSx8J5PqW7ywq2Fs/nhFzdhXG/FwwzycS98Mi6QP2IVbW166sZZCdqiI/nNNjwr3WRXhtPWocPu7+L2HplYwp20hquxBbjw3aMnfK8FySnRMMAJJ7155963CFmrR7LIOPxc5DwVxN69PakcrELVLtOQasetVWgLlzAVO7GGOmnfEkJeTjgbjvfD1sNCY6IezUZfavt68I+XppOafT0vbruG9s0hbL19OPojRuxfB/CPPWPJn6lme3k3Yje2gcHF+POTsNo12F0qPE/rUY2qZ0xKDqY8qLu/g/YINwC+B8uxBroZFXABh1Fh/5p+RDxm57aZGaz74ys4Ax3c+dpGUGD/9t4kvtOGIVdH1OY6Tg1cyw3GZvzUbXRMG0Rzkjexzx5nWo9sno4ZwJ5eBpqdHuLSmfCrpyBqiDqTGC1B+JEKZiylYJ6M7ztejDnTirTfh9gdd1FytYrKUX6EPmhBbVGYnzWDD88OwKUDRQVfvjWUhkEObveu4dV+62hJdqCp0xD3ooPmP7TTONABwIWnkxj0xFzURXqo1dEcI1P6rIqma9u5Ju4UFHliKFez8qsRzBv6NTY/6Pp6NcZyJ7cs24oc0UFweCOyHe576X6svip0dSokDyeGEi3aRgn/0wqEWfHO1iEF2fDJd1GWLiHbZAbffhxNkIXkN1pRFeuJmFVI3UgbbYMsvPPca8T71OFzTqI11o33a2Y6QhTUHQr+0Y2E7ZRpPRxI3oNqsMtUNJsYGllI+Y5oALz1Nr4+l0zdYCcd+SbiNtjQrPej/NooQrvW8M4n4+ny1CFkh0Luc15sNfckTmPk8Lg32Fbfi0ffeh9rkBPlxFm6rCmh4OaLezSNuOce/vLFtVROt1F9tQ3r6F7sXj7on0XWHx69OHsnkiLh107sVN15xGgJwo/g9nQxIWEI/l/rUVldfNXDi6tv3Y9ik9HVqgj76ALtXYOpm2hDneuB7owH0oBmOkIVWhJApXfS+2/zWF01BMkp4zQoFMzwQfWBP9hlbrnpazzLZTqCJaSkNu4YsxtFDd6feWFt1fHJN2mE9K1i3LQsjGUyKz9Jx25S2Lp/E81zWvnT/imYvCwY3vDB5qvQMMhBczxIvVpQl+uwBrixJFvxvMOMVGagI0zB19SOebSColL484T15C1IRin0pCbNlwt3LOFkTjSmI3q6vCNxzZaHKH8+gYZebrSNMkfOxBHxjZOGPi4az/nTFK9C2wJuhwqVl4No30bOvdCTjq5WRt+UhbnBG0ntJjxDxmVyUTpGT2uURGuii8paE0HHHTTfmkrbiHY8PGxUHwijR+YtBKk8yTqWwK7mbvhFNtH2ZSzNqRGMTj9B4pq56GqtxD2SicuiZlHqekqmSDQlKwx8ci4Ay0aspscxmdnlg8W9zwRB+LdEQiQIP0LR5BWMPVxJ1gtL2LXmXZS03pwY6kVgRBMx62ux9e5C6XgVj/TZgbYJwvZZGBxezAcz3mTJ9OVc/f/Ye+8wq6p7//+19j79nDnTO9NnGGZgYOgdBERUxF7Q2LAQwVhjNDfRJJaYa4yJuRoQo7ErKiJKExRUkN4ZYGB67/X0utfvj8F8CTdGvD/u9d7kvJ5nP2fO3u+99tr7WXPO56zPe61VeATO6aV+2WAGbZQsmfsKwdgw/Ve4AHh13UyCUeDKC2H9zMbu3mys47tomx3k0UmryF4TpMtp5fBDI4iuCTHz4n1oSX6e683CdSIW23ED7l0JhO/vJmQPYy8zMOXcMuLesRJKC6AEBMZqE74X0zB1CmSGl+5uG8IcInMd/Ll+KjrXwKSGYaNg9KOLSM7qwZGn0XiuEST4Y1WQAnOXZOuFv6dlmo78glbijoCv2MuP73iPon9rwGrx0/xeDuE7ukhfpWddZTHUWtF8Ki2XBEFCXLnEPKELER0gab0R49o9uFMVQn4d7ppoNJ2kNKWZEU8vxlansmrrOCan1uL8NIWOK718Ul6MEoDlHyyj40eTED6VpQX5CEuI1K8k0Tc0Ubz9ep7JH8qxW4ew852RpMU4uK52xvfZjCJE+LtIBJo8u1uEMycSEEWIcIbk37+TOWmlvLBmDjnrbwOgfayVqkeGk3C/5PiiONrHGpFRIf7y1MVIBbzJBsp7k1nw4j0s/Ph2Pto3krCm0FcITecKPugew7rzn4UDdmzJLkJRGlIBU5uOyQv3UvFFLsk2J7G7Dfz22Bzy//0Y+m122iYaCRsV1h4ZhvSpPL/yQjS9xNgrYYSDzm2p5K4IkzKvgWO9yTTPDaMawpg7BbEVGnp3mHtvXUnYoyP7LYXU5D5af+CjoTIZ7RkHFTctJWQFTSfoqIknukIQjNFAga5SgaVV4bI7P2eQzkbxxBraNmSQuKEGszXAY6uv4vgjOTgb7cRW+PF8nII3XkF31IahX/DOzGXIgAJ6jdH3HaC/Ig77TjMd46BiyTgGre+hNKcRY4+CqVtQ/7tCND34EiWDPtPY0ZaDpU2jYtrrKLoBD9Kli+7FmSWpuWwZG1oOkvsXsK05yKdFqxmV1sSCE/WgaYy6pgzxSBxv53we6SWK8L+SSMrs+yPytCJEOAPyH9iDNnUkjSuGkfPTHaSn9/B0Tx4hKyQekKz7fAVZRW0oQdC36emd48Uz2os7RaX1eBKBWEnSbigpbETuiebELUsxt6gkGpzM276Y8fPKyIjpo+g3TQgN4o5p7OnMJHVHkDWD1xO0CtKe1rHx8FCcuWGs47voGC2I324gYacOJSzQTBq9wyTeXjPGPjDureTfstfh2pSM3hIgPaGPnLk1tE2R+KNV3r9+FvoOPa5UPa2ViaQsNzFj9FFqO+I5//hc7LUafSUhVJeCVARSlSixAUL2MJ4hflYunUHpbxZz6FgWeodEc3vQq2HCUWFqLl8GEprONWDq07j+vvVoeknq7EauW7OYmP0GLhh2lLWHSlCCgtLry0jZJtG5VOovjaPh9Xx8yWHM53XQXaRibdaQWV4WPvMBXU0xWFuDjHxiMVFWH/60IN1DdRh7BU90DWHcgasw7K9ixE4/rSEXRiXE/Khe7ln5IV+eKECqCoNfXcSctFLGjz/xfTetCBH+VyCEuFgIsUcIsUUIsU0I8Q+H6gsh7EKIV0+es18I8ZQQQneaJlUI8ZEQYsdJzQPfUNb1QoheIcTN33B84skytgghdgshzv8v3+g/IBIQRYjwLQgNat8ahghpZD2hUbFsLP2bUnhxzXl48gJ0jhbkrLmdusZEPGkallZByKNDCEnp9WWIMFRev5TOi/zUfZTLPTeuomjZYnQ+2Nw6GL0+zPb6HOo3ZlP1w0ziJ7Yx8icHaKuPp3Ohh6IXF6N3SXRPdmKL91DwlpeeyjjMhX14EwQ5t1YQXalR/FgDMfk9TCyuImiD48/nc8tXNxNbFcJwwMa5Kcep/zCXmsuX0TNUoDZ3gQDHXBfPnv8GnaU6tn8ynKwlCgkmNzGHu4k9qBK2arinush7N4TxiBkkFGS001es4cqUqC6V/b9YSvXPh9FfF0Pm2oHnlrJNYKsH55VOXjoxGXOHoGlLBsYulWAU1DjjyVkusbQItn86jN5CdSDNV6fRM1LD2Kni3JqEP1Gjp0QgGsw88dY16HtUfAl6HPkaJUktxO/Qo+lACcDb783EsyWR8t8OYUZUOak6G7s/GM7g1xdx5+oFDM5op/YOUAtcJO+w0z25F0N/5GMwwv8OJKBJ5axuZ4IQYjTwNnCTlHIa8BtggxAi5R+c9iqgSinHAhOAqcBjp5SpAKuBw1LKicAMYJEQYuEpGr0QYgUwBYj5hrplAOuBR07W7YfACiFEyRnd3HcgMjFjhAj/gOr5L9Ab9hCrWvhl51A2/HYqPcMEwegwi6dt4i/vzSEQoyGS/cR/YiJsgtjjPpKfqmVPYyb6AzZyL6jB+/MUms614I/VSMjvxv9pIo7BIVS3SuIBScAmsF7Rhqpo9K9Mo7ckTOGQZkpiWmj3R1H21jA0PWgquIf5idllwOCQ9BaLgQVXSx2oqoa7OQrVpaCZJCT4mZZfxRdlQyAsSP9M0Hx+GKGToEikT2VscQ37dxZgaVWIrgnTl6/iSdeIPiEwX9JO+5EkMj4L03CeihIUZI5upqYxEUWnITqMhO1hFJeKFhMkZaOetqkaliYd3kIfsXEunC4zhoNWnrztVR4/cRFCSMYlNfDpxlGETRJ9v0JwiAfjEQuxFWFapgqy1oVonqHH2CMYcekxyt4vxpcgSd0WwpGpw9wjSbyzlhOb8ig6t5LeJ7MZ8+ReVq+ZgKUV9j+ylDlppTS8X0IoqJLzvKT+HomiSPwdFmIPK6h+6BkKuSs9tE+wcugnS8hbfsf33dwi/B/mbEzMmDHMLu9/f8LZqhIA9xd/+q0TM54MSoSU8opT9h0DPpBSPvJ39MOAMmC4lLLs5L6rGQiSkqSULiHERcAqIEFK2TJH4R4AACAASURBVHdS8yDwIyBLSimFEBZgipRyoxBCAguklK+edq3fATOklKNP2bcO6JZS3vBdn8c/IvLTKEKEbyD//p0UfHEzoz++j+qgixOuZJyXOwcOmsO8cGAasSc0pE6ybMLrFC4+ChJ0/V6OvVGEYY8NX6JG+a4cai8zUb5wCTVXvcCYpEYW3L4OYQuh8wjaZoQRGjR3xuBcnoZrupuoKh0Nm7NY9/5Eyt4axm2LV+NOk3hKvehaDTjyJZ0X+FELnZjHd5HzEyd+n54tFz+DraiXmLweZL8Bb1iP4lJJ2qbSU6hy3vCj2PcZ+fGYTxk6uIl9ewuwVwmia8KoAY2oBg0pwNqm4V6bQjgmRPttXlKKO9AMktrmBNLX6JBSYC/oRbWE0IwahX/y0z5RkrBHZfS8IxhrTcxMryTcZsaTEebD7tH4P0+gqy6OvX8ciaVVIFJ9TLiwDKQgY0M/nSMU9C5B0yw9IbMkrjzE8VeKiL2gBb1T0FOkJ21+HY5rHRzbnou5Q3LwcC7OdB1PpxzAXgM/ve9tSp5djHVLItFWL5kvqfh+6SDYbibzqjJqLlvGvl8uJfbVHSQM7aTmCjOesR4AZk4u+x5bW4QIAILwWd7OkHOBvaft2wPM/gd6H3DkNL2Zgd6erzXVXwdDp2gygEIAKaVHSrnxLNftv0wkIIoQ4RuofruU3OsOUnPZMq49soC+mW48/WaCyUFi410YT5hpuyAIKtz16g/Z9dlQEnf2IvUqjik+AtESU6eCtVGgpHvI+XghOatvZ1xUDQujK4jZYURoYGjX0TtMQqeRH/3kA2xfWAlOdDL38h2YOyWhWX08u+Yi8kY1onl1iDBkrQuSFO8gOdpJ8PMEqm5JQ9VpnLtzEa7jsfTWxlJ56VIanhuMziXonesmbmobV8fvxjHGx4vL5lG5NRtrg4JuXhdNcyRtN/kIWgWmdBeOmxy4Jnq4eex2wuVRtNTHoyT7kCEF5w0OrGUmbs3fTt4PDoA5TP/jXmRUCEcehDQVUyeUO1Iwtyvo4n0c7kzFmR8i7oBC5xjwJsLQ9FaidD5io92cWGhBCQpCZomxWxBXJnjqj0vpHhWmbUca2hgHnjSNhr4Y0n6jkjVuwGtlTXfSM1wy/JnF9A2Bj7tGYmmTHN6VT/+uJNJ/XUlrdzQ1Vy7j7qrjlPx+MWN+sYgNLQcxqGGqrnuB3OsODvQojXdj6It8JEb4/vg+UmZCiDggGmg97VAbkPsNp+UC7fJvU0xtpxz7+vXvlXmq5kz4pnKST/YwnTUi//0RIvwdque/QNU5r7Kh5SDTyi4juD6RimdKSU/rwVphwGII4h0UAoeOmsuWMWizB4NDUP8rHZpJD0ISSAnhTdJwjPYT7DNh6FLBoPH7P1/JeUfmM+mW/UhVknBYUrikk6KR9azrKmHWwp1k/EFh3fsTSfpBPa52G/aibiqPpSO8ysDQ+YdbMT0bS/uX6fxm8V9QQoKS1BaiLH7sxd0MKuxgzJM/onOUYNLsI4Q6zbS0xzDLHKZm9l9Y9MOPuOKibYhpvfS7zAhTGI7bcOSCeWMUfyx5F8MxC6uWnUMo24cwaijVZggoOOujkSq8VDmZipfHoG810L8tmYStBkydgrJVRcRWBqjako17cAClwkr/iTjScrpw5IOMDRAq8OB4fBB6EcbyQgyKRyV5SgvV81/A1COJv7aRBa/dRXZ+O4EYjexH/Cj+gV+7FYsMVNWkYLisA5M+RFStQlx5EKnABfFl9BVC6g4NXWkfX+0toii9jWd7s/njjdeQfmE98pJuRj6xmC0lHzLn0htwXT2BihfG0bF4EjoPZG4Ifc+tL0KEs0qCEGLvKdvC045bT776T9vvB74p4LB+g55TzjkTzZlwtsr5VnTfLvl2TuYKFwNGBirfCzwkpTx8mu42YBHgPbndIaWsPk3zM+BKIAA0A4uklB2nHNcDTwHTAQ04CNwrpXSfjXuJECH//p3kKXdgznTiq40i78c7mVt2hI+emEVrdDRRbvCuTGbmLWVoCPLfvoNc6SX+SBD/TC+3/GUzP9t4NcYelXCBh7yng+S/WEX17XlU3GzH1qzhfy+ZzekpBGM12iZBX14y/r0amkGyV4BhjsrPrnmP5ZfN4Nq3d7L+z1OI0ST2y1sJv5DMiYJk5LUSGQrwo89vIP1ImLaDeRQ8WMmO43l4GgyYJMhBXjp9NpS4AE+N/eCv93jcm8pXL4zFMSbErRO38u4bM3FnhDGnu/AVwU9+/UNyb6yldkMOo3MaeC9301/PLd5+PW6jBfsXcVAYIr5MErSCL04MpNpSFVJ+Wc10WztvrZ1OTKXG5Pt2s6G+iEGbg9RepTJouUrdpQpdjYVM/EUZnQ15GH4VzRztRkQxTEqoQUyX1H+ZhdUNNfMTsTbBTZfsYOPMwdTdno8jzoQ8EE3WF31UPmhArdbx2MG57L7p90wbegv+8hgs/YLKlETKatLhehUq0nlq5rtcPbqfkl3XUbbqDa6umYVtSjdsGsSGojV/vc+IpyjC98F3SHOdKV3f4iH6+rvTeNp+I+D5B+f8PT2nnOMG7N+iORPO5FpnhbPVQ/Qq8KaUcpaUcgJwCNgkhEj+WiCEuAR4EpgrpZwCfARsFEKYTtHcDdwATDtZTi3w4WnXegoYCYwHxjHgTP/zWbqPCP/i5N+/k4ql4zC1K3gbokjeBe4rx7P60ZkoYUncZhMpl9Uz9Yd7+KKygOb7cok9KmidbMWxyEF7dQKPlc2l8GUn/rQg95ZuQv6un7V7RxD1fDsAfXkKPTN9eDOC5L3vRfELUmY1kbQX9A6F2OxeVD/8cvulPL3+dd7ZOx5PusTSpdH5eRp9eSqZKT0kfaYnfoeehNR+Jj+8i86ROo6sLELp12FtkTjyNcZkNWDX+wj7VB7acwUj98xnxO5r2duZyVV3f8bU4SdY89Q5xB0PkZTfTWKUG6/bSO8sH0crBuFN0TjSlsqIpxfzZFch5198PeGwQvJWhag5bQi/gvuafnQ+SfS5bTiyFFyDg9Q9V0idN57X5z+PphOsqRqGv8LOoEcrsNTq6S7RM2PMUdxOExW/GEaoMoq+wRZ0HQ4ybq5i9fPT6fFa0LsGJqmUisSZp/HBw3OovC+P+GNhPA4TceVhaq6KpuDfffz+2lcIdpm54IH7CO+IxdYAcmw/yv4ori7dR1FRE1sv/D0fd42kOuhiy5iX8MsgzjuTCJ43BmY1kbNqIaMeGxiSn3//zu+5NUb4V0NK8T+eMpNS9gB9wOkjylKA6v98BgA1QJIQQpym55Rzar6hzFM1Z8I3ldMmpfxfGRBtkVK+fcr7Z4AE4LxT9j0CvCGl/DqHuOyk5gfw1yF6PwOWSCldJzVPA5OEELNOamIZcKj/XkoZOpm/fBq4VgiRf5buJcK/KNXzX0BsTid+v8rRu5YQe0TQU6Rgv7ORtssCtE0QXHbfZjrey2T9+rGkJvbTNMtK2AjujDC5sd0k7lLgkB3nv/soym/md1suwB00YE70cHDrYAqGN5L+pYeYLSb0dj/23zYTtmg07k7HcbUTqULCvAqkAvOGH2LBr+4nfpcOUeCi4wofejdMvXo/9cdS6RoJIYuguzqORk8sgTwviQf9xJwQdI8JEV0hKFszhLKOVIRLR9irEmPx4my201KVyOYSK5V9iXSMl/gX9eDcmkTrnlTSkvpQGkwYov2Y0l2oqkZ4aj9/3j6dmgcUgp1mtOu66dmWQnSFQnh3LO2zgnTsT0bvAvsRPV2XeMk09/DIgtvoGwKpsQ6kCkdfHYqmg4TZzRxZUkL1rFcI3tuNCIFrkKBtdioHD+bSPTbEddl70Hkk6TldaAaYOvkoM3/1FZfP2UHXMBVLhZG+ApVASpCm8+O47/0F5L/pI/+eY2h68KRCoNKOL0nj/cOjaH8niwcb53GoPY0lXdMYt3UxdzdPo3tkDFJA5eujsNbrSHxhBxVLxtG6qogbZ235vptlhAj/E3wGnN6LNObk/r/HpwwYqIeepvcC207R5AshYk7TNEopv8sEYJ9+x7r9lzkrAZGU8vLTdnlPvhrhr4HMaE5xikspgwyku752ig8Hkk/TtAMNp2imA3r+1nF+AAgz4ESPEOG/RPX8F5iTVkrbh1nYWgY8JJ5UQfkPl1B+YhDjsuuouu4F3qocQ9gkEBqYfxNN1upebK1hYnN6aXZFM+v+bQSjJJ17kmlelU3e8hCde5NJXWpEyXNhUMM0XGDG4JJcWHCMqvcHE39AYcV1f8DdZsXYI6h7fCLWCV2s3Tqa3mLoPcdHoNUKtRa8iZKNW0qR1hD2SoFnkpvoEwoNzlikFNTN1XPR4i3MG30QVyb4EzTC+2JQfYKocgPOd9NQ3QrWBpWqN0fS1hoLArr7bPiSNcLZPuall4EAy1Yb5o1ReOrt+Kvs5KzUSH3TxKyxR7g4owwlCJoeEg8EEQ49wbgwjmFBnHlhgj4dq16Zjj9Gz6DNQdzvpGLqFuTdUIG1RdLvNdE9XJL7/h04fEaMvQJveoie8UGy1mlYavWsu206vSPCtB5PQjNKKp8u5vVtk1lXV0zIIgmbJffdtJKi3/bhzgyj+gWOPDOekAF/vAZSoPMI7JUKsXEu4o552bWnEFerje3PjMP2lYXaewejqdAxyoC53IRvhIfetQUMvnMPH438MztuGx2Z0TrC/yjf02r3/w7MEUIUAQghLgRSgT+dfP+EEOLI1xkdKeVRBrI3Pzl5XA/cAzx7SofGega+4+87qbEDC4EnvuMj+Q8GAquZJ8sZAUwDfvsdy/lW/rtM1RMZGJL38cn3OSdf/5GLPfcMNZL/51T/OrDq5ru51iNE+BsmHLyS6rdLKZ5fTsikMPS5xWTPrOPe1jHo+lQanxlM/tt34Om04kmR+BPCODJNVM+PoX2cimdPAhZ9kNV1wwhFhdG7BXffsZLq+ToCqUG6SowkvW2m7EQGgcQQnXP9bG3JxRcPiSuPce2y+/nj7Dcxn9NJ6o4QPRVxxB4TMMgLnUZEGMydglCOD32mG3O9gYRDHoJ9RlKvqiM3uov4OBdCwhtfTmHTyrEo+S50LoWgXYIAZ16IqKYQQ8fX4MoNUT3zFXSmEFdM3o3+qAU12YuiaCyvGY25VaC/oJOcGyqJyevh15e+Tf0FOppnKPw5Yxtrmobhzgsy4pojtN/iQ0gQYcEvp35EaWkN5hojSQe89OXr0D/URl8R6J2w53AeIbPAdTQOLTGANGhkxvSh6UHnVFH6dEz8zS48g/3UXWIBIYk/ILA0K3QPU0nfJDCui2b01BOMmnWcW6PbqL88CREQhCySkElwuCmdyROOYewDf7xGIAZ6u6Oom2cma00QS5OOnmKBGpQgBP6L+vFkhgibYPCjLrp7bSAl1xxZwPmvbuX8rHGRoCjCPzVSyn0MZGteF0JsAX4OzDklo2NiwMB8aorsZgAhxB5gF7Ad+MUpZWrAxcBIIcQO4EvgRSnli6deWwixRAjxxcm3PxVCfCGEyDylnEbgQuDXJ+v2EnDV1/MfnU3OekB0Mqf4CPDwKWboM3Gxn6kmeNpQv9M1p9dn4dfu+rA74ruO8J/Jv38nxqVxZCT20n+1mSsf24C5U9L5RhYfbxuNocCBpkJaSTuqU0VmeknfJLDf3IQIQ87EBhLKwtQ2J+DzGtD3qqRu8/KX+kkIv0L0IQP+OEnzdAV7kgvVGiLvjyEuyjxKINuHf0U0MVVhHnrjZro67fjiVG6ctYXcmyoIufVMHH8cY7eCM1tDV2vCts5G2ChpuE/DnOghqKns/HIoQkg0vSQpv5uwSaJV2QjEaNiLulF9YGnUwY87mJd0iKFDGsl/axFC0Vi3YiJ6N6TF9xMKqDgcZpz5YbRVCRzdNJj+4/Gs6BhDYWkDhiwX4362iPbGWGwJbo6+MhTlQBRzJh8E4NHNl3JF8n6i6iV9P3ET1RCmojaFqFroH+cjYY+KK1tD00uior0goGpLNpl/qSTjsyAiLPhg3WTQBBljmknapqNrnIYrb6DXztLqwx8ruC5pJ/ubB1G8dDGxlWG0mBAjJ1Wgu7wT2Whh/6phOEf40bsEQbtE0WtoaT46RxpRAiB1EJjbR/UVJkamNHHJuP1krXUy9J1q/jD+XSYdChA7t5JPrp+MZVM0RS8uJu8D3/fYSiP8KyABDXFWtzO+tpQfSynHSimnSSknSyn3nHLsASllrpTSe8o+h5TyppPnjJJSPiilDJ1WZouU8mIp5UQp5Ugp5dN/57qLpZTnSCmFlHLIyb8bTtNsP1nGtJPX++S7PNcz5b+jh+hJoF5K+cwp+87ExX6mGv1pRq7TNX+DlPJFKeUYKeUY1Wr9e5II/8Jo0SE2tBzEtbCP7Kgeyh/KYHtvHr4EQepNtVgGuTg/uxzTD1tp25eCtUlg22nBG69QVZ9MONdL5/JMOktVkj41EP25GVNxH74EA7o/JSBtIdwTBppm9pogjm4rihqmbWIUb26eypD769hU/DEtsyShIR4MDQasLUHeW34OB74azMjCOnxhHcFoyahRVdhKu+kaH2bKuWXod0cRCqrU7xmEvsCBa0ci8QcU2uviMLcL8t7rJ2kPJD5qwNgtMPRD8940ltVMpddnJmmfpGLa63hyAyxa+BFXDdpHWnIfMTFuFL9g76NLmXT+YZQQHFlfSMOGbEyfReFNEJjjvQx6FEyXtzPl0gNU31mAMclD/H6VX314NTf+dA2WZTG0zJQoxjD+GIH1qImcWysIWzU0PXiPxWDoVLn/qo+ou6MANIm1WWDqFBha9NRUJ+PMGFhDzdiuY9C0RlqnWPHHSv7tLzej3xOFaWw3nkSF84YdpfqNwYg3E1CzXbhzg8R9ZaBoSg3WJsHUvCpSPjZgPqcTd4ZGVC2Ed8Vi6FdofLSQZ1P3Uv8AHB4ludjq4ZeJx2j4xSRWr3mduYllZP5qO10lZkTy6b/XIkQ4m4jvK2UWgbMcEAkh7gWKgAWnHao9+fqPXOw1Z6gRDHiNvr6mDojnu7nWI0Tg13PfxVph4OGOEvaNfo8vDhShJvg50paKqVtSviOHRUO2sHbNBJq6Y9A7Bc7RPtIuqyOu3IfwqkR/aUYqYB/TiS9OwXmuG7bE0jZBoW2cyrnDyhFCMvOCAwSidZgaDdg3WTH1SnKHN1N9XyE5628jKsXJM2PfI2wATS/wJ2hcet5OHhz0CUfbUgnFB1mR9xnKynjSNgm+qs/FPKMT824reW/24um2ECz2YHBpiLAg55pKTtxnwpGt4E214E2W6LwSnVvg35hIS0scfrsg77070FuDPL1hHmuvmkRnvw3PvgTiygSjH13E1s9L+NllH2BwwC3Xf4KpT2Lsk4h9duoujSFmgYdtH4yk9n6Bv8WKVMHSLnipcjJdw3RkfAKizYTOA8l7fBzZUEjSNhVrpgNjnyAwKMBLT12Cr8BH3WU6rr9tA4VXniDpgEbuexqBOA2MYc69cB9VVSmYpnWR+4ETX6KG0ED7LB7nVC+fby7FkQddpYJAmwV9tw5PisB/TzzeJMkrmVvpvcZNZ1MMw0tr6RkXZPTFR5hx0X66i/VcXTOLcEil7vGJFC1bzIwFt5E4uZW5V9/KrdEDWYPEF3ZQNeOVyOizCBH+STlrAdHJOYYuBK6RUoaEELlCiHMBpJS9DBihx5yi1wMj+H9O8cNA+2maJCDzFM2XDMxPdKrjfCSgApuIEOEMefmSF5kf1YveDXtKVWYcvQRUSV5KJxm/FWg6CNnDrPzRecybt4PUWAfGid3IgELF/kyap5vJWRmif7Bk/y+W4v4qkeJryjHssxF3PIjeKdAVO9ixcgSJK01U3V1I62SB4oeeyQGcmYKuDzMIZvrRdepx9lh5/Dc3kVTaztAnylDSvHy8fgK/bTofn8OI/bCB+bUzcacLWi4MYdpho7vHRsgMjXPjUNwq0ZvNWJt9SFVyeHce5uMmvFlBGuZCMD6E0aFha5T4EiXmGgM944JYGxSsW61UX/MC9RfHExvlQfWCP07QO1QjZNV45qUrB2a3PjaZ1llhzD0awREu9E5ovDaXRTetxmbxE3Nc0DfFh7PUh7omlscWvEnTLIHqhwcXv0v1fJXAEC9KWBL9VhRSgHDqcGUI6NNjS3fw2bAoDmwbTPN0gfbTLiZNOoax0cDhX4/AflyPWBFP0nMNmDoV7PVhfFOcXFx4GNUrCEWFESFI3iEIxoUJGyUt58QwbEYlM2++Da/DhP2YnspP8lAMYQ58OAydCOMcFmBO/FG0sCDv+WoeunYFpnYPjdWJdI6w8GhnMZXPjcd/4VgANrQcJO9977e0sAgRvjsDM1WLs7pFOHPOSkAkhJjPgAnr10CJEGIMAyPDppwiewK44ZS5iW5nwAz9FvzVgPUksFgI8XVu6wEGjFqbT2p6GXC93yeE0J1MnT0AvCOlrDob9xLhn5/8+3diEkEAQiedZ3WNiRT90UHrx1nUXG6jd3SQ2ktfpPYmycot47He6MV9OI4HJm1A00uiJ7fTOcKIrV6h8OVFeIf4uDVlC668IMMfO0jiwRA+rwGpQOdohXkvfYG9oBdPZogfj9uINyeAsU9jyN3V6DyC84YdxdSn0dIcx7aWHJJWmNA7BAeO5hBzwEDyXg97txcy6sJjJG02EFMdwmLzEz+9lWEXHyf/vp10Twji+YUTNTpI2KZx/pU7uWviJnR9KheMKsPQH6K3CHRugaVNojh0uDM1+ovC5Ky9HU9WiO7DiXhTNRbevpqCoc0Yu1Uy1nVxfvExVFVDcau0TVCwbrWh80iY1suu/hyCW+LpL5TIXgPGWhPOLHhgy9VEV6iEjfB8zTnYKvUYjEE6L/LhSlMxTukiplzBmx1gyPNdeKqjqfjLGEomViFtIYw/s7P14BBGzTqOK0XFMSRI57Qge5qyEGP7Uf0DVsJRtnr0LsCgUTy5Bs81/diP6wjl+NB5JPuO5VB3kYq1wkAgGtQAGCvMpH/az/ZlY8Cv8IfXLmfvOX/ixEO5vNY4kfXr3kZ1qxT+4DivHZiIsUulZZqOOWmlAPQXWCJBUYT/FsIoZ3WLcOacraf1BpANfMHAomt7gL8ZliGl/IiBoGm9EOIr4HIGXOy+UzT/wUCA9JUQYieQD1x2mon6pwz0Ju06eR0nA8FVhAjfypCn6wAYadQoWraYuPIQvWsLMDUY8GTbSd7lJpQcwBLrpeCLm1HbjEgBx36RhQgLlh6fxrAR9SivJOKPlTjHecme1Mh9YzZx51sLUTwqn2wcQ0+RDvt2M4kHgxh6BWvbh9HXGIMlyc0zu89D7dFjaQtSd+dQcmbUsbMlm7BegCawGgMsenIFc+dvJ7pch84t+eErK9GS/Rz8qJiOqSG8cSquNhvXZexm19E8KpaNRd+hx/FpCrMKjjN0SCPra4tZ8cR5mLoER54YzsinDmBtHpgEUrmsi8vP2YUpw4mxU8VWqSd+j0o41Y8WHeSZvbPZULQGUeKgY3I8W98fBYBUJLHHwJMmWXjXx/jKYyh7cxjBcU4Un0AJCnJfbSKY4Uft0xF9cQuaUeP8tHJ8CRJ5xE7M52b6i0O4DsbTOy5A/HY92gs+Jkw8TuxuPb6wHkuMl/vfWQ6qxBfWEVsRIC2rG2OTnsKkDtwdVhouAL/LyH88cTUx57UyenAdlZ/kYV4RTSAGLi8+iGYQ2I/riapWsbRJwibJZ/c+Tcgq6R5pZ/TtB3lo+lq8Q72M+uQe5k3fy6KsL8hZczvhuCC7KnKI2WVA54X8P1Rz7fEW5qSV8uDDb7Hxg9ci6bMIEf6JEP95wNY/L8aMDJn243u/72pE+J6wtCnop3STePHAnGCO9XkMj29h60cjkaVOsh8PkvJiM3tWlXDk7iWMemwReg/4YgWeVEkoMYip0UD5wiUsaJjK7uZMghV2LK0CfxxoqiSU7cO+00z/OB/Re0xoOnAP0jDnOWBHDN4RXlI+NOC3K4Qsgv6xPqRXR1ZuB/0fp6HM6cJxNB7NIDF1KYTMA2XGf2YiaUEdlduysTaCMxdshb3EWLzUNSayYdYfuaX8BsKvJdE+RRJ3QMGRC8GYMLYaHTnzaqjcnItUJUOm13Di8zwMpb04OmzkrJA0ztKjhCCc4+X5ce+weN3NJO0SeJMUHIUhFJ9C/AGBIw+CUQOL1qbMbKK2JQFzuYmQTaJzCfQucKdLhk6sQaeE0aSg4fV8+gsgc3Qz7oCBzuMJGDLcxK6wIhZ00NlnY3BKJ8eOZEJYYMvq5/C4d8j5aCGWBh0Zs+up+yqTR655j4e3Xkbxk13kvNvKl015BAI6pITojVYC0QKpgDdJYugTHL1rCWN+sQhPssA32MdtI7fx2uqZZP98B66rxtNxmZ/ozWbChoG0QumNZfjDOg6tLcLYK9n/yFKmlV1Gl9PK8NQWdEIj1dTP0ykHmJNWyqLKKpYW5NN5x0T6B//rfI5G+Pu0PPMs/sbG/185qpShcfLGt2edrSoB8HTpin3fsnRHhJNE+tMi/MtQdu+SvwZDFUvH4dqUzKfHi/j5je8ij0VR/W9Gap4oYvOdT1O8/Xpyf1BJ31w30Re2YihwUPBykJBJUrxkMdF6L16nCV2BE71Tkro9QCA1yMjsRtLWNJKyzsCBny0hGAXh2BDWFXa8SRrFg1oJLOihbwj0j/GTuMmI/bgOnk1k6HXH6GmNJu4YaFFhdB5QgoL0pD6c2YLmFTloOugrkkRXgPmdGN4d8jZ7Zv+RwXorrceS6BopuGLibnqGa0SVdIMCsZUhjtSlgYRL5u3A/XAaqdsDuD1GdLYg7WMMIAeuFWv38NSdNzBj3FF6L/Zgm9OGrVoHEt587HeE871krQ/hyQmi/6mdrNRukvYHMHYLPNlB7Be18tY1/0Hza7kc/WwwzS/lEzYJwhaNmupkOrujsOX243MYufVXq5CvJGHabcOm91Nz+TJUeojq8gAAIABJREFUr8D+pp3ClxdRe8mL6Mb30vluJrm/P8YjG6/EdsJA5ZN2PvlyJOG9MeQ97CLcYqFrbBjfBBdxc1qYe+4ePHkB7m4Zy+iFB4lqkMRuN/L+i7MomlKDf2M2Y3+6j7yUTpyz3fhjwTfDyc5PSqh8eQiGfnDkQsmu62g5kow8bOfoh0N4etAaDo+SfOIZGAj7kw9uoHHFMNwz3Fx+zq7vs2lHiBDhLBDpIYrwL0H+cjcbVr3BlLt+iH9BL/1H4omqBcd0L3pDiPNyjrO6vATZa8DUrhJTqdFVKoga1o3rYDyBlBBpGxXaLglw78jNPFc2naBXj77VwKjpJziyegj2Wo32eX6yU7rp9ZhxHo+jYGw9roCRprZYDPVGVK8gamoHRXHt1Driad+Whj/HT/JGPW3TNIoKm+AqP6637DRWJ5KR14l3eQruQYLkac00dsQR9qoIr4o0akSX6RHn9nBg7HJyP7uFtI/0tE0S6DPdJNpd/DL/Y6aZAgz7agF6fRhvfRQlo2qp6k7A3W3BdkKPu8THotFfsrG9mKrKVHLz2yiNbWJxwhbO+/ABpE5iaVLxJWjo3ArhAg9amwktNoix0cDx25aS+/4dDHmunc5n9fQfjsc6tBffgThCFom1URCwg25sLzfk7+bLrsG0vZmNc7Yb3UEbky49xO7lIzB1S4JX9tBXH0PWWg3rT5s4djgTVFDdCpmfBGgfZyThcJBBD1eyfX8hqlshbNU4f/whtjTm4a+yE072Yz5uwjfUi6HCDAqMnF3OjiP56Hp16LwCS6vEfa4Lw64onMUBsjM7UZ+Mp228CXudxuQHd7G2ZihJL5vRO4OEftmL5TaN8scTKbhxP57LxuPIUjn04JK/+opafjIJT6r2Pbf0CN8XZ6uH6Pq3Z3+78DvwTOl7kR6iMyQSEEX4pyb//p1saDnInPSRbGg+wPzamXQ/lEnUk810eW1kRvWyoyaHQcv19BTrGHf5YcpeKKGnRCJCglBcCEuNnmCU5HdXvcZn/UP55LMxpH8eJGRR6RijEIzWUOL8KIpG4gdmNL0gaBXEzm/C96c0eoaoiDH9eJpt6BJ8pL1poH20npBVctuFn7G8ZjR9HVEInzKwuOvITqIvrKL/+gk4chR8eX7MJ4xIHdgmdJJic3KiLYlQqwVdqgf9fhueNA3VPzCztcXiZ9CPHIxY3cj6l6agd0t6ZvtIie+n/7MUXEP9KDoN2WOkYFgT4oFopFHFkWth7k+/4J13ZmKa1EXw8wSMvZLC28vZdnAwlkYdny9+mh9UzKf2YDo3z/6Ctc1D6TqcRO4HLipvtGBOdaEocsBQDkhNIMOChM1G3OkCT0YIERIocQEuGXKIlftGo7hUTJ0KgVhJyg4Na5OHyrt1yF4Dqk/B1CUIjnEijkURjNJQfYLYkZ1oUiCWJ9AxI4i+XY+5qI+oN+305au4s0Oo9gCiycw50w9zoi+JpuPJZBa10bI/lRkzD7JpywgSDkLvEMGsOQcIS8GmqkKsuy1M/cE+aufFUPXHZOJXWugaIai4eSmzy+fx46yN/PjQlQy64ii6QelU3JXJ0Ak1LE7fzB3rb0EEIyN7/hU5GwFRcnGcvO7t875d+B14duS7kYDoDImkzCL8UxO/LRaAqjdKGfHUYtofz8P1cycPDvqEaclVNDw1mKxXFdomqPhHuXCGjJTcUXYyhQT6Lh2+ZA17Ldzz6fWs3jsSY5dg8lO7CNgUTJ0CU4qbJePeQn/IhuM6J+2TJaoPWj/NwNQdwFfkxd1jxpTqpuDBbsIGBX9yGAS8vPZc9Lowwhgmb0VgYM2xp6Kof3QSvfM8eNNCpKzXI0c7yFzvYEJyHUcOZRFwGNG5BHeXfI5UAJ0krgzUKjOOtiiky8X2h8fTVxpEKiA7jITeTMKVF8JiHxjHYB7kpPGLTDSzjvoLbIR/0M2q52ZgmNCDXBuPJ12jvxBqHXFkF7Rz5fwvGb/+XjqcNrSEAK9sOgen10T2Oh99Q2wk7lJI+w8DcmcMtq8siAYzxnIzaav1OHMG1n9Dr6HE+9EbQhy+ZzhIECk+/CUeUnZqJN5Xg+txN9YoH0Meq0Jkuim7dwmBDguW0V3cft4mREjgDegRyxPwXdGH8KgDw+1/a6A/R8WbrKHvVQm79ch0H5srBpMZ1Ys0aLT22knbGqLKkUh0pcAXp5C2NUh5XzL7OzJQq81seeAZ1m8bSdbqfqi00nGxH32hg1GPL6KuPZ5nFlzH0Ylv0fjwJEJNzeQ+tINVBRswiDBLLng1ssxHhAj/R4kERBH+aYktF7yd8zkdYTdjc+pxZWnw4w60d5JY+Ke7+ODDqcTdX4+mV1AKXaS+ZaL25cFUOxJ4Yt67BFKDmLoE86bupXtcCNWrYG7WYe6UfFg9HEtHEO94N8+Wvstd795GMFrirbFjblbxXN6POzdI23gzdruXl2a8grLXTsVTCRj6Q4joAGMnHwegqyuKq0r20/SjEKax3Xz6ziv4U4KYjEEK3vDTNk3i7bLQPMPOmm2jiapWISwIG+EP6+cSiJHYT6j0lEDQJjF0qlT+WzHc04k+yo8/RqCZNTpmB1BsQfy1Udj2mglW2PHm+vElGQnEh4kzewgbBNoXcfRNCJCyQ1Jx41KmJNdQ15jI6zsmY2rS4zsaQ3JSPy9e/GcMG+3UzzENzETdHuTTd14h67VqXJmSnNVeTD2S9nEKht6BhWCL8lpIjHWi2xVFwwVmhDlMuN9ArN1D/D11VHYn0u204i+LofzfczDut3Fd7QwMSR56e2385dhE/BkBzCuiceQKokx+DEkepA5aploJWcDcrmDsFqj9Ksnx/YgOIzt3DMGc6CHhAwv1FwmWFrzD8AVHSHunkr48A0Ux7cgP4wlm+hm9/D7sVQpL0ndi6hQkxDq5tmAf6txuYjeaqb8zTM6GW/nzLc9T/cwENrQcpCHk4jd5wznf4mdOWilS/Z4bf4T/s0TmIfr+iAREEf4pMWU5if/zDgDmHlpA++N5ZG4M07wnjfi1J3AWhBAlDqbFV9JXoMffakEqkLGgio4v03h4zTVYKw240zU+byzAXKdnySUvk/NqA71zvJyfXU79xSoxdg93Lb+Nc2YfJK5MIlXwDAphNgQxtOsouvgE2hdxLF6+kKBVYjxswVTdgeZXOfJBEWGjZEJ+LZ+8NolQUEVujGe5M5aSIY0EgjqqFylkrtGwl+u465ZVmNJduLI1TK06Bo+pR+cWKCGwXtiGkuUm/rAgmBEgFBui84s0ElYN3JetWkdqch9J64xoJsmMG3dj6BOMzq/Hb1eJOaJQszuTvtIgztwwSreeH/36PXI/vYX39o0hcYseS70Of0IYERK0Ncdy577rMPVJ4sskyZc00DbWwPDd1xL1QZC89x2MfP4gg68/gQiBJ12SUBZiRuIJMqL6GHrZcSwtAulTQUBXm52yhjRCB2LIeFZl6PQq4pIdeFI0On6eQ/LrZgYPakdWW3lowno6xkkypzXQ90UKQkDYKPGWeAkVevCkasQfC6KEBr4MrE0KF52zF2+nhd5ChaWzX+PGozdx6LVheMZkE10fpGZhHtffv5605D7C8UE0FQZ/eRNhMwRXJbL87ZnMTK/AlS4IdZjBq/LkjEuYMaWM4c8s5vbMgSnXRj22iJq3S5HWEL+e++731v4j/N9EItCkcla3CGdOxEMU4Z+Or31DABN+cgftM0LoO/VIBTSDJHkXtF/k5/qS3eyZnkDt3UNJm9ZEXVs88Z+ZEGEw9YbpLNVh7AVjv4YjS0HvHph/Z+j4Gsq/ysXYI0i8oIm2TYMIxEhCCUHshw1cd9unvFI+gUC7BWnSGPKck+dXv8S5n9xH6ucqYb3AkSO46NIdrPp0AtEVMOjGGo61JqPTaQQDOiy7LYTMkLm6h/Ubl5Oz/ja2zn6WeU89iDdZYhjeh9tpQldnIqoOgvYB43K4yAW1VpL2ajRfEMZaZcA2rYMUq5Pj7Un4u8wISxgZFtRe8BKT7ruDrku9UGch//dVtFxbgFTBmyhJOCQJWgT9BaB3CTyZIXQOFUOfQDeuF5fThOmYGZ0PvONd6HQaSXYXDW1xWGx+PC4jn0x/jsuX/gRjj+ScH+5i5d7RoEpemvEKt22+BWEME7fFiCsTZKEbrd6KqUvw81vf4ZEP52PoFwRiBhatNXUq5J1fQ8UXuQDkvtdN/ms1fPrxWHypIYQ1xO8nvsfTP/sBLReGiD5gRM7sRW6JxT3Kyzn5lWzbWELUqG68XyXgzg8Qc8BA3/CBSToNnToCKUEIKcwdfYi1B4eT8rlK76VuAl49qWsNuFIVUv64ne5bJ7L38aV/bXNdYTcJqpXZ1yygbp6J+GGdhD5MpG+IJPIj/V+Ds+EhSiqOl9e8ef7ZqhIAz49+O+IhOkMi4WOEfyrk/8fefUfJUd2JHv9WVecwPd09PTlHSaMsFEZZSEIJiZwRQSSBTTBhbYMjGK9NMmuMyEkEAQJMEkIogoRyThOkybkndQ7VXVXvj9m35+3bfW9559iL7defc/qc+bP61r1dv/nVvb+fBC2P1rDgmlUsunAl9/5yHVk7dADk7Uii6TR6pmukf2Pig3fnkLcpiaDCwKf5eDYa6T83juwQaDt/OOugD2t4FyRw1SkEJ8Qw+AT6ni0hURAnd1kr2u88qBODaDoN3YCeyNQIP3afoW7mm5hzQ9SMOsvSdXu4/NEHGPnEAIaAQt98GVedyglfLs9f+iLL7vqGY6eLMByxceOIPZRm9bPs+l2Ypvfje3z4Yb1lwdPc2ng5WfsCSFGBSKMDhyNC0cYIhrCGMsuPOipE0mvGOCAgaHDTlF1ERsbweh3UezOJ+0zoAhIEdEyoaOX8hiXEHQJCo4X0sf3UPlJC3AmBMTIAaWeDJGwCtdc/i5yuYW3RofcLGKcNEGx1IPQa0cWg+tJabqreQ9RnovNwDra0KNEWO8KAgTKdGUuPRsIu8NHRiej8Oi6YcJRfPXAzxm4dhhYjilEgURoj7xUDNy/ZgnFA44W22fz8wvWcunMNyTQFURZIODRO1BcgOxUs3Rqul3o55cvh5Rv+RPZOkdsm7MSnWBi4LEJ5oZdAmYrjdTuJmiCN577GtmMjGT+/Hu0TN7EMFedBPf6pMZA0PHt0qAYw9Oopf0vmT3n7sNXrERMa1NoxNpvwXhhj8Q276XpgOopRYNn0FQDc1lHD9VMu4fKm+bScb0KKwd7xHxA4N4K9UUQfTP3Mpnx3CsJf9JPy3aUyRCn/MP7XqsH9t9UQLIbyV3vpn5FF8uJB/E1OMvdB6PIAF5ce47PnZzM0IYlgUiCox3VUxF8Gml7DWubnzfGvceHXd+D81sjg5CT2Oj2hEoU3lz7HtVtuw3ZGT8IGCbuKZtS4bfY2Pu0cw7djPwLglvYZnPrDGHpmaeRuh87zVAyOOAmvGSkqohqGixl+fO2TXHzgNmpnvPkfvtNdXZP5qmkEZfcOUvubbCx1RiK5CpfM3M9Xb9Zg6VXpnT58LF4eH0I4a0UrD5O21UrWZ020X1OG6dw+QlEjsR4rJSO6ael2U/KKQN94E7bFPQztysbs1bjtR5/wuwOLcTpDyN9kkHkoTusyPWJcwNohYO9KIsU0in9Vx77PxiCPipK204SrLk77fCPWTvBXamg6DXNeCINOId0SJRQ3MjWrlW/fmETBZU20fFJKJEdDisNTV77GT15YhZiEtFaFQIFEIg1imQqe/SI19xzgXs8O5m67mzkjGmgOuClL66cxkIH3m1xy5nbQsS8PxQBVTzXTfHMp+klDiJudWHsUHPs7ibwq4fskD3tHknCmxLmr99Ibt7Pn62p0ZSEy15rpmCfiOiEQ8wicuGcNo3ZfS9RrQUgK2FolQkUKaAKjx7fQvr4U2Q7RHAXBJVP1zxGUU/Vs6jrKjqjIXLPK0nELqft5KRfNOEB71Mmq7F384NP/ved1yj+Sv1SG6NK3lvylLgmA5ya9ncoQfUepf11S/iFMnVr/b3+3vj+GjBf2kMyR6Z2Xhb8CAkELF8/eR9QjEm1K40woE/+MGJJfB6pA/maNQBmMrGkGDYz6JH/oWcj4knYGJyoUfQS3rNqAMSvCDXtX0Xz+SyStULQxhNkrMnNiLe81TyTyWTYAc09eiFmSQQNbo4QhoCCYFOSgASksIibA5BWR82RGGiz/aTA0ave1OPUR0j+1IpdkYq43EhkVQxcS+XDPFB7/wUv4KkWcJ0TCZQnSNluhIkz5wzEGZ8jUPlSMqgdfwIJRn8SaF6TlRC5qQqJlmZFQscq87DMwPsDgeIWnT52L1R5D+NTNhItP0nSphOqRSaapxN3gvyHI4Eg9Xx8aRenCZqQWE6Ouq6V1qRHnpD4iuYAImlHlqopDrCrfQ0u7B99pN3/K28eRB9fQ9kEplgVeHCMGkDOT/KJ2BVMuOU7Gsg5WPvoZgbEyxkG4ZMZ++idofL5tMue9/gC/mvYJC5yn6QtaOblmDJ1Hc1DMGg+XfkzOt0kyqvuo+3Ex0dI48ePp+MYm6JoLTasKaWn3IKdBxzyRoXEqX7aOpPGpkShWFUUR6JwjYukRCecJTL/0CCWf3wJH0jB5dXgOiNgX9lDwlUb+NpWzm0tRJbD0aEgxkXUzXqThRicPNR3l7aCbm3ZfD4AaCGBtl3gy5zD+mQP89KmbUqfPUv5Lqeau369UhijlH0Z6vYDFq2IIJNFtPYRndzp99xaCqqFY9LScb0DLiiPpVIqfgd5/kgm1ONDnhpGO2Il5VCw9IhdcvZOv/jAT6UovFel91D1fTSxDIGkCa5fGULWGo2qQkW4vuw+MwJQXQrcnjdCYOFmb9Kg6gY8ffZxnBqazftMMjJUBzIYE/Z0OpIAOfUggnqnw2bKnqTaY/913+P1ABX/+/XxKb69nT30Z0tDw6z7FlSDNGcH0UTreOQlcmQH89S7KJ7ZTfzYXyZpE6DCRdCewnjUQLkuQfkxP0WWNnN1YhpymMWpGE8ea8inIHWS8u4NNX56DnJlkwogW6rxZJJrsVExupXFXEfrwcJd6bc4Q4jYnqh4MAY3BmXGcu41IK/oxveikf2UEudWGoSjEiMxe6r+swDqjj75uB56denyLIyT8Rm6Z/jXr3pzPtddt5tVPF6APCHiOyhh/3E2n30FVhpczH1YSydZIZiaw1RkIlSbJ2CcRzhVwNKsMLI+S7QrQvyuHnDkdDH6cTyRbw9oJxhVehDc8eJfH0emTKE020htgYIJKwVcag6tCuF+0MjBGTzRLHa6I3QhJs4DsAG1SgJ+M/pLr0vqZ9KvbufKur3ju4FwKPpToukpGSYrkZPqIf5jF4Kw45486gS9hZufJKp6Y8x6X2AL/7j6WfnQbzqIhDk16n0W541l0MsBg0spvMk9Q9u7q/7Y1kfLf5y+RIfKMytAuWrvsL3VJALw0eW0qQ/QdpTJEKX/3/rB8Lfsue5KMy9uZ9LNDXPjHLTSvG8e3teU4nuik/Tw74VwDWnYc+34ziZCB1qUWgn4zqkklPmjGEATVonLOxSd4++BUwrkCyfcz2be1mr4ahWBVgkSahj6iUfm7Mwy1ODnQXohxQEQ9lUbUo/FIzcf0LkpgDCpcdPIGdv16GsYhAft6+/CFqgI7rngc9ymFeZNOsSNSyeuBTMrfWc05v7ydCQeu5KPHFlB6ez0Hdo1A16cncz8UjO3G87UBhznGwBiGqyqLGrqwyNnuTHSDOtQBA44zkPmNHn0IPLt0SFGN03tKyfvdbqzVQ7R8UIZzvwFf1MT+pyeRcVxDiIsMxS2MyOwld6dCbW0+JZ8E0UXA3q6Qt7KDQKXC3KsOkLAIzKtqIFgMkqjSOUdkRGYvYl4E4VAaR4+VwiQ/1mfSuWnKLvqmKRS8NNyn7e1183GfTPDq6RrsLSCna7QulzBICpGQEW/EzvIbdiLFBSyOKNGs4ZN10mV9yC6VaIaI2SzT3u7mvAv3Iz7iJv2MDBVhgkUQiRvomyhgPGkm0WnlB8s3EigBQRGQ7SJvT3iV8b85gnN+N8biIKY+gWimQPauQVQ9iKLGSw9eQun61fgr4PW3F7G0+iQzHt6LY7uZEb8P4zRFKVx5lpLcfkrNfRz5aDR3T9/8H4IhgKaLX+DQpPcpe281iBKjTe04dBHm3HZrqiFsSsrfqFSGKOXvmnFQRB4ZRXfWzJTzTrK26BtubJvFkbfHkLu1n8jTMr3f5uI+peC9NIorLYLwZgY95yo4D+sIlGkUjOumtcuNrsuIffQAgVNu7C2gLPYhH3ZS/KmPwO9ivDVyLStrr6NnMI20nWZ0EXBd34ZJShKd66XnrhoylnfgW5+HuHwAAP8JN8YhgUiOiqlPRDVALC+BsVuHJkDhpiiNt4jMq2rAqovzzRuThxvFGjRUHSxbcIDP60ejKQKubSY0CQwhjfQj/dTfloGpXyTuUsnfqhAo1uFoTtK2WEK1J7G5IkSa08g4KtB3rswvpn3GDWleyt9ZjRQVmL/4CN++M5FAVZLc4n7iH2ShiRBfHEA76MA4bYDETjeqHpLjQpi/tRHN1Ch/uZMNuz/l4rMLOdmZi1RnRdVrnL9kH59tmkoiM4H9tAEAQYPkTD/iXgcXXLOTddtmkL1Hw94SoX+cjahHIJqtohnV4b1cIR0ZB0SimQJiAmyLexjYn4XsUtGFRBJZMqYWI/ZWbXi/lwSGEQHC/Rb0aXGUHgtzp51k5/YxaCKkNUI0S+D07Wv+bc7c0j6Dw948hlqc2AoChDrS0Awqaaf0BEtVCrao2O9r59TZPFwH9ARKQB8WEJIgqFD0VgsdlxUTHB+n6bxX/q/zs/Srm2g67xUW5Y5nU9dRvowYeeSnN/Dtv7yQyhT9g/lLZYguWHv+X+qSAHhl8hupDNF3lMoQpfzdarzyeTxHk2yb9Qymfqi0enl6qJhdO0YTztNo+42erkM52Kf04SuXsFni9La6GBop4DihJ/ivGYRKh5cfnbOVO1ZsxOezYhoY3lwrbHWSO7uDM9c4cBhjnLvxXrprMxHazEQzh3/3Wr8pon57GT131yAmoSajGaNfwx8009+WTiJz+JTYvkufpGxRE0mThn5AR9KqIVSF6B9rIX2vkZ3NZXRFHdg7FZQRYTIm9bJswQG+aKjmnZqX+PnkDSTsAtoFA4RzRGrvdoEIlm6N9HqBodtCyAsDdM7Roek0TGlxQr02zMVBoh4B+zEjj2y8mOtaZ3P3ko1oEux/YQLRLA1bVoiudjcHH36OQDmEe63kzm9n64TXCI+OMf2CY9Bgxb2iAzkrSe2Pcilft5q2teWIkkosK4mzVuNAfxHuCV62LHyamiuPYAhoeI7GiHitaDV+6oJZqHaFoUqJjvl2wvnDY2jyiriOSBhbjOSU9xEqFCj4vA/Xsk6Cm7MxDghoehW1MIorI0jcpdI3RUHOSaAUxEiecJBdMMjo3G6kqEBf3EbClcRc5cM8qBLNS3Jp44J/mzcjrd1E92aweNoxjk9Zhy4okn5UTzRH4/El79BxRYL+l4owtRkouLoJZy0YB4aDoaQFlBwXgbEyBov8X87R+oUvUvHW7YQvmQrAH2fMwbZ+HzPvuo3y+1INYVP+PU0DRRP+op+U7y4VEKX8Xfqfrx1iTolbCmeizR9i3bpz+ePWReTuUkimK1RmeElaVeKbPYTLZWL73JgyoiTsGnIazJ59AlubwFhbB3/YsZiX3l6K5jMgp2s8c8MLxDKguSsDz2gvDQeL8OyVyDg8fKS9aE4r8Yt8vHX90yRsKoGxMpHZIbY/Op3cH5zFbotizIiSu0mHrUNFQuCXhZ8ilYagOMwdi75COmpH04EoA80WDp0qJXlLP0VrRGJ/zuKTY+NJRPRc+cUP8egCyLMDWF9OJ5ynsnzKEcyFQfyVoOgFri3fzw9H7AANMvcKSJJKZUUXckMaOX/Yh61DwV05QKF5iI+6xnPguqcYmhsjadYocQ2SnT/IqN3XkvTINF/wIoHX85n32P1UFvSSZQxQOL2DtkN56O1xpp1Tj7VNJGkWoNaOkBSwr+qkvdlDbEMWZXob33w+AVUPl6/5kowDEuK3Dg7VlmBIi5O0akRGxBFlyDoQp3hBC76ZMdDAvyMb/aQhzlyfweDnecQ8GnEXnDfhJL+Y+Dnp5hhVr/qxtuogKWKyyMSzExifcXF8XzkJp4L/8UJch3SoqkjaD9opfzvBT/O/oPKb65j649vZ1DuKh657jwPPTeBZXwEI8NlPHsN1QuMnf76G2eVnkWSNRJrK8ZPFDEzQ8BwNo04O4KxXabjTgLnJQN3M/7gR/n+nFyTOXPscnUsVAL448hXqnAk88dgaNnUeQUz1PUtJ+ZuRCohS/u6M+JdO2n8+nSVVsyi49Qwtj9RgXu9ANYAUE4m5JNCpyKqOtDMSoWIVVAF7m4Zw1I5qUUhWhzk9lEX268d4cstSiip7ePiGt8g4JGLuEfjpI7eii4LltIkJGZ0IeVFsXQnCeQIN1z/H2UOFxGUd1xy4CXLiLB97DP1hGz0XyAQfyCW5LYOiy09g7YwRKBa5vuliJhkNxMMG1HYrH3eOI2d3DH91gmAp6MICGfsk5KTE5D8eIr4oALLI5RMPYuyVeOye64h32HjwyTeo+lMXx341AVmWyNqvop43xEufncefuyeQe043rgP9WIwyvjcKsLVC89ujcf6gDf/RDA5fUUXr6RyW3P8jirMHSK8VGHq6iJ5WN7GIgexsH5Wv304kW8DWpVBfn8fW7kpa+5xoOg2tzUrtOyPRdBCZESK7pguDTyTwej7NK15EF9aofuYOYvkJEjaBWx1dhHMFdBGN5ROPIocNiAkBQacip2kE7wlS15rDuKIOBBUss/rQdjpJawYESDsL4xfUcfKJcfz2/csIywY2blyHqh+eC55XLAgmhYpfnqZkYgfmjAjtC0X0EY14TE+WOYgF7RhSAAAgAElEQVTs0HPta/egP2bjhYefJvR8Hms7agiUwTPrl2PyClx68gYybmnF1iZw9rFRdC9Loho1XIdFnCcFEAQc1ij94wTOKWsla04nK858twJ6z/vy0PfpmXz4cvbGFJpXGJlmGu7tkTZqgFcuePGvtFJS/h6lKlV/f1KjlfJ3pfHK5+mfnU/+1jBqMMgHZVvImdpN77lJMo4rkB8lWCCQv0HiVHMuaed34zkAk0c10TdZBQEkewK9IUloaxYt942jcKOK6X4L9++4glCewORrjyGoIMXA3Kvx5d5xZLkC9NwaI+7UKN9+I+fNPYLcYyE+YMZy2Mzp+8egGkDXaqJ1qRXFAJLTydlbJKQYXJsz3Ebk1zWfYO0UaG/00Hi9wFVT9mHqF5h1wREUEyS3ZLDrV9MQdjtYPWM7EupwMclpEvqAwAMnLqH23hz6bojQMHstXXMEQiET1y3bTs/nhfijJia+V08oaiSwPERaW5Kllac4VZ+PySvQfFUmBp9I92wVuz6OfF6Alb/9jN/M+5CcTB+hmBHTgIB7YRel99cya3wdQ0EL5r027I0iWl4MZeEQd9/8EQUZPnKtfmS3iqqHqldvx+RXcdUrGNNjqDqofuYOij4eZML1J9i49RycB/TY2jQ8G40oVhWrQUbXY+DsYAYJu8aFBccJViSJnRcgODFG6LwQde+NoPfCGPZmGDjpoXTLKgx+QNToWimTvsfIzi/H0f1VAWajzJFL/kA0Q8Rw0sLRt8bQM1XC4IdIZZyLvrwTy+ouzh7PJ+0sZJxQWLJyN4oqcqq2AH1IwztJRJNFSj9KEFkSJP1snMFRw6cB7S1wfFslrfXZPFiw4TvN2Vebp6MJ4Dvt5s7TVzFnxkkAxj55B4cmvc+qjbfgOfzXWC0pKSn/L1IBUcrflZKNNyNoGq3LLCw6OXy6pyq9l/K1Cp3nQmHmINr4IP3jJESfnmhCT7BI5OxgBmUfyogJKHtaIdJnJWGH9DMqbUtEmh7Uo0+LU35eE20hJ6o03Kajf6rC9En1hD/PRjueRsWUVqRmE1vOVmErDFD5Rgx1hp9Yhh57i0bSrGHqFyh6t53wjAokrwF9SMMthbipbSa//uJSQsUqYlykaL3IxrZRJCyw7WwVgTINfUijZ6qEMi3AJ4/OZ93RyWgSyDkJZJdKuMOOuVvCuCONko03c+zSp/G4gpxjaSaapeH3WfjmlzVcX7WPeLcFxSCy5f0pw4FgXGP64uOYe8F5QkJF4IXxw699rrEP8O3Yjzgx9R2O37eGHaM/5pmCTUSSBhKyjujUMOFCDdNxMyMyvDzx3sXof2zHJ5sx9EuMvOUUurBAxxKVjoUaq6t3EvOoRHMU6u6y0/jISNTcGEmLgHbhAN4aDc9+kbZuF9Y2gaDXhpYXo9ToJW+zQDxmoDK/FzmqJ1CuUPmrIEmLQNKRhCED/vEyZesUTGaZ6+7ciFwUZ/JFJwgdc3POO/eSnOXH3KcRLNIYM/cMhqCGZ4eBkj+rxP6Ui2pViK3wEyiU+HDbNIKH3QgJgb5pCkJFiKKPwVdmxGWLoBhFBsepDB7KJGdTN7devImbZ+9g/dBkHuiZ8F/O2c3j1nJm5XCbj4FWJ3s7iyh7dzWf3/0YAJpBJVAspk6fpfxrL7NUHaLvSyogSvm78culH5CzWYfjrb2MnNnEnc4zjFpzB3WPjKHjh0mMAxJDETOcsCOXRXGUDWF91sG0FcfxNzqpfuoE8vgwV63dhJAQiOfLeJfGcR0XMe21Yd9mxSAmEQQNa08S9zENRI2Wp6qQ0yBWLNOypZhEYZyE30goYKb5HlCOOxi6KoQxqKA6E7z+oz/Q9rSdnpUxVINGcqmPWz+5hcO9+RgHREy9Iqo9Sfd0Hbl3hbF1aCh9JpT0JPN+sBfFpKHbnUbcIWBqNeI4A+eNOQWKQPFnCgm7hj6oMWPkWSbuupVQzMg976xCK4yibzcy9mfHeHnjAnZc+CRpx3qJZqvobAnkdIGG31ejmGFonMIMVyMzTCK3Orr+0/F2iGZOdueQk+FnXlkDtlGDxF0atZ9UkaiM0D3LQUNXFs56lZ3Hq1BMYG7Rk7lb4uU3l4InzrvL/0RBUT/ZDzYyobidqEdjqNUJaQkSNhD1KvFzA+jT4pyd9xqPvHEVg6MkzEfMDLxZiNRrhPQEvBBFPtePe78OHAlK1ml0zjKRTEq8sWYpT89YR6M/AzkribVDwKhPMjBJofKFHjJNIfyV4LyunY5z9WgSmF1R+DYdo09DzItgawVLfoiqV6Mk4jrCd/gZqtYY2J1N6/kSph4JWzv0LMzhtbWLkQSVLz+Yxix7/X86dv+rC05fxajn7gANNLNC/qWncTQIzPvofgAun3yA9Tc/SdPva1JBUQoqwl/0k/LdpQKilL8LurwIb4/Ix/7uXqRRlZw8WMKyS1eRtGkYfAnsn9uIZSZxPGvHVaeiBQz4WtIZui3E9kPVqOkJPj01Fo8zyA1pXszdEpdPPIga0RFf5ic0KUrm3iFa/S4a2rPonaynf9nwRt/umQIGPxjbDajjghTlDJC7TcSxz4TSbcFxRsW8MQ1ldT/EJG799T0kkxK2bVYy90PW7w3cs2gj/jYHmgSKWYOkSPHnYRzvhIhf5GPE8wNYzxj45Mtp6KIC0SyNUD5oAihG2PfWBHQRge7pBsrfGiDuEqh7bSTGQzYSp9NAE1D6TNib4Yt94yn8UmbutrvxzsnB4BdxbDUTKVDoWKLiWtSF0RXlx+4z/+W41818k11jP2LL7nEo290kMxKERshkuwLYOxRy39cTdYsYnTF0EShe0MLAOI3CRS0UZA3xq8mL6Rmyc2zTCA4fKUMriIE9wQ8n7iBUCPZdFvKe1qF0WZj2wGqql9YjJiD/hROECgQqJrdirjVR15aNe62VgSlJTGeNDFQbQQDxQBrW5T10JZz0DNlx79fhXtGBUZ/E0qqj9bIcdr4/kbEzztD9RSGjp5+lc6GKcCgNeXKIofkx7hzzNUPzYjjW2ThztRWp04T/hBtHvYA4zo8lL4RpQMPWqTA0I46cpuGVhzfE/6Zh2f8xS1Tyya0A3Fm8jWh+gqcveQ0EjY6f1JC1awhdWGDJ0qtZv3sqT/Scx7IFBwAoHvufB6gpKSl/XamAKOVvXvl9+6iftRZ58WQ2dR1l6EkFw5DImZv1OOrB9+Mw4VwBa3YYMaERtwu4DotoVoUMW5iMAyIkRM4fdYKuLhfnNyzB4IP1JyZyU81O3hz/GroWE41XOhmqc6HFJGztGprXhOOkHtxxYm6ou2UNkqQSeieXpFEgMC2KFBPwTtUYmhuju8dJem6AmFsg5jcCEMoTab9XZWNNEQC6CT7KZ7dQul5hYIyVPafLCTc66FzkQROh9KMgqn64u7ycoWAeP4jr8g4cTUnmnHscVdKou82FmICBaQm0Gj+mPgGjD4y5YaThU/4U/rYBRI3BsRomL8Q8AhPHNiJZk3h35jI+r5O3g+7vfA/cFQPEXBppxw1cOuEQvh3Z9F4cR9ULhGoimHbaueuGjxl6qRBzj0htbT6DETMTt3jJfcNIPHO4F5jxpBlDi4kNd82j/LE6YvOCdMy3UDC6h/6lcbwROyOWNRD90I25TyPwpwKiuQr49cScEsZeHY9c9xbmAZVYXgLzrH5iH2bx+32LSQyakNMEREGjtzGDSFESxawRHh3j9JeVxM4Jc+xwGTqfDucZBZ1OgS4T7/1yMTl/NpA0i1x37jfkTuzG2i4QmhtGUUSUYw5iGQJfv/gi9iMmFCN8dOAcMg8lsD+VRlgx8rwvj6cGS//dmC2ffIRvYvBo3RLunLGV5zrm4f7WQMKhIUZiSBEB9ehp9q94ij1txex96hyGNlQQjBupemngL7N4Uv6upFp3fL9SAVHK37Tyd8M0vzOWF/25mHrCzLvxZuKfZpK3I4Kt1oCgQTBiRDWAethB5xwj8fP9pDfFWT7uGK1H8hgaAaYuPV81jSD9oIHmARfhQo309DCvHpnOlWt/hCaCcUjAWu4nPTtI/8wEmk5D0MBUa8bWrlHyya3odjkIFgv0L4qhDRlIpqkUjeyhIHMIQ5uBUL2TnCd3I/l0CAqEKhIIx+0oo4qxtUhEwkZavyhB748TX+qn8FOB8gntJKxw6s41NKy0oYkapskDCIqA9U0HHYPptC8W+GbLWPJrOhFlgcyL2nDv1pP/KFh7VSI5Grp9dmIuAdEps+P4CKQuI/qggKsuTjRLxW0MozckiXkU9p0qY21HDTe1zfxO96Gvx0HCqRI6J8od7p3kbwmgdZko/FED9t0W0hsTPP7pBQysiBAeE6NwI8Tq0nl/40y8k/QUVfWgz44QrYqTtKlsfesV6n5VRUnGALW3rSEQM1L52wgDYQuHaktwmcIMTUjSuURBjIloZpVohkA8O8mvn7+WvY89j+OEnkh8+LiZ1RGj9EMFe7tK4qlsXMdELph8GNmhoSkisaoYWe+bMHeLGMoDBIokIgETjrPQNQd6J4sMjYCtv5hJe6+TxAI/CZ8JQYBYoQzjAsw7dQFZ+8KoRhWAwVtDqHqR078Yw2DSRm0459/Ga/Tea/AlzIzSh1FUkf3+YuoOFuG4shNVr9Fwaw62Do2Hmo7yw7bliMfs7H38eca4u7mtdCeDE904zqQeZv8/Sp0y+/6kRivlb1bjlc8TKrZSctUx1jx3IRu/eIe2xToUo0DXbAtJGwSLBCwmmVh2ElO/RtKmEuq10T/GRMPtI1A9Mjl7FNBAabGRtceP/msHrhMavtZ0bpu4E3urRu453Zi9GucV1lHt6cHcMlxp2TSgUrSwhVCBQO52kWCZQlqzhqHBjGFIQrMkaWn30NbtwjxuCPfoPnrvnI6gCrzz8yfI+loiMTKC+ugQS67djdRhQk7TSDiMZNjC9F0foWtDEfKoKOXvrCazop/MgzAztwldUES+cRCl0YYYFzBV+/BuykdJS9LsdRNZEqRnuoOhKhE1N4a5X+OhO94m+yMDWflDLJx/BNcpDd1DvdRMrcMsJYh3WXGXDuE4oWfTyM+5ImMfH4dt/9f7sD+eoHnJyxRs0lBDepbsux3x8UHszSLHPx1J9dWnab1Uw1gRwLzHhjBkoGuGRCJdIeFQkGQw3yFh2WkDQUMXFhjz9B24jgs0fVNM6Ue34T/rpH2Jm0hLGvY6PYMxK55vdTgzgwgKZH0tYenVMHh1aLN8AAgLBrmg9AQDExXUIw66ZhrpHysMb4KeGeeT/ROxdIm4PAEEQaN7uki4PEHiTBpb7nkcQ4eBoWoVR6Gf65duR5Og6sFTqFEd4rcOKiu6cFij6Pr1yC022r0ueqdasbVKYFQI9tmY8bu9LHlsB1u9VdT7MhmxayUlm27i6vKDeKN2TiesBH0WGt6swuAXaTuYh+pOoBVGcb65n9Wv3cGxzSOIVcYYsWslrxTu4tVfXkjSIhCYFSV/u/pXX2cpKSnDUgFRyt+kxiufp1aOsO0Pz9Dx0+kkbFC27UYeXroeVQf5W4I4a1WqFzSQmxYgrU5H5cp6qp7tQReQ8I9K0jfRxvmjj2McTKCYNJbNP0D97Rbk2QF0K700XfICz++ZS3BxiNgb2cTcAhs+qqF2IJPbr9yALiTiXRan/kghtavXoErDndz1V/by3PXPE8+XKS/0Ig3qkHqN+LrTuLNsOw/88D2kGFx36nqUqwZRBozEkzp+n3WU9DH92FsZfmh/kYfcYiNUoiCKKqpZI/mhB+36PjbsnIS5V2Cw1k3Ck0DvFwk1O4jkqBSV9GH72kJ0yEw0W0N2Dj80fSPgT/90Bb2XxPH2p7HxxGi656o4TREahjxs2DwZQRluJyKnQ8mGW7jzvZuZZ+7jnu7/vLK/Vwlz5ad3Urp+NeUPnUbnkFFVgebtxeS8eZJInsK3teVUrjpI8mQa+tBwKyDLCB/2Bh1CQkRO0+helI1pWS+iTuPhK99BN2OQgckKdbesYfrEelSzipyuUTymCzld48mK97Fc0434qQvVoDHurmN4Zygk8mWSR9IpX7caX1caW7qqQNKQ4lD858Hhdw6AY58JXUDirhs+xtfowuMOImhgadZj7RCYuvEeZI+CySsh73PxyoGZKCaNxodG4MrxU7i8mYa2LOzGOGavgPu4wOLK0zBniIQNPDuGA+b1n83k5Q0LaO7MoMLRx5MT1/Prmk94MKOevncLuf7rm7h54i58o1QyZnZj8AtUFPQiNZvxf16CpUdDSMDWuX/k2IxXAbC9v5eByUnKrj6K6bP9lN+7N7XZ+v8Xf+HXZalXZv9vUgFRyt+cvK9VSj+8jZEGCyvyJlPwlZ9Hb1yLIGr8YuNlnLh3DQ23Gem/KMKprZV0+h0EKxXq+zOpvS+LsveDZO0S8VdqfHZ0HB3zzRR+NZwxMLfqkeN6KtL7GP3HOxDiIvFBM4PLI7iXdOKa2YOyOYN1jyxBFxQwNJjRB0Wqn7mDuENAMCp0tWTwRPsiqss66d1QQMmETjRJw31A4mc7L+Kz/nGkN2ho73oQPnAjKAK9B7Op/Pp6HL+zYQhoBAp1qBJcs3AntmYJrdVC3lYYmJJkz7gPeXDxx5RdfAZBERBCw60+yIiDAAObc/FNkzGlx5g8rxYpOwJeI0mLSvtiMB63YLLISIM6dA6ZfQ0l9HnTcNZC9h6NpOVfowZFwHNE5UXfaMJJI4fiMjOOXwzAfd0TAXi8byb6oIC9WWT7wWoqcrwkfCZi2Ul6rq5G02tMqmyh4bVJxDMUfPOjPLnsLUJNDsQk5FR5ERQBdcEQ/ccz0TT48a7LCIbMiBGRqp3XscB1mtySfhJ2jRxLAK0qzKWbfkhSFfFXgJYhU/fwGH4653NMZ4zMPv8I6acF9M4Y/iMZlK9NkL9piLPXONEqwngniRz52RpsbQJ/rJ0HGsifeZCKQwD4p8QQEiIjf9ZItFgmUi4j6FTsLSIdCwxk3Ztk6NlCpH4Dbf1OVq36AuHKPrZ/NAl1l5NYVhJfFZhb9SAO1ybKzfKxY89ofvHYjbz8k4u5sW0WRr9G8XsiP3bXkrVHoKPbRdaBOG5TGKU4RkTWEyyGaInMmoFZGAU9S8fOp291DTZPmE1dR9nUdZTEeefQ/rPpqaAoJeWvLBUQpfxNKb93L3GHxO8WvQtA2y+mY3u6l4efWokqS+j9IjX3r6bgM5Fkv5lEZYRoXI+lXSKxx0XVC37OXGsj6hFJOyMgRCUMPhioNpG7XURUwJUe4tgbo7nkqq8xDEoIioDZlCD+Sg4uc4QRV9Sx8Kc7MQ1oVM1v5NGr3iKeoZJ2cTeipKF3xDGISeoOFSFosHnkZ4h5EZxXdyAGdBzYX8nA0hjBIoHkcD0/5JwE+hNW/GUmoh6RzMvbMM/r44ivgMikCLZ2AWt7hJFPDDDhwJX88cWLObm7HKkkREb5AFJJiLR9ZqZOqcfRrKDFRewbbBz7eBTlj8TRJJg4oRHJnuCpm18iecZO8ecyTkeYdFeYEf8SZcTqUwyOktAMKtH8BKZuHe8+9SQvfXYe+aYh7j97Gd1nPZS/fTs7e8q4r3siH347hUS6ihTXcB8e/rkoLuuluKIXSQbPPolTWyshJiFFRfT1Frb4qtGMGsFiFf2/uNFFIXo6HVu7gBrRkb1Zh2eDkepJLXw5bQ03pHkxSAqiLNDod5OI60CFoW+yEQojGM0Jhip1PFM3l1imyu4PJiAv91HkGcLcK3D2Bh29NemYewV+OWEDq1Zsoezd1bzzwBPYzTHyt6nY2xXiQyZMNf3oOoyIEZH6hyowtRuoLu0Evx4xDu5jGoaXw0g3ezEOCuiP2PjTl4tR3/MAoIuBYEtia4Pi9V5K3xsCATq7XKhWhfSzMsF8HYd78hmsFog7JBZffRODowSkHgO9k43sOV2OwZjgwZFfohigvKSXuKpj6YjZPHPoEzL3Bdg1ebghbMVbt6MJ4KpV/m19pPzj0kgdu/8+pQKilL8Zz654jcDV00hfu4cHDwxnKmpXr+HU5koMy/swNRq56ZJNmPuT9I3VITpllJiO1dU7ieQphCtk2pa5sDUP7/WRzwtgzg1hCGgkzvUj3dRLuCRBYdoQmZe18eaOWZj7YM2i14nWpdM9X+Hu/M3ck7OZ3T+cgvECL2f6M7jEFsB1XMC8ogezJY5rg5m2N8upfKGPzAMRxv/uDkymBGebszD3iugiAoaTFqIFCcZdfxLBE8d22oD7VBLXiQBLbtqF9qCbwVo3bR+UkuUKEChTaV5hI7pGYWRGLzlrDiGVhZC9FsK7PFRm9eGvVNhzrIIZD+3DeVTHwFiNcL6C66VeytbHOHSyFKHTxD2v3YLBL9A33kRiowfL2+nILjN7vqlGTlcpXa8g2RP8+KoPuOL++5kyp5Y3N8+m9+s80gt92CuHCH/rYfcTU5DCIiUjugkVgbzCh6oJ9G3No6UxC3mFj1CBgOxQceb5uXvxRqxdGkcH8nDk+zEVBfGvDhIul1ENGr6JMo4TenxVIqF8kZPHiijRD+9f2l79CaUfRfDvziL7CwMmr470Mwr5r+hw2cOsXLWJULcNnDLuBV3wjZOejQUUX9qIvk9PNFtAH9L42fZLeOvNhagWlZW/uQ+TLonpni4W/vYbdPYESUWi4YbnyDiqgSdO7q446YYo6cU+Rl5Ty6i7T3K8JQ/9710kHBqmAY0p0+rpnyfz+erH8E+JUZnfyz/96F0afmGn6QonTz/wHC5PANdBHYN3hwkVaEib0olnKoTyRHqmmFHLoijZMp6jCcyterIcQR7cdDmqSWVseidJTaLumXIu++cHqF9tYfYT99G3uoa88d1EMvXseuYFzrwxkdZfTyfjSOoh948s9crs+5MKiFL+JhgCIostcaxdMj8408DK0fsYteYOxu2/Cl0MfAc92Fs01v3pPJY8tZ14WQzjSTPVpZ2s+XQJ5e/EsNUZME3vZ8rVx8jdDrodDoT9DkL5AkuKTzPwdQ7ZX0uc2FFBk9eNLiyACv+05iay9qkYvDpu+fJmrtu3ipZlJvQvuIl227irazKyQ0AoLSQ0ZMG3Iox/XpT2C7PoH2/BcyyK60UbJkec8qWNSFVBNBHEkMSOk1UoQT2GoEbSLNK+yEFLxE1PjY30WgH/OJme2kxURxI5O4n0qJu9B6sYsVshHtHz0pKXuXPlJ7S/V4rngEjWbpH3D5+DoED5uA7shQH2txbRfJEJMSqSvUcFETLO7cI4pOGrTuKdKCCn61DzYjRe8Tz9Y4yIrWb++YNL+OVvX2H34Sqy92jYavpI7nCj7nDhbFAo+WE9+qBAjiVAMifOhSXHqW/OwT6nF71PQhJVoqVx0MCgU/iwcwLuYyE6WzIINKXjftNKqN6JvXa4f5n9lIGrbt1MrEBGUKDp0hf+3Rw4u1qi+H0vmghZBxIYbu1h8uOHGNyTTXPUg2BLUpg9SFttNpoEqgE6Xy9FFxGwdGn4K6B5xYuI04cQEgKDYzTaj+bSuquQulA27g0m1N1Oytetxn9hGEGAwI+CdIYduJ+0sKehlB17RqPFJPylBjIPqLhf3svBnSPQNLjgyC2kHTTR6Xfw062XoTtrJlEQxyrIZP5MYnCiMnzKcHwPsQyB4rJekmZIb1QQm8zoug3seOUlyhc2EXstB82kYhyQ+KptBAd6C9GiOobOSeDJ95F5JIpvlEqhfRBNgsrXb8dklYlnJdn3++co/Sj2Pa3UlJR/XKmAKOV7V7BVIWnWWDZlGVveeZVnKyr56jezyDyUIG1tGu5TCZJmMPcnCcyOsqV3JIT0mPs0Rqb1ICThzA0GJlx8kkFvGgfeGkf3TIFItkY0RyVp1dj+7DQuveJrzP1JksUxrhp5CEoiJC0QzRzuX2UYErDmBnF8ZRmu87MyjLFP4ouvJxE4J0bDKhfIIokOK/eN34zZq7Fg1R567onTdoWCcNRO42dleN6wEC1KoLllbpi8m9zifsTlA+hu7EXVw+8KPiPm0rBd0Y2twUDeqF5MrQb0/TqaLjJgaxXZuHEyzm+NvNQzh2fr5xCoidI3TaV/nEDhxyKD4xX63ysgccAJrRYKNiuoFpXOCxOk1XiJvZFNX43CrPF1qCaN7hkCtkNmFlyzimhNCDR46NL13P3mLZi7JAavCuMLWAiVKMTToXNZkj3HKhBU2H2wCmHQwLufzsbaYGDgSCZJi0r0oBshokNQh3uMhdfloFj0GJwx7E0i5ju7EJLgPilj6hMIjo7z6scLsNUbOP/aXf9hHhgtCUIjXXinwEC1nnavixOXlWCb3E/LQiPGRhNdh3MQM+KoeqhY1Dhcs2nSEIFysLUJVL1yO4FuOyumH0JQoWxiO3FPkt2HqvDWaMPFIz+XmVnYhNRkInQwg1hSh7/EhL7XgC43gn5AR3hRCNkm0vjWeDIPqjQveoXImXREWSMcMGFr0eE4qzG6qIvbf303gd/HwaCi5cWIvJ/NuSsOEXonFykOfVdGKJneRjIvTsXa2zl5rIieWSq6QR0mLyROOEj7g53cbSLWMwakt9x0zTAz/Zx6emsCOF/fQ8XL3VxdeZDsogE2RExsfv/1//6FmvJXl6pD9P1KBUQp36vye/fStTKOLiLQcFch5e+sBqBvokjv1OEqyJoooOk1fv6nV5EklbZBJ1ml/fjOjbLllRqKp7eDXqPpyZFIJoVwvsbUKfWoehA8cZLpSZJWgS3dVfRMMbB81An+3DwWw1ErjgU9JLNkZE+Sgg19zCs4Q39NkkRhnDHZ3RRuCpGzS8O93YhzxCCevRJiQuDP3RPon6Ry5N4JiF+nU/iehKDBAze/T/lDpzGlxyjMGeT1b2eirs1kqNZNpzedeEmcudvuZuTsJvQPp6MYIRQ34JnRjVo0XOhRnhZk5YrtDI5X2He8nFiDA487SE5ZH/nbkoRu9VFa2cPglARydYRkbpyWCwV0PomKfC++A5lEPesav8kAACAASURBVCLl6xLs31yNJkDlywOEczUarxXJes+MtVNgp78KOV3F4IfxOZ1sm/EsgiKw5cbH8GQGMPXqqFjcSNOlL+A+JpC0amgiJG0qTZe+gBSH3LI+HA0Czup+BsdotK1OkrbJysIb99A26OSiJXsw9UaI5Gjk5Qwh5yWIZql8+cIMFG34dNzReJzSzavghJ3OucOtTRI2DXXQwPj1jYRjBrquqyaWn0DJj6H1GUmf3cOxU0UMnpMk3Ohgxfx9aCJQGWbW+DpkVUfVGi/dQTuTxjQhJASsbRLhPI22RUba7i5j7qKjlMxpgTc9RLIEhCRcUnkUxaQRGzSx/5+fg24jvnKJ0X+8A0SN5GIfIx/qBRWybm5mIGohminQO5jGqIe95HuG0Ec1uqJphIoE4pPC6A7biSQMiJKGUBRB02lgVEk6kwRnRrjp4q+IufV0LUrCVD+DowW0/8HefQfJXV6J3v/+Quc0PT0551EcaRRQRFkISSRhggCTESDAhgUDXkec8NoYAwYLkQUiI0AYBQQICSWUNcqTc06d8y/cP+Z9b92t961be6+9q3t351M1f0119VNd/Tt96nmec44ArYFUGjZMxXvrLPLe7+NnabWYXkjlqQdvpuzde2lavX70TtF/QqMJ0YUzmhCNumAqf1PHju4aNE0klqVQ/nIvX137J1p+PwtnMyxeeQzhyiF6Z0hIWRH+2rUIoc5G5gYzvZ2pGAwq5mEN6QELlhYj/VNFSl7QMHkFaraPhZwY5tMWPIdlfv2DDQwdykJQwSCohENmnK0ag4czyfjGCKqA3tLBlnMTMfbLOGrMtL5YQfRXQXrmCgzOVhjscTEwW4H8KJEXcrF1SLTcpRMq0jA80sul3zvIU69fx7f7J5D6oZWOPjeZRcP0zldxNkH5swkAJhR30fFOCU2rTTxx8zsYNqViN8Yxn7RiKg9w7/i9vHZ0LnlfC5SU95I9uZd7S/bw2fiNdN6cZEFOIx1HcrGlRjEYVGSDSupxCUNYINMaQJd1rMv6iGSaELSRgHj+oRRWL91HTu4w1ge64JJhvj4zlqItSeTlgxxsKOHDQBWuOoF75qwmuisdbVyIpq2lFH9xFwmnQEqdgGVAp2C7RuXra3G2agz47cRTBQaHHNg6RRzf2AgtD3H67vFMzO7mw+PTqLvTgSEs0NXjhpiI+6yA96IkU//lAYq3ruGm43dQ8IHEi7euR4oKRPNUyua1YmuTOBfIJh4zEJoZYUnVOeRWM0UTuunucSNYVRwZITw1Ap/VVREYqyDU2UgxRPElLLSuziJx0s3J78pxnxUweXWkuIAYF2hYK9P443HUnShg6PIoYy6rRy8Ls2n7HIqquhFiEsWb76a4ugtHu0b20g5Um0a2I0j3VYUoNqjbW0x8UyaRiVEsx630Pm+m1+fAENE41ZmLYbIXQdQxBMD/RTZZnxjJfteEHJCwtBiRHEluGn+EN95fRvdSlZRjRizGJKoRzMM6Xd2pZHxlwDcWznuzKPnoXna/8gqmrUdovHE9/WqY5JKpwEibilGjRv19RhOiURdEwZcq9T8ZwxcRE6U31lCx9jCCqrFo+8OUvu/Dfm0P37SV4/m5gWR+nEvLztPi9WCs8mHaeoTx5Z2oDXYCRSJaYxvRbAUxCR3LbORvGcbRriM3WjDNGSS2MsDDn92Cu07D1q3z1RuzRsrUzQKzLjnD5B/UYOmWqX9lLNZzZqSoQLB0pKFgZ58bS49IZq4X0aIgGDWUuMTADRFuv+ULdL+Ryy8+Sn/QztcbZ5K2tAtDQZjeWQI5nxqxGROgC8xbc4SS9U2k7jPie6aAQNlIT6M3rltJzCMw9EYh4WIFDrs44C3FlhLFXyzR1udBeyUDg6CSJtmwWuMcGigipQ4MX7tIxGWUuExqXYxorkLduvHsvPkpIl9m0r1YhfFB8r8e2Yn5fMPFDBzPpH3YjdWUYFJZB13zjYQOp1FR0MtHHdUEL47StCafaKZOZVY/yWlBrI0jPXceffh9kjaBrpuTaEboma9h/dZOsjrEjVVHiGbqpK9u5/ycjUSzrdRtqsRVY6RifCfqmBCmdhNyUGJwbhISIoaQTvp+mUSDk6E1YdZsugclN44xPULw6XziHp2zB0vQVYGSF3S+rhmHqAgM/S0P0W8g5aCR5HE3+Xc34NxrwZ3jR4oKnH9oPL6Ehftv+BxLP6SeBldLgkiWADrIkZE+Rc03CmgOFa3Lwrm+LJRBC5YBgbYTuSDr5H8JkfU5+EtFlmeeZUn1Wdr25+MfoxLLVkimaAzNSpKV7se5pJfAOQ+3jz2IoID5pJVgv53Snwb554feIe4eOZaVH+xFyUiSPb8TNS6x6f35TLi0jokVHURydXwBK403vYhvgsbVk44TSxURFEg1R5g9vZaKt9bySONZKt5cS4Zk45u3XkNKCCzLmzqaFP0nMDrt/sIaTYhG/cdLj9N+qUT6hH58qpXeh2bT9NQszv0ijanjmzH9ZRjtrxk8NP4bmq91Yq43s71+PJouEPRa6XlkNq1fFJPMTmAMQt3zkxGSIo9e+ym2Lp3SN1qIpQqoFh2v146434W1W6R3vkrBXQ3E0kCOCqx8+FvKrf0c6S3g9tU7MNVaiKVrmId0frN0E/1XxTC0mQiVKfS1pSJ2m3HUmDB1GBFPO3j1g0tBg7+drSKRlLEu66O1KZNURxg9NUnnChXD4w6KSvs48Px0vtpZzcx7jrPyN7to+P6LlLyv0fWETnBSHPfpACVlvbx69/MMxWykvGNHl8B22ELv1XGe+uv1zHzsXrKflBk8lIX3kig33/cFYqeZFePP0PNgguVTTzGwOMHS1x4jMCGB57DM+KweAgUyWbslMk5E0YuiJDps+HdlcX5vCVpFmHi6Sl1DDtHtmUgNVpJOnbxvFLo3FlOV041i1Um/soOn6peS+9Z5GhZsIGevyqIp50g6QG+x8e6+2SM9kt4uoOrp+/CVykQzdLJWtdH3SSFlmYNoFWEyJvWRctyIGBWJXBbAMqgiqpBMSqgWjZTvTDw75UN6Z0gUXtRJ2W/PIBtVBidZyS0cwhCEUKGGq8RL3g0tZF3cRcvGcrzVCr7WFKJZGq33QevXRTyzfSW+6gTesdB6uYGkXefM3S+QubSTpHPkq+g+JqOZddQ6B0WbFQJVccQkZO4T6Jsm4b6/nWhhgvf+vIyDmyZhCAjoVhXJmaD56pcQYhLxjzMpdQ2imXROB3Np/55KxrE4qAKJPDePf3sd8UyFnGk9tPV5IC4ysC0Pz34jZ3+wjrqPKjl/qJikU6P0GZVxL96HbtA49PvpPHT/JuSwQDBpIpQ0UX/Li1xiTWL0CdTE4yxYswZDQKDxzSpgtCx/1Ki/x2hCNOo/VFqNQNPiN7jq4sM4lzfx3BPXk5wbGPlnQuTcFxWcOVJM/1SJ1397BWp+jLR5PTw0eSfsdlO5LkHmoQgmr44QlvFNTuCukZCiAq/96iqGZic4/nQ10RlhlBSF0twBolMjlF3VwNiKLk515pLSoCEmBD5+awG775lJQpHZ3DmJWLYKmoD1qj5eeOJaGhZsoPS5eioruhj3206mz6nFvaKbjGMqsXSVeKqGkBSwnDczr7CRIZ8da7tMaHsWl44/i73OyMBUJ4Yn3Fj7FXL3Kmw7O4HXP1tCycf3sHPja0QaXYiDBurutdJxOJd7nv8BiXXZ6GsGSNpHZqk59lmJeaDvYpX73/kYXdJxfmvh+X1LMPoFLFIC7YSLo/35uA6bsHXruNODmP0atQOZ7P7np1HMAs1XmVDCBgom9pB06BR/GkCstyGkJMjfLhDK10k7raJZVfz3BommC5zbUomoCDS2ZpLYk0b/98Yw8c/30b4cvjk5FnVakNtWfIOpX0LQIH1PL+E8jcAYhbK3h2jdV4C/QsX3UgHyGRvdbR4iF4cQFIF0R5j+qTKqEazf2Ll5/j6CRfCbxpWk12j4Y2Zqnx5D0mvCX64TTcqEqmKoThVdFyiweundm4tvnI5kU9BtKrOn15K/QSZakMRQEEaIStg6BbLG9qOkJ/nt4ASa2zJIP5HAlhrFcyZK7k6Yt+QU7bdrvHDxO0jlIXyVIrde8Q3Bp/O5eHw9vjGw8PojJNw6ztNGbIetjDvwfYyDEp///CnOvDke3Z3g1MfjePiir2m5VsTcJxN6LIA8LCN7ZXyf5+Dcax6p0ItBSlOc4s/XEKhUcJ+DjStfJJxn5e7V2xCSIr2zBP7w3jXk7g7T3p9Kxzsjw2N/0lfFC3evp1BWMQ3EOPvAOjK3mliWM5nGP8+8YM/2qH+M0T5EF85oQjTqP0zWuH4m3HeaqxuX8snJKQjTJuB87yDxqIGibXEqf3iKWI6Ko0Uknp9g/W+fI2WfGV/UzJ+PLyHuhofe+xDll15iqQIbVryEtclIoFzHEBIYHitScfsx+qcKyAYVyS/TNuBGaLNw8ngp9ccL0HWBvrkaKfUa+lwfTfeJcMhF4sNMEMAYEBg6mMXaX26ieOsa+ldVIIsaTc94ONJWSKlzkM5LBH689HMKdqhIcYGEW6ctlIrQZMXWpePsUDmwcQqaDP5yCBSb8ZcYaLtaR/AZyNmn4GiUqP7dfVj6RdQUBYwaud8msQxoDKyOYP2diyXLj6OaYPItpwG4acZBHtx9E5Z+gVA+IyUp0/yc8uaiS5Dn8BEs0QiUQDBkoXdVgrDPwozXH+Y3P32d1PJhDIMykY05OFqh/5cKJRe3kbHdBGsHAOhdlcDgjOPrcqLYdCz9OoVbAtww5TCmYR05qnPx9cdxNsik5frJWW9kw7ZFvHrbC+TuSlL/axeuegFHg0y4JAVdBEERmPPYIeQoZOyXUVURPTtGe0Mme+96ig+ueQ7/nBjvnJ2OowUGTmSy7y8vETqQTs5XIn9a8j4ln8RYU7qf5qWv4z4m4293sfPzqWQdTiJnR9CGjQiyxv7aMgyP9zJxTAfn52yk+XsvYb68j+U55xj3kw7OBrMRDBr5T9Tj+MhB8CchemeJ7N8yicqfDfOjt+/gpsojpM7s5a3ai/j2pZc5tHM8qlnju95iNAmStpHPviq7G3F8gGzZTmBeFEutmWCpwvq3V+I8byDu1gjtyaB0ejtqTpxopo56qY/s/SqL7jxIMM/EC4s3kl7oJbAixIO/v5+oR+S1DSsQkgLz55xBk3Wuf30HaZ+bOfbEixRvXcPWDXO58293M+3jh9mxeSO/HRxDz7IkE46JNK1ez1+veINfrth0oR7zUX8PffRS9YUk6Lp+odfwH8aUn6/nPPLQhV7Gf0llDx+k6LCFzusz2Lr/M4q3rmHcr7ppvLeA0j+do+75Eq4eX8OpeydQf4dlZMK5O4H9pBmjX0exCqhmiKXp5O1SiKTL2HoSaEaRnlkyCY+K57iEvxwUt4LslXE1gHe8TvnbQa5+exe/P7CC/M9F9qx7meLNd7Ny+kkaAuk0nM9FtymMfSpI/C8xTPdKnH80DdkvQ16UtC1mVv54N+83TsXwtYukA/J2Bqi/zYa5R0Kc6sdljeLflUXO3jDyQBDza0FqvyxHnxQkFjIim1TEJgulbw/QcWUG0UlRxuX1cOZUIbpVZdaYJprWj2GwWqeoqpsen5NovxXPMYlIloCoQt7XAYaeSOA/40E16Zj7RQovaaVldxGGai8plhidvW50VaTsTZVompHuBXDRlAbOfjoGZ7uGvT1C6+U2jOP8FDwUZHB+HtE0gXB1lMwtJobHiaSdVgnf7MfX78AwICMmBazVQyiaiLDbTXJuAM+7NrrnCWTv1fnZHzaM7OpYwgw9W0TnMp30fC8DnSnYmg1ce+NuNr+0gGgmLFlxjC93TsHRAs5rutGfySB8nx/fWQ9XLDnEjg9nkrTrpE7rx2pIEns5h9jNw4SPp5FSr+G/KozFlIBtqcTdApoJUmb04T+QyZO3vMWff3QjnUsFrJ0SsaoIasCIkBCQIyKmQYG4R+dnqz7ihaaFxL9OJ5yjo3qSZOwy0D9bxX1Kwt6jMjhexj57gOi36Rh9Os5ru2ltzqCyvJuGk/nkfaOx9Hd72HBmJtqAmbTjAsOXxJCbzExZXEvtxjH4xuqUTuiiyD5M3ZPj6b8pSjxgwtZgJJ6qYywLEO22IyYELD0i+mw/EzJ7OPnlGBSbjhwRMARAlyHjkk787+cyNE1FSAg4WiTC06KsHn+UHc/NxdavkPOTRn6T9zmvDc9mvLWLZxsWM1yfeqEf/f8yup9+lnhHx9+VgbgqM/WZL9/wj1oSAF8ueO6Yruv//8MKR/0roztEo/7dVf76PDu6a2i7pwT99STPewsxp8TQ/AEyp/fS+1YWGdtMfP3aLBp/KCOFREyDIsY2E5FcjaEZCveu/YyYR8feIdA704AhojFQbSKaJo8M7cwME1/pR3GpyF6Zqy/5juDSMNXTGgkX2vnDsWXkbpMYrpQp+fJOUs7IbD1exUDYRlrxMBZHnM4nJbq9Lhp+7SLrWwk1I0HJjTUMXR7lZ2m1cNiFf2aMBdccI/K7MFJYJPNYkmi7A9+3WYTLkvTMtTEwN5PzO8t5/Y7n0XU4sPg50j8zk3Rr9C5M5/RD69B8RrrfLsaaGyI9y0/jK2MIFgo0rV5Pt9dFhjOEHJLQrxoiWpAkmqVRd5cV8QMPQnEYdLD26lyReZKizV6K3F6E59PRFRHZrNBxv0L/dBFjZoRDJ8tQrCAqOvW3mXGf09E0gfZnHQwsiZN0QuGbIj2XKLx/yzNUPX6StKfMfLBkHaZhAcWm4WtPQTnmJu+DZqIBMz2zBaSIQDhT4rEX76SnIZ2BqI3uiwXyiwcY6HBj7JeJ5KtUW1tJOkGKwdZjkzD3C3DZEN7PcxFUncBJD7ZOgU8OT8O1sBdbNwyeyiAQM+OtFPG2uhEV6J+hc37ORrwDDgCmXnEG42QvvkOZmKYNc5UtxPAYGTkgknYqSaorzKLJ59AdChmT+/CcS5LMSLLut9dgf85FzKOjOhVks0I4V0B2JVBNAl3zR8JihXuAeIqOd7xO8MMcxLBEQ3cGpvwQSavIO3XT0fvNpJ0QiKYLlD6vogtwsKaClOYkhpBA1zf57GsvoWPZyLGg2RUn7tEx+gXm5jfjaJZAh0iORiRg5uh3FcQ9KqXT2jFO9qKaQTVBTJEJXRJi7F98GAIi8RRQ/QaOL80inCNgf7STxpfH8GjbKrZumMtblfmkXlY/etH6/zKjfYgurNGEaNS/K/OASHJCMSuWXId3gpO78/bwSsNsLDvt1D41lo7mdIQtqQxeHsM8rGFosiBHBKSpPtAFjF4RRJ0t/VVICUCDx1dvYrBK4ok1b9M/SyWenyD/9wLi7pSRKqKowIfHpyHX2Bn8fTGCpuP8zoLh3l6EmT4IyRSvbsDglZme1c5ArwtOOok0uUj5xIb5hJXeJQppu4x0PT4b234b1zUvxnrxAGmeILs+mcrQ7myslT4W/3EveOIYg5C1SyLrQIQpa2tQx4R5qX8B4kkHi159jKGJAo8s2oZ/jM6YV9di6pPwjtOJ1zvxnUxj/L1nuH31Dkp33g7A7gmbMXoFhvudLJ9yGmejiGBWsdzcgyyrZI/rJ3p5gPXrr6R2rQOrnOCfnnmXx2dtJ+c9I9IJB3JYoCxjEM8JiT/d8jp900UqyrvpW6gQixpxfOBEj0v86fuv07c2hqXFyJNdKzj4WjVdC6zcf/YmYlMinL/hryycdhZlXJj6B4txnjRi6RMxBAWCJRqGoI4hI4pZVpBzI/Qez8LaJmPyCkghkR8du4bs/VHkGV6yd4mc+tE65A89+CcnGB5rZOklx6n553VYumSCX2Zx/OcvUjmjlcEuF+ZBHXRIOnQQ4VcD4yApEiqA71pKYJebRIpGMGxm3ulViLO8JNMVAoUy+sdplFoHQBPobkqna4GMIGsMTIPpfzhK3R0vIpg0hDYL8RQdz1Yz+Ve1YG8XSbp0DhwZw6rl31E4roehixTs7SLZnxrJeMPC8PfCFHqGSTkvMLhopFFk50I7qdP68RwTSftFC4VbIyDA4qJ6sveI+MfoSEcdKHYVQYUvj04kd1s/l8yvYepFDYgDRuSIQMUjJ+AxN4kTboRpflSzTjRhID5kIfl8lGRBnLL5LVx+0QmafliGyQdbKraTcAmE5w0gLxmk9YMq5p+Ksixn8oUNAKP+l40mRBfOaEI06t9N2cMHyf/zMcR9NbRfnobn2BA/ff0W4mdTqLi5jpTTMhVrD5NWE0IJGbj0x3uI5ycQNAHH+06MAUi4NX40ewddHxWjGiHjaJCnX72GpEtj23AVtjaZ7B0Ghn8VJ/eqVhzZQTxnRi5HxydEaV8p0jtTYs5tx2htysQoK3iOi0QUI4pdYyBmZ+xjzRRs82MsDNG/Io44x4tnv4HhS2KknUrim5Lg+MFyBtrciO95MA/rFGwdht1uvvrpPCxnLDhbFVz1IRSbzOEN1egdVvY0lKFNChLLSzJ3yWne+/kKLL0iyeIYjjYdQRFQHRomn8Duk2Owigm0hITaaKfirbXoEogmlf1vTyE4O4K50YR3ew5ZriB3F+0h3GfD3qVi7pY53FjEsz+8gU/uXEL7Ko2zP1jHzjV/pO+tIvyl8Mvf307hthjNx/JJPWwgc4sJb6WIwRHn5+euJNptR47C+c8qSb+2g1hZnEDYjBKXWHHjXbxWsA/7t1aMfoFIjo7Rr2MIj9wPqrrjDAmfichrOZgO25HiAjfdsBNjQKf81V6cX9noWGIhnpDpuyxOTTxO8tphBFHHtHSAowP5lO2+DdWsI6gQ15PUHi3EmRkiliaQt1NnwfxTzJ92js2vLMDglUiZNAjtFpIOsBYGENssdHSnIn3lxtgr45ugUXh7A6/sm8/UilbEqIDnpI44YMSYG+YiWzMT/3wfjpQItnFeUiYMEb4mwNn6PIKlKlp2jMJxPXxYM5XWpkxsnggnH1vHvudfwnayG7XFTn1nJv55MWxnzEgzvIgqBKNmvON1Qne68VZaieYqfNk4hoFqAXuRH80IUkqCouUtOOtkan/qZLHrHEfOlLJ4fg2GiX767p5Gw/ftxAoTqCddAPib3YgxkabOdJqXvM6Wiu3YpTiOdsi+ppXi7XehmuFvXUfIfETl3NwN7L9yDABaioKz1IcxMBruR436nxl9Qkb9u7B1iuzormHw+1MAiKfqdF6aRuFng+y95U8c/a4CV0uS4dtn0fqIgGFYZsOBuby94GXkST6iaSLBCXHEuMAHP1lOcHYUxa1Qf5uFSVefAx2+OTOGhFOn99LkyJveYyHvRzG6lmoYBmTenvUqjgaJ1KoBttZUgVFD/TKNwYsTdH1RiOe4yNm9ZTQ9XMlt725DO+dAbjcT6HbgXRDjg9kv4a0wIHllBE3AWSfjvzKMd5yOf1wKweoYS3+3B3unxtAdYfwVdhJOiSm3nEK1akhdZuJDFloue4WuiAtLf4Il1x5G6jIjx3WMPoHdlz9N3K1j6pN54f3LGVvSTdKjYPCPdIbWEhKRHB25wUpibBSjX6etz8OLv76GlitfRjUKJF0aDleUvhkGEr/yI0QkJjx3H1efvZXBWQoF07oYnh/H8Os+bJ0CMY+AvSOGq1FDOm9H/iQVz3GRxIwg9k6NwKt5uNxhVEWCuETrcjMlH92L8YoB8pe2UT2nHn+lTvCiKA23vEhclTH2y4SzRQxBHSEJ+68cg29RlJ5l2cTSBVyNOitLz6IlJNb+5EHCURNSv5FwzEh4VwZqyIBi1TEP64zZdh+CIhAYsiGoEHNLtIZSOfzZRKbdfBJDUMB7Ng0tL0ZKo0bqa3aSKSoWR5yECxLpKq9f+gqRu1IYO7aTmkNlOMp9RNNFVLuG0mLnF298n1CpQmVaP/52F54njCTOubhh+iEKKvrI+puRwR25VBb2IsZE3G/bmfKbtVT/9j4KPx1C0OHGqiOIksqdt24jHDQjzvbisMSwlAQIjvMweHGSkk0qkqxh9AlcUlCLefoQKTstqA+mICV0rOfM/PzUFUghkdpfT8Tzsg1br0rKeYH0TD+2Tp3UszqmIZGXr3gF/AZuaZsHwGRbG8OTNbwvFyAPGZDDMGb7WjSnhRWrbkX3jVRv2s8ZybiyloInDoweof0fbrQP0YU1mhCN+ocre/ggeTuGmPH4WrQrh2n+l1kYAgKptUm0hlYuPXEH1m6R4bEG5JhO5rtm1NwY6YckHnzyfrQjKfgvipGSGua6FfvwF8poPiPOzBDm9Cin+rMRkwIIoDg07piyH+ltD023ZtB5RTYIkH9RFzlylHCujjdoxdxpgJiIs0PBfs5EaGycGx7egWrWSXhU/nnPNRiCAmJCQFAErCct3PHcQ4SKNNQUhdTTEM7TiUcNyBlRehZomJrM7OyrpG+hQkVaP847OxmYIvLNqbE4GiTU3BjLp53it4NjaD5YQNtKM4NxO84GGJwskHTqLHn3URSbRtGnAcYvrWe8qwcUAcWqUzazDXnQgKE0iHkQil4VmHXfUewHLfSviAMQvDaIp0bAZFBwtOooL2eSUuAjUqAS3ZbJ2LIuhsNWsjN8NH5XiKtZwRiAhpuNDCyPM3P5aQanaARKIR42EiwUSdoFXpj4LqKo4TkmIaqgO5MM+exkWwOIgo6rTkD3GynfuJZ3i3dhrfISLtC446EtxMdGGZqTjRqV8ZyNkbTpKNcNsXXrDARZo3epgpKUSJvYT6zdgSaD7JPRJUja4eopx8ib2k36XgORAgXvWOjalY959iCn/lpFpDxO9ex6xvzCy+CVUWKpEoIiEO23EimPs6T6LLfvuoOW6zOwG+KICQFVF8g4FqV6XAuGkuDIF9WkcXJXBfZWicbrHahmnQ92zyb5SiZ9MwWimTqRpJGp0xp47Zk/41zVQ/mNdez6fAoZk/t47/Q0ksNm/nJkMddMOEHiuBvT86lYP3XhuyUIgs7gRBOqIiJMK8IMPQAAIABJREFU97OtZRy+1hTiHoHa++14J6lccf0+lEYHqlMl/fFm2m9Qef+ZpwnMj+L125i99ih9MyCal+Sur+7EVStx7G8TALjO7ueyGcdJ2gV0SUczgt0Toe4uKzs2b2TiN152dNeQ86cD7OiuIXrVRQDcuWTXBYkLo0b9n240IRr1D2UaFvHsdxN4Oonni0aOT/uAp7/3Jrn7osQe8FL/bDXhk6noMhgDOr5ykbE/Pc3U4nYq7j6PHNU5+8A67K4oseOpfPH8XPSFXiydEqbNKaS/bcH+nou7V3yJyRHH0i3x8fpFDEwZmf8kJUC0KDQ3ZrHkwH2UTukg/UMLyYoockDCWyYTydWQB4y82TiDt1f9lZydAmJYwnMmSfrsHlJPiSg2yLq8HYCbph4ivCqAkBvFdspMMmoAHW679ivKnINY2oz8ruAzdozdwqc3/Blbs4HozDCCpLP3oym8s2kRP7n6Y3QRDrcXYPZpLFxUgxSHZKqCkJpg/Kvn6X+6hE92zyCzaJj8b+LEf5MNOuRfc4bqm07T80Cc3Z1lGJYNsqSilpk117Cs8Dz+y0NM9PQwvDSGeucgs7NbsWSF8FclaThciOmjFPQNGaj5MSLpEr4JCi1XvEze+waOfzCR2dPqUOw6+ZslUs8rWIY01rz+AFqnFdP3+sjZqyD3G1lQ2kDnj0ppeamS+KUBMg4IZH+nUrzjTtZW7MFVK/D8R5dz+djTaDcMsbzqDM2rjCTSVPxBK4nCOFVFXZAQKcoaoq/fhW7Q8SzooWBKF3m7Rjpqf1ZXRUdNDjGPQFl5D/a2kSabwmceQrkCxCSGf15I9/Ic9DYrvpVhdLtC9rciskllV2MF2V/JKFYdf9zCtPm1qIfdoOucOFtMdMiCZVDHft5I9gGF9Jo45MRwNghoNpVIuoS9TaRiWhuRTVm0vlrBsGom8Ek2A1E7qbUaXZ2pTCzoxpweZVxRNx+emEasMIH1UBP//PON2DY5mVXRjGIBJS4T7reR8oEDRLDMHyB3h8jYyk4+3TyXZOpIVeTJA+W4DplZ+tpjaL0jx5Xbv56GqEDhZjB4JaSEjqBByUf3svjcFRz7wxR8Y3RsHSLhXA3PqzbmVNWz7Pxl1FRD+ca17OiuAeD6J7cD8JO0ugsTHEb9m+i68A/9G/VvN1p2P+of5v/tkjv/VJQvfj6f3pkSilPD3C0RLU5grzeiWMDZrBMsEpi18hR9MQcdm4vJ3dZH/7wMJt11mhNvTsR1dTdt9VlIYRFXAxiDOkOrIphNSaQdKeRc30rjniIuXnaKPLOXDYfmIIYltJQkNleMFGuUvpOZiEmQwwLWPh1/OZTNbKPuZAFGr4h9+iCFrmH6IyNVS52NGejSyN2e5qtfomLPLWS9Z8ZbLmPy6liGNTpXaCMVcnV2Us/qDI8TME30EW52YekTMQ/pDM+Lk7rXRPiSEEqTHcsYH5aPUxiarFMxpZ3mvYUAmLwC6SditN6toQ2bkCIimVV9BL7KIpqpM3feGQ7/bSKOuf2EYiZUVSQ2bGb5lNPs/XAKkVyN3LF9dA2koA8bKdihMfd3B3n32znIGVHunbiX9/+0jJI1dRzbV0nB1C7a+1ORZBX5uANDEKKZOkWfBuhe6OKeuz5n3cbLieaq2NokktODmA44cLUqPP6nt/j5n25nwq1n2VtbDlGJtCMS7vookSwT3htDRHvslH6UYHCCBc/ZGE03izjOGom7dZJOnceWfs6G1llEEwai51NQrCM/7taiAOGAGWOrGSkOq67by0d11SRjMqn7TAxNUUnL9+ENWKHHzG8v+4Cn/2U1kSyB6JgY9pNmbEv7UDUR88tuOheJ6DaFrJ0y5jt68JjD1A+l8/sJn/LohjtAh4Iv/IiRBMExqXQtENBNGlJQwjIgkLTr5H8ZZfZfj3BkcRa+xeVohpFkG6B/qsDUuXWMtfey4dhsPAcM+CqhYEeCzruS2HfbUM0C19+xk7c+XUzSpVH+ZpCGR4xkeAIMeB1oioieEDHYE5T9OkbTTR5mLz7Dkc0Teey2D/n9+9dROK+NoYiNwEkPCY86clF91hBOc5zuIReZH5kxDSdpudKIvVVEVOH3D73Gj165kwlX1JJmDNMZSaFzYwlHf/0iy0tmUv/kZMoePkjjn2ciJQRU43+d+P/v7R9Rdu+ozNKr1938j1oSAHuX/Gm07P7faHSHaNQ/hBwRCGwvBeBMMAdrZ5ikW2VKVRPGAJy+9AUSTh3VrKMZoGxJM3t2T6T562KCk+PU/tRFKF/gxIaJTLr1DL5PczH1SSxbeJzhuXFiN3qx77QRCppJLA3wcfnfsFUPcVv6Xk748kk/IGPKD3H5xFMkap1096XgagB3rU7CreMvA/OAwMDGQjK/A89ZlcF+J+2vlRN7N4uOlnSevWQjGfsl3IVeFqxZgxIz0HeRRLgqRqAMuhYIWFMjJGIjuw99MyF7Rg/CLjdlkzpJTA7jHa+TtdXI8AQd4aQDzagTrU0hnCNgLAzBPRZsnSOfWWxmiNY1OsWZQ8ieGO5z0H88E6NfxzwosH/XBKy9Or6j6cSaHTxT/SEGr8wjGV+jmiDlnMCeiZ+SutOM7lCw7Knlk48uJv2YgOG0ja29Ezj85Iuc7s1GSVXoGHBj32clHjKx7NqDCJqOrXqI+odMODo1tvZNRLHrZJYOkpweZFZBKxffdIzOpQIP7LyZhWsOsb+xlPTdRoxDEoFS6J1hJZIhkpfiR3Qn8FaYsfWp9MwyYz9vZPp1p1AtOrpJ44XaBThMcZbk15FMTyKmx7D0isQaXEg9Jq6+Yh+R0gSfvXMxJlMSsy3B0V+9SN5XAsONqSwuqyNvp0pHMpXhRTGieSr4jCSd8EDJbob9NobGyYjpMey1Rkp/UEtHTQ4LPXVkX3WeFzsXEi1Mkv+7A9TdZeP8g26sD3SRXdnPmEfPUzqlA/vCPrJm9tA3w8r5YBataytJOASGJgjsf3Y9zvoAamaCE515vHV6BkZbAt+iKKpdpWuhkSx3kFAhGII6r+yfjxQDW5Gf+tvtaBEZk6SS7g6S/pWJtP0GMjZZaP21EXRoCXhwtGs8eWo5ziadurpcfGc9iBUhBFXA2aahf+Gh51A2SlJi7wsv0bbChJaSJDAxQXhuiG2+KkpWNBNImPnyqykkVwt89cunWXj2SqKLJ5I+ZpAd3TV8vOo5in/83eioj1Gj/gejCdGov1vZwwepu+NFYgkD69v2MfRPeST/EGTq+GaahtO4896tTNzyQ5SiGLYuAa4e4nRjHppJR5fAUmfCes6MVhkikSKw+8RYfBMUShe18MW31WTuMOIPWEEAQdLJdAYxCQZem7CROWaRzeU7GJit8NG0V9hybiJSXCB/k0woXyByjR+5OIRq1dFFCC0P0bsyQddiHdcJI5FMgViawKGVz/BPW29hYHEC77CdpFXkioknMY7z84uLtiCVhRBS40gHXAh9Jiz9IuZBEW1dBoHJcerrcshP96LZVGJuAVeRj7QzCqaCEBRESVvcTTRgpn1VJnG3QNm8VkymJDdPOMSzpR8yOb+T5Cov7vMgxSFUkeCzG54mkilQuC2MeUBk7Z6b+ebGp7j0vUdJnduLt2rkiGlwtoKxy0jts5UggHccxMdH6fsqj2U5kzHucmFvMKAMmvFVKRTlD7DtbzMJzQ9jfcPNqgk19F8Vo/eDQnQJBk9moOsCu0+NYfu+atIOi1Tcc4TN5yZhP25hcLqGXhbBEBIIFanElwTYMXYLUwrb8S2MEneIFDxXQ6gySVAxgQDp30lYTQn6t+TzxaaZiEEZdIFIrkrBlC7MQwLvnbgIV42RUKnCqYveI9nsYPrx6whniuiiTjBpZmi8AbOgYD1pQTdovH7Zy1QvPc8rbRdTmDFMxqIuRFFDsUGRdYhxF7XwfWc9w3fMomVHMcga3Y/OJu9LAUuXTM+2Aga8DtreKKS+NQuzrBCImQiVqBxtLEKx6ARK4YeXb2PZ1bcwPNGFrglkvmfG+Z2FZExGarJQVtGDFBOwPyTjPg/DkzQcDTKRoiTZjiC6XcXSYUDRRPoGXCx4+DsGL1LpWqYRHbKgWHU6+t0oNw6jNdtx3taJs05GSUuS8aYF3aoQLBDRZbB3gtRtYvLv70PLjmHsNuJIC5P9tpm6aUnOdWUxP60BW6eA+o7EzU3fI9saoHOhxLDfxqQ/3MfjxTNoeXLWf3+Gy98JXqjwMep/oI92qr6gRo/MRv1dmlavZ0XVYhITC2m+xkDlo6fZ3niA6iOrCUeNKH1WKh45ijZ9PK0/BNteG6bL+hk6lU7G5D6EV9OR7+6jrSmDws91OhdJZB3U8RdJJFw6OfuStH9fpWiDiOXn3XR8WkwiBf7w/Q1cYYv8q7VEtARLz6wm+lkmmkFg1Z27eWvXPJbMPknrRVEEWab+2ancMe9b3q2fRtpGKzG3hGF1H0OHskbGblzbTdeQC7M5ybzcZs7+bCL+YgPBItDMOmhQMqkL73t5ZGw6R9OPxuFshsxbWun6qJhwns6Vyw6yZfMsEqkamlFD0ATsLRKKBeQo2Ho0hqoENGlkVIguQDxdBWcSdAFx0IAUE9BlKNs4TNMNqSO9Z9arNN5gxtYmoc3yE+10YOkVkaOgzvOTqHeSTFFH7h2VDNB9JhNzv8iZB9ex9LrbGKi2IqgjDRq9EzUMARFDQCCarUFaHKHXRMVrQwxNT0OTYGh+AkdKBOWwG5NXxzdu5DXJ3ARj/hhk8ju17OkrY/BQFrVr1jHx0I2YtrmIZIx01rb06SQdAjmrWjnfkgOqgBATydovYAhr9E2TSZZESfvSzMAMDVubRKhUAUknY49MwikQmhOhaB0MPhojcTAVQxhuunsHj6Y2UbFhLYpNxzwoYusaiWNyVKd3gQoGHVudEYBoloYu6UyY1EbDrhJiGQrWTpnUWpXOSzUsbQY0k46gCsQyFdKOjvSIUh0qolUhP9NL12AKis9ITskgsU8y8S+IkpkawLsvi6RDR1Qg6dQwZUW4rPQMn+2YSeoZnf7ZOuX3H6L+lelYWwwknCPz6co2DtFybRqxggTSsAExL0KqK0zgQAaKVSdrai9WQ4LG3nTUgJHCz3Ui6TKG1X10t3tIPSpj61PpWA4IOiknDZz42Tom/8t95H7UTMeNJTx41ye8+pur6F8RJ+8DA4FCmZTGJL9a9wrzzPBFxMQzZWMRZJkv2o9S+v69/8HR4z+Xf8SRmb0iS5/011v/UUsC4MAlfxw9Mvs3Gt0hGvW/rWn1eib85T5Cc0sJ/CiIbtRo+sUk5v7wHjKurKX8oT7K3o8SunIqrZdbSUYMOLoVtA/SUdwKvm+ziKSL+KNm3CclOpZKWHtEih6uI1ysknFcY6DaiD5sIphn5GxtPpEZER6/YdP/JxnaHLYz4dMfMFCTibVfI54Cn7y+gOx9Ovs/rsa1z0PLE9Oxdkhse3IB+iknwTwZY1DjjsIDZB5RMPt02ntTKft1DOlLN4MJGx1LZey9Ks7xQ2SN6afyD8107skfOS56K4uCWZ1MubeGs3V5BGZFMfoEtrWMw9KvU/nSEO5TErZ2iWBFknP3rcO6uJ/hcQL507oQVbB36Lhm9qPbFXRVxP2dEQRIZCW597Id1D5qR1AF8j+U6Xk0iblPQkpArN1B+YRO3A0qjkt7MXzjQjXpOOplbpr1HX1Hs5g/5wyRXJWdUYn+qVZUC/imJYheEkSMiZj7R0ZZCBkxrh5XQ943KoHxqQxV6fiXhZGMKjeWHsXoh4RLQLcrGH0Ccq+R8z9MYdOOOQRiJtTyCBXf3op2JAXfwijRfAXXgl7CeQL+6ji+9QWk7TXgqDXw+opX6JsB3XMkLAMgt5oZqgJTv0TCqXNk5TMjCd3dDQTKNaxHrPhLLfh9VqLZKllXtbHuu0XMrLkGOSogRQXs7Tqh5SEEDZx3dZJy2oAQkmCGH/v8fnR3Ejks0ry9BMf0Aa6ccZzqy85h7o8j+WVUq45qhFhOEnOfTPDSEJfMr8GRGcKz00zHmSzUYROCKiCtTyP1+k7Edgu95zJQJ4ZwNoMUESjcouJ538pgwo59/DDRdJHiyh7828qw1xuIlCaYfnEt6Sd02q5Mw12nYfh/Jtdrmkg0YUA1j5TZR5MGmg8WcNuEgyyfegpvuYGkVSD0RRaLqs4zfFGSzu8pSI4k1lYD2de2UvbevUSydMJvmonkaPzLZ6voXZqkcINI+0pILPSz6/WRZGjFkut4pmws3ltnoSsKy3JG7haVvxu6QNFk1KgLbzQhGvW/pfKVIco3rqXgkz4Sdw9jWe+mcl2EO1d+jWt/G4LJRMtdpSgOA/7vB1Hy4hAXiaVIJFwCsldGjoAcAX1nKr7xGro7iSZDiiEKNgXFLGDp08k8COErAmQckCjNGuBXu6/irUDav1rPVbYQzd97aaR/j0UgZ18cd0OS7vkQzdRYkXYaa69Aap1CwiGgTwjinxEjWCDxx49X0TNHYmiCQG6Gj4nv1BPLgLObx5B6WiDmkvDXpeL6oYB3UQmx3CTf3vQU9tddKJrIvk+rqSzvxn7MwsrVBwDI3tFN/R1pZO7zEs3QsHiilH54L0On0jGMCxB5LQfzgADXDiJsTMPSaMJ5ykjadR3oIrSseJWXNi9D8Bow+cBbIaMeTUGo9hOeGUFPTdDUm464pp+EKpH1xkm0lCShyTEOr51C3e0vUv/H8ZgGJR7YeA+mSwYIlSgjuyY1TlRPknDeyMVtodPC55/PInyfH9UgMG1WPbTaMJqSvLZlCYvuOEgkX6Xi9mN4ziSpmtMAJhXFpiF95SY3zceUgg7kCIjtFozuGN3NaahGnYcv+ppggUigFMqubGDNpnsw5IYRVUhb1YHnjI5lQCBWmEAeF+DG1fdjSonR8EkF9iI/uVe08qOfvovrsBk5I0r9+Tymj23mB6W7yF/Qji6B8r1hEl02Um9pp74hB9+UBDfOO0C434b+fjp6RMLaLRAuT1CSMsRnNZM5uXkcbSsspNXoJJ0a2QdVcguHyJrbRWLAyt7OEsItLnzLIlj6RNDA0i3hvy1IU2c65S+0UzKpi1RnmOFqFVGFXW+8ysBkkQNtxYRPp1J5TR39X+QxLb2DaKaGEJPoizrwVookq8IYgypivY2ME0k0rxF2u9EMkHJnB4P9Tgq+jPPm9oWc92USnh7FP14lPD3Koc1V2D0R9LiEJKvoMtS2ZdN4w3rM43x4t+VQNb0Jyxgfxe8JtK6SQQPlnBOA4r/djS6LdP14NsmrvTzbeuC/V6LZn+nljatGexVdOKN9iC6k0YRo1P+yptXrUR1mjD6B/8bee0dZUaZ729dTVTunzrmbzt1kaDKCAVAUEBNmHRVHBNQxjDPqONlJjjkh5oAB4wyCCEiSnDN0pnNOO+eq+v5o57we35k5c87nN/O973CtVWuvXfWrqqd69X72ve/nDvS7+VnpGjrOkhkY6eQix3Eankuh6cFxhNNU3Hf6iB9OIP0LI+bUEPLV3STUxlBtGuE0nVC6wFsRYUh5J49M+TPBYWGO/2o09mMm+i8NwhV99FQIwkEjfaN1qquykR0x9vkGA7i3hcGvhQGY+NASrJ06eYtrabpVo+OmCEpKmNyNKr85OBf32CgjHj6GPxf+OPYTDC0m5JCOuUdgbRcsmL+D9mMZPJp+hNuvXEvuZ93M/+EW3v/1YzhK3NQuTMNTJCEFZZY0XMGoh4/StTV7MF5mWQ6+MWG2dRZxWdExTj2YRtLJwb+XiAtMm52Y+iTmn78XebuLUIqEZgD/rlS6punccf1qghmDyz6aVaXwy4WkH1ARcYGvSMVwXi+aEU5OeRfjCSvJyX4kWUV+KgVdF1S/UE7epzJyp4mGy6yUfnUTjrtaiKSqJFZqpNn8gx3oi2KkHY6R+5mEHAaHEkbkBXA06fBpMqmLG9l3sAQ5KAj2WXHVwObWUgxuiYbfTcF0fwfun+VhsMTIK+tCzO6jf30WVSvL8Y6OYCl3Y9xvp3yZl7hd57lVc7jk+u3oRUFOtGVhLx/AvN2BrRV6P82lc7o+uIwUlFFPuEh7rJGIz4QUg1BVApWVOTyw5SpiDjDvt2PukDm0p4R32ydRU5uFtUMQ2Z2MnhSl7nAuUkjC2GFgzWvTKX/Rhz9bYO5UCE4NIJlUZiRVYWk0kjizg4RqcDRH0K0qUlTHszmD0FuZnDPuFKHTToQKRb+NUDynnhGjmzh51zLCJxK4ccxemq4fQlyT6D+UhrlTYdKlxyhYtYhoWpx4VObGi7dwZEcpUZdOazABoQkMAxKhV7IIZccxH7DRdo5CzpYw2t29lA9vwTs8hmrR0H6agrXWyOlbIOWIjumiNtSITNougWKIoymQ9JodYVTRNUEoO47thInJP1qM9SMXDy99lzzbAK63nLTcHOf8CcdwZPm46ZLNlL+ylOwvBXU3JJKzyYfHY+W+8hkUrVzM+vYj+Kb38vvR088EW/8bIoSYL4TYL4TYJoTYKYT4u8tsQginEOLNr885JIR4VAihfEuTKYRYJYTY/bXm/r9ynXIhxGYhxHYhxEEhxI3fOn6uEKJRCLH1W1vZd/Pk/4szBtEZ/lv8JfiyZpGJk3ctQ/cHeLa4HD0vxOe/fRxNFxh2OglnxDH1ywAUfNDN7ieWIx1yMLAvnZYLZFzVMjddvJnsC5vQgzKGXybyxNNXMb6oidQHThNO0UlP8JFu96GlRrE5wmhGHWOfjBqWqfWm8queYZhFjAtPXMe8mou47ycr8ZTC/hNFWA5bsFvDpH1ipvkalScnfgCawKGEcY7rJaYrGHyCQDYMufw0mhHePTSJoooWCj++nZP+bCofSGD79ydw3c9+RChiQM2IEsqJk1zSh/vRPL7cUIEcgeplQ4ncPABAV30K7x2dgIhK9EyN0z82EYNf4B4R59SSZSQagky/7iC2uZ3ExvqJJmpIYUF1MAPVocHDSTiqDShGlZhVQkuNkrpX4vbiHRR80g+A0QN9DYnYt9povkAmbXEA5yET3lwFNT2CqW9wCabl83ySD0mc++AuqtoycDaqZH0p0TvCQO9whWhWjB3thVh22+mr0AhmCDp8TowDMuowP45qA+EUgW2FC6GCatUJxQ3U3yBhOmSnY8BJZHcyvtIYnhExCnN7GJraRdwONbcmYC/woEuw5vXpxGMyLkeIwMlExKx+3EM1VBMYUkNEs6Ok7xJEsqPsqi3EmhDCPS6Ca0Qfq+Y8S2lxB7GxfhYuXEt0aAjNAI2b8jEMyJhm9PLabc8xvrAJOSeIs1ZCNcK7P3wC9x+iOJs08j9zE/MakRWVZ96+lJhdx7MuE1tnjLrrDcj9Biybj2M+u5ee2RH2tOSTO6oDssJU35bA0Zo8KtsyGPryUoQq2Pe9UWRe0EL35mxuungz4bQ4Ww4Ox5QcYuizXgpeE2y+fxqWDoFmgLq+FIwDAiUgSNzbDpKOf1SYRy5ZSf0NEp2HM6huTcdeYyDhlETHNBsAC8fuYvcTy+n7cyFyn4H+EYJInwUE9I1QMNWbMe+3YWtQ8BfGUU2C/nkh+uN2Vu2roGOaoOgplfofD+WqwsO88cUMPrnpCQx+FSkKnVMcnJ71Ou23V2DulRj726UYtmbSfOfIwc/t6ug/e3o5A/+aOkRCiHHAe8BNuq6fDfweWC+EyPg7p70JyLquTwAmA9OBX3/jmhKwGjim6/oU4DxgiRBi0Tc0dmAD8K6u69OBS4BnhBCzv30vXdfP/db2nRfUOmMQneEfpv6a5axb/S4F677PfVO+pOCzRTQ8VIG2KRfptIUb5i/iyc7zydjpw1WpEE3QGGhz0TErjeHPL0UzwcZb/oizTkJT4JUd5xB4LgdbRgB3sYXJCw9zaHcp92ZvYMS0OhLMIW7J2oml2kQwaMLYLzP6vBqys/sJPp9NX8zGNWvv5N7CjfS+MoSH91+KeagbhI4ShIHTSbTN0hGSzq8evwnJJ5Ni8JFyP2xwDyd9VityWPBR8Wr8+Spl+R2kW70oQYmTT41ENsdpe1AjcKmXB0etJ+8DCcUZxeO30DPaQMypkTu3EbnLSChqoDyvk1EjG7li5GGSD0sUvafSM1kjXBzB1qRwa/M03qsZz863xhFYm0H19Lepu245w8Y3Mt7ewMQxtQx94RT+IRoF1x5l9+PLQdIxBjScUoieCYlM/MkS/GcFEaoglCYwDUicXjiEwOQgpnndyJ0mYnYd4ykLn9z1GP2jdDY/PpXy7E4GymR6xkmERoaIOXSurdiHP2iC8waNuZe//zwev5mYQ0ONyYjpA6Qcj6EaByt4SxGBJ2RG8itMv/IQsR4LYoIHKSCTtlMh8EYWeysHlxSdhW7i+xJxDu/DMyaKYozjOZbMkEmtGD5NJKesm5zPOnDawpxVVs+oe48ijBoPTlxHuMlB+TMhpA+SuWTtD5BvN2DeZeetF+eQtMGMq1biosv2EEuK09uawAN3L+HQ7lJivRbS9/uQ4nBf/hRiH6ehyYKam5ygCgpS+3Gd1rA3CbwjonRXGHHUKiSdAMP6BHpbE7AeszA2q5WmqgyGZXdy+srlWJoMzC07gVAhVhQilG0n8HI2JRfWs6m7DGurguKVuLzkKM2PKBi7fHjzDASzdaydgsCABWmiG3OfTvOCHEreiILHwK/eu5byp/0kVEFZThfKtH7cIzWM03u5cMEeVlROpPTNJXj8ZmxtgpvnbkbEBOF0lUBpFAT481XsrRqWNoXeKXGMpjiPbbiYjXOeJGubRt1VduqvVnj34xmoFo2FP78POaqRMr6L9Gd38Yue4QQzdRjn4fDDy4id20HO73dR9+RkGi42nvEU/ZP5F3a7fwhYr+v6KQBd19cAXcAdf00shBgBXAb88Wt9FHgauOdrIwdgDjAGeOJrjQd4CfipEOIvA7sZsABvfK1pBVYCP/1HB/5dcsYgOsM/RP01yxm8YAp4AAAgAElEQVT+3FIAXjx7BRWWBhrmv4zBD4arg4w+twZO1bGzrgh3mR1ncxzjgMTai57GP2Sw+rQSgHN33En6Xi/+fI3kQzJtMyDYZsfeFmX/i2OpmFLDncev49SWEq7O2M9IU8egJ6rbzJBpzURVhbKEblrP11l9YCz20zJPP3gtrtoASZvN2Fe6aLj4FQ7/dBkiOcLSaZtI2mxm1uLdJFYKts4dRv+4ZCoHMoipMjGHzthld6OkhGnzuJiXfBRdgu/9fDW6Jgh020h92cqj7y+g85YwsqIS7TOTuTuMFBHU7R2CFAXtkIvm9fn0PpdPUzCJcIqg794g5s5BL9nsq/awZ9UoxEEn4WSQz+8F4BO/k/YVBfxi8+Wc+Kyck+5MCv8cpeaVCYx8ail5Gf1oi3p4qflsJBX6zw9jt4UxeCSSKlWiCRrDz69B7TfR1ZpI7qYY0VQVocPcnXdg6pPwFkr0Lx9CMD+GatbR+0yUTW7k4FgJdIHjHSciIcqLnTNwbLaROawbU5WF6OFEWs9T6KkAS7eOEhKEKhOQM4J8cWwEyfkDxI+7SDksiDoEgQVeLh97CGeqn3Ny6rB26hhkDSFrGI7aMZZ7+eGQDQCYlDg90zPQVyVjkFR2thZyetbrPH7kfOSsICNer8RdCq4qhfqbM/CVqkz83mG8RSDN7sUkxSks6qJ8mZeOKTKqXUWXdOruUXjqyjfovGcqfeNVus+JYe6RsKQFqT+cQ8QlMM3rZsjHgiEXNJI//zRmj0rbewWk7JMJjg5xsDWXxOMSJxqyOffEpYQKohwfyEKoUPpYmI6zFAbKJI63ZHG6KY3w0BBCg/Wt5QR6rHT9UeAthrFTa8jaMkDmlwo2UxQ5CkoYmu7R+eGstRSd10D99YnEbILK+izi25PI/EpQlNhHpTcD4wE7UrEftdPK6KtPsPLNmShBCWurzJoZz6EN86Okheg+P4Z5Uh+pWW6CbXZ0AbM+vw9f9qCxh6xjHtfPmDGn6Rut011hxvaIk8bfTqHA1IPJLUh508bIJ5fSs2QKs094/8MQWt9+5IxR9O/BLODAt/btB87/O/owcOJbegsw7Ruael3X3d/S5AJl39Ac0nVd+5ZmqhDC+t99iP+3nDGIzvBfknoQZp6az7XXbmb0vmu50BrhttfuZM6sq7B06zTeXoZDidC5aBxSmxlPCUQX9eM8rTP3s3sx9wjOXrqI3FcrSfvMTMOPJDJ26vRVqLiqZWwtMj1jTKR/r5GekJ3hqZ1EE1Wud/RRahhcPqi/ajnrh66hzefCYQiDBMkHZIZdXkUoSWLoi5V4i+DWX/yZbWEofncJC4Yf5rVTU7G3xfh0/RQGhuvkfNRL18wY0bfTGdiUScY+DWmch1jQQLDByS+OzsfSJXi+6lzS1ppomP8yvaMMhDPjpDgDaHV2HPUKDfON1F23nPR9Gkhgm9w72CqiQKbp5VLuueVTzCsT0CXIXGvg0wPjMHoGqx2Xzajn/VFvMPnIAq6wewnP8aJ8nRUVXJ5F02wT6V/J2Gd0YTdGUDWJbJuHvtE6hhoLoSNJmPvAc62f3C9VDh8opnx4C4ZeBU++AdknE07RKHoyPhh0PCxAx0wVa6MBV7UAodMXstLw/mi0dgu+G7xoQYWq14ciR6FvTwZTLj5G5e3LqL3xRX518UcY/Tpxi86cC/Yj6mwYbFG0z5KxdurYumKkXN5CKGjiT7sm4G9wsfGjiVj6VWKqhKHVhNEHFZktPHHLdbgvCtDUnURgro++cSr7/zyScMhIydabuWvUVizmGCkGP1K5H2+ZyvCz61A8El/uH0X21DYG+hys3DaVYmcPrhe7STo5mDJfWNIJHSZeaJnB0R8vA4OGudXIF4v/SDwukTKsF28xdDUl0fP9IDXt6VRvLyCUKBN1CuzXtpOz0oB1mx0lDKmbjTS3J4Mq6NiaQyRZo/5BI5NmnCTnnBa0mET5c0Ec+y2kjOsiGlcoeTOK7dUEyA9wcm0ZHeckotzShfpRKpEEQcwGxv123mqYTLvXiVTkJ3NBIyIkE5voo2esoPHVUoJ/zCYwPMKQ5MHaVqdeHY7jgk7koCDm0Lnq4PeRTtpRO61kfGHA+EEifJRCenEvqy95CtkVQ7qoF9UE1gYD7n4bvodz0BTwjwrTMstKfEiYR764nMzdITonyaQei7LpJ0+wcVou8L8qz9c9OZkvrnjiXzb//FuhD9Yi+i63/wohRBLgAjq+dagTKPwbpxUCXfp/rtvT+Y1jf3n9a9f8RzQSkP+NfVOEEOuFEDuEEKv/ypLad8IZg+gMfxehCvY8tpyGk1n8NKWKoxPfJ6arIIHqNGPviHHqjmV0XptMYEqQcdOrSTuo0tWcRChNgAz+khgDJQprT25h6D0nSE/wEciUSTkg4xuiI8VAtQzer31fFqOdLSj+v/6vWZbUzcLkHUwaWYf1yk56flaAIaiz5e2JSKV+frNnLnccvY7kY/D5B1Oh0oE334ASFmiOOBuryzHbo3iKJEKjQ7TNgIDbgqnVSNIJQazRTjhVR97uom+kYPhzSwmOCmFMDHNh1inKz2rA1D9Y+6Zo5WLiZoFeFMD+vAtpmI/g6BD61b088c7lBDMkooka7RfEKStpxzMxgr9A5fjhAi5++37cB1I5Z9EiAm4LRrfAstpJxClx7nnHGLg4SFddCjW785mXc4KdVcXcc8EXXHbpDnRJ5817nyI/qZ+WmQoXTTtMw1f5GD2DRSZzN8VBAMdriaSqWPbZQBWMu/gEti4NJTVM79E0pForxWNb8PbaSN6r0D9aw1ME4bQ4BzpzKVh9GwDXO/rY/fhynKfhsx3jydwZJzZgpn9yjEiiwJdjIKbK2A5Y+HTeszgaJEIZGu1nyXh8VpKP6Uize6n3pFB3nYHkVVZyX1XIT+5HmFUCxVGy3zcwu6SSpzZfSKDexYp3zifsNiMFJdpeLSZu09GNGorQcCQESd8DGw6NpLo3jWCGRNoewcy0aoQm6Pgon2nHLmfWyEpUs86TPedit0boP5JKyaPVpO4dDOB+ftJ7qBadvvMiqGbo3J5N7ygFf67OLQ9+hmoAPSaRO6SXYbNr0Aw68R4zexoLaN2WixDQNDcBf65OutWHYYMLBPSMVjAethO36gSydFo7EwlmCLyjI8iRQS9RT4eLB8vXYdnmoPJ0FklHJIp+0I0SEPScHaP5WhXXARP1h3NIPKSgGSEUNZA0tZNpM48T7LERKQqTPbSL3tECXRJ4LgrQ2ZTM5e/eh+oz4D2ZjKM1TtlFtRitMYY+dQKDX6BHZFz1GhcPPQ6pEequMeCqA9UkMefEjVT+voxxhzVaPxnO7KwxfLXgcUoNNmyFHuqvOZOB9n8gKUKIA9/YFn3ruO3r18i39keAv+Wlsf0NPd8457vSeIAm4Epd16cBTwGrhBBX/o2x/Y85YxCd4W9Sf81yfjX/Q2ZnjaF0ZAuzs8ZQsP5W5mWP49SSZaz/9G0GSo0UbriVYFkamR8YObCnlJa5OsIa5/ZbV5P/p8FCe7EJPgpW30bzj0oY2JSJ5aIukk4FSBzahz9v0Fta05nKDy5bw4rXZ6Pmhin85Pb/bUzfT9/GJRvv5ND2MpymMG3nmEi8rZnwVD/iqANju5EMpw//pV6kKEhRWHrPnzD4YMbwKiRZJxpR0Aw6hct0MnYIfjplDQY/9E6OM+e8A5jKPATGhwZ7belgrLWg6/DeBzM4XptD31gNKT74C9q3wIf9KxumniDioBOnI0RfbTJKGIwze1ECAoMtRnVNNoZWI7qiUzKqhfypLaALekcqOI8Ntm5IOezBPTOEUwmRsMZG5jaYeE4lH702g6S9Bp7cN4v3Dw12LF/yk7spdvSgmTW2rRxH5aJlBLO0wS9cfwx7s0TrPePIXaejzOjlqkn7OPTnEXjzZGI+I6pNIzokQrGjlwnlDSRVhQa9bid0FL+MtDERgzNKwerbqHhkCQADZw12jm++VsXcqWCtNaIEwV0KW0f8Ge/QGFe9fw/BLB3NHsfeLCj5Q5iOWSqO5S46K9NAh85zNPxZRrr8dhKTffxi2me0XRtje1shuGKYeiUiyTpKn4KaEqV7WhxLp0TCEQO1VdmwJRHVKDD2yAihE07RcbRE2HFeNq4a8OfpDHyVwc7VoxkyoZXa+WkYP0zEWO4l9qGNiEtgGOXmqeKhKLkB9LAMOiRVaUQSdeIpMf6470JMV3aRtUEi/H4GZjlOWnEf5SNbiPeaESO9ZGUMYJrYT0FFK0eOFJJ3zWnqFsuY+yCcolPwiRckMNUP9mhL2W4knKwjh3ScJ438ZNV1BKf7MbUZiNsEtXcXYm/RkYwqyVtM2DpUsr/S8EwNE0oX/HLoatrbkth8fCi5BT2Y6s2or6XjGtVH71idaJ+ZlL0yFAdwVisUTmih9YYYx/YVkfm2iY2rJlB9y4sIk0r/CMHGDyeSluxl0qg6fPmCax/9HPnVFIQ1zpFrSsm54iTzT/Vx/qs/BsD1lvM/5oUz/H+HhvhON6BX1/Xx39he/tYtA1+/mr613wQE+esE/oaeb5zznWh0XT+s6/oiXde9X7/fDHwEPPw3xvY/5oxBdIa/yl8mvRd+cSXqeRWsK/8cAHOTidq3KpidNYZRjy9FxHX0oEzfYj+BdBklIDB2Kbj2mrkjoQXfPT4mlZ/m5fErSMryoJok/CUxunpctNyv0dvlJOWwIFwcJsER4sMfX4TJrbP73Of5/jlbWRc08a4vGYDPg2buO3kl9mojSy/+gkDMSKwwRPc7g81SkypVMvaotPQloOuCSJLOjEsPctCfj688xlc7RlCU0YOsqMhhQc8YK12T4PmacwlmaZQWdZBh9PLK6BUkbDGjO2PAYMXlxLU2QgVRTO0GjG4Jc49O3TtjyXR5cV3eTs3dg1l30b1JCBVMAzrqmmTkMh8ThzRhaVOIpsbBpJJp9eJ+Mxe1JIiY4AEGM8eq77FgPWjlz9sm0jNZw58ts/NYKfZ2FccVHSTvMiLCMggIZkjUfK+IjB2CuAXG/mYpeetUTP06A+UW9HMGCJRFmf6b3TxQtp7Vn07F3qYRt0DCMQO6ovPcWe9xuC+bgYiV3CfqMHfIKBGNvIo2PGWDNXKyNskMjFR5bmAIxkYTdde/iMEUp/C1JhjvIZSm46qDigNXM7y0ldQjOuZyN5YGI5oRGi9PIjevl6QHGhlZ0QBmjbnjjhLMFES/SsH+cgK/+/MVWKwRvANWFIOK9axe8j8Pk1gJVmcYZUAhmBdHaDp5azW8Q+Ok3NxEzKXhO5VELCNK/VVGTt9dhqVfI56gYghAQp2GhE7l7zKQb+gm6DNR15TO0QeWcUfZV7T/eCqxVhuWZgO6BJ4CidKXuxj6UBN6WKZnwIE/W+b+n7zH8e5M+tx2Grbmc/qKlzAbY7RXp2H8JBHjzTq6VeXUzkKs9gj+PJ1Pr3yKG1euI54UI/VonFCGji5AHurDfGUX3tI4j17yLsoRO/GiMLoEFWdXY7uhnbLsLgI5gujCfsyr91E/8w0+vflxHn7pZhL3Gyh5PUpLfSoxl0ZPhWCgKglnnUTabpmbf7QGrcmGFIWBsIWi646gmXR6bw1SOquekneWYDptRvELoi6dyKo0qt8rp3LRMp46PpPuK0M8MnkVVT9xMv1YmK/6SwkXRJidNYbtL7xE0crFzK6cR3JZ3z97Svq3QOefn2Wm63o/4Aa+nVGWAdT/jdNOA2nfCI7+i55vnHP6b1zzH9FoQOPfGXY9UPx3jv+POGMQneF/w9wrMa/mIkY8s5TwdQN4fuhn/M+XoE8dTd70ZvSQMljhVgL38MEMl/iBRK64czMpE7uIuTS8k0MUb7mFhD9Y8d6WwqJ3lnB3yWYa5yskZXgov7OOWL2DsmUhfPN95H0oE4vLuIsV+kfoTFp3D4c8uSzZ8j3+8NrVFKy/lfXukXi8VjIubOGFzy+iqSMZ3WOkf7SGVGlnoEymZbZAr7GTtNLG7ZeuZ2tLMY035CC7FZS8ABWJLeiaRKgoimeoSuIJgViXiEgb9NKu+Hgm39u7EPdQnUtGHSV9ViuJC9r41c9fRzKrmAYEkTSVx376ElpEpv5kFr6VWeR8onB76xSiTh00gacEBsbFUSsdxHWJmE0n/0+gdBnZerKMgeFQ/GgEyxonkSSdiTcfJj3dzZgFJ6i/ejkiOhhvcu6YStzFMr6IEfd5IWxNg94Mx+xOKn/gRFnYRThd5ZolXzLs18cZsrCW6Dw34coEit9Q+eTT6fzqjevJPbcZoQ0aXjE72JoUfvzaQnIdbjo8TjZXljFp3nE6pspIQuecCae45fIviTgEyDrPnziHX169kpc9WUin7FTfk8vIjA6unbuNhLoIsR3JTE+uo+eyEPLGRIQGwSwNdGhpSqEioYXQA+kUv6WytnI4c67dRWRcgOY5ELdpxI8k4EgMDsZy7Uyh+QIzvRM1bOYo8YQ4+cVdKCEIJ8oknFCo3TuEvPUa5IUY9kgf9mwvKcc02i+LIgUlpFl9yBGd9UPXMOQ9iY66VCYWN9Jw4auUv7qEGbYa5BAY3RKqWcfSrRMoj+B9DrouLx78ZhI66PDApquJ7U8kHlLQFJ0Rzy7FU5+Iq1qie1qcUz/PRLbGMQ31EPCY0RWd9rgLt2rFYIvhzVPQZR1fIcRjMqOS23GdUvjxmusIZamoAYULb9rF9MRaOnZl07omn7SDMSamNVPzygSK31/M+sAwhAb+PAhlmFF8MnJIYPAI0kd0E0ke7GG3dsYwZp93iEgSxD5PpfaZyUhhCWmHi9Z3ClECgnGzT6GadDL2qQxUxPEMUyl9cwnKUTsOW5hlP7uSEfntbOgYyoeFm8jP6WXWCR+fB83kr4mxfugado/54Eyw9f9dbAS+XXdo/Nf7/xpfMhhAPfxb+hCw8xuaYiFEwrc0Ld9Imf8SqPg6Rf+bml26rgcBhBD3CCEKvnX/bKD5v3yq/yZnDKIz/Ce0xBhxm06mxcuJu5dxaPwH3FW8ma2/fIrAz320bM7DlhYgkCWITvBj6pHRFbB26ryxdgY9HjsphyToM5H6uYm6G4xUPWAnmqDxmyNzBmsTrUmmYtsA1nZBz3gnSe/bGCgxMCu3GnOfjtEtkbxPwR8z4Uj1I00dIDHZT1yXMZw24w5ZSBvVRXluJ+lFvYjEKKZ++OPC1zEMDDbA7C+XWbH8QiJ1Tmq+n4qtTRDpt7CmaTi2PVYs9UZy1+kEswZrETkdQQIxI1k7IpgO20g8IagcF+eC9Eq872Vz597rSNxqxlesAvC5ZzTEJIQmiCQJWi+P0xZMADEYOH3/patAFZRMb6T27TJsbQL13l6euOItiAsuPn8vXVMTcM8KIYcEjZPCeHams/1EGZXRIMX37kFX4PC7I4m6dJ4dvhLVaySSqONsgAybl/TcAWJvpmNIC7F83znsWFmB94FsAg0utLwwdQsl1KEBNAO0bszDe4WPcBpoMgTKIoTKwxRae5G3uyAk0/JQCfGUGKfbUzjw6Ui2nV9I39QY546uIt5m5ePucfx+yzyiJSESy/s5tL2MLZ2lyFsPMWROA6//eRbKSRtGr04oW2XMxDosXTrWRgMbOobSNMfG4lc+4cEJ6/ho+ySUEzYUVxQccdTyAOMyWpG8CtEEHVctZG4d/LW8ZvaztO/JQorpJH1ejadcJZ4ao2+4AYs1QtfMTPwtTjqmCpISA9gKPYSjBrquDlHyzhKEqpNwQmLv8WJG7LmeSIrKL1vnkbHDQ8oJFWuHIOVYEHSBe0sG/ePiyK4o+c8IYg6YVXGScJZK9lqZeXP2cu4VBzH3SHgmh5ECgwZq8joL0ZMuil9T0RRYsvVGXn7pYmrOeQvfxBCWTgk1P0y8x8wXx0YQzNYxeiU+nfcsDfNe4dH0I7x5egqGUW78eRot18b5/MgoZo06hcEj8czO87nq5s2kHtGJOCV0CaLJKqFMlYHtGcTsOppRp/GWItbVDCP73BbS9/sweAWaUSdQESI2141pAPY05pMzpY0Lf/0VtnoDrkoZ04DAOrWXmCojRzXqelIIv5dB8XuLuTDjFHsHCniueSbm0z0AzJ92GXJyEqU/O/4vnK3+b+RfVqn6D8BsIcRQACHEHCATeOHr978RQpwQQpgBdF0/CfwJ+NHXxw3A3cDTuq7/pf/LF8AR4N6vNU5gEfCbb9z3LQaz1W76WpMNXPMtzRjgjr94o74e49XAsn/04f5RzhhEZ/gP6q9ZzqxhVcy84DA7vhhN8buDsSOf9Yzh7N/ey85Rn1LwZiPxE07CqdpgFoOiE06PE7rIyz3z1zA6u43eCg3NrOEpktg150ly/qRg6ZAxHLUTzowRShGsfms63pFRjD6dqF0ilKazevUUUm9txNKt0zdOpc3jQt2XiPmTBML7kvnq0woimXFsL7h4f9jbeCJmvNvTyfnQQEJdjJ89cQvRtDjZW2JEXToJl7SRtz6Klh4hqWpw+UtZl4C3REWOQMdUGVM/PLzgI6K7kvGty6Brggl/YRz3+SFqn5/ExjunoVzZzSXlR9Ev6QNbHMUZZd17U7CmBjD1SESdOsnJfvpfHELWdpXk4zrPrLgUOSDRvjIfXz6oJgh+lMHPTl7CQ2d/zoFfjmf0TSeI+w1YO3WUjHScDRrEBJe++0Mc21Mw90AoXaf05Q5u3rOQ8ue9xJLiFCysobYvleCmNLz5ErEeC5JPIakqRvOFVqxtEqaTFqQBA3K1jaQqjZN3LUPa40KXdESFB6ISeA2seWcagezBTLmOKWaM7QZsRyyY+nWiJVlYGo1sqy1GS4hzsLIAa4uCFpHxHUxhyIRWOo5k0PC7KTRsKCCWpGFv1Ylf2YcckKj7qBT1QjfWqb0EogaEKnjq9Cyerz6HirH1hHLjFF53hFEFrUi1VnY351P2QheaUcefLfDlytxTsonLVt6HZgT3JQFiH9uxNcqkpHsJDFHR9iTiKQaDV0Kzq6iaQP4ykZQVVmJuM8nHdBqulhh98wmmj64i41kTMypO4Y5a6JzuYvvzL+EpUwlkm0nfpCBH4PFzP0CLyqiPDCA0aJtvw9wu035JjA0rJ3NR4lHCqRpF2T3ImUFSdyoMlEM0K4b86x7MvRLpWxUcF3VS8PltWI9bMHp1HPYQ5m6ZuaOOY/AIwhlxFv/0bkY/tpRZ1y9EfJxM7FgCIilC2loTyoDCgRWjSapUeea8d3nrxGT82RI5t9Zh7hUMLW9FN+ow1kv6Po1RE+pJPa8ducFMfVUWymO96AZIqBJoYZmHhq4jkKOjDZgwPOikPZJAaGSIwFkBDNP78BxLxiCrDNzkx7LBgW+eHzk3iFWKcuhoEdU12cQbmylYcxtNT9ipfriUmkdGnokp+o75Z2eZDd5TPwhcD7wthNjGYHzObF3X/5IVZmYwyPmbFtbNAEKI/cBeYBfw829cUwPmA2OFELuBr4CXvxnD9LXxdAHwPSHEdgYLOd6r6/r6b9xnOZAH7PxaswL4MV8ba98lZ7rdnwGA3BGdNFdmIEcErlqIJAps7RpdZ2lMGFmPZ1ofkbkTGCg1oAR05Cj0jtVB0RkxoonjtTkYOw0UTG2m3evE12XHlBhGknS0k06kKOgKuCZ109WRgMEaw7bThmaAkVeeYmd1EcZWI3pJgKQ1VgKZEuKsAWIxBdN2B6oZmOLGqMTx1CShuuK4jhoJZOuYBgRXXPcV1f50alaU4Z0exlhlIZQTIzHTi68qibyKNtq35RDOipO5VaLjXA2hCoQzihaTkdwKZ005RZIxyKr9FSjOKGpUZshKiY4pBoQOlYuWMfTlpUSSVXDFyFxtoP+qIMemvsn0o1eTZAnS/mk+2R+fpuaeAuKJcVB0REAmewu4b/SR7fLQsnEIb9/2NNevuJtIRhxjl4JtdD+mlYkMlAviNp3bZ3/Je8tn4ylT0S0q6VlujLLK0vytPP+Tq4i4JAaG6Zh7JFwNGp5CCUuXjtmj0TYDCj6N07Iohlw56J0zd0vIUVCNEMyNI+KChHw3M3NqWLV+MoWTmmn7YgjGs3uJb0zBOzyGsVtBiovB55XAUSdjOb+bpB/JVN7tBElnSF4vzVXpGLwS9haYtXg3H+2chBQR5GxS8RQa2PPQM4xdfjf5H/dQ9ZCDP0z+hFduu5z4T/vpD1gJhYxouiBlrZnrH/qC5VXTSHUEcK/J4siDyyh9awk5m6LMe3oLKx+bTf5tNeyvLMR62oAShkCOhnFAYtK84xzqzCG+N5Fogs4LV7zK7RtvQXZGEc0WtJwwUrsZg0eAgFDR4DKpvdKEmDqAv9mJNcdPsMUBQkckRqHHxA0zt1PrT6N+eTmxK/txu20kbzERv3SAgsQ+QhcGqP/5aEadVYskdNqfLmbJbz/m9aWX0vA9HddeM86WOOKubrJsHv6Qu5rZe5dQedYKynfcSNW0FZS+vQTVrJM/op2Bj7PxFYC9WeAeF0UEZSxtMuZ+nf5xcey1BqLOwaKPZ918kL0vVtA7XsXgllk8fz1vvXYhclhHisPFdw5+LtyXG5i7pZIXq84m2+Whc00epn6dQJZAioG5T0e5qhvvtnTeWPQMS3/3A373wKs8c95sembmMjAU4mlR8nN6sdyhUHVnCiU/2Puf5pC6Jyf/C2au///wXXS7t5Zk6SVP3fpdDQmAYxf/5ky3+3+QMx6iM/CzOZ9iuqCRvPUq6aO6MPp0QqNDeIolUvbLfFi4CWEwEkpSiDrBU6YTThZk7tBxVsnUbShk1shKRp9bQ6s7gYDPTNaQPtRGO9mJHhYvWEt8RIC4TSe0IQ1Tq3GwmnRARwnqHFk1DGObEaNP8Pi4j/EUSkTG+fH12sh5Rib5RITwmCCaJpA/Tib5qABd4BkbxVw2GJT89r6pHN5ahhwBqcmMuUcnY6uMpy4RJSho7EzG0qODrFN2z0kMbpmS8jaKlmmkb1ZwFXlFnL0AACAASURBVA9gV6JMsDdw/9lfkLTOgmxU8Sz1cvn8HThP64zYcz1xi07RsHb0gMJ5D+/izuFbGb7iTkantFF5OgtHm8rpZ1PRZR17nYEFYw6SfFii/fIooXonp/fmMXpuJQufuYfokAjGLoWEGjC/l4h3iMR9V65CyQ2w5erx6AKQQBg1onGZ7r0ZfNw9Dl+OjKM5yu/mv09gSBz31T7yZjfimR3AXShj8Eh03xVCU2XC2THksMB1WiP1SARDAPJLuqi94kUSrSEeyzgMBQF6VuZx4bW78Z1MxlekUvZSENfYXiKFYdAhb41G0WW1dHW5aJqfTPmzPnJXSxQ4+zB3yyhBgaYIPtwzEWO/RMGYNlqvj6Oa4OwHf4A20kfTpalY7RF+suo6Sh6vpG9dNv5GF5eUHUM0W+gbIbgrsYlYjZPWnkT8uTpzqueAgMYbdT7+zQX0TIvjX5SMudWAEoTkk1HGTKwjbtU5+uYIvO0O0g7HUEKCn/36+xgTw0hCZ/bMQ9jtYQweQSg/xrJblmNxREATBDM10p62oBt0HB85BkOHFB16TaTvA3/cRNWKcnpmRnE3JKJHJdzlkPoHEzVrS6heXk72lhiBu9Lwx0yEEyV+tmEBclhlZnk17ooo5/92G/cVbGD3qWJ+0X4RSY4A4361BOMeBwWfLWLWzMMofoH6ZDpibh8Jw/v4/h2ryV4roys6yZVxLJd1kZ4zgD7FQ0ItZC5o5Mv1FSRWh8hbC2pemOc3n49vbJhwsuCGe77gzUNTOLRxKLX3FvH8yosxGWIYZJW4GfpHQLA0Qsn8WuKXDtB/MI3R8yq57ej3SDoZ5M4D11J9dw7mazupvfFFppfX0rMxm54nZU4veAmAT1r3IBSF+af6qL9m+Zm4ou+Af0XrjjMMcsYg+jfn6FVP89jbC9Cmj2Xrq68Q+CwDOaJTd94b5H/cT2iel88CVq481oylN06kIML6BY+jS9A7RkI1w5g5lbySu5MPCzeRYA1hssTwf5EBGgy8l8Mbr8yBRis5m2KEMnTypzdh2WMnlC7wlOrIEbB0Cy6//iserb8Q1+Ru1E4rSfsMdE+w0nSREbXfhNkYI35FP5YbOkg4YmBS2WkiVS4CJVHK7jxG3Db4KzdjXCeSCoEsCVOeH6HCpIJGPGeFeWvmK+xZOxItL0TwhWy6J9joHy7IcnrZuqqCXx2axysvXox7bgBdF4SjBj5/axo9MyOEGxyY+gXeN3MQmmDvneOoDaUDsK9jCC+es4JAmoSy14EuDy6TfbJrIqoJdE1g6hv8uO2pK8DZpCIGjIMxSImCzmmDntqn37sUSdLpHZ+Ed1yEyybvJ+cTBX/ATMyuMyu5El2GjqkmHvvDdciuGJEGBx1eJ4WPqQTHhFD8An+XHTWg4DxlgPwgfaME3nt9pBwP03Qqk7mX3oT/7WyKt95Mxkoz/ePjbH5pMlIURFKU6ttsBHakIvUY0Y06wTvddAcdJBwwocvQuCCJjikygbiR/JUdGD3gmN+BiAmieVEmJTeid5mIWyDqEMSbbQgNAr1W7pzzBds+qcCfr2LplNjbk4+jCexDB6h4ZAn6kBAGY5zsbSpVJ3MZM60GPaDQM0bQMPcVst9oI27ViTkh4pI5dLSI1CM67slRbj5rBxGXTM7GIMF0wUeTXsZhD/HlhgpuKt5LOEvFYI/yh6aLSH3LijyggA4NlxqxZ/jJXlqHOddHcVkH9kaJwDUeNr4zmUA2lD0ewNQnYaszcvY5x6ldpBBO0TC0mGi+UaX5ogQ8y3IJXOAHHQJZJtqCLu6b8iWf//5cXm0/G+tpA1/tHU57czLp1zQRHB8ESefLTWNRzTopDzXg9th4pHwVF9oqCaTJ2E8rtMzRUZ5Npt9r5cTkd+kZr9G4OR/VBO5iC+Y1+zDWWhCaQI9JWHp0VneMQvgUYk6NeGIccz/0tyWQbvYxZFUvmbs0UrYZOdmRyfDUTpShXnafKGZiZjPKb3uI9ZsHPYTvZDB30jwOfD6CmENnXu4J/hywU7NsIldPuBQ9HueYP+c/5hShnvkSPsP/mZwxiP6NOXDlk0x5+j5yfreLLz94g9Jt3+PwT5cRswqKPlyM4QU3OVfXcu+eq3mlYRots2X0kMzsHXehTOtHlPiRw7B3XxmveTIo3HArAwELoX4LcStoRh33jPCgEfRZkPaFUeQiP30r8ohN85K7ph9NAYNPx9Kr8WFNBT2H0glsTsPgkfAWQXiyn9wx7biqZPxHk0n/qeDF0vfI+KqfbLObuF3DYIvxRu0mko4J+jw2+jdncv/9Kwmn6Ji2OAnlxeg7a4AFww9zx/Kl2Ft11JBC13gJ34QQCaN6Mcsx0GHxyO1oMwcw77JjrLTAUSeZO7yDhoEE0QSd7nNiKD6J07cLDvxuHFKhn8z7o/ypvwL3MA3NAAafIJKooVtU3OWD2Upxq07W9jhWRwRdBjkjSNqhOOZ+jdS9EmmHoshhCPVZKF5URdIOIxtXTMZwVyei3srIcQ08tv5ifEMHA99jNoF9jwXVpeLut9F4iRNJ0jF6QPHKYNAIZuionRZeufZF9ld8yECJibSSXlpnOOgZryFJGv6FHqzJQXznBok5dKzHLOQXdREsiuIsGSBln0yiOYT6dhrhZMjZFGDIag9SHE51Z9A+J5PIeV5aGlNQUsPIXUa2/O4szppyivxPeii5rpqEakEwV8XQp/BK1Vmc+MEyRGIUR5NGTJPI/KIN29suLl68DaXaiqKotF0bRTerdAcdyCGJEzc+y296y9lcU4qrBsIlYUIpEuZOGdeSZiSDxi9ST9E9XtA10YoyrZ/bfn4vwbCJWG6E+5JOc/qyl5CrbDRvGkLLBYLEkwLNpqInxJiY2UzPo4VomsTi3K8wBHSsH7qIWSFvaiu1PzFja9eJOXT2fTKK3FUytmIP6eM7URrMOKZ3Y1/cRqTdhj3PS/d4iYbeZFbdMRNPkYQ3YsbSrWPplLDVG6isyqH0lz4kSxyjR1D8vo9EYwj7fgvL2s7j0kOLkC7q5fg9yyhdfAjrwSa0FhvDdt1A6YhWogkaP563ip6pcaYfCxMpCqM54+Sulgid78PwkAs5JGEr8LBm9rN4CzWyvxTsbC5g7cYP6bwygtmtcePQfezZVU6k0UFCho/t60ZTt2cIt037Ck3RCVzmZcRnrWgGcNXBL1JP8fxtV5GU6wZJQj2vgv2vj+FNbxrr248MlmbwnzGK/icMxv2c8RD9qzhjEP2bUnzfHq6+agl3LFxFZEM+x6Jhas5+m3NuX0T37CiZO3TcT+axrmkf5T/vp/9IKuYuieRDMuPzmwgfTEI+7CBrQzfmPB9PvXk5YsCA4SsXGbn9JFWp7FjwONqAES03zP/D3ntH2VFd676/qtq5d+qc1Tkoq5VzRhJCCCQESAaZKEAiGhOcwAEbGww2BiOJnBEGIZIQEqCEcg6t0DnnuHfvHKrW/WPLZ3C5591zznvc53NtfWP06O5Zs7qrulavMWvO+X2z+ZI4tk5cS6jDgj9N4rL8s/zi03cR9ggz1hyiY7pG0GvAXgtLV+4mc3eIy+YeoSClh0J7D1ELhBNUlGfdXHf6Jip+ZGXTsTHMGHeOaEThh9UrYlmZiIIvN8rjL65AjoB+YTeSX6Htocns68wnMDJA3+wgifv1ZOyLMqWwFtepJM5+XYyjXmPz/bMJVjjx5mpYmwVhh0bLHDtaUgShCFKOaswbcRZ7Dayb9DYdE2WMB2xUPupg70dlmDO9+HOiGFwS9jqZxIN6kATCr0MrCNA6U0fAZ8CTrTArv5rbn/wQJQyX/HgvfaUG/Okaxg49Rxpz6J8RJG2/h+1DPsU4zEXo3kTsNTJosX6O+MWtBFIEqBJJyR6M/RJOux/jgEbS8C4IKqgZIUgOsal/LCOeXoOpP1Z+CyYJhEGgqQr+kwmIEw60VjNF7/qIr4piM4RQ3Dp85Qmk7GpnsLODyT8+TMSuoXu8m+b5DqJWQajGjsEtCPoNKNYomiZjcEu4CmS67h7EvI1HSDL6SFzezF0zvyLlmIZ5i53BL6zhysGn6LosRDCsJ+v9bnqHKry9fRqphyM437RhPWjBecKA1RDC1CWzqmkOrxydSnqym6OPrUPqNfDNz/9MIDNK1dEcrIfMFO26EVPBAN48leCxBLrHa9w+ZA/Dctr+be07qzWMbhhU0onJrWE/p0doEs0/KqD9+iDSaRu/ePt6esdFMd4Y6ykd7Oggyekl5bpGTF0ShgFB+LY+BvotNLckEs6IIDYko1wbwJzlIVTuRAnB+SlvEbHq0I3tp//zDFyzgzhndKDpwZzsJ+H1XkyWMDnrz1N7rZ3D740k/YpGgj9Lw9tnIbIzibGPrKb1oQmoOanIKkjH7LS6Hfxt6bOs/8sVpA3q47grGwb0DPltD+2TFKLVNniiD6ETpNs8ZCgCzaLhS1MId1oY+8vVqL1GukbLbHhvNpZCN/FnJAY8Zq5YvJ+5c0/w2tbZbFr2DNZNNt4/MB69LzZLb9Kpq3j3reewvBZP3apcdJ4wyesPsM9dxOLqBQx9bg05jx642Gz9/xL/IJbZRXAxIPqXRO3y9YjJI2m8zMIdzlaM8xp4YMXtjHh6DbtfeJG6ua+y99kXsJ1sZ9Kpq/jNjg+IpEYYvqiCgQI4+0kpebMaCKRpVN2WhHzAgRIGS+4A3hwN54N6WhZHuf6me6lb+gK6GjPaUC/3NiwjfY9E5i4fm2uHceubd2GuMTLZVkNmbg8pOw28++hTHL5hJHXXSXz14XiqD+Ww48BwfDkqkpA435xGZG8ixbkdmJv0tN+Vg84Q5cXCDcQvbyH9YwO5hZ2kHfBhGdVHf3kSCaclAikaWTYXekM0lkXxCRqXadT+aTArLvuGiD2mlt1fbGDElGo0W5TAlW7sxf348iNYTxuJy/bQM1Lm4IYyeidFePC5Vag2lfHLT0G3Ec0A0mHHhWwQ5F1TjTcbioa2ojjCqN0xQbzMjXoSzkfY80kZv/roGvY9s553901Gnt2HwSVTOKOewifDTCqoZ9vHbzHiqTWEzjqpuM2KLiAYXtqMuc5IQ20qhgEJxzkdplfjMfYL+moScNT46WhIZOn4oxgajBSvruHzr8ZReEU17hVeBjwWCv/mQbGHUXQqamGAYJqKtUnGXRhHf6GO8rpMIJbValuYwTetBRzvy8bSJtP6cS4I0A9ISBpEbBJFPzyObb8Zh93Hgz/cyIoVO6i81YJFDvP19jIaehLYfN8s+koU3HMChIsCtAadiD4D/jPx7P2ojEipH1O3TMip0DFRxl0aZaBIo6I1jfR5zdyVth3Jq+DalUb+VzczfHQ94168n4QsF3qfhJAh6tUTqrKj+GVsjQJTp8K6jy7l7IlcPvTayft8FSG7xMDEAJ17MwjZZTQDmBoN+DNNGI9aObd6LWnTWmPMu0Pp3HndZ3y+dwx97jiqjg1C7xNoOokpqXX8etIn1F/6MqZmA4k3N/L6iU8JdFhJOaGRPaWF/K9upmu0joDfSMKiVmRZo7M8laQzUUS5neNtWUzNrqPimTyiSREiFqhsSKduiQnFpSP3ijp6p4YJj/KiuPzYa8E/KIphq4NfLLmR/glhvDtSqdhWhByU6LgkncET63FUQavbQd6oVpr64pnw9o+RzFGSTvkRZhXfPC+FG4KUTK+PBWebnKTe0IAWVtiyYTKn/jgKRxVc++r9yDGCJsFEgdZjxG4IMefZB2mdA3nPnMX/Ox8tHw5lV21RrBz3/BlGHJcoeO8OvrzqqX/QLncRF/Ffx8WA6F8MHy19hsk/uoMvN75B5c3rGPfz1XzYcpDaNTLOebEAaMjza1hYuRDHu170ssatT90HEZkkgw8tO4jeI2jbFJudlb5PoBliP9v4uQPVrBG3rgd9u4FNb/yV/C9vIWdqE9Gwwrkjuah6cP3CTySikP9yI7oArL/lKjr67HgWe3i6cy6tc5zozFHCdoGWFUTnkXCekRGKIGOTAXlyP+1bBmF0QdMCO9EuM7N33EucLkz3KJnGtkR6h1tQZBHTH0qUsBf3E1YV1Borcr0ZvVdD1mlETTJvHp6MHJYw9UaJzHaTY+lDCil4O614vGbi0wf42e0bUHY4yf3MS8L5CCtGHyaUKJDCEhkmF5YOmfhKjawv3ViadOj80Om38YcVbyEeSYR2E3VXr2fQjCbarg3TM0zP0qv3oOR7mXnLKhSvjO9cPHIUzrekUXeVg46f5bM9oBCxQnRQkNyiTlwlUPlNHrox/YwbVou3NIw8t5fWxVFyb63C0iZTe4+M3q2w6chYEss1ejako/NLnGzIJnLOjvGMmco7TGh9RiqnvYkka6BKDBSpuAtkvMURDC0GzO0ySSckhl93honpjbxQ/C62ZpXAJC+BkhDSYC/WhpjQY8/tk9AFBJPSGnnq9WW8uXkWjnM6nvhiMdddupuwN7ZIJl1+mrQPjChNJirfLUWKSvxg0W5MfYKkz00M2tJPx3QNc4kLR4WOuCYZoUHjkSx+sOluCt4PMWrxOZzxPip2FRAcFMZzJhE5DP7xfqxVerTMINUr12FrCZM0tZ2oVaCkBvjph9ehDCj0j4+gNJkIx2v4MiSUIOh8Me0qb1GE4jdWk23tx1Kv58pFB3hm86KYNtanZkRaCN98L548jc1bJ/DrL5Yx5tercdRonK/KZMnZlSTm9BNIlKkrz4SBGBNODSv0f5pJxtsGlBC0XKliHtOLWmnj670jSdxtJP1LHdYWgaFNT+YuDUuBm+4Xcsl5X2ZB4XnOP5BAwhk/hl6F/skhKu83kbzbQHisl4RpHYyaUEPSNc2U12WiXN1NsMpBusWNzRLEWQEZn+mpXSWjt4ZxfhRHy+w4zrWmgQBvpkTXa7lMG1yFtzhM12gJS7fK9EUnaJ+lkZnfQ8EHHtAJak9kYZ7ZzfhR1VQ8U8i7g98iO96FaDPRNMGHa+EQ7k7aQ+3y9WzyjGT3sqcuNlv/F/CPoN1fRAz/R2j3kiTdDTwLzBJC7PqW/VZgNTE1ywBwhxCi9jvn/gxYBoSBVmC1EKLrW8f1wBPADGLy3ieB+4QQPv4D/KvT7gvvP4gudxDRhiaafjWZvA1d9E5IJvFAJ1IwTPecbOIrfFStMlDwlkb3fQFSnjLSel8U85c29P7YWkk43kfVz2PzAFWvDmOHnnC8ipwUQovKOPcbcQ3RSDglE7FLiBn9RI/GM/uKYxzqzCH6RRLebEE0IUrR62HqlsYmuzqqJa6562te33gJ1mZB/xBBQVkL1c2ppKf1E303FV+GxBu3P8P1R28h2G0mLa+X7rPJJJyR6C8FU5/Ep3c+ydInH8I7xY/ppIWwTRBOi5K5VaZ1oYqp2YCzUiNsl3DPCCA1mVEzQqQkDeA+lELy5HYSTH4qdhWQMrEd86VNVD87FsUrE00NY2gzUDSlgar2FFLiPfTvTsPSIRgogOSTGl3LAqS9Z6Lj2hBSrQX9kAHCFXayx7eiPp1Kx0Q9pl5wLmqj96sMvIURpGCMqZVyTMOXqlC0opJTO4tJOC/omK6BQUPyKgiLSlpmP9IbyQxc40E96cDUCwNFGolFvbg8ZvKf0qi6x4CxzkTagTAts/QoIQmjCwaGhsl7T+DNNCAJCNslkpY00+uzcH3+EXoiVnb/fjLhOIm+OUFEVMZ20ojeK/CnS0gq+HMjpG9X6B4jkTKiE/euNHzFYazxfrz9FtK36eicALkj2jDdJuMbnIwvRcf8e/fy4UfTCKZHUWwR1AE9cQ0x7R9ftobeLRMp9ePcaaZ/qMBeI2Ps19Ct7KKtJhlTp4Iy2oXPY8Jx2ISQIBQPcgTiZ3TQcTaFglEt9Pgt+I8lUTSrjsi9CVTeYsOQ6ufUlFcZ/NUdOA4b8eQJaq5bxxO9Rbzz+iUMurye9g25sKgXTZNxddr46bTPeeLofHLeUmiar0NLChNnD5L1S43qG5yQHiTOEiJ4xknCWUHfEImCZypZvKeSp09eglJtIZIfwFhlRucD74gQu2b/hfW9k9lwchzLRh5nS/0QNE3GtNsWy7ot7Ka7zQkC0ncoeK4dwNdkx9Qlkz6rhdZ9WYTzgpjPm1DCEHIKDAMSphk99DQ7MbfoyNrho2a1woziaqr+NJTO8WDukvEODpFwwIAvC9IORmmZrXDzvJ183DyC/rNJGPskfEVhiMhY63RkX9pARVMaUr8BJc3PqKxW0kwDbP9wHIHSILlvSdRfK1N/2UsAjHh6Datu/pwXX7sMc4+gfwjkP3SAbW0nmZ8xCvjnpuZ/H7R7c2GGyP3j/zrD8f8LKpb+6iLt/j+J7z1DJElSBvDAv2O/AngcuOzCxNpPgC//rnx5weceYCUwXQgxEagnpob5bTwBlAETgPGAE3jp+76Pfzb8vZ5fcW8GnuUTEbJgy86N3PLTT+iemsrvvvmQu37yAc1zrGycvZaWWSbYFU/bdAtBv4HwpW665sYaj+uvTsLxjYn0jQYsSX4cNQJ7tYLUZCb9Mz1KCJKPyEQtEjOuO0KgyokShC/2ltHT4sSfITCVuNH36Ki51oxmEKh2lWCChEc1cdWSPTiva8HaKCN+kcjogkZcu9PonR9E08PVn92N/oANU6eOXxZtxl4rMbDQiy4gMfSKCi59/SEsl3eQn9qDzgs6v4S+V0f3aBldjx5TD3RfHsSfLjE2p4lbL/ua3Dcl3IdSCBcF0F5MofuvuegC0HkkjbYPSyhd24+jBvLeAXstnK3IJielD+21FDL2+PHkgX7IAK0LVRzWIFGTTJwlhKVDQjro4N4rNxOM6mi8ShAtDKDpoNsTh7cwwh2Td1E0rIX0/SpCgqHXnePoqULkqMSvf/MKprYYEwpgypAaOmuS8KXKWE0hpBEDbHroSUR8GHlDIpsnraNzgp3s93XcvGwbjYsVbI1g7oSBkigl+e00rBTkrqrC2hLGlymo70zkkuxKXnlnAV+8PBXDqnZ6xqloAR2lT3oITfbQPzQ2i0vTg6lFT/tcFed5SDT7CTsFtgQfyessKH06HGddaGaN+rMZuNfK+Ne4cFYHeO/cGO669jMwaJgtIQb/qQclCJoBVKtKuCiA+aQFf6qEwSUz6vpyAskye0dsQhg0Zi4+TiBgQNdiJGwHX5YgGidgrBvjH+NJG9pF39vZ9HXbCeUFqTiQR8dUJ2gSpj02Jh27HmODEUdjhNJxDSyvn807r19CUnmY8ydyMFzZhf9YEqb3nOh7dDy//koGZ3fwhxfWkXBawnLOhLffQtfEeCxtMklbTNxatB/7qF66FoaQNIn2a0t5r2UckiTQin08OeFDombB6QfWUjfvFWZsvp8vXpqKqdHIUEsrJcldBHvM+DIE6csaiDOE0dtCoBMEnRJJVh85Q9oJZEdp35VF1s4g6SkubNO68A7ScFbFSlkJZj+YVLJnNyH9thd9s5FDm4fjS5UpfM+LEgDJryNilbCM6qNnuB6hwKvlk+lpSEBIYJvRicEaxtShw1sSpuWzXERQYeToWgr+EOV8dyrbvhjL2bvXUvqEl6hVoWS9n9G/iQm4hsZ7+fP2BSSei+CoCcR0uIDB69ewre0kDb+dhCH7P3xvvYiL+Ifhe88QSZL0IbH5JOv4VoZIkqSjwG4hxI8vfK8HeoD7hRCvXJhl0gb8Tgjx3AWfVKADmCuE2C5JUjzQCVwlhPjsgs94YiqZRUKImv/dtf2rZohK1nVhesWDb3o3HR8PRj0Qz467/siEr+6l+OajuH44iUN/WMfgF9cw6Ff7Cc8fS3+Jgefv+ysVoQxebphC9P0UekcJJo+rwKxEuDrxMHcdXQH1cQgZtIwgd5R9w9svz8faqqL3a7RN1ZG5O0rLTB1aVhBDjRlNETyx/C1+vPl6LK0y3iFhHCcMxHVqeLJkUg/7aftRhLOT3uFsOMCKkzdjN4Xwh/UkPm2hZoUeKSohRSSEXuAc5GKgKh5jr0zBgjpaNuThmnxBcO+UCU9RlDFD6+h+Mp+QQ8bo0ljx5Oc8sfsy5o0p5+t9I0kolxjIjzHDlEn9ZN7SSfurKbh6rVhqDCy5Zg+KpPHNQ5NoWCqheBRSDwlGPnSKr3aW4ayAvmGCnK1RWmfoiTg0UvJ76S1PJupQMXYphFJUbGkehJCIf91KxySF7C/DLHh2F6+/O59bfrCV50/MoHbOawx9bk2MkbYvStMCBWu9jD9TYGmVsHRrZK+ppvyrEpZcuZcNhybiOKcj6XSQhssNiNQQc4sr0Esa1QPJeCMGeo6lYmsAV6kg6Tg4qn10TrQRtkMwRUXnlYkkRcnJ7ab9cDqpR1RarooyIqeV6q0FOOo0OheFyH1NRugkVKNMyzURhCoj9+hRAhIU+0jaaEbv12i8XEJnD7Nv2vPMee5BMnd6aLzURlK5iv/mfu4u2sV7K+fT+BBIJ2O6Qd4MHYmnvWQ9W8/+L0YQdmgIo0CJD2E6YUHIMGxxBQ1ri3HU+qlbGkfUppJyQMFVAuGMMKZ6I8HMCKnZ/RgUleb6ZKSQjC4gkXBGIF3XjRASXdVJ5H8UpnalDBEZJIGpXU/K0Sgjf3WC+c5yDniL+PDjaQSzw5iaDKTvC9FfbCS+JkzbVD3Gfin298sJk7pDR8QiEXdVB51H0ohkhcGrQw5JLJxxjM37RxPXohBI0Uge2k1Xr53UzUamPXyQ94+MQ28LI9VYCKdGSdul0DFLZcOc9fym8XJaNueiBME1MoKtQk/p0krez9/OmF+tRl7ci+d4IpmTW+namkU4XlB50zr6VT+jt95L/DEdcZ0aPcMVDANgmd9J9P0U+oeAlBkg6TMTPWUS1devI2/zKszNetRhXkR9HNYGUEIQdkjoAoK+0VGcaR60XQkY+wX2hhDeLAOOm1ro3ZDNyFvL2b1vGAU/PsjAFwXYL63lsfoj3PDavRTMrmdiQj0rHEcp0FvJ/+h2pNA/X7Pv95Eh0Rf2dwAAIABJREFUMhVmitwnv98MUeVVv7yYIfpP4nvNEEmSdDkQAbZ+xx4PjAGO/t0mhIgQK3ddcsE0Akj9jk8nsQFuf/eZAei/7QOcAFRg7vd4K/80eOWKFxHNbZxsyKbmzxPJi+9j7e1rmbznTuoXvEzo0nF4Fns4HIqwfMkumn8xmYalsf/pO5+5i7+8vJStw9+mt0xD75E4WJ9HcVwHD/1lFVXT34yVeUZ2ovl1rD85ncBEL67lXhIebiBq1+gq02Mf2ovlpJlosZ/KW9Zx/75rMfbIhJIEkqJh7tHoGR77nUpIJVJtZ/ihH3Dd0z/G0xNH+7kUbC85qLvSyOCSFtILu5FSQphbFTwVCaTvjylnl1dnYfAIzOdN2E6YCCYLEo4r9P42FyFD58wos3+3l7++eQVSVOLAe2VocSq2pjBxI/rwlYa4qegA1Q+V4G52YGjXE0jV+OTtaWx6awZ9q30MeTzGEDIMqBx+oYwxUyrxp0pIAmY/tZeITaPkRTfOJc2k71cxdCtkfxkAvYbYF4/hCwe9Q3RkjG7H9SMPr2yahz8ryoaGsQwf1Mb4n67GUaehBCX8yTo0g4ZvkEbqiE4SKiP0DZY4erwQx4Quvnx+CoZehbIflOMqMPKLyzZBt5EdO0bRFbKSaPIR/lsqutIB+ocKVKvKyHtP0TbTRuqz+xEy5H8UIeJUUawRunZnMH/BUZoXCmz2ADOTKglkqQSv6yfzQz0tsw24CvT0DtWh6DRMNUbMHTK2JpDPxRq+u0brSTihQJOZhb96AJ0PCtdWEcoP0XK5StoaP6/8bAm1P1LQNIlgUYjGqzUsPSpt023sPDaUiE1DCUkM2qJBsxl5Sj+qEU59WUrXJEHzAxo6nwQmle65oVhz94Aee4Ng1aRv6GxIINfey/DBTRjcEgnlgq4JAvultfT025A06BxrQmeO8sD0L5CiMkIWhJwye14fx8+eu5ltf5mKrV6QndVLMDtM23QD3hyBO1fPI9e+j6lXYBnbg+zW0TVeoL+ym76d6aj5AYrWhhk9shbneYmQpiOpoA9fjkp8aR895Sno9CpxrSF2PjMJyaDhsAVIO6xS+GaErvGQvFfHDe/cRc3BHNTJbqIW+P2MjQgZjpwpYPif1uCaHsRzPBFbPTSUZ+AdHCZhdBeLqi5lTzAJ2avgzYWOCTKqRcTWf7cDIYOaECHuoIVgvIx9cC8lr6xGbw+hH9PP4LRONL0gMt+NoyFEIDX2wiyFZVJ/q0dSwZcu0TLLRNKtjfS/lY2rBHacHozeI+PeUohFH6H1J5O5/el7OX/HWs5WZbG7u4hlTzxE0a4bsTQpFxloF/HfEt9bQCRJUhzwOy4McvsO/j6ptv079g4g/8LX+f9JH3HBBvxbYNX7LZ/vXtdtkiQdlSTpqOr710rXFt5/kJs/X4Xt6zgmF9ZxYNnTnKocxO1vrqH4115+2T2UiE1BVWXuf/Au9q0ZH9voG/R4czQmrzyOMr2PUVvuQZg1hAKm02a+7ByCe0SEsiPLWXvjesoSW0GVMFvCzCuoxN9n4dyuQqSwhNEl6O+zImkg2k3kfXIb1nIjgcIQUp4PEZFj87jiNbwjQjTPtRFxqMi7nGgGiD8W63/pK9WhmbUYy+cviTh2m9D0YB/cS8cEGU9pBNmkMvyecuKmdqMaIJIcoW98BMNDHfhudWGtNPDm6QnIYcARwTc6gBSUaZtuwFUfz5SSWu6Lb0DLDIIGEbtG+j7IfqMaX65KxqOC2j86STkMvnQ9/TODTHHWEkjTMPbIfPTXWTgqFJBlWu4bg/W+FsIZERoXmZGNKv50DU8u5MxrwPNhOu7KBB68+iMSjyl4/EaSTV5sLWHa50cxDEDXtCg5nwvSDggyrG66yvRE4jWUpBD238Yx584DlEyt55uDQ+mdFOE325Zi6pLRjAJvxMiZDwdTsuo8xq/tKGl+UCVOPTuSpEta0aaVESkJ0HVvAMUZRl9hITzUT1vAAaqEALZdN5n4Qf1cn38EV4GO6ZecRjVIGPsEalQmmKyR8Y2H+PMBglkRmq4QRId6CSRLqHEau379Z1au3srnJ0bE+pA69TRfMwh3roLJHCbsj0kSWJ0B+ot0GDyxjJDQgVzgxZ2rp2BcE/KX8aQcj+KY0IVQBGMzm7B0CCQJdC1GUk6EUQISQoJXt83GXq3j8cwtNG3MJ3NyK51zopQ8XM62tpMYjBGev/y12Do/bOHl6imMGl6HbXQvGbfVIkUFA6VRekcJekcJWqpSmDfiLM5KQcYeldSdnTxy4EpKVp9F90Eig7ap4IywJPsUk5acAiAcb+TU/iIG5vrY80kZPQ0JpOT14jqbiKlLIvEjC7XXGugdKTDGhRk4lYjpvjY0Y0yVurdMkHRaw14Lvx/xEYnnIrxxzQJW/HA7Iwc3Yp3TiWOfiVB2GHVJH+mDu3hq2vv4vkqldmcei+P8GLO9RC0aUbtK4Vt9CBnMcSHSVjag9OuQ5/YSsUNwfxIZE9tI+cDMQE8cDRsLQJMwfOGgfaKJtEMqfWOi/GTuZ6x8ewvWVpVApoqtQVDVkUz3eBXNpJFwTEdcsyDxzgj9GzOxN2js+8kzXLLiJv48awO1ZzJJWbufot/6cc6Kbd/CeLHj99+D+J4/LuI/j+8zQ/QYsF4I8d2ABiDuwufQd+whYgPj/is+EfG/1vm+7fM/QQjxohBirBBirBIX9++5/FNCi4+wre0kg59s4djRIgYiJpbd/SMuH30S69ge5Bd8HJ7sRLmlE6rj6B2qEHzUjWtSCEuHQDVrfHFqGL6AgeLbjqDr1WEa7MLcJQg9m45sjhI8kcBdL93Bjk/GULp+gEmZDVTfWYz9rB7b6F6UdD+uYRrCr2PEsnMgwFatw9YcY3gp5VbsST4SKlV+PucTkncYiNgEuYWdZL5TSSBZYG+KsmTBAVQTZOyQSR3WhS9dR8EPq8ia0UxkZxIzZ55GCspoAR1fnyult9+Kf2iQX0/5mPqFL9PtiyPb7iYUL1D0KqZeQUaKi9JH+5HDEpoBhFFj/5FSCnbchIjKICRySjvoXebDMy2fuCaFijV2wt0WuiYIJt99hDhrkPVvX4YwqQgdIEFkrpuSV6oIpmi0bM7FUmNAPyAxv+Q8iaclIolRqo4P4uEfv0vWTpVn1y/FNVignLBx8IORNM0zoO/SMzA2SHzaAD946nMGrh/AoQ8SHBLA2K0gAFexhU1fTaKiPQXNpEFYjuku+cDcKZMd14+nMMqBujxcpYJZ+dUYEoN0zlBprEzD/3M35hNmvC4LQpUIpahoPUZONmcR16jD02Gj4nYrgZCB57+eR+ohPwc/GEkwSeDNBmOlGVO3TCDdTP0akPQakl7DtjOO5FNRil/1sqTyav66+xJSv1HwZQpmzTlJwrkI/nSBstOJudaICOh4ZOjnWDoF/lQJ81ELWds1wu1x/P5Hr2BSojgaI3SP0tF3Kpkxw+o435tG37gIV488hq0BOscYsA3po2e8hrlLImqCn7Qswp8ucH2QidEawj97GGW/XYM4Y2f1nusZdvsZhAKeGidnvylkwGei6y/5hOIl0vbIDNqqojmiyGGJk8+NQjVINF2p0TEnldLHXeypKKJreoSm66MoHUberBrPob+NxLbPQsssPZpeoLVYsLYIlkw8Qp87jsTTAnuTRselYYxdCkKB3MciRAcFaf98EA2X60nbK+Es6KNtliCQLNERddK4TKPyfjOvb57N+Y5UlmafxNasMjy/lVxnH31eC4+8cT2aHnI+dVP4zmoM39iJz+vHeVZH+6xEDDN60O1xcL4lDSRwNTnx5UcwDEDLsQxaF6pkb5YZsvw8alKYQKqEPydK5xiF1N0Kf62cwW9PLyTojGXTesapRLvMOM/pSDgl48mF3rEqA2VpjLnpNL1DJeadWU7bFDPvdk5APyAx+JgOhODyzHKG/2kNdUte+Mdukv8dcVGY8R+K7yUgkiTp703O/0950L+nZozfsRsB/3/RRy9J0nef8rd9/qVRu3w92cM6GJzbzqjfr+H+3VtJ3yt4IHsrgQSFz46WEdyVRP22PPyzh9LckhjrnRnppvNIGopOo/+SANY6HSZHCKctQNOvJpO7OUCo3EkoQSISJ2M6a0YOxZSbH7p+I71l8Ww/X0rVTWbCThAfJVL0Uzd5Je3YK3TUuJJI3y8IT/Sg3NKJ6ZyZhKkdqAfiMXWF+GDlXGb+6ABxbRKujzKpu6cYOc9H72A9H5waQyAziidLxr0rjcCiAc5/UoL/hUy8I0LsqC5GlxwgrlaP7NKzYthR4spNLIhrBGDtsHc5tz8fey2YD1rpHSnoKE/lqR3vQnqIqE0jda/MoG0qQ7PbuXX0XtIOQNeOTJI3WJjx6H6yXzqPFJYoeclD8lGZrZvH42m3YXQJFFuEwrl19A/XMH1hZ7b9HFJ6EG9OLPhKnNHOvndH4yoB+3k9eaNaeeLpH9A2VYckIH2/YMziM5j6BFGbyqipVWR8pifL4eY2RxveJjt7Px9J0Z/CLL5yPxZLCGtLGJ1fwnzICoogb5OKwSVja1axtAsOvV1G8ZtBrNYgBZtC7PmkjOm5tVir9OgGZNra4/GPDCBCMvkvgs4rk3JYIuVTE76cKKmD+pDiorw6+g3MHTLuAjPp+3wIJaZWPG3xCTJnNdN1XQB1wIDQJJYMP0HfyJjC9IMb30Oe08w9M7eh9wmWzDmIVQnROUGPalMZKFJRTYKSl3z87NMV/PAnm7n92i0Y3ILQ6j7Gj61igSVEmbOZvlVetBEeEssFbc8XUpLQRe5GiQ++mUjULLFs+W6mZ9by+Nz3CSUI5lx1hJ5ZAWwNkLqpBvNuG750heAsD4NnV2OP91Ma14GkweTJ53BWQrTNgqtIwTS5h45pGl23BFD6dahWlSse3IH1B21IBg17Y5TzDyRgqjdCVMZyyoy1pB/dNw5UI7jHB5k/5ziaVQVgIF9i68aJvDnhVTqnaLiu9aBvMmJtESyeehQkiYR4H1EzvLl4LZ2LQjw9ZCOJxxU0I5zxZWJqMCJ3G8jZFiT+UwuvbpxPyCFT/3k+J8vzUVWZ+Gkd+HKjVN4eF5vFJsH8rArcpSpRC/Q3xqP3CBRFAwFyUEIKy7iGRykY30T8UT0tcyUOVOWzbORxDLERgUStgs4pGp5+C0GXievv/wJTl4703TILJp1C7xF4B0HBO/0M+U0TLfME+z8eSSReg5eSuX/lJuINfsKpUT7bPZZ5G4+w+dHZ/HzVBrpUH7XL11P8uusftFtexEX8z/hemqolSXoEWAIMXDCZiAVIpwAX8DBwEFghhHjvW+ftBtqEECskSRpFrB9okhDi4Ld8GoF3hRA/lSTpSmKss3QhRMeF4zogCNwphPjfvnL8KzRVF95/kG1tJ3nf6+CNWVOItrYRnj8WS2UXgcJkAsl6+oZILFp4iC2fTURIYOqFyAw3hp0OtLn9iD3xpO33UX2jgaI3I/QNMeMZBJcuPEKdN4nzB/PQeyUCWRHMzfrYW40ODGX9BMudDJ9ZzbmtxaTObKV9fyYVq9byvtfBYy9fRyA1Vl6yN2g4anzUX2HFOrKX6I4koiawtghcJaCaBAlnJHpmh5B1GqMHNXPkbD7FrwTpKbPiyYNIYhSjI4iosgKQclwjapTwp8h4ilTS9ko89Njb/KL8CsIV9hibzaqBXsN53EDYDo56DfOtbWwf8inTy5cQiOix/9lGzzAjwWRBODVK9ucSnWMUND0Mn1zD6ZZMJCD7dR1N83VI6UEK0roJ/yGdS/+0k3W757Jw/Ek+PzqSxOMKli6VnuE6Uo9GaJuiI2lcJxIw8GUantII5oQAhm/shKZ5CLVbsGR5eXz4R7zXNYFD9bnoDVEsO6wEEyWctRq+5W78fiNal4mhZQ3oJJUzrRlsm/w8l2x8ADUhwgvT3mTN4eswnLWgGQWlM+qo2B2rKps7JIQOkq9opulAFnqPhGqCaJxg9cJtrN02j/izEr1lGopPJrFcEL62n3hLAIchwKbCr1hWO5eTzVk4d5jR+wRf/fEZphy9ifDJeEzd4LyyFZ2skWjycaQ+B+sxM9N+cIw6byLeZ7LoHKcQsWv86dJ3+PGhq9EiMnGOIP42K+ZWBZ0/xiKruW4da1on0uZ3UNWdTGDARP2Clxm8byWhDkuswV6FuBaZ9MWNeNdmxej4Tpn+6UF0BpX4zyw4bmpBQ0J/l4nzD9lBlUCTsFbpST4dpv5aICLjLNfhGhbFXqEjMNFHfmoP2iPJVK80UDaknr5gHOr6VFrmCtALilcdoerlsUg6gYhKzB5Wwd7tw7E2gaVbo3OszKAvQ9TeIEFIQedWKJnQQPcruagmcBdAwfgmpifV0BhMoMzaxFtNE+g+noreIyFHwZsXxXlWR2jWAIFuCwXvR+kvMuLLlBg2u4pznWlEIwqyoqE/bMPepGFwR+kvNqALCnrHRrGneolEFQIeI0qPATKCGM+Y0UZ7yEpw0bklG09hFMWroNqjjBlSj0mJcmTnYMydEp7xAQB0DSYig0JIvQaM2V7CIT01s16j7PE1uMcHEUEF2adAUgi6jTy44DMscohfb12Gziuh80mEhgW4a9RO7otvIP/rm0n/TI/1g0P/11Pyv5em6oJMMeiJO76vSwKg+upHLzZV/yfxvWSIhBCPCSFGCyFmCiFmAssvHLrvgu0QsUbof3soF1hmI4GvL5hOE2OQfdsnBRj0LZ/dxPSJvv1wywAF2P593Mv/rahdvp5Pl/4Z95ZCyh5fwzRTK58f2UJ0zhgsp5p5bc8GusYa6Rsqkf+7Exz8wzgY6uHp617F2qYiH3IQsYKv2omQIGLXo+vX0f1QkIhFImeLny9qhlB+dhByBOQyN5JJRciQuTdAqCAIgLUFml4pIlAQxvtuBrrhbvK23spf6ubgHRzC4Lqw5CRw/TpAOFHFVZ1AJA4iDkH3eA3ryF6clRLdE1X0zUac280cOVPAivGH6BtqJTR/gIhTJeGoDuWEjWhuEL1HImKW6F/iQ++Psc+MLpUXF84j/c8GRs+o5BeLNoFeIz7Zg2t4FH9BmI7Jgra9WWwPKLRUptBXlUDzXAODr6kg4tDQ20J0jlWovGUdg8a2Ur25iIQvzcj1ZhouV0g8JRH16un02Ljx2Y/Z9LtLGLRZY397LkN+00TiSwf4Zt2LrL95LQ1XCyj20V6dTL/PjKckgvOUnjhTmPD0AcJBHY5KhVS7h1fbplG3rgSlxcQjI7fgGqIhR6F3uIQsCTLf1mPukKn5Kp/q3mTMh+K46vcPojqiOBN8HPXnE3fIQjBVJWoRVHalIKkS8RWCYDIMjAnSdCALOSqhmmHwnGpMRW7Wbp2HrV5GvbKPhFMycW0SfcMkInsSqatP5XRzFkPWraHyk2JMpy3Er2ih70o/Ew7dgvx1PM4qDaNLo/1ABs37sjj+TQmiz0B8ZYQth0ZRcSab5ssEzirQ+WR+9/hKhNuAYlT5xdAtvLVwHUix9TF2ciUAX+0oY+CxLAI9Fsy1BorfWA3lNoQikFKDmLtkDAOC0BPptC2Kol/VSd/UEIlfm0h7y0TnNJXaE1nUNKbS8JgRQ4sBS62BuFo9y1buouuOAPKADnQarmFRnBkDnHp4LTq9SvXpbJrnWIhL8VH7YRFN59Nw5ykgQX5eJ7VPTcTQrkfq01OY18nOihLU3CCh+QN0j5K5cdEOmu6IUj//FQy9CmQHqN6bi3+pG3OPht4roT6azPb7prL/gzI+uHsBHd0OND1kzG0mkCJIPKHw8D0b0M7ZcGYM0HZnGEu3RsJ5jZoPiuGknRVDjpKb1Ie9UcWdJ+PN1IME7lkBlo07iuVvDmbnVJNw0IAShAfKviQy0kukwYorYCb5ZAgpLJO1K4pkUjnRmI03YkQJSGiz+8n6mx7bYTOWjlggacjyoVbaKL6nkbyttxJfFQZBrHRt0hCahEgM82LNVH5zfBG116xHFPoIZKgUXHcCixzmx+2jMdaY6JgkEZ4/lnPXPkfJS73/oB30vw8ulsz+cfj/U6n6t8DKC1R6gFXEmqHfARBCaMR0itZcaNCGmJ7RfmDHBZ9+4HngR5Ik6S6Uzh4ANvxHlPt/ZhTef5D5GaMYajCzKm8vKX/dz40509gVkNFtP0bd7QX8pmM2kTjBuZueR/s8CUtHGMN+G5/3j8JW5caXpcYk+iWBeVY3PSMMjJxSTTCsJ2oB9bF+wh4Dpk4dzkqQDjrAq0foBLbHWmJlgHfs+DLBfamPuEoDM+8+iNUUIiF5gI6KFOynjUQGxyqbS3/2FS6vGVNSAEveAIwaYNWiL/nBtP1AbCNP3ylj7JWIrwpiadRx4sahRM0SoaCekuJW+spUAukq6R8ZEAokHOpEOWHDlykhtiXSX6Sna2Yqtaskqt4uYf1jS9F16ylN7ELfrzC0sBVbvULawQg//+Uqfn7JJ0gCDC6J8o500vZKZCa6CSeqTHh4NQPBmBigJ0fiq5V/RPHLqHowdOuYN6iCJ87Op3OqoPFyidLELs7/NIeat8oo2nUjN3y1CtmoMj23Fjkok/hmHKXrfOh9gkBYT8BjxHTOTNQMfT4LTe/l41/mxtIm8cjBKxBWFX+GhmoU+IMGmucq2Os1dAGIM4YZGBrhuYeeJ+GwHoc5yEv7Z+DPEAidIPG0RDSiMG7BGeSIYNai44iQQjQ3SLgggBKEs+3phCodkBEkaoaEP8XROzGCPy3GUDLN6MFxWs+Oac+x9Ko9yFP7MU/tQX+PBbXFgu4bB+FZbgZyZFJua8Dgjo3SEDkBBpV20jJHIf0bKTYJXYD3igGGTK7DPc+PI8tN4hdmOqIOfvjJGvx5EYKTvBzfXULJnh8isoIk/LIRc5KfS5ceJJIYRTMIDL0KWr+R7M+6cRVD8w1RSp8YwGoIEb/fiGuBn75SHSNLm9DMAmu8HyEkTMNd+AvCBFM0Ptgwk9mDqihZ20VGVh/5RR0k/tnCkHVriEZlhpfVE0pR8bdZ8eZpxOf3oUzvQ4pK+MIGskfGWib1HplxCY2IoILUasLXFUf+tEZePzsRwwkrtzRNJfmERt5zgqyJrYiDTkx3tSFkqLlBTzBRh9El8GYYkNtMOGqgtd+BpEHf1BBrH76aUGYEV4cN26c2ALrGxQYj/+GG13lr91SqW1MIrHRRuLCW4OVuTAu6SNxi4tCvx9F5aZiukJVnH34eWZV4Ys9CHNviSD0iuCV/H03zDBAfpvPGIEqnkacnfIBJFyGQGUW31Ul4TS8ZX3TgGhPCVGPEttmKpMLo7d1Y4/003xil7pJXkcMSuR9r4DKQvVGHqsnMyK9he0Bh48QXeW/RX2l+ZDIRoXD4d+NIPRph9pRy3Pl6huy4jfMPOi6qWl/EPwz/J4QZnwH+XhZ7RpKkDwCEEJ8APwe+kCRpL7AUmC+ECP79XCHEs8QCpL2SJB0ECoEl32mi/gmxbNIh4AjgIRZc/Uuidvl6uj4p/Tc12FGmJpSdGWxrPcG9z97BtraTiKEeqseFqLxlHaXv3Ilyu57oo304FrTzdW0xN3ywDV1ykLQDPhzVEm5vbFr9icZspufU4M+LYNOHmDKkBikK3XNDBJJFTOyg1MvpQ4WE0yO0zRGMmF0F1XEULqxlxwsT8Yf1eMoT0cwqT9z5CvbdZqJWwamBbOTTNoK9Zq4vPIJy2M4rH85n4yfT8AWMxDt8jH3wGBnPHqZllhlHnUbtCif+TIEka1S1pCKpEhnF3Qx9+DRyGGpvTCPn5WqCqVGk+b2kLWkkaUUT8fuMuEoFCYe7GDmlmqO7S3EM7+VsTSa+cX5MHT7cBTLvt41FSwoTihdEqu24CmUaW5NIPKqgGkCnqNxz48ekHolwJJiBfUgvtuVtbF/5R55IPcnZSe9Qt+QF5o0tp91vp27ZCywdepKoT48cUJhdVEWhpQu9VyKudoCGxU7CNgmHJUD2JwqMdZN9WQMRVcFyRSeedhtLbtmF3GMADZSQROYujRsGH0KOSkRW9pG7uI6es8mMG1zHfeeW453jo7kznidm/42EMzBz9HkGciXiDliocyfRNU7mwOujsTTosR4xk5bsxlccJtXpwdwhoboMmGb08PW7r2JsNRCJV8n/0Ivu3QT8k33M3HEvn7w7jbQrz3Nk9Pu0zU5kzrRTRC2g3+0grk1w9lgu3lyVJ659C0kSdLptmNtlMu+pJi7XDXqBv9fC6fJcot0m1N0JmFe28+Ej8ynYGMSZ4kGrj6NwUiNaQxxSs4m6/kQCPRY2b5lA0gEdq5d8gaaHuiUv0PmkhGYQmEwRan6YRH1vArff9wkZiW6QoeqrAizNCsbNDlJfMxE+EU9cQgBDn8xfV62n4p4hdM1IJdnso2NHFr1DTQQyo1Afx9nWdJIPy4j/wd57RsdRp9vev6rqHNQtqaVWzpIlW3LO2djYxoAxyRgYcrQZBmY4MwMMTGAyaYZkTIYxGWxwxAlwtnGQsyxZOedudatzdVW9H/q8866zzrnvOnct7p2Ze72/SR9apf/TVf+n9n8/e+tVjIMini4HKdYQumGRIscgLS3pIEI0O8apK3LIyB+kdFIrgiJQ1+mm4GWBYEGcI+tHY13Zif/JIKomsGD5EdoO5WBv00jbr2Pm498RzBYY++BJ7l6yi9F3nEWptWMYEvjhhN20LwJrvQHHWT3ydR56J4uoqTKeSyI8vPVWrB0SUocJi0GmeX0xWX/W0deflDhCdkk8P+1Tqr8r5Qd77iFlag92d4BImkD3bI03GmfgOqlhNMsUugZBhcfev5UMkx/RLuO4tou+gSQGprvRGRWsUwcYmB9FV+nni09nMTGzHUHQGPP0KoyDIvbHOzB6RaJJIrHqZHojdu49eCs/uWMVk4169BO8vFo7G+3efgYr9Zx+ZTQZ6xth0EjSGQOVx0WkrP97JaEXozv+cfhfEt3xz4r/EzVDHfwwAAAgAElEQVREJZ+GkAaGURqa0e/OpHVzIVnPf4dYVYZjdS9NQ6kMDNixnTKhiYAIrks78YVNCFtS8M6IooV0TKps5PixUhx1AlzmweezkPuxDl+hHuOSPlRNQPe3VEJuEeuSHrrr08jdrjFQpUPVQ8p5hf7xIg8u3crmUck0fzQG3VkrsWQVfUAk5lDRZYXQGq2UvNPLso2HeW79VeRN7aBvYy6KiUT8w3kDYjwR5SBkRhAbzaSe0+gfJ0BumOQdZgZHa9y/cCfv1k1FliUctggDbU5SqyWMPpWYTST79iaavywmlKElNBAjY6AKTBvVwDhHG2t2XooYExBlSD2bcGHunwBXzTvCeX8GtTW5FH0u0/9QGH+3nSmjGzj/eTnRacNE/UaemLGZ3x28ggWV53kj98B/qsveCNxx8A4yNhiQ7u7DvKiZwbun4R2pgQA536i0LxZwHxAIu0TEBYMET6WgCwnExwbQNIhHdWRsNbDx2edYdPIOTGtTUG4fwCAp9B7NIJYhk7ZPj/2WTqz6GA27ioi4VIxZQSw7bXjGK7j3icSSEjX1djkwdemIuhRSTyYmg1yTe+GNNDoXqbiyfFS6urFKMbYeGgsC3DJ7Pwd/OJkRz9fw1YFxaHoNc0aA8LAJ52EDoUsCZLxn4skX3+beDfeQu1Mhkiox48dH2LxtClIkoTUDMLcnfGz0QQina8QtGkaPSO6uAC0/htiwAXFYhylvGElSMa9zEk4TiDk1dJV+VFWgZvr7LF76A+ruNZNyTIf7xlZaPcmIokbu3T0M/C0Vz5ANrc9IyegO/DEj/TVp6P49/DWWltDH1G0sQ5nix+0Ypsg+yN6mEqR6C/dft5Uzwzl8c6aCrNxBvAcykO0ael9Cd2XqB98ohbzSXnwbs7C3x+m8XuaS0gsc+WhMwqwxR8Z5Uk8oW0OKCIgxkG0alh6B8becZvehSsy9IjGHRvZumYH7Q0TqHJgGBEQFghPDWKwRhrvtZO4WEe7oo6vFRfHHcRqX6xBscTK2Gui7Moo6aECKiOTukmm+XsCV6WN2ZgM72sqZmNnO7rMjsDYYcDQq9E0QMXoFIi4NJVXGYI0xLqeDrmdK6JwrklLiIdvuY5yznXV/m0skLfF/6wNQeWMNoqBydHslpj6QYhqhDCHhyTXGS2DIgqBT+W7Oy0zZ80PuGn2QN0/OQBA0qvK6sOhiHNtTzsQ5tdR8XEE0Bf74g7/xZucset8rYPzKk5x6YQzcPEB/rYvyZ1rovaIQb8W/zv70fWmIcv648vu6JAAab3jyoobov4mL4a7/oij5yWHUZBldt5e3vlkLQODPOTxz/1v8suEoXb+BwYgV4WMXI/O7+eOqtwmUxAkUxWlpcBM9lMqqh78gbaeR4k/inNk5AtWsMDQ1ivUdJ5rPQMfNMkPjY8gb0kh6xo6gJbKj/GETjjoJxSTgaFIRx/g48MJrSJHEs8CxP5XMD42svGkLaALXLtuH3i9i2WtDHxCo+5WDP+28knhehPEp7VTdUIMUhYxdOrKeO8QbD73AwtknGZHVi6VXQBdW0XQaBe5BBsdqTJtay6u7FxAasKCqIgOdDsydOobmRfCWS0TSBM4eLSTtZARNpxGdEkAI6kDUOLG9goBi4ueLNyIooJg1xj9ygpBbRBcSWH9yPA2H88Eu03KFAevHDiZUNtEftqEYID/Vy4iibu5y9HDJqFq+rhvxX9ZntgmemrQRT4VEZ7+Tjsemo+rA4BPZfPXztF0Ooytb8OeLhLI0bG870AcFdGFI3mQh6WsrWkxiYJzA3KP3Et2fiOvwnkhD/KuLUbMbIC4iX+Ml8lom504UoAvB9bMPE5clAMSIiGIU8M2I4K9LwZAcwdGosmTaSSruOYd13CDePRkE3RLoVY6O/5Tx9jZ2t5eQuQ/suX729xfTNcvMluNjsOT7cRcNYF9vR/DoCc4LIpyz4y3Vs/LTe9EkaFsicvjpNWxpHIWqg5IFTVha9Lw8fy2aBDnfBrC3K7gm9vLAwh1ESqI0Pigh1FlZPuEYN8w9iGVzEsN9NgRVI2VxF3JhhGh9ErEWG4Vb7qHuPjP29AChDIHOjQXY19uJn3RS+5cC4ptd5L8lYmsTudDmJvlHGpo7SsFvj3PTlXt4/9LXyDANc+bHq6mZ/j5vln3AN2fLifsNRNPjvPrlZezfVYXeFkNWEuJuY4kffRCsXRqhTI2yd4L0+W2JUNvROh4Yu4fDXfn4R8RJatHI2iUhJ0HaCRWjB3QhsLfCcKFKntmDao9z7Y17UEwaLdcKFKcMktQM6SeiBHITsSZX5J8jbwsgwMBRN3qPRNtCI4IqoAV1XPWLr0lxBhBlAcWoYXm8E1O7nmhcYuPOKQS77CxKPoul0UCwIM59v12HrcLLtgeeRk2PYm42oDVaOd6ax0ClDiEO4X0uOtYW8eGmOdx8+04Wzq9GF4Ipt5ygzpPO/uMVxBwqvnKFiEvA3KeRPaMD20cObhl3GH2jmUuO380No47z/mfzaVrwNpoqUL+9mLr3ypGiAsf2lDP91momLT7LL89dSbZlCNdxL9/uGot60yC9vQ4aV6yhbXUKqW8cYt3VL/xfdYSmcVFD9I/ERYboXxCNK9Yw+cT1zMuq5+yyXLouzwURTjy+mmmP3E/SR4fp+fF0/JUxmi97k0VZY2l8ZhoNN7+aEKQKYOkSCEwNofWZMOcN47SEiX7qxluRiLCwdmnc9PB2Xl+3mGiGzISRzRw/U4SpR0fenDYA6muykUIi6WN66RlwoEYlko/qGS4CvV8gf/MQPTOcCbF2WQzRp4O0KOqwnuXTjvDpkUnY6vXMuqGab7eMR4yBoCQiGhq9LiK7XZgGNX79i3d45sFbaF8oYWsTGS5QEVKiZH+uZ/wvq9m2eTLRLBkhKlJY3k3rySwUh4LzlB5dSKPy3rPs+W4UxkGRtNNx1PsH6OxOxtRqRKj0E+mxIgVFBA3kdBnHSQPhtMQbfeEN9Zw8UczHS1+iJ+5gqTVB5T/eO5oNTVUkWSIcGrPuv6zT+ViIVfU3Yri0lbSDTk72ZJO81kb3DAHNHcVUY8bRrNI3Ed6+Zg0/fGUVRo+GqgfXDe10b8lDioLrmnZWZB3loK+EvXuryDiciGN47MZP+fX26zAOipg8oC7wMsLVx7HqEnRBkZHTmzjXmUn93HcZ8/QqwukaFAWRPSb0fgkxKpA8uZehw25O3vsCRkEPwF1tM1mScpqXW+fRWpeBM2+IQMiIPGTC3KEjqUXFXyiiVg0jnbQTSVcpHd1O62AK8SYbcYcCooZkk9FUgVG53fQE7MQ3uxiqVCn4Ik7nPTKG4zZcl3bSdiYTTQIhNUrytyaiKQKBEpmKF3w0PGnGcNJKuCpM4yXvMO2R+/nL71/hp4+s4vfPvca976/EfUSh7VqFES9GaH1c5PyMtVS8toqK+fWcaMhnREE3dQ1ZCAaVu8fv543jM2le9NZ/qNX4366k+slXCakxxv7tIeTUOGUl3TT2ulDjIvZqE8FsjakzznPgTClpOUMED6SRvSdE52wLoeIYSWcMIMLwuAjCgAEhI4LptAUpAvqAhrBsEMsaJ3tef53S91eScVhFiqq0LZSQogKL5lXTFkrm7MkC0Gm4Dwj0TgFzzjC6vQ6S2hUiDhEEmLTyBHs/H486yU+4z4K1TUfa/E569mcTc6pIGWE0TUCJi5jqTURTVKSwwK1XfMu+0Sb8N02ld5qGZlFYWHWO3TvGIpUPo6oC0UEzgiyQWdZP/EM3A+M09MMChtFDGLY4kW0C4lwPUVlH6gdWfLf5iZ1K/veoDw3ToMB1N+/m/W1ziNsVBFsc0wUTkdIIlvMmDD4NxSQwZvlZDn87CjlDpnnxmxRuuYfsbSLWdd8xvGIq3XNUbE06sp45+C8xgfZ9METGomwt5w8PfF+XBEDTjb+4yBD9N3GRIfoXRfLl9ZwcB/HWdoKzA0Sdid8/8dt3aX9yOkIcij7UWLJgOW2/mk7xTw8x80f3IafGEaMClj4V+wELGeV96CWFyGduCu+8gLlfIPmCihSDTf82H3MfiGGJcz2ZjHgnTCQrzuqSj/GELcybdI6kRrgl7zusJ8zoexMbasmHPmIOjdizAe5+YBMGn0ZSahApK4S+2QSSxpZPpmMY0LH3oWdZnX2YWEmY+JgA1m6NEms/w9WpGIY1BqbG+cNjt2HqD2MYEslf1sRtc/cidpnouCbhbaIaNUzteko+iDCwKQdTv8jcqloiKZB6yo+IhrNWIFIYxVuiY8BvZVxxW+Ioo8GOzhVGkzQKJ7VDLHFLCKpAxn4vZw6VoJkVft1yFd/4RwLwui+LDU1VpL9uxuO3srR+8X9Zo37VQktzOnnfWTnSmk/uLxX2vfIaDTetwXrCjKaDgbEC6UfhoWdXEUnVSK4NYbi6D987ORRc1UT29c00dqYxELdzvesoH1//AmJMw96qsbZzGpo1TiQrjm+EQuiCkzZ/MlJIJKkBzpwoRO0z8Zv+kYQyNabOO8etI49gdIUR4hBLVmBtGnJZmLOx/+/FaM+BSoZVE70+O7fM2o+30wFA6XtR4lYNX7FIODtOwdMaWXtDTJtcS11TJravbDy69AtK18Zw53iZU9TAdaNOMMHZxhNlW4gu8KP3iDTfKKDEJdQpPlo6XFw26wR5OxRsR814ZsWYeO0ZplQ24huVzM0jjxIslqnK7WLJnGsIZIvc+8qDJB1p556PVoKW8MXS9xiY97cjnJ+RYEujLoWgbMSWHCLFGEJvT+SLff7qJfxl5if/oU4j9t1Kyc0XmHbqWsa+/xDuoyr2Oj1D7+Zy86ijpOwxYvQmNvq24WQkvw7fSReaCI3XmQiXRzB1GAjkqwyXKLi3GZg7/SxSsxnn3B7UOUMoV3p5aeRHtC+PM/rZVeiLhtn38mvETSIzJ9fgHDnI1sNjOVObixQR0IwK3hFiwvwyYMQ/Nkraw02J78xMmW1HxhCqiKI/kISgCeS9VU//19nESsOI/54TprtgoWnB28j2xNGkkhfhrUOzaPpwLIEcEc2iYG4y8HV9OYop4QBu32YDk0LRFzLhDW6++9Or4IoiRgUmZbQjWwX0AY2x7k6sphidCyDY5EAXgviYAGknNIJ5Cls6RiFFBIS4QNZGPWkn4yytPE2wJIYmCoQnBzm2pRIlNwKCxphnVlH0scr8X+6n90fTE1SJpBEYEWN710lKPwx8Pw/Oi7iI/x9cZIj+hVDyk8M0/2Ea5TOa2Vi6jaKdd5KzXod5wxE6HpuOPCZAdqoPw6WtzDkdZu26+dx87Tc84aql6rubsK5LIuQWCadpxFPipGT6WJRTy2c7ZiSSqfUqk0pbGPppDoOPhwkdd2Hp0ghmC1i6NWw9Cp3XySwoq2XHuZHYaoyoOhh3ZQ1HWvN5aMw3vPjFFSxbcogNm6ZR9EE/oZJkem6LYNtuw3FjJ3Z9lLq+dDhnJ5IXQxzSYRwUSa1R6LhcwXHaQHBqCJM5hrDfSdmyC4RvszG0WqT7Qhq6kEjmAYWumRKu0X30NrkQogLWIh/DXguiTsVyxkwwW0WzxkEES4OBUE6corIexqe080zGib+v6ZinV1F8TT3BOQP8tukIKzb9ENEVxXLcgntJO+37c7G3aNz+081441be3T4Pa5fAY/d/xK8/WUE0S6aksJedFZv+y5p9GnDwVvsstldsBmDUS6tQTICgIds0DD4RR4NK4NphnB/YCN/uxXchhSVzjtMRcnL6uxIEFaydAr5xUVZN2s3ahsnoJYWr80/z1uFZ6Ad1lE1rIcUY5GhHPtFeC5pRxX5eD7O9SIKG32/GvcXIwGiB8hnNnKnJQ4iJmLIDVGZ0c7wlj8ZL3gGg8Mt7aV72OiXf3oHrKyP29iiG1kFqHs2gqKQHz/ocgrMDOLdZGSpLiL1tkwYI70swep5ZUQrWCgxWGUmvDtN0D2R+YcBbJpExv4P2ASdLSmvY0VxO2GvGeUrP8PQQ5Vm9nGvMxlZnIKlFpXeSwMjJzSg3S9Q+kkv5i91sObCB2WeuJsvm49yGcsTpXvwDVuw1Bgw+Dc84FWurRP7lzXR+Xsjw9BCOPWYEBRwrOpGeTGboiRAPlXzDdbYenuidjCdmZXdDKRkuH5Kg0eNJQtKpGA7aiaRqGPwC6cejdNwrIwgaoqiR8rGVmFUk7BY48/BqinbchSM5yKtVH3DTllUYByQiGXEwKeiMCnpDnMsKa9i4awrkhklP8ePdn8HIRRdoHkoh+TkbQ8VGFKPAJXcdpuaucmrvtfHYvM08vWUpqkkjf0QPrXUZaNY42Zt1xE0C5tu7aTubyZWzj7Hv9UlISweIfJ3G8MgY4rAOTdLI3g2d8yH9kED6PS1c2J3wo7J2aYRdArJDQzGplFV20HA8j6IvwzSuFLCcNiNM95Lj8NHtT0LRBEJBE1KLCa0ohNppJuWMQODKYczGGOEjLiIZCmJyFNthC9YlPUTXuRMatllexK+T8U+MYGgxEc2QES1x/jhlPc8/dSPBDJEzP1lNxWurOH/fapaUz2bp4UZkTceDya2E1BgW0UDxx9+vR8/3ie+LIcr+/ffLEDXfdJEh+u/iIkP0L4L/Nwyx8PFDiP+eUGOqN/H5y3+h5+HphEpi3F+5n4GAld80HeeItwBzn8bBmekUbr6HrCQ/vTNVFCPEnQqiOU7oqItPayYwamoTALlZHo635NFwswnDRylkzexAUECMgahA740Ril4DmxRFb5YxzRnA1qFx6Eg5Jb8O8krNHDQR1u+YhjYiSN/MNKJJEtoFG4E8gchrmXxZup1YuxVV0tCZ4thbRLIORrj595vJzxtgyg9OYDxtQal2kvPWOaqb86h9MIPHS7YiygLmXoGhu4axdAkEowbMHRKmfpGAz0zFswEa579DqCpM5gHIyPZSlNdHzKmBUcVtHub65CP/YV0jLg1v1ELLU1O5+ZMfIcQFSu+sJesbHz1+O1IEzCt62Ng9hprhTKQYyDb43ds38vjyz3Bl+Bl+O/t/WLcv+8ezNPPU33+29GiYx3qIl4RJKx+gYn49dz6xkUiznegdXvIdXjL3a9T53NQPpqGmR6mY2oxpUMVSb+TzPy9EPpFMytNWdj80Hf2gDl1Q4FxtLvUvjUTTIKe0jxHFXThaFCJhA8MBM09PWUcgS8ReNcjZM/lUPD/AxAn1xGWJ9mEntmNmCjfdQ9GOu0g+I1H0+X3QbcTWGaPpKiNNt+Vgr9fR1JBBIF9DHjIxtDgIIky/7DQDnQnvHEdzDF2nkeZrJGwdCg0/0KE3xNn/0mvkLGijKrkLuzXCwZ5CVFXA1K5HiGvoay3UfVdAyhE9oQyVnqVR7M0C4V9kcP7RHHZc8yxdS7KZ8OuVBGN6qttyCY6MEqp1knJEj/vydqy9CvZ6iax9AdqHnITcGqrXiL8ILP0KPT47QyMsSO+nsmVgNLc0L2bzl9NwGQPoG810tbjoaEhnYl4bkX4zl9+6n5QajUBZjPZ7ZBxbrUQ9Zizb7AwVS3gqIZSpUvbeSgxtBnxtDp5qvZJrZxxBUCAlZwhUgewP9IhHk/hy72Ti6TFy39WhvZNOODtO4ydlKDtddM41MfX+aiRZY+PXU2i7zImpV8cLa5fhruwjq6SfQNTI5HH16HsNiIpGIEdk+NMsxkxoZN9rk0g/5KW/10E0RcOYFEVQQDNoqDoBe7afuFGg9mAh+tFDWMcPINsEkmf18INFe3AdF3Eaw+jCEHPq0TwGgrkKt5QcoX9tPoZ1yaS8YWNeyQVUCWi1kDaqH9UAcVkictCFNH4IQRZQgnp8I+MM+Gz45wcx92sMD1pRLhmiYK2APgiF6zQurzjLr9fejOP9wyhTE56+d16/nUVZY2lbWckr717Fg8mtFO24iw5F5q/eAkoe+Y5Xl771P7jb/g/BxTCzfxguNkT/AjD3Jcq0veskgtHIl6XbWXjtbShGjfublzHj5moemrqLr+6fTag5iZ/+2yrqvilmaFqUpkcqMXbr6fg6jwXjzhEdFUYMi6SmBFAsGvZDZlo+Lyb/SxgYtlL+xCCGQYnRD59CViRkm4ClLzGFZTVHUQ0iB5+bTDymI/2BMLqIyg1zDuJ/QSXJEiFtXC/6YYHD09cQcQlc/thupPJhnrrxA3qmCYx8dRW5uxQKZ7QR9xvwjY4x4flqntlxJb4vs2h4dCT2Ob3IVo3zz5WihSVyvlb5ySd3kL1b5Y67t6KoIsmXdxEOGTH4IZyj4HIN03x9KpUvrELXZmLMT0/iHbYwwtFHPDmOs9pAV9DBZKP+P6xtxmGFb0dtQM6PUjqlla+ufY6WR8fTNc+BctyJpUdjhLOP+nY33aEk0o+pyDaNyOgwv9p7NZHdLoKZIk/0Vf2XtWvypdIUTuOS2+9m1EurCF/pJ3guGasliud0GmcOltAWTcXaJuKpTyHw8yw6lsZpOppLsNMOPj01nRlEk0V0QQjkCGTN6qDxegO9E01o+WFUPVhadag6gUdHb2dv1Rdc7j5Lx0INY7WV26sO8fbCuQSKFSIxPZo1Tv9MN23+ZFKdASKb3PgrY4z8Qw/Wc0aGZ4Wx5ARw1gooBhFnnUC0MEowR0W0yuxY8QyWtCBlGf3IaTLfHh2FPilG3KIx9PAwSl4E44BEyC0hWeIATDi+nO0Vm/lr5jE8PQ7kLWmYjTKWSQPIdoGsA1FsbQJDMyOozjiO/SasS3sI5JowukP8oXsxQxNiZN7UwpDPilhvQQtJ6IoDFN5ST/CNbDpvihFzQN8EG6ELTmJpCuuXvEg8Vabv5jByXRLDeQI9C2UuvFvO0foCokVRPjs1AdMAWNp0SMMih86U4qzRceyBcZQ+WEPSGQNLSmvwXx5ANyRx7KlXSTuVmJxLL+9HkxJRGVJqlG5/ErvenkakLELgdCrZX0l47wkQTdEw94iUvKXQMVdPIEdEiImgJZpyXQD2dhYRS0pMpZkGNcwTBglnKTjvCBHclMFAm5P6v41AiEPH/ERjYR5UaX+3hEAeaEY9tloDzgsQHTLRcNMa3PtFumdrpFhDDE6NU/K3ASJ1DrwtyYQnhOj1JLGueQzBLIHmNSOIuuPM/P1hNLMK9jhvbliIp0pDvs5D1w9ifLt3NEavwF+ue4fCJA+DU2QyPjESc2ikvm7FMCRS8fQgI94MkbzJgqaKDCyMYOjW80jFLtSfDRIcGcWfq6fS2kk4V+aXTdXEWm1M/en93JB0ihtru4iNDXLm4dVUPb8KxzEjZXorO66ZiH9rEU/fcwu60EWx8EV8/7jYEP2To+yJ05z90WpGrl7FoqyxbGv+jkVZY4lbdRh8AuHFQVZnH+arUU6GSs00rlhD4DYfkRyZ5OQAql7jyRWfoAvC7m9H4/rKSNoxqErtTkxZLfCimKBjvoRaY6d1RQ7GIdhRXUVHfTrhTI2hMhguAm+/naYbBYZKRIwNJgZnZqPqBNZvnEl3v4Oq1G6GgmYUk0Z1zI61S2PfaBOcSuK9OdMoGdOBbNNou0xEuluPaJPR9+vxxKyoFgVloRf7rzoIbXdjGhAofUvGvU8i/xe1pE3opXuGxJpzswgHDfTtycJmjfDYQx+QnOfFe85F2ok4+iBIEYHt+8eSmhSk+Y4CRhR3Ybqil9bG9P+0vl2zJSY+uRItqGPriK1ctveHRNxxUhZ3EclUmLTqBLuqR5HzhYRRihPIlDCU+BFbTaCBfvYgtvm9bHp31n/67D8MjODw2M/Z/foU+u4Pc+7B1czNbUDTQeZvRJ5Y9hmjpjWx6Z1ZDJcqpB+BhvslCnIG0AcEcrdpGDND0GkmPHsYw6UDWHo0mhoyMGaGUEwwIqsXYeQwqXO76Zuq0Bp1AfBgciuXTziFbIO3v51L/ytGjL0S4YCRtD0GvAsixDamEfsinfwbGlk4+hwdV+cSTdFQ+k3odzkYnKiw7e1XkaJgP2XE2iGS6/ZiFQUsRpmBkBWX289nV7xE0jcWrF0CUVmPocGMbNNQDJC8y0Ss20qKJUzZntso/uYOdIM6VAOIokp2kj9hRjnCSFJbHGOdGTSIWwVuyD1O8o4LiNV2Gp+q4NXZa+n8rJAphS2oBsj+RsBuidDqSyGYISJJKpZejaTWOPU/eBUhJnDzGz8md7OIVmdDNWiY+zTsZ42EXQJCUIfg0eP61gAChDMV1lz/OpnfiihGUA0SxzdVEihUOfjSJEZndRF3KEx8ciUdl+gwl/iIfZmOqtOwdAsk7THj/rVENBkch03YRg8yUCWhHE6G3DDZfz5I+Ak/9lZIPx7BWejF2qNg7hVwNim4nzUSylQRVAhlCUQPpZKc72VoZj5pJxJRj54pMn++YS3Fn0RxX9NK7ySR8BV+4haNiMtEoESmf6JK1k6RUS+ton9xlLQjIuqLbpyn9HT+UUfZ690Uj+xCbDajBPVckX8OXQSS72hDtMY58PMp3DNlL6ZGI7JdRTWruP5kwmSOkXomYZvwyId3UP11OWgCoqyh5kRoWyyRNq0bIa7g+20YZYUHUVJxHDDhmtjLUzuuxvJjI+YLRmJXDpEqBbA26fnJrx5A7xPxF4gsevtn/HX1dVyY8x5jnlmFLgTpLx9kUdZYtn77OdGN6fgKjZy942WKPwv/L332/mPw/U6YXZwy+5/DxYbonxiNK9aghkIsuOlONt3zNJErJrMoayz1L07B4I0QKI+hRqJMPnE94asmM1wgsLj2ctKeNlFQ0Ifr9yZ+fvUXPPeX5cStYBoU8I4QGM4T+e7L0Qhxge3j3+De27cgRgWMHoFgXhz/qBhZXwtk7BNQdRpGj4CcEuebBX8FVSCSFWfm5aeIpCYmXorf6cByysyucxVE2uwUrvNzz5a7GRyroSsqQIpC891FXGjITEwgOWSkdyJkfmHA6BU48vEYUo7rsBhkTp3L5+tHniFuBU0S8YwUGLzOxuCwFfJDTM5txXzWnNgQHVr/a0QAACAASURBVD6eWHcTnnYn2bvj+Ap1FN9wgbzNPjL3a3S1puKvcOB7I5fBk+nohiXK99/y9/W9o20Wpn6BsFvAne9hzJ9XoWsxYU4L4f0qC2OvxFfHRyMFJGJ2kboL2fjKFSxbkognqeh8OtSdqTgfM5JxcJhl9Yv+Q/3erZlC0c478YxRiJ9NovyNVWw5XUXcrrBtywc889Zyut8sIlCg4jwn0jtDQ1NEfOuy0E304rkrgN0SIXuPQtRnQvsilYGZMsmnJCrcPWTO6cCzJp+cl3WMTe1g1rhaPqkfzytDuYz73SqGZDMj59Vjzhlm8EIqjkaVO8cdRLqhDzUuEJgTYvSdZznf42bfpnH4y+OYewWarl/D0IQYzUtfZ+QnDyLd0IdpUCM4LkzXoIOlZ27j6rxT9HY7Ce9zceeLD6O/po+hkXGM25J4bMWn2NpFhsviBJYEyN2lsiLrKHZbmFG53dTf+irhDI2kl5OoPVCIPghDo+N0zhGxdmlI5jjhdI09nlLa38wgnKnQO1HPQx/diX5JP12/L0HJjhC93Us4pkfbkEpqTQzrbiu+Uo22JVDx+iqMmSFiTo3FT+0hf1sYvV+k4KYGhitiLL7uMO7iAaomNOOpBOcVXZh6JX7y4n0oRiFhQDrGhKBCyYdBfvjoZ4TiBiZVNTJ0SYT00b24XzARWeindGw70VTwVqq0PwGF73cyNDqOtC4V81gP2iQfv52wgaFbpmH5jR1p6QCds00AiPf2kVYdJO3hJlofSLBIlAewT+pHlCFyyMVgpUT97QYKNmpkZHr55ZpbqfjrOerqs5CiArG6JDIOQ9YTDbj3Shg9El0LVCKjwugbzQyNgO4ZUsJJ3Guh8U922g7ncOnCaipKOvm0ZgJoYJLiJO8x8e07b7L+pUsoWdBEcqEXQRGov8NAPC6h6BMBtpaxHtKn9CD5JTpultk88xWKKzvp7HMy9KqE/bc2tA2paE1Whgug77Qb53kR1awnnKEQOe/kW18FZ3+0msHR4DqjEM6XUQwa7hcPUvWXVeQva8J4RR9tv5pO8NopjP3jKqqffBVNhCuyJ2D6Uy+NK9ZgGri4jV3E94OL36R/Ulwz9zsCasLEu+l6PRYB9rz+Op7NZZT+6Du04+fIzvGwuf0I/T0O9r76OgUbfTQfykMKRFFediP9cYDD/mLEOIQKEhR/LFsmtSaOtVPjp8vXM33nwxwaKkJQIOOvB7lu6lFcGX6650DPrEQYaiRNBWDpyz+jpLgH5xkd556vQrZC32VRzj+SCYCl3oAmabRcnYRmVUg+KyBnOgmnacQqQpi69BiSI2iKQNNgKqZBGUuPhmyDyEI/va0pWDp03NpwPYYhKH3+POY+gbof51GcNkDe6xLffT2KuAWimTItXxUyfnYd9gYdPVN1APQEk+j7TRzvTQEqHm1g3i8O0jsVUsf2YWsRUFpsFG67m/I3VrHnyEiMHo24WaPfY8dw6QDGIYGUj6wY5g9w5/LtmDt1CNlhPJeHcZ7VkVXWjxTTMHdKOC5AcFqIjkudBJ4KcupCHpeev5JHusdTsvt2bHusFL4n4C7w8Pjyz0ie3AuKgGiXmfjLlQSK43iqQBPBtqyHrG9B36PH5NUY9lhxv2BC+NhF5xwJW52e6BU+RJ+O2Xcf5XRHQrfUPV/hkbc+ZPv2iTj1YaTDSbxZP4OsDS0cqC3h4ZydWE0xxJhA8Do/W/40F+39NMx1iQ1596lyHhm9C1t7QmyQtaaaEe+sJDdnkDVD2SQ1iNxXuI+4CTSfATlgwDtsYUvXKBAglKVSuKyRNEuCwfDPDfOn95fjr4whhkTyUry0XaWy0zOSfIeXMxdy2RIyoUrgGWlAigoEx4XRDUl/HydP32hCccdw6COoR5xo9jiRbBlrV+LeiDok1IgO/fspBDwWBqfHMPaFUBYNobpkbp+xH4MXpON2TP0CH344n2iyAcWscbIlF2lIx/pT47ky+ywNW4vJ2x6jtSGdmlWrKbymMaENytLIvboZ2abRO9XOU1uu43xHBsdOlWA+YebO/IN0zTQhx3R0bsvHfVTGnBVAOOKg/p4sbp+2n6EycP9Oj3jYweqfXU/lA2douMmEr9pF3Krh7bMT+iyD+jsNnD1YgumolUBxHOGsnejONKIujauu349xCPQDOrpm6fDtd5NSK1P7o5HkbBMRq3ysWvoVPUuj+GUT6o2DPLR8A4JZwVhrJuqOoxg1zOVDWDclkbtBImmLDUqCbP9mPHWn8shI9XHqZ6s5dS6fwXEKI95ZiWdWlDNN2Xg6nJg7JYSYSKzZjmV5D0tHnSZ0IpX25jRM/SJlvwmw/MTdNJ3JJikpzPCODIafDJD6xiHiSSquUxrOWnA0x3G+0IUxM0TuzhjH/jqO4q/vAFGjZ4pIbv4ARVPb8GwuI/NAEIMYJ8M6jPuoTNdsgaT2OIuW3YKvJOF11uFzUPbuSnRhGDOx8X/DU/l/Ey5qiP5huNgQ/RPCdUJg/d4pXJszleC1U5g4poG7xi0FIOWKC2zvOgmAdXETI3bdg6nVQOGGe2lcnsSImc10/gb6x+no9Dk4sGkMaUeHyNgtkdSisLDyHPqAQiBHYPVfryY3Z5Du35aQtyPChdWTOfGTcYT3u8jcA86zInqPhK1dROfR8cx9b9F2KIdArsaO5/6KYtbQtSSOjoKjIxT8rQVDRoi4RcNWZ0DQYPDREIpNwWyOYenWkCM6bDVGYlEduz58m74ZceKjAhi+cWBOC/HRvc9zvj6bQIHKqWfGEsxRMfeInKvPoWu6CQQYOa8eU6ceKQqndpQjRUFO0pDtYNXHGPJayX5eT+1fivhk+0zGj28guMsNizzMnH2WuybuR/fvbHv6+lqy9ss4DpgoTh4gNCFE3wSRgd4k1r69iEhJBEUWESUVf7FKX7WbYIaIyaNhb5eJ+w24Tsfo99qxNOtpqM9E1iS0PhOBPPAVG+htTeHP7y3Hd8ANehVtyIA+qGHsk7C2C5h7RFJMIXyFEjnfxFB1UPGMj/6xJgI5CfdiQQOl2omz1MPW+pGMzO7hyszTIGq82zsD2+hBDvUUEqiIMdRvo/nOAka8FOanv15JVNah5YaJn3DSuyjG4WfWELdoyD4jqcd0rKtIR1zej84vYdzuIG1CLx0N6ewYGIligtuT+jBd08ucCTVIVhl9tY3AVxlMKG3B3iJSc6CI2gOFzBpXiyt5mOSZPRhsMVSrgndtLkJEoivgYHHaOQS9yuOv3EnZG334q2I4pvSR+4EOxaZStD7C2d+N5tOnn0VTBE6/VoW5TyPrKz1JNXos/Srp1gC9i2NYUkL4rh+m9C0ZmzNM5dvniZ1IRouJfLx+Lv6RMlGnhjjDy49vW0/bMhWjV8BoliErwqiiTta1jkHVQ+vtKhn7RSoO3MLpswUIKuRN6OSOrAMUf+RBF9JIPwoVOT2snLOLYG7iBUG2abi+NJO6oIv2+TpKXQMES2ScYwb4YOscZl1yhvqbrIRGhxn/y2p2H6pk6fTjOOvBWQdjStvxlYEQFjF3C8Sn+7G26HBO6yWcrqEYNL5oGMNNd+5EHxCIWxMu55N+e4zhAjORu7zcXHqM1RsuY3ZJA3XfFRCO6TnqL0TfaiTiVkg9LvHeNauxm6LoIhodyxQu+8lekqwRxIIgKSUefF9nJJhTg4pol/nNdR+jBXRY64xcPvEUxqmDYJPJqOql85ybWl8ijjLz24QG6vzDKciyhJYSQ69TCOapvFD+MX0bykGFYIaIPqjRukKlZXUZeSleWu9S8VYIWM6YyR/TReoZDekvLnxRE/5TqVz31k7K7b2syv6G3W++gbVTZMRj55A8ATS9RqM3lbSldTxxzWdk/OUgfS8WUbBZ/t//sP6+oV00ZvxH4mJD9E+GM8tfJPWrC1i6RLZ3naTo386TbgzQc/0IloxbCCTGor23TUv4c+T0sfWup7lrxl4KNkdo2FVEmi1I+rE48vFk0k/I9P1OQYzDwBiRM8+PofO+GLZOjchCP53n3QQydXjLTCTn+Gi+W8PWoeIdIRGcF0ROixPIVyl7pZ1VB25m963PoCsOcOmZmzEMCbiPKVhyAgiChvnjGEKNnZQziSMHe1sMb0syUpJM0G9ClMH1jRFBhXifmeJP78e9T0LtsCDGNYz6OCve/AloYOkU6V4ik3FYIz5lGEuzPtEUFIY5UVOIe0YXkTSN/M0+hirjzJ56Dr0f2r1O7NUmGpabMNcZUbMjVJ8qxtynEahLxhczsfb8ZELlUd64/E26b66g954IxiGNUzvKSdtkIm5XMXTrMQxroAnoDAqxiD5xXd0CcpJG4Q/qsdQPAGB+tAupzsrRVX9FCors+nISuoCAoMCVP9xDWVkXpYsbSZreR+Y2PeZOiZ5ZiU01mgqiDANhK2d+shpfsQHPKIHepwWCOYnw1tRTGjlXtBC3aPyqfDO5riE2lm7j4eQWXG5/wkV7hwttQyqWRgPNS97k3hVbGfXmeSLXDCF8nYwaF3HP7EIL6ij++H50FX4QNQanyvQ8PJ3J6a3oggI13W6uy61GGhZp9LgI5qiM2HcrwzszOLS9CrM5RmxcgLgF4qr092iUeE6U/Ucr6Olxkm3zoSgiuQUDeKo01ix+m85+J/c6EhRPYHyY8z9PZlbFBSLb0umboKfwizjt8y3IVpHWuAV9nx6uHcRxYyddlyoYfRrD2RKN/S707Ubs6+1oJx1Y/thDYMiMrElcvWw/xAWiLgX3PgkBMK1z8s4TV5F6WE+gRCYW0aEpAs3bCwmGjdy9YhtORxDZIhDxmtD5RawVXjoP5PCnC4tpvCkFs0fFfHcXNS1ZrH1nEfqMEGuevhpLd0JPF/lbBqpZpdGTmvj7O9PIGN/DtxfKGDG2Dc1jJM/oQRcU2PPOZIQViTDaU7V53H3ZLjSDhr9SJvlTG+GqMEZJgcIQdy34lqRNNmqDGUSTNZ5cvB5NB1vWTePDPz5LijnEhw0TST2jUf1RFUJ+EHGvk727q7j2igNokoZntMrtnz5AYGsG3XNVdCaZrzpGomxORW22MtDuBA0MhoRFxZj8Dk6Hcpk7/jzBQpnq58Zh+iAZfbuR3pNuDH6Ry91nST8R56onviaaomGv11H4lEzF473EtqeRdgzueeEhUv9iwdQjkbOlj/6rIqQcMNB3qUzzkVysR81suuVZzH0aLaez6J2p0nNXhOE9bjQBXn/+Kr5aPZP7dt3BknELOfPwavbuHI3qsKCkyPyqfDPyggk8XbOInoenI1tEjG0eFsw89feJ3Iu4iP9ZXGyI/onQuGINr/vKqH2yFOvcPhZljaV3mp+Xs79DH4KCzX46fz6d5mWvk/zeId71p+Ndm8sl237Me1vn4fhDOwVr6nmv7ENkq4g0foiBSj2pfzAznCOiFIVJWldNsj1E33SFaMTAsllHCGUKpNSGsb7jRB3W0zdFQz8MS0vPJBqROAzMy0WLSiw9fSdyi434Z+kEihXal2gE+6yoUYnqxnxUncbdj27g9L+tpvl2jdQTIhmpPuzVJjzzI0RcAtHJAYyDIslnBXxFIq6RA/jmhAmfSiZ7X5i0nCEqltVhbDISvs1LtMeCbrKX5TfsJi15GNEqk2X1IcoCTf+mI6twAIMYRzFB/EIiYgRJQzWCGtAjJscIuQXKXumg9f0SzAdspO/S88O19+G6uh2hOomYXUC2qyjGxBvVosXHMAxrFL+lEu8zU/aSTFr5AEcffQnruEFOHSql5wUDYkjE92oeqDD7Nw+RelLgjTtfJmVCH3KSyrvHplOcNED99mIEQcN4dzdp87qwZQYomtXKtcv2EXVp2AxRrrhwGUOzIgmW6CsXuqCIOmeI4XyR8w3ZpI3r5Wcf3k7ngRzG/W4VpWtXEtnrwrhKj+myPnzzwsi2BEf++odL+OLgJKwfO9DmeynP7aGt3o1hUCLlrADHHOg9Oory+0g5H2PrhVHox3kxH7bxcHIL5IUJNjpQk+IYv7ORuaQNxagRCRuID5r51R0f0OxNIZ4cJ+YAfasRU1aQ3CwPZ3sy0XpMdJ53I4UFaqLZf/c4MlhiWE6beWjqLvadHkE0BYoWNNO2yEDcotG7KMYte++h5H0PNmOU5i4XY8raGJgbI3tZC3FZIpYhM3h5hHfvfAFPxMKowi42na/iD+7TGFIjZO2GoTIBw5BAfPkgXfMg5BYwt+lRoxL5ayVCBTJF97ahIBA94GJwosJjs7Ygu+L4/WZSalSsqx2knVDx50uYbomRvktPcFwYo1Hm0cc+IJyhEa8M4hkloHOFmZjZTtnKI+iHNcw6GUOLkZZvChg9tplP/7SQWIrCcL7G5qr3EGXI3wiv7b2EguJejI4InpEiUruJ9r5kst/Vs+V38+ibFefEh1XY2gR+t/kaIpkykUyFQr2N0CvZ6L920DcZmOclK8XPiGvriKfL3J5yiLytYO6WkGJw8tHV7LrseYR6K/1tyXjHx5FiYG3R8fEPn2NCRgejijr5f9g7ryipzmtbfztUruqq6pxzN03OGREUCBJISEiycg6AkGUr2JZl+8jZsnICRSuiCEgiCEmAAJEzNNDQOefu6spxh/tQd+g+3Jczxj3H9rmDOcb/WKPC/uvfc68115yh2f3s+tsM3incA7pAz3TonyjgPqfjbADLhEFeOjUPg1/li6cuQbVqxNJ0Bie5qfl9DlkHA/iLRaSoTjjLSKQoQcOdGShhmUAhFKyXGDGjiaV37ubKQ8tJ2ASkmMDwFzzkvyRz6XWHqbtjDfbruokt9CNGROzrVUq23kM8N0HtKjOi18BvXruD9rtU8q45y533fs30h46gvp6g48oUynfd8T+bFF1omf3LcIEQ/ZvANCSyIHccX/7iMpquew3TmlR+0Xj6xxR7dKj7xQj2PfgsC5beSuy7Yp46vYDDf1mDyR3F2i0QVQ1YN+g80z8XY0AlEjFSsPo0LVdZ2f7zpxEAKc1Nb68L2S+R6gpy7DcTUU3QMddK30SRlPMyul0hOCnCgb9MwTx9AF0CdLC0GdC/SGPV5VsZnBvDXS2Sct6ApUNG8sqIBg1rt8BH7VNYdPlN3DTmCN5KMD6Xin1RDzlfGjEEdSbmt6NUhZmx/CjSOB8Ad4w6iBgTkEIJIjszONmej2XSIJomkl7qwd/j4NDcLPrPZmA/ZqE7nEI0W2FCYTt9J5OiYOPMQaSogH9iFHO3RMKmI0ZEpBYzZo9O+7ICQvngXtIJwIyF1XTsKSBcnCDjRJDsgzrpWxrAoXCwt5i4TaB/nAVbp0j3LAee6gym/WEV9tddaAade8v2YSv14SsRuWHZLjyjNfrmJrhr7QNMzWwlt7Kf5kVvsTrvIOEChd6mdCIJA91DKeQ/qVN7opDPa8dj8AnUNuRS05bDG9PfZ+jBEMLCQeQIxGIyJo+OtdmAf3s2sUyFgm1hsg74sLcKyCHoWJxFb7cL1zYL11++l6q3VuBq1Mg4LNI9RyNc66KuOxNHg4RSHEVUoGxhExMuqqV7Vz7tt6lo/Wb0vW7cl3dRsulejGeslI3rQPYYCOVrxFSZzGMahnNWpJDIL/dcR6YjiPO0gcVLD6CZoDB1COm5NGpmfMjDC7dg6RaZO/8kL+6/7Md9rqkiwRExXtm6ECEhYvLA2dp8BBWE4hDu/Saa5r9N65VphNbmYjlnpv2jUmSTQqvHjfG0la8ufQXbQSu/a15KjyeFcyeLELuSmqjai97HWy7hbADFqhM6kYa5VyJzThf62ACGfgPaowPkFg0ydPlwjvmKWHXbV9gbZT556HIen7WFW8YcZnCUQDhTpmeqiHqRD+VDmb6pOjZ7lMRxNz2KE7UgiusbK4ZKP5omcmYgh7p/TMJzUZymowWcv2cNcjjpGB51i6TUyTTcvIap6x/h/qu/JbDchzEjTNehXDRNoGROC5pRJ/8TAy3XClQ+fJb8r0V8Y+JE08FR4cWSluz1TvjDCnqmiXhHK5R9FiHP6aOtNotss5+LR5xn8dpHabtSJ5KrYukRKN91B5dufZjM4xqYNCypEWK5CTIu7WTppw/T8NQI+t4pZtwJiGSIlH5xP1JAYtg/fGhGnd6LNIZG6gx1p7C06hR9E034SkVun70HRxOMXn6a4s+gbaGD9DMKvkqdhE0gPdeHYtGRLCq6BO0LBM4eLGXjG3OQTjhIOEAtiNJwewbeCjPb1k9hzDMr6fU54JATzakQTJjI3yzhPGmkecHbiFlRzAM6bmeIGafifP77hXz97WRCq/NQ+gZIdYYY8+xKXLX/U9tFwn/xuoD/LC4Qon8DNN7wGpGSOMWHLQwONzD9keXsfv0N/l42mgW542hfNwrXBwcI5hoZv30VzT9PmqhF/SbG/2Uloqgz7eYTbK7cyrHaYk49MZ7eyQYMBpU7j1VjqfIyfdcqlIhMfK0RQdZIPQ3mN9z0TjUQT1NRHDqF0zsw+ZJjyTkZPmY/cYAi5xBCVhRdgli6RnhRgFe+XkRuphdbr0okS8c+vR85LOD8wUw8BboP5VB7j4P19eOwDx9CeWSQNEuY7qVxAiXQ+HoV6oCJ00+MxfZlCil/tvPFq/OIVUWY/+5+Us8rqIrI7LxGvP12POfTMPVJdLyTTcbIfgqWNiM8lU5lRReflHyP0S9gn9aPt9nN9ct2k7PFSGx4BDksJHO++gRSbuokZ2+Av934AX1+O76rQuw8PwzLhEFMPTJ9k+x0XxUn+KEdXRXwnUgnnCsQytMJVsUJZ+sUbY4gXTlA2xId3aijIhIYsCHosPnFOeTtAsljQJfghzcn4w1bqPhgBTc0X8ywYZ2YMsP016WjN9poudqNq1bAutuOpV/H5Iqix0Ue/Mf9+FqdBIIW4k4d9xYbwSKIVMYwzx3A4JVoW2Sh7i473rEJYqlQvrSe9H0GQrkCi1NOkl6t0XtFDO8waL7qDeSIQOZGM9K8QexHLegCtK0r5dy6KiKFCQrel5AyomQu7KB3fy72RgPu2T3Un8kn82gy9mGEu4dYSlIzIiaS37NjdwHhPJ11xyciRQQSmkTrEokrJi6kOphPqDRBazCVytLuH/d67gdG0jMCmDwiWWUDiJcOMuyBEyTcKnWz38fs0bh89tXYunT8paBLkHdTM2qPhWEZfRTNb+HGNx7G1ZigriMLg1FBM2uYPP/n4JdiMLQojLlfIJat8OSta2ltT8e+1Y4hIDD0dS7RhEygUKQ/YufdPy9BF6FjnoHqYAFfvzgbd63OwNw4rjoIDVmIvJTLnMk1ROpcDLukkbdeXYK12sLYldVYTQnEFgvpvxLJ3i6jRyTKPxjikpormXxdNQa/gC6Cf0SCYXtug/QYb362ENtbLor/rmMfO4gSk/lZwTYcTSKeETJIOoc7C+m6SMDYYyCWmyB6LJVo2IgYEfFOi1O+1ourWqZvsg2A8o+ibDk/ivPeTFLH9ZO9U8J1RsQ7PkHD3HfJ2y7QsUCHhEAsYsBea6TtbA5pp3UGxkq4a0Ns/mwGYlynZFg31m6Bob8mcJ4XKP84QUq9iLFPpjPiIliewNals/69uehLB/n+1HDa58vE0lV8xTLmfpFQroC6OQ1HkQ9NEcg+rGIYEtFlHe8oBV2CUFkCfdCEZtIZmJFA0CBtUSdGg0IkR2N0WQd2Q4z+8SL+SpWSjfdR9pyCv1RAX5/O2UAOilnAfR56l8UQLWbEj9LIeXY/x/5jzf/sStEF/NNxgRD9i5Fa6eH+jumUfKTzev4Bhl1VR8rHBxn14kpa/jSd9t/MwLQrhcF7pvPan15E7jHy6LhtNH1VhuSVCebrzC85z/a9YynZeg8nF75M4meDrLx+C/JhB3955WbY7WZyaSv5WySGO3swn7UgKjo9UyXSq1WkkAgaNHZk4BmZnHoKf5XFtpdncq43m9RvLXDdAHk7NdTzDswDSZfo/jEy5iov8gdp2DrBV6GjTQgQT1OR0yNYTHFSbWG6TmfR+kUpM8saEWMCcYeAlB4jnGlg4SM/0LbAgq8cJFljbfMkrv/7VsRuM5v2TMJRY0R1KSTsOuYvXPQPOTjbmIegQTBhpGTTvUSyNIySiskj8tXbc9j3wmuU5gxgmzCAtTdZM+4actI32cG6/kk4P3MQ67Ei9RjxtrqIZSuE8sF+1EJHXSa2OiPxvDiqOdkmeHz61wgq9E224j2VTvOSN1kxZzsvfrUYBAiNipL6jwN0XxtDdahYewRcDXGU005EBQ6dLqe2LjkVJkUEEEAa4+PYf6xhzaMvs/qJl7im4hToAiYviAmBW0Yepu72NWTc3cLqG95gWkUTgbCJlEaovXMNk8Y3gAipNSpnOnNJu6UN2/QBnrz+djov07A7otTetYaK91eQqAzTOxWGelOwDGj4SwU0A/hHx7E2G/CWGVEiMp3787D06MQnBvFvy0Y3anRdojJ9cTU7WypwNseRokkdlckjkEjRcNdA6hEDqllH+FM6qSdFQu+Z2b5/LOgC3V8VEUoYmfbYcib/ZgVDFQYkUUOx6gyezCSxN422x6dwx/S9SauAconBGdkMXhollqli6de5KvMkclDg3I4Kon/NJZaq0bpEIDfTS7jfRnbRIKFihde8eZR+dzeGgI5tT5Koogr8csdPmDm8gVCOQNEX/fhHxwEIV8XYMWIjB59+jVhqMsD0yEsTCGcL+EtEbM4ImgyOs0Z6p0js/WEUtvakzio2z0/6ZZ0MxS04n7UjxQVal6bSPxHytwnUPmpDezqTcy+MIpapYgjqIOsUpg+he42ow0N0zZLo+o3Kpfl1jCnu5Ok7b8E8pHH6Z6spL+ql+KdDGAtCCKqAtcmAqIDttBn3OcjP8dA11406fwhNhoQmMTDOytjCDoS3MuhtS2VwrID96h4enfkNJZvuxd4c5MP5r5G5VyY7w4ejXcNW7COSJhLLUGl4QMI8qKOaBDoO5WEI6egfZxBNE2i40QAL6g3xqgAAIABJREFUPMSzFGRRxeCRcTbF0UVYVnQKe2YILT2OnBkhlK9h9MP0xdX4qnQiESPGFjOX/H4v5kEBZ12SvMaqIhQUDmDJD/D60jcZVtqNGAfLQyYmZbfTeP1rnG7IJ/QTE646HTE1hsErUXufhRdufhuLR+XUzkoO/W0NUbeA4wcLW+v3MefRg0jpaSzIn8iEP65Ajgj/s4jRhZbZvwwXssz+hah6voP4uwLbhm9K6oV+OoPEHB/Xl5/gvR8u4vbZe3h/52ykiICjBe59aCNH/CUMxS2cOlZGwXcqsZ966O1zkpYWJHQonZoVqwGoeH8FKc2gX+HB2+xGzoyQCBko2CzSM0VCl8FZD1es+oHl7kMsfPYX+EckqCzvZoy7k6OPT6RtoYxxSCRaHKfkI53eKSYieSqXTa5m+/6x6KKO65zILQ98y+qtC9BlHUED1aVQtEEgkC8z5s4z7Ns1CkOFH2mfk8DoGNnZXswvuJn012N8eX4scq2VuFtDjAkYAgLOmb2on2biaIvTM9WEqAAzvLDfhTIlQCxiQI+LpKSH8Hc5yDwo0TdDxdoqYwyAr0pFtyl8Pe9lnuu9jF27xiCXBkl1hLiraD+vvnQ13uEacmYE+bSdjFMKVb87zb6OUsLtDlznBSIZAo5WneCSZAJ4IiazZMRpzvuyaNlbSEoT+MrAVQeuO9rJsARperGK3qmw8Zrnue+xn+Etl7DN6ie2LYNTv0hel/s7pvN6/gEALju3hJbeNHLSfMTWZtN/cRxdFcjcZSDmFjAPavRN1zEOSqRN7SGwLZvo5CCWA3YCFSq6qCO74qS7A3iOZ5JwaowZ3UJ1cz62mqR7MDrY20GTBVLaFIxDcTKeauXYzioSDh0pIqCZdNJPCIxaeZrDG8YQT9HJ3ZMgUGhgcIKKYFOQZA3TKSvp1QlarwEEneHPBzj3cwdVr4RpvcIFIkTLo0gGDUHQSfiNlH+kkPGXFj4q2UnZjjvRFBH7WROKFaKFceznjdg7NSIZIhlXtdP/ZQGhvCQRtXcrJGwSXRcnBeho8PuLNyAJOn85s5BYswMEqHjHy+AkN4FiAaUyjOI3kls8QHhTNoKmY+3T6J4pUDq2k+iruXTPFHCdE/DOi5Kb7qWjx03TZf+g9PPljBnXnJw0SwiggSEgknZWI1AgYpo3wNSsVr6pG0FGqp/gziyMFw3g6XDRfNUbrOqcyqm/jqNjvo7BI1H0dYS73/6Kx3deh61ZRoolK16aAcK5Kk3Xvg7ArOpr6KrLwNwnAZC/I0Tgd0FMkkprUya3Td/Hh99fhKVXJJaq46yDSIaAZSDZlopMDWGzxsh4ykTPdBtSFIx+Hc9okMMCpe91oksi53/nxlpjJqVFY9wjJzn5zDg0A9ju6GJgcz6RLJ1EqoI8JPOThXvZ8PlFZB1L0LIMhJhIwbc6/iIZW7dK75Tks7Rq0ZCiIgafQKRAwVUt450co7KwlzaPG+mYA80A0XQNRB00ATEBtg6RSKZO6ed+2q5wYhoCZ0uCrlvivDP1HW75bjlSQEJNTZC9Q0aO6AwNkzDOGMRpibJz5FdcfMc9RNJkBA32P/8aMx5ejjGg0n9XGGmfk8W37uXkpem0rKwinpI0zwzla/9t5/l/SZZZSb6e8+SD/1UfCYDWO351IcvsP4kLFaJ/EcofPsiWQ5sRL2lnbSCNb7tOcvJXq5H3OlnXNI5hr/vY/OIc3KcF6m5fg68cnv/iSua4zjPa2cVtF/9A5hNNuK7uRFdEQgfTSdiS5HbaY8sxewS0y4eI709DTI+RCBuwuKL88pn3MfoECid0Mjhe5YM9s5j51SMkUsCSFqH5SAHrDk7GM8KILkLOwTgFG0X0Xw6gjA2iCzodVzlx1gqQohBzCXgUG8NebCP3Bx1XjYBoUum4WMIzQWHXuUrUvCjhXhuBSoX0jABmWaH//gibGkZR+LZExcVNpFYLmIYEpi+uJro5i8FxOqEcA5oJXl7+GhZjAkNQJzpoIeWoGXnAwCNV28jfLtA3Q8XWIhNL0/BPjaBbVUSfgdv/4xEOfzQWUQFZVomty+KvxxcRLARTThjbD3bSzyh4qmSOvzYOSFZwblu1lbhbw9UQIeI1Y9rvQOw2M83eSNPBQtAFAsUCJq9AoFAg+Ho+Az/NR9BBs6nc+NIjDFwbwTKgY33FRTRDp3TbXQA/kiGAbcM3Mb6wnY6GTAruradgnYT9nBHP6KQIVrEIZB4QKJjVzhW5Z9EF0FtsWPs1Kkd0cMfMvaRvNjNQnUn5293kfQ+NnjRMtjiWXh0xJmAe5SVhFxh9yxnC6RKG/iAHasvIOKFhyg1RPLkDzazhrYDvj48gVKAihwW85UZibgFTv0RuphfVY6Jgq4feqQaszQbMbUa656ZhTwvT9lsBS7+OGAfBY+SyivOkbbJgckfJ+EsLDW9UMeGPK0AAa52J5XdtYvbiE8geA2ICrH0JomnQ0JiNmNBRXCrRNAF/oYy/SEQKJpe9Seb1313LS3++nkRdCuYBkYyjULvCia8Crr96N/Vz38VRJ9NzPpNoGkQvCRC6zYccEWg7mI/jwXbydmrEXAJ5HxvItAY4dvEr3N02i/TyQWq/L8PUJyEoYPSLaMOCdC1UiE8NEN6fzrc7JmCutlDg8FLwRQ+W99wIVoUpJ66jIZBOzzQR2S8hhwRCeWYe37cMd66PFbcnw38VM6hmqBzV8eM+2DtmA7pJ484bviVaHqPnl3G8B7MQn0kHQWfdJ3MoHdOJYbqHKy85RPoxP5FsjUAhyWzCATPazlTCOWYMcwcI5+oE8wWKJ3aQtztK5+I82q/OIXuLkfjYEP1XRjn+wjgSNoHETzzkWP3E3MnKcHahBwH48pOLUC06xse6EWIiTde8jmoSSD8ZxrKyCzkskNIIOBPYWwQiJXHEiEhqbQz7WROF9iEy1loQFFBHBblr3i50g44UFZgyvZbAhCjOMYM03Oxg0hVnSK+O0D1DRolL3PH5A6TmeVGdCilpIbxXh3j7hefI/9aHt81F+6kcRh28GcUqErgugGdE8ja2/7nX2PXmmxT9Okb1o6tZnrafL6q/pWbFatJP6ky8/jTlDx/8J5zu/4+4UCH6l+ECIfoX4OGFW6h7exJz7r8PKS2Vmx2DAMxfdju52zzkXl1D42+MzF11kGi6QNmOO5FiYOkV+GT6aL78xxy+eGsux9sKGPzJeAy2OCVvN5G3R6H0u7uZ8+hBImOTU1vZB6M0zHuH4sJ+4jEDryxdiqVfp/l8DnJA4rIp1aAJ5OyLkWhwsHLJVtILvWQubkdUoHOOgdSft+L5Og/3RisAHWvchOcF0XWIpWms2ziLmify6FoWR17Wj9GkkHZSQApKuA8ZyUgLYO6TuXhcDUVOD9F3kkaOC8vO0THPSPDP+XgrkwfyUNyCd0Kc8k/CXP/4d6RN62HVm8vx+pPvjajjaFcRVXhy+zK6p4sYXFHCVTGMPhHLWQtCWMLcKzI0MpkTpZYkJ4Ds3Qru7WYyx/ei1drxVakM3BrG1aQyMF1BO+Zk+PRm3qqdya8Xfkn9LUbGV7aimkAuCdKVcKOkaCjlEc7dt5rqR1bz5X1P03WZSuVrtcz41SEcWUHsXRqJmMzjj61l/lM/kH1Axdhq4rJzS/6vvXC8tRCDV6Tv2VLaFwjJtkiHwIJrbsM7L0rgmgC9ATtrP7kEV5NK5RvdRNJFGo4X8M1Ts9EkAeswLzWPZ3DDn75GOeUi1mMlWCCg2HWUw25yfwhw4IeRBBYFEWJxpg5rwv5AB/atdgY/z0dyxtElMPgliqp6iFTGsPWqDFtSh6AIxD/JouzTOLX3ORFG+QmXJFi89H8TuwMupuW14pkeJ5quU/ZZhMcytxMoFDEccVD/VhXBAgHvCI07xxwgmqGxrX8EFilBxTsDBCdHmP3cAeJFMVLOGrAOaDjPykRGRbjlgW8RlWSWn71N5IUVrxO91YO1TyGlEY49+CI9szVyyvqRq/x8/PVsAKofXc3ESfVoBp2iP6rcXn6Q/KmdxHPj1B8uom2pRsGmPvIer+fstkouevVRal4ahf9oBuZ+kMNJYmzt1sn+2IzzhBHO27n35q/RjDo589up2TyM1muz6bk6TsYOE9qGdGqbc7D0CIix5E8zVCVirTei7Erjmd2LYO4QsTQNKQqDYRuXX3o9I/bf8uNe+KZnJLYaEynmGNX3vUzvRCNIOnG3ju+DfNQfUjn1yDi6ZzvR7QpKSRTTkI6tXcIQ1OmaC0MtbnL3KESyNdr35dN+iQnbFT1EsjTEO/vQOqzoXWYGroiimAV8AStHOwrImdGJoSyA5VkXlm6BSLaGpU+g5UABacdEhr++kr5rIwyOstDUnU4sUyXuEtAjMs6lXUg+mWvmHML0m25yn9nPsXfHMHBLmPw1p9AbbXzxyjysrTKuWjjUXMxd4/YjrEvDWS9w6PuR2P7URf7UTggasHYLZNhCiFaFYIsTodrBgu0P8ev1a3Gelyj9IoplUwrtC0Ha4yT7YIJFpdN4eaiIBbnjUGsbAFj02i+4Mm8ys6qvQZMF+pY50GeO+/dun+kkg/H+K9cF/KdxgRD9kzF5ah0PuNqpejmMedNhlKpCbmmZS8WHKxAOnELo6eep5kPkfGhm/f4p5O/wU7myAdsYD7nvn6X7puHMv/0AvtEJpGYL1z32HRZzAs/cYtoWiVw+8gyfn56A2RLH0iPQOcfMsD234f0iD2ONhdp7XXjG6Mh+Ec2g893JURiyw3Tcn0BJU9hy/1wUVaShKRtrZ1JbVLO/FEGBu36zEWtWiPhxNy57BHOrCWuXSN6uOJYuGc1voLfdjdJoR4rrzJhRQ8wtIL6XTuk77Rz6YgzHzpTSPwFGZXVTvyyXeLpK6+UyWeN6YaKPjoCLqp/VYnmql7XNk0gxRbnyJ3uRz9pIPR+jsqybgbES6GBrlbC3CSQiBvAZUGw6oco4VS8NkHDq/GzxZlLGDzKntIEfGsoZHGEgY/1Z+rx24jkJzLkhop12eqaKLJ+2C+esXrrfLQHg7SeXYmuROdudg2lIx7LDTnfcScWHERrmvfPj9byv9mZQBTYfnMAXO6fy9YQ3yV9Vj2xU+MPqW/i0aQJ3P7OBvN1xmk/msTFk/fG1x2JxjDUWdBEC+TJlI7oweXV8IxTqb7ZQ8fco+gknoSYnpmmDWDvCVH3eRu72AXQZYk6Bz//8NP5BG8Of9vDCpsWkn1LRrSrRLJXxM+rIPhij+VGBgh0J5OMOWm7KxybFqW3J4dqfb2donIrluJWcAyr2VhAFHanPSOclOi3vVWDv0Im5BBpuMiBGREwGBWOvzO8zDyGHda66aQ9tITfO4yakKFS9XEN9wk3CrhMsVTD7NKLZKkJC4B+751CyMc6p84V8vW0ytfenIRtUPtkwF7nLSOIiP1G3SHBahIp7anj1u/nIYR3VAGJC54+NSwieSiOYKxOcH+SyM9dTvjbOdQXHMe5wQkmIkk33MvbvKzGKKohQf6uT7weq8ISskBApndSO85SRn27ZRJ7Fy6obNhEqVhBv60M1Jx+no+k6qkUnMC+EvWaAuXccRtAEXj4xD3eNwOC6fO6/dQtpc7vRlGSVUJfAfdRAYHQMoz/ZPoplqih2HWuvRkqtTLDZiWlQZPay4xjfT0Wra8JtD3Nt46WYemX6gzaCpQq+iJl5p68jkqeCJqBkxDHc0Evqgi7a5puIu0Eyq9iOW8je2ETa2QSesRqWLonLp51kqMKAs04k64jKxHnncZqi2FtFus9lotrV5KBBtRXFCqbTVrI+sNCzLw+3PUzHPQkcHSp/XvQpgRKNjJMakSyBc/evRj5rJ3aZH91nxNYiEc7WyN0hMLQlF9WhsvXT6ZxryKP+/QlYBjTM2x2cf3k41h6BqfeewNWgMjheQ241M8bSRtrRIbzDdeSgQHVzPrfmH8TSIeEfnqDz6yLEDjOuMg85+2JYWoxkSyHEhE79rUYG58SxNcuYB3SuePp7Jh4M8dLmy+n96Qy6H54BQPE7TZQdMfPRiPc58JdX2XLka9Tfe/hl77j/T3PQLuD/FRcI0T8RJRvj9EfsfBc24B2RgufO6QyMtnDsmxEY/AIdv55B+13D+NmqVex+4w12XfkstSvM1P1+JLZ3XSjDi0GHjfWjcWUFSBTE+Pzp+fj77fRdHiOlVmL3ZxMx15kJB0z4hqt8f/ffyXvTgG+kSkqLRmbFAHJAIJGZwHVeQIiLWM1xEr0WKu89QtP9Ar4WF6nZPmbdeBwpJlA6tY34HD/rh2cSa3YgJiC0M5NZi04RnRKkdZGRhF2neKOO5JdJrQEpodP5RAUFl7WiixAalY3Zo2NMjSJFBcKKkcu2VHPp+LOQHmPAbyNlnQNdF/B/nkHsQTf6ljRaPW4+2z4T96wehh4O4vmwgGhugoIdMYyzB9AvGaJwvYStQ0Sx6Bj6DLRem0X5+4O8+tESBj12Dq8bQ86XRuwX9+K5cgRCvQ2TI4ZenYJuSD4Jv3FyFpKgMzg3RjhgwuRVsFw0gOM7G4F5YVSLgCduo+UKGxOO/uTHayr/JZWMAxK6qGP0iFz518eo/2gYYp0N86DO70Zs4ckdywjmGbG1ibzcdgm/6RsNwH1/e4jP730WQQP/1AihN/LwjNIR7QlyKvoJPBUjUhaj8YbX+NuIDdTfauWr82Ph1SBrFr9NzCVw2XuPcWj+iwRf1pBKgnRenaBog0DRZpXjbQX0TDPx0ZS36J1kxHpRP9EREfa1lZC+x8D2lbN45uJPUE3gfLSNmEugqTkrabHQIXP0D2vwjNbRDCBYFe5duJ0JWR2oZp2Rm1Yx/L6zrD02lc5dBfxq1ce4a3U2nRzL/dvvRA4LCAmBgkfqEJxxRk5sweATabnChL3RQMKtQHoM8ZwdRgUwDfdxaXEtg5MUstJ8dH9aSkqjiK9CJ3NOV1IHdCSHkvV+BsfqlGYM0t6Rhq/cwofPL0JUdfQ2G9YWA4HxUfYfH0aiIEbZ+A6at5QSOZlK6jGZ5v5UgoU6H/ZNZ92pCbz90mKmjm6gzDmAeVAgfokPOSRQMr4T6byd8w9lsPHARBSLjtBjIrQgyDUrdvLKhssZ2JtDYY4Hoz+p6cnZ3A5APCWpIdJFHb0sxPAHzxJL1dGsKqMW1bK7rYyBMQINf5tEbH0WNd9UUvyVn8CgLemxdXUNoc3Z6JKOYUDG2GEkoUpIT6UxZe45VJOO8byFdx58gY4bygjkyTQtex3NAG1hN+GZQXxToygPDHBq83CK7R6Mfh17i0heyQCWAZ1F1x8gkq1RtK6H/rEyaDC0JxvDKTvZP2tkXd9EdKNO12KFlGaNyndX/NiSN/dIxCcF+cm8/WgGAf/IBKWfaIQq4pi6ZWSjQuIOD94ZMUSfTHy2n72fTCB4sx/doBPPUHj4y9sZ8V4dQiKpKdKjEn/79Frc9SomdxTFnvx/RWJG2hYaUc06T3UvwFcOhiGJ4g/BMbcXbxVMtTay6Z2LyBvfTWKOj8nXVzPpdytwrothEeMsH7uEinUruez6O9gxYiOfHZiCXNeOZtX+LVtouv5fuy7gP48LhOifBM2pIO06jvJiNquO3sjzf3qVxT/bzZ0Pfs3VV++l5IN2VLPOvOuP0D/GwJjDN3J/1XwszUYyqgao+uUZ5PNtuJoSfD71DQDc7iDRdIG8b0UYMBHK10lp1Vi6bC+2lChZBwSuPnMH7ZcaydktMOfRg/gOZ1L82wMUfikSzhHQrSq2953cM3cXns2VpLpDOIp8HJv4Gd+cHUkkXyH8Yh7a2RQqjpgoGd+ZnJKKwZ5vxmLfZePxxV+QVq3jf8CHmqIA4C2X8JYbad9WhL09Sv8YA5YBjfiQmdq71mCV47z1/uW0BFPRAga0Rjv9kyDH4aerJZ2ui1OJXeYn4x9WrBVefDuzGRqyY/JrTB/ZQPc0M976VMzrXfhKZRJ2GPbr06SO6SdcGqdjYTqR4jhPTP6a3EVt2NYdIrI1GTngnDBAostGtDDOg7O3Y+nTkdvMBLZmI3iMWM+ZGRhrxGaMEyyCujnvkf9VF28X7iV/Z5zsX+ms6pxK5Xsr6JplxhTQsLXJpJ9RyFhzAN+UGLFMlYy7W/j1iaUYM8K4z/rxj4nTF7Dz/V9nMu/Oe4i5BCoNSTLp3GdmYKyAlhlHC8sMHMtCfzsDQdRZeMXN/P7xu/jgijXofSYWZZ7l+WXXkn0oRvGMduateYzeI9lkfGwFvwF/oYylcRCp0UK4PM59f3uIWKrG/Lzz2O1Rtk5dw/D7ztJ4t8gLv7wRZVSIxm9Kufy6AyDqZI7txdqnM/n49YgxAXe9woiibnYsn8nbhXuZMbMGc4/M4W0jWTL2FIIKNziGqHyghtzvZER7grhLZ9TYVureqSIlJcLZE8VoRtCyYmgyNC95k8aL30E1JnVhsZhMrS8LISHSP+TAP2TFO1rBdV5gRkYTPdNEiqe103qlk6o/1NETcJCV4+WZ361BXTKEMaCjC2Cb1U/RRyLmbglTo5nmIwU8e9+b2Dp1HG0KlgN2xIIQ+2rLqHw9gWYQOFRTxrGuAkxDSbf0aHGcYSl9SRIUF7B2SChuBXuriJKQ2Ng+Glc9XLNsD8FPc7Bd3oPi0OldWIClyYTi0DH6BFLzvZiO2bk2/Qi2Dn48beW9zmREg5i0MIiUxnnw0/W4jhmRDCoNz09jym0nsLXIGH0C0vAA0R0ZRDIMNL9QRe4ehcTwMBNNRhQL+Cp1Stffj7NRo+ZIMdcOO4nUY6KrNY3rb9jFyWfGMTQSTD6d0MZs4g6B71+fhqVX5Nwj6SBArCyKYtcRVTi7o5LajZXoskbBeolAQTLEWUlR0U84UUcFcWyzse78eMJZIpZ2A0M/CyL5ZNKm9JLwmunvcuFIiVA4qpubKo9i7dWQJRVjv4SQECnZHOWLfZNR7SqhMVFKP1M5e88rdM2B9M+sSGN8rL/heaoye7F2CoyeXU9n2ImeEyV3YjfNNwhkWYPIIYFVr64kNjNA+5ls5hU2cOTzMRz9wxpqPhnOhqMT8V1aybFrnqP5Kgu3tMylcuVhvj79PZXLDyNlZWLr/De7DV7QEP3L8G+2E/7/RPnDB6m8+yieu6Zj9CcoX97KyjM3seW5OWwe6abQNMiWA5uIO3W27J3I/geexWGOMWpvBEebTm9bKgHFRO/1VTgeb+cXFbM5PGktQx47YgI6L0nuekufgHkgwfotM8n7g0D/RIh8m4mlykvvVTH2/2Eq0UyF7kdmEM6UiDs1DNY4nQs03v/iEob8VgY9duxrUyjZeB+C18DDF32LYkm217Z9N4GEKnF21Wp0CWwTBvBV6bz5p6Wk3tOGp93FsIou+idpONqSpfZwaYKBx6JULGwk5hTJ3i1S8u3dnO7JQUxA75YCBKuCVB7E1iZyti2HvO0CS+/cTTxmYHCEAVVNZoel7zDRvTTOwcYSdBl0WUf7ySCX3H4QKQbtHxQjvpeOYcBALFVH9hhY89zVDH5UQP17EzDMH+Dg31+j1DWIqV9EDMiENSOanGy1X3rbQYq2JDjz09UoZtg16ksSKTql2++i48pcRh+6icFRJsIlTg71FqHkxXDP6sFXIpG3K8jgbSF6vhxOUd4AjuwANWcL0RttxKMGmpemgCYQO+Wme45G+3yZ/AWtjHp3FenHBbzDdRLpCqKkIwYlNAn8JRK202bqHjITvsXHrT/cS0qpl5er5xJ/NkT0F0N0BxyYB3UMQYHBERKmfgkpClPX12IaEhBEnVAejJjcwkd7ZxBscrJkzS/Yd2gEjhMmuq+NowyaCZfHCakmsrfLBKMmfOXgP5mGoIPRp9CxoQTxj/0AHPp+JIwKII4I8M22SWQeTzDywM2c6s1DjmiInWaevOoz3ipdDzoEGlxkDuvHNWIQe0qEaIZG2fd3AlB3xxoMzhhqQuLb4ZvZtuRZEhED0qCB0vIePOM0fvjzdFKaoPFkPurwIA2PDEPVRLxBC3cduJPRmV14RgqodpXAkQz6xxuJZqsYxw/xp2s+4qmWRYy88yxtV4ic+uVq7h55AGu9iYbrrSTs8PisLUj7nHjnRYglZFIz/HyzezxZY3vR3AlW3LYJc0cyS03rMTPQkMa1v/iOj05NZv6qfcTWZyGosO/JlzB5wJAbQpPBZkwQHR9m9dxLOPbkGpqveJOhJwp5btXruOpAN+gMjtGRvDISOtkfnMZwyo6jSaTjKifKxAChyjiaJmC+pJ/BkQJDFSKFv60lfbOZHREpWV3JjmHulTj49GuUjO+kJZyGnh9hbFUb/5FRw/7nX0PPixJ3JPUkUhQc7QrhHI3Rw9somNtG7iYDCZdKsETh1mt2EBweRzCrOB9tQzMmp9WkiIjRB6V/VXDf2EHp3xX8FQoGf7KSoxl0uprTyf4hmX2o6QLtp5J6QX+JyNCgg5KZbYhRgZbLzciZEexNMsUfishP9DLjVw8gB0X6JoqMyurmqq9/SnV7PkXXNHFqfwWxp3Kgz0RrZxrrLllN3XdlJOw6ynQ/DmsMe6mPYdYe1jzwCvPuupecnR6mjmpk34uvc+mfHgHgeFc+yvZCFuSO49uuk3S/kUrB2kbeX7r6n31buIB/Q1wgRP/NKBzdTfNfp9Px6xmk/uMA73zwMqrXR6Y9SOhKP992nWS5q5PRz6+k6rVBKj4M0qtqBKMmdr80jb5ZCrJP4uxXVaBD79sldH9expKrbicny0s0DYR40uzMPKgz7ukT3LDkB2rvtTFt+nkUKxi2upBkFQBzr8ysG44TLBDQnAp6qw3ZI2MIgNBuIXuLkZ6r4lRWdGHIDfHmP65AVHTkiMDcy07S2pDJsHdWkF4dI9UaIbuqD89Ige7Pi3lw9nbmZdSBCO5D3cSdGpY2A6+P+YCa/aX0T1Gxd8RwnDERHrQSc+uEJ0ZI220iFjJiDOhoikgkVeTB3nc1AAAgAElEQVSzDXNQE0lX42i7g0i6wMAkjeJ3RF6e8RGaQSfjqEggbGLj9qkk7DpZKQHGPXoS06CQbNkUJjOqIlkCI4q6mZDZzpQT13HkcCWKTUdMCHzeNJ6B2XES2QmqvXkYvTFGvbiSWJbK6EM3YekWcRw3o872Eey3YZg/QNdMmSMTPsN2xozt9w6iaTqND0rEG1IwbHLR2plO9LwLg1ek7o41LBt1As0AyBrxVJWyzxVsHSK19bkknMmJJ92qIoYkhA4zzjoBQQdLnw4zvWRm+sh6UkKPSEzPbcG210bbsTwcxhhjMrtxNsbZsOJp1DFBotkKvgUhPtkwN5nFFpSJ5Seo78ug6ZrXMfeLhAtUpLBAyTWNZG8wUjSsh8nDmtl6ZAw9szUCbSkoTpWURsgY38v2tf8gODVCzxdFACxaeASrOUbGu1amzDnHz19eS7jfRkVaP4agwujpDfxmxzL2RHJ46zcvUDCmm02j3ufB8u8J9DhYMvso9Jn4JOBm7FMrESUNc40FgAWfP0purgc1VcHzRT55O6BrsYJ/VpSr5x1CGbBgCAhI37qIeSxkbDFxZOso1JIIl044i3EI7rrlG8SYQMBr5Y9v3kzriTz2nKtECotUfLiC7x6azcO3bkAzJ0ew73N2MfvGY1xU2sjlJTV4G1MxFIXwf5fNzGGNPLNvISXzWrC3R7GV+pBiAmt2X4o7NciW92dhWtbL5HnnGPHpg3gnJPP8ah5YTUd9Jo+M20bNb3N52lP245nw+B/uo+8ihdRqEV0C1arxs7V30/Lz0RQ8dwzfhBg1TxZQ9ks/tnojiZhMqiVMerWGboDznix65qk8+O79qCYQu8yoo4OMenElPd8WMMXVjPWolfi9dso/Xk7Jl/eRudGEerEX7+QYFo/G1D8fQcyOMjetjvYhF+K9fRhcMco+U9h931Qkk0rly3FqDpeQGBNEUEFLS7Dgrv1Mee8ULUfyqV1pZfKYRhJ2iIWMFH6jYU6PMDBWQM+PsqT4DAa/yIfnJ+Ns1EjdY6R3XRGCKvCTBXvJdgeIO3WkX/YSTBjpm5PA0QJqfpQmbxoj/tDKXyZv4MzJYtTsGB2XGMg8AllZPm74/CEY78fSLyAcS+HivDoq0vrpiLv5LjCavvEGdIuBoZkeSjbdS8q1XTTcvAaTQUG+tI32J2awIHcc68a9xZbj3/KH0gmUj+3g3wIXRNX/MlwgRP+NOHDds5iv6if7kEp6tUJ84WSWN12HPmMstS05nJ2+Npky/b8x4ePzNDwsc9Xbj6EecDM4Vue9S9/EWO7nV3d/inUgGQ1h3uDihg+/w2GMIYchtVqgadnr9M9UqA9msva72Txz6SfsO1+Oa04Pgblh4h4ztgc7cE3rZeuxMey+52mKPxdwjRxEzYkRS9fJGttL5spmnHvN1J/LY1xeJ4oFEhYRbXSAfRvGM3ZkK1JEYGCUiaaudBy/tTFhTi1Z17ay7o/zOeYrZNbEcxjejZLSIBDJT3DDxgfJ3atg8EpY/tSN9ZK+5GTb1G6kZjOeMTrlbyYJmxCWSdgFUpp1ppS3QFTCfTopWrW1SXSvjPGr1+4i7YzO0HCw/ODA6E0aBY50dbOrtZyCy1uQI8kA2bgr+du2bSzh6Fvj+HXlVqSowGc3voCtXSDQnoKlwYSl0UhjRwZ1PzWiT/HhqJdQj7oIFStEMnWcnzioLOvGfyINOSJwf8d0gqUK9beaSLhUUvaaqZjSSuYtrVTccYzSKW0IFUHKP17OuoOTMXkFLE0mKh8+ga/YRDhHp3iDTtFmDf/YOFWvBAFwNoCogrMOFAsEPVamZLbSepUTQRWo+Y/RWAa1ZJxCwMG+6koGxpioNNgo+20IOSBR+oyGIQDaEg/2fD+Os0aiAxZmr7wP1Qy6qKPkxQgkTHReplOeMkDni+WknpSQghK6TSX1hIQhrFPmHGDMMyvREiKBqRHWB1PY3jqM4Mk0umdJ9DxRyu/PL0ZyJAjfn4rn5yHa3i+neekbLLP7GWcysXPkV6RLNt5pn4nRHeWrwxPIGdHHlbZeEnbIcfu5+cYdlO+6g1WLvsG1XKHquSDeUQrBnCQR1PtNfPXdNGxtSa8eTU4e9O7v6nFO68NQa+X7/aNJpIBBUEmpHMJWYyIyPoxqU0k7YEBJSzB55nksv+vi2Y+uweCOYenXmfbYcg71FpFqCLFzzTR0AaIDFrRZPk5tGIHskanvzqR/nBXbp07yx3eRelwk5VUn1vm9WP+UwrFvRiDFkq7SV161n2m/WI6jUeKZLVdirzew7u/zKftsOZ2zLQyO1Uk9JnP092vIr+rF3igTL0q2WZveG0bzgrcxdxtouCcXW5fO9LJm6jszGbwuTLQoRmhvBoUbBdRRQdJnd2NrF6Dexv23bcFdq/DKNwtRLOCZkgxatTfLRN0CIb8Zd1qQ3sUx1p8bR9rXZt6omUn6h1Z+GP0FGe4Anp+HaH9Uo/yZOE2PSEyaUYsSk7F16kypaOaz45PY/8BkdAmemvMZp3b+L/beM0qqOt/3/uxQOXZVV+fcTTeZJgcREEkKKANiFsUEYg5znODo0eM4enTUUUfMOiqOCgZ0iIoEJTWpyXTOOVXOe+/nRT333hfnPmvdc8+cM8+61+9a9bL2i71X7f+3fr9vKEcXArlbT8wpEe+wkF3ZTf0l7/NM5imkONi2W+mZoTEwM46zMUHSpvB0xmlmZjQQz0zycOFO7ij6CZMzim+YhsGUoL/ZReuqUl545nqkmED6bgNiHKztMVyPiLy8/H1GZnYTqIyCAJvOjOf4yVK2No/k89oJWDo1tm/+CLmkiMz8IYTn0rnssus4MflTLPs8nLv7deo/Gs99l9xIyaY1rKppg0vbKXvo0P+/XWg/4z8VPxOi/yQ0XPsGW0PFbKs/wL7X36L92gT67Ufofb+IxnsEmha9w7RfrqXgZZGHuyZgmtPH/kenovUYcU7vIVScxNwhsuYv69DtcfDYvuUMDpdwHtfjK4Onvv8F3qiJOVcdI5IhMP+61ZgbdZw5Wcjwyc088e6NiD4Z03NOtFYzzrMyAH1nPRRu1pi54RH614QYOu/m5Rmf4qiF7tOZdAQchHJh5UWH6Q7ZSdg1ei9Okmy1oEow9EIhtlYNg1djbnkt9ddYONZSQPSFHCb98hhnt1ZwaPcoar8rRVRAMKjk7Vbxr/FjHOnldGMuvkMZpOX4mOhuJeFUcZ0SaH8wSdIsQFIgMDKO/eZ2zn05nPR8L9HFfuRJQ0xZcYqT0//C79d8QOd8BVWG0Mwglq5UvcapwVzUczZadxSlVhbtEhVzG9D7Yeo1J4m5BJ598kaSuTFuePtBDjz6MmmnRYo+60Ixa5guGJE6DRS7B9HP6ydaEcVTMMTYi+uI3jRIbWM2qg4SFo3vDo0lb6eAoAig0/DNiFJiHWAwYqbzlzMY/LCAdHuIRbNOkH5EIpytEslOEloynucfexNzp0DzCgFdIIE1LUzjVU5Uk4p3TpRYmkBkqR9/mYa+W8ee9jLGzKuhYItKyxKBmEOg+Js4/iYnOq9Ecoaf8U+vY3CKBzQQnx8kWKji+LMNYU8arvMJ8kv6WPfcRirnXUDfLyF1GWg7mstfF6zn+xOjWPDYj7iuaccwJFD2fhLPUT9jHj7JyU9HE3eArdqAIMA/n11C7r9KOBrA1gS94434LrixWSPM+/wowTOu/8/6pPBHOciygvOMTE91Jl1KnHN3v07HgIPfpNdQP+cDHkhrpn1ZHq1LXXgOSbhqYuTtFHAPG8DaAsHiJKIC2qVDLJ98lPabKyhxDFC4NYCQFJAi8NpXlxM76Cbu0FBiEhhUXGfCHF/wCkfb8jlbl4cuCImIjqFRGpooEN2bzravpzEwKYk5P4CpQyY0ZCLvey/6Mj+ubSYiGdB7eYz+oIXoYj/tc2R66tPR/Usvnou6EFLyOU7eOgpBAWu7wtXz9vPw7Zsw+BU0Z4KqNS+Sdk5Av6yX8r03440YkaNgsMTRzAo0Wij9fC1KRYjs/Ul6ZyjUrR9BSU4/jr9Z0XfqiblVwukSFVm9dFVnoV/Uh+ucxitbLsfgTSDGBBI2jZ45SXwrguhn9eOdHiPvS5mhPhtmS4x3pn5IKFvEcMBG7wSJCUevYV52Dba3Hdg222i6yk7aNjPnvhiO5ayRvhlJDl8owXVYR8PVBq6ce5jHji0jaUu9C8SYQN/iGNYWkbZGz39/5mfuex1FD65qEVGn0n6JjOO8zIuDJaxwHsXYoeOPt1zH248vJ9phZdgHXso8/aQVDMF0L0kjbFj5Cn1TFfQ+ATGhsPX7z9njH8HxhkLM54yMW3yeqSXNZO4XSJ50op2zEbrCz5Rf30WkxM1gtQd/kZ65G6pYmFNJaFYfCU1B0wSSjc2MHtfCs++kjBL/bXL0j4Sg/X0/P+N/HT8Tov8ELJ51jIU5lfx1eA7l+1YB0DD3fbQZ43Csaqd+zgdM+c1duHc1oX+ml/0vTsHxrJW4XaLswUN0N7mx1cnoghp6L4y9/gyeHC/6qYPYlnbhnNSHFBLp6XJytC+f9NNJ+seYKF3UyPA3vdR2e3A2KGgitM43YG0TCBZo9H5egL1eoHO2jBwSiNY4kKLwyMab8S8MUbG+C/UbN4pJ44vvp9O3O4f87xNk70o1pEezFAZvDRK3pk68M4NZFH8bQ+02MnBHiFNPVBLJUZgwuwaDFyKL/BCSGRghc3zSZ8TjMrdMOMg9136LLKnY5Sh5w3p5/rdvMjWvheKr60DWEHQq3434FsUA5jedJJMi2v40Dn81ljH7b+HJ52/G2KYj+6BK5ucmVBlKPtEwXz2EJoKgQFptEkGBsx3Z+MoVVE3A1qLiLReQdQoJq8bozfeiygJaexeOOijc3I9WEGHwjUIiP6Xj+d7AQJ2bhs/KiVSloxuQGXdRHXMuOUXeDxqCqmFpkdANyDRc+j5nhrKJ7MwgUhmhf7KK/7ssTj5TiRRP9YGhVzH1xrjtmztJWAFVoG5VSiOVfSiJkBQQBI3j97+K2ZCqRSj8Wxjj104yDEFaloGQFBisVDD/cyeCJ0buxE4igyYyD/sIZYnovQLXZleh94q0LJaI28FXokMUNH63+Vqq6ovQBwT0PgHLiCH+ufkKCrbAt6/MJvB+Lu/e/ipt9ykESm3U+z1E0zXkMOjn9aMmBcS9TuqvSUUHBArhkTs/RzWqROM6Pv/DArImdjM0K0rx5jv/ze/CvbqFRFwmlK+RPb6bhkQaS2ovY0FpDWUb7mLW6V9QvOUO0moSRPIU+mYkabsjweCNQSJ7PHiOBbFkhgiNjCFvdbJ59xTCWRqnvh1B93QbWnaU0Jgoer+AOMVL0bQ2dD16hJBE85VmJu25m2RMpnSDQuahEDpzPJWsboDLrz+AKgM6jWhEjyaD7ZyemlttJOIy+qBKzJNkckkL7rcsKIpI6ZRWsCeoOVFAW6MHexMUTWuj6So7cZtA5yKFjd9dxHOfXEX3FAnPLj2Vnz7AwASFqRkt6E5ZEHan4VrWzricDowtetLOp8I9ExEdHbNlHOdkemcnCL6bSyBfIGnWUGwKCatA3Q8lAORY/YSyRZKeOB0Xm3DWgOeEhrFdj+NLK5KoIesVInd4sblDhHwmVu9bjbNBQb60n1hOAm+bk58emkbrcoW+eTHiaQr+pcGURkkAZ1aAeaPPM/bWM1xx0TF2tg4nEdYhRQTSbmpj6dKDWA+bUGd5QYARb67jttaZXFG3CO8YlcGZcQ7Pfg0xDu6zMd77aBGfDk0FoHmpCWN/Ap1fIFjuwPtiAVcXnSB2wUHokhC3vHc/+kEJZ4NC+yVWZjy0lq93TuNfpn+NmICj+4bT8MZwumepuKZ1kyyLkGaJMFCpIWgw7JVGbO1JzgVzeLD+PNGlU3jLV4QgaKgXjyc2u5uc5w/wZOMxMuekip/Lf3ca4L/ehfb3FlT/TIj+XfiZEP2dUfbQIWp8KTfTjs5qyh4LULJxLYunLmHnpr/g/yCP4i13MH5dNef+pYC/lW/D1J+k5+Eo1o2HScybiN4d5cRDr5FY6CM6I8jJz0bjesKAKKp0ncjC9oyVzCqVwi8Fho57sB5rxTcpRseGYs7fZ4cmC8FsCcOASPopjVCuxvL5B7F2KVx3z06k0iDa2AB6n4CgClg6BGzfWeien00wHwRVQCoIEclRaLoGQtf56J+oIrpjhNpt6IMag6OhuyuNyK99CBoEey20LhL553lf0Oh1M2XVCcJ9FtJOiag6KPlyDWqDlaplw5hjrqVq/Eae8Jxj35ivmGNS6YtaOdedhSarSF0GyjbcxcxlJxATGrkuH6F8hXC+guNvVrKvb0YbHWCoXOKtF19CTELXDANtt4/C1CvwwK1fMnhziJhTw3DSDCIc/3As4SwRSzskAnpcZzWwJPGXqVx4aSwD4zS6nxUwnDbTMwVMM/t58vH3cJ0SqLjuApGCBO5xvZzfWs7Br8bROVOkfa6A+3yCpEVlTft0/FEDwQIVJaBDDoikn45jO9sPAty6cgdzR1+g/cEk6ccFCrb70Tmj2DKDRIMGWpYKpB8V0bqMXHH5jcQSMp6KfoZGmHnl8dfYVjOSpqVvgzPBZ4v+TN3uEkzVJvp35GJwRhkcbSd/Sx/GGf384a9XE8tMoskaa67ZCipor2SQWaWCJhCbEOTsva9j3uCkdTCNgdUhFGMqtfj6b+9mbE5nihw848HcLRAdF6a/3QmagGlBL3JQQFBhwfzj/H7TStJOi+S/LJGzpoGOs5moAR1NV77Fba0zORuP8K4vi9IfVqMXFRJDBhyjB9g08mMWmBOca82m9oERKOlxtDczcGX5GBilI2c3LJ1YTdpWC66PrRgGNervlQn1Wrh81BkGpyRQDRrmYV6Y4mPyDSeZWtxM0QYR9+kEFem91HVkkDSrGPoldD6BicWtzB9xnoarZRpWmtBaLQgJgYRNYOsnM5AjUFbYg/mYCU3UqLzqDMYeCaXHhK845Y6q3lVBMEcmHpPp2FZI3pcy5k4R9zGJhEWgtjmLqXPPknldC6JfRsuLUHxJM4IGU+8/ipYVI3ufyP4/T8YwpKHJ0NzlpupoOUmTRjA/1VvWtOgdnBcgmK9BUqT7EgUEqKhsRTcgE8rTiKep5O9KcKo5FzmkYbDEodLPike+Z2i4SMKi0T1HoWr8Ru4du4doQibYYSdzpw7PHj3d00QGe+ygpSIS/IV6jE0GTNYYhn6JP43/FN/IJNF0jbRXLey6UEH1R2PYvmUy9k9sTCxvJu0cxF7IZtORSQQLVO6u2MsPl72IYQiOfTKWyOPZiGkxCMhcdnI1Nbetp/3OBAm7RkgxcOtVOzCW+2hfk0DTQdr9LXTNkHjj6CySuTE0DfS+1DQymCMRyVXomqWi5kZ5teES0uqSZE3oRtGDrS6VUF7ymkrss0zeuuJtGldKDLxnw1ckc2LDGF7vuIS9b77F228sJTfDS+s9CvPOBGj44zSmGSX081vo/noET57Z899XZ/NmnvwHnig/478SPxOivyOMhQF8N06DS9vZ0VnNqFfXUXNPJjgSdPyigMsuv55QjkDGPpm92yspv+MIxV/fSdQtE6lxInk8CL/q48Exu5AEkWiNg/iAkVErz1N/rZXBLgeaTqNzlhlvmUTbfAlpeADLxgS538gkbAKiOYmzBgquasTappG4cZBEmsLXW6cjxTTe+XoBr034hKKnk+Q9d5jFSw4RnRNgsFIlc18/cbeCJmmIZ63k7gajPYb7NTOaXoVOI7Z6CTGhkXVYxdiip6vfgbFPRIiLeKpEnjiwDDamc/JP48jZJRDOFIgWxzB1SiQyE2zZv5lRetO/uXdjHR1cmPkR1owQilnFMCSw88woeibq6N2ZR9oZkRUzqghlC1h1MThvY8KVZ7j1Nw8xcGmUlSv3EixNEnXDa68uJ/slPUmrRjRDReeJcNO67USmBhkcr5C7U8TeFGH4c36yD2gY3BEcpUOpqVWLmpqeJWR++dZtWK/rpKqumLXT9mB4zUVkeBS9V6Pij81UvOOjZbmGJmvs+a6S+H43GUcALfUC/+GDd+i4LJOeSxO8fmQOjQE3ts02gnkC27/5GLHGiuMvdi4fdQYpJJJ2IYS9TqD+OifhC07UzzwMVKo88k/rUEMyxVvu4MzcN7jp4/sYdWktpx98nbw3T2O3RIk5Bbou8eDzWyj+tJf84j5KNikMJS34Riq0zxPxlkpYTxn4cMr7LK+fz/6X30C/z05o0ITtii7i+9LxHBM41ZlDf6VA2pMtZO33UvKqRs4uAalbz2C1B0GD3kvi/LhhIgsXHUWTwFtmorqxAMOAiM4rMezju7jec5Drq2/lmaOXYTAmOHWyiJySfoIRA9dcuJ6SnbfROO894g4d2dt1RJ0ihY4hcl44QHiVlx/fn8zAWI2eKSIJu8Bt4w7gOiGxfe94hKhE1n4BnaQg/uTg6IfjOLZnOM1LJLxlOmq/Lsdyysjw9YNE8xIUbBnk+MFyrFKMkvJuzMV+xMIQ1jaBwLgYoTwVc7cGT6UTKE/iOaFSvWk0ej/gjhEYnmD+9JPEMpP4Lo2gDenR+zT6x8ooJojbBQLFKrdMOkDVztHUVRXy1hVvY//RxIXWLOwNsKe9DM2rp+KBswTzBcLZAr9YtRfHfiOSJ4reJxD1pITeY19YR9wpoBpUCv4GKAJFlzbjfy2fhEtBkzUemL+NpuUiOZleLL0q+6a/wY3lR/iobgqxdAXVEwdZY9Lv7mJLz2jCTXZ0XhExoTE0MuU2MzXrISlgbRa5/P59eE4lKXYPIo4M8GT9UnK/F9B7BWJpMpoqEJ/rQ9FD+cPnOHa+GF1ExTtMh/uoTNmnQV6vncUTnZcTztEwDGl0zjAhtpgw5wbp73AAoDVa0ErCbDs5mo0vLCBW46D08TCJzDgXDhaTflLD6QohtxuYXdxAoETFPy+Ed3QSnSeC66SEyRKnt8nNDf/6N35TuhUpAckZfu6dv53GFSYq7jjPHdtup+grlZ4WF6NvPoulR6VxWwmTH7uLzCtbWZR9Dtv3Fja+sABjn8jC5ato/2IUN5ZVcdMn91Hy5RrqX5xG85QIZZ+G/otOkb+zoPpnUfW/Cz8Tor8T5sw4Q96Ks4gJCFwzjZiWYMayk5Q9cIjGBe9y1e0/sG3rJ4SKkgyOAXsjdPxqBvYamZW/2YmSHaPl9mH0By2MMaYC3j695k/oByVOfTuC/HFdiCEJ5zkBKQbKpABCRhTdj3YaPyin5+oIggqaIuKsi1D3QwnT7z5K+EA6I171oRg1Wq5V8Uzq4VfnV1B/nZOG5yezfdM0Yt1mbp+1h55/FXAfkygc10nMraLoBcTjNrpmGtA7Y9gbBAy+VBt2x+IkyYow9kMmLB0axuwQkRVepEGZoeHgvK2N7ukCjmm9GFoNeE4maFr0zv/03g3/6SZqgxmMeWkd4SY7ubshlqbhqNYTd2pEPKnR9672cjQRLnw6HF0IjuwYTd94AUHS+HD/RdjqZC5aeIqsA15ueHsLZZ+FKa1sx77DwvotCzEct6Lvl4i4RH7z0UfEXo2h6ASMP9mIHXIT3uehe5bKsdUvwX4noQKF0IYcbCcMvPX9pbQsETA0GPGOVjn3+zx6f6+SnzfA4kknSeTGiaVpRDwiOcX9oMGIt9Zh6VGZPKwZvTlB+7Ecemcl0WT4JmRmyy3/SvsVSY68MgFri0jUYySUDyVfh5h88QX6ZiZ58bIN9E4UUy82RcAs6smb3sHJtjymPnoXrfeOQf7QTWB8DO/kGJluH/3TPLQ1eFj//is86TmLITOMJkCoMIm1U2VV1Wq+LPuO5waG4RuR5JEZO7A8asJzMs6uZ14iFjBg6hY4Xl3K0Cg7DStN9CyLYxruJelQYXiQ7O06cq9s5sQzExh38xkyVjcj6RWs7RpqUQQ5LPBY7TICXjPXjD6GyxrGeU5EEjQyHQFaOtIZ9kaCymfXEXHLmPoT+BaEOPNTGXV/nspQl51AsYqQFBBLg/jLk2x5+hKCc0OknwBrnh/vVcFUcnWJQiRTS8UnyBpyWCMwPEEkW6XpKSPpB2Wal7vQJI3NNWMJfJyLWuXEeMiKf3oExFTHlrM2TOsCIwZXBF+JRN7fevFNiOE4YAQNdp4Zhf28Dv1pM1JYxDsryu3XbEdIprQalnaRr5vHkntRO+YKL8/deCO6EFQWt2HwqRi/coIKFinOxIXnyDiR5KumsWiyQOHbIuGiBM6yQYo+7SCcoxKZGkRICvROSK3O6nvSqfxVNbJXQooKfPXgfApLenGZwrRfrnDD9ffwddtY5uTXYxiQkDv16Lt0+OeF8b2TT8X4VoSkQDhDIulMEktXkKKwd/GL/MvdH7D1T7OI3jFE98dF2M1R+k5kErOL5P/+AP5CCVO9gSxHgDuX7KRq8xiEuEjW/Q0UrWjANwzafq0h7krDGzdT+LcImbc3pTR3DgXhgANbZpAxL68jmR3njjE/IYRkLr73MKqsMTQxHet5QypuokjE22XnyKoX2d9WTOnGKIWeIYZ9FCfuMxCeF0Te7cCR5+Or6+bw6Gu3EcwVODfjY9756HLqr3+D/WeHpYJArwZEjaiiQ7mlnxXX7SX35kZqG7IJq3ocDXHSPjhI+qkEdTeYyFtxll2VaRT99iDD7jmMFE0RivprLf91YuufV2b/MPxMiP4OyN+lcKijkN7Nw7F9dogDL73BnF/ey72ZuwC4+O41pMsBfGqEG6YfJPOwyuAYjdwfAsxbdYhX980Dr45FVx3CtNHBb+9Zw8Qn7+Kq7++m9E/1uM8m6TyUQ8M1bzB4cZyX1r2JyRBnQkEb2Ve2oFvRi9Zqwd6iIMgqcWeqNHPr7klkH4yiSalGc1Gv0NHhQtVAignY6wXiDg1N0vi4djJpL1qRoxq9ASuuMwLTf1lFwqEx6tJaEr0mzFf24CsF6eRNDmAAACAASURBVJpenMcMJKMyvkkxVv/6G7RTdgKdNgwDIuYuga5vC1EtCrp33BRuCdA1Tf6f3rsJT92FKGqcbMsjnKtiLAzQcSkoNgUxrpF2HhR3Au8lEXyNacRcGjnf92Gf2425U+PKuYfJ26BjzKhWxCQc2DaWmoeMPPnjlTRdaUF9yoNnfx8Z43o4/cDrqVBGq8C969fS1pdG7+IY8285iKc6kepvapcZ893dJGyQflREjmgEJ0fQJI2c3SKKUcNxQWLY2wnSzWF6jmRxuKcQRI30cb0kzWB+2p5KtzZoOE/0k6YPk/emjvRTGs5qHTG3ysNf3szC/fdgssUI5gloIvjzZeIuhc6LrXhvTyftmMwb1/8C11kNU3qYpiveYuT6dTR1pJOMSfjKBGJpGn0TBcpfjVJR2J3SSrWlJgOLvnkIAEURuH72fsS4SO/SGDRaKN5+O++emYG5Tea1z5Zy4X4TPRP1jP3yfuQ+HfKcAeSgiKIXcJ0WEEUVfkwDFdz2EGJSo93noHe8yFR7Ix0+B2KDiYF5UTI2GzGOH4SPPBiaDGz5cCaxpMzQ5ATCqx5M/2RGUwWal1jwjY0TKBJQf9WPY6cFrSjCyFGt6PtkTN0iSXeSREwmv6SPxM0DxMM6BsYKPDL8O6LdFmpuXY+mV0laNNScKK58L/55YZBVJk2rJd5hYcSdZ5m26DTWVhGhzUTEI1D0WSfBAhVtwIBrr4H0kxri0/3MX3CcucV1ZB6NULahBZ0pQcaRAIXFfVjP68n7ppNIjoLrLAhdRt7dsIjwsDi5W3sJTYhgNsTp3ZGHcjCNulv0/Pp3H3GyNY/4rYNU/WE9pi6JrYcqOduXhbdU5p6KvcQv8SFGFQ5c9hIVrj5aV+QihwXkM1YcZUNEC+PcsmYrqiKx79OJyGGBiXMuIMZVzPfrqN1TQvoBHf4iI329drb9NJ5oYQy1KIK9Hi4vP0vfeIELLdkoJi0VeJkVwJwZImmBy46s4Vcnl8PyAfrr3SRsAv1eK4IKHzzxIrVvTSY4Mk5arUL3rjw+rJuKMNmHvVbi3I5ybsnZz7Dn69E0gaKrGuh5p5iiP9ZRZBnA3K2xdMoJEEC/xcmIK2rQtet599wMNEuSM94cCnYm6b0sxj/f/jGF2+JE01WQVaa8/xDRbgu9E80kFIlIpgFDlw7nVxasi7uJnnARyzCz8cHnSdg0yvetIlScAGBEWQcVM5sQZJX3573LkvSTKJs8fHhoBpHZPSBqHKmUaLwJhm6ZzsxnDuEuHmJHZzVaMqWO73h0BsW/OUjr46kqkNJP1/5nHyU/4x8MQfu/KNvbkJ+v5Tz8wN/1mv/tX8PLQ0X0J2ycWFJAsr2DGy60s8rez4i31hHNi5OVO4T3UCZ5P0T47rP3mfLruxgYr5F+TCCUK2Ds0xicHcO110B0sR/1mANEiA+PoIRkJJ+M4k6g79ChH+3D8oWdwcURVFUk72OZwZE6nAu66P8xG3nSEGWufk7UF1Je2E3fxnx8wzU0Z4KiTwRiDw3iMYW4sKcUS6fGYKWCrV4mMS1AIi7jcgaJJWTkrU5GrT7LgcZS0rcaCBSIpNUp6AIKLTeoDH9qiIbVWRgGUs4wfY8OY79A1KWReVShfUUSqz3CjglvcyCawwqrH4DnB0v5vHkig/UuGq5+gxub57D/7DD09hiKIqLXJ4kEDMg9eowVPqTdTnwjk9jPy5gX9dDd7WRMSQedHxWjGAWm33ycPd9MQFBAUMEwlAoj1AUF4hODCDUWzF0wetVZWp+uoH2uxDvL3uK2b+6kZGwH9bXZPDbnG97/3ZUMVkhY2zXEpMbAmFQeEBqYuwQSc3yUpg9w9lgRpZsitFxmRj/ax+y8BnZtmUisKEZpfi9uY4hpzkZe23oZZRNbaR1MY2x2Jw3vVDAwXkXTa6BTEXQqRR+LGNt8aK2dNDw2loQ7iRAT0XlF4m6FnN0C+//0JtMfXkswXyRYksTgiiCdsGFrU5Fv7CUc15H40U2wNElatYRnZRvJ32fifqqZs1srGHV5DWe3VPDuHa9y/b47kTv1eCb04N2XRbg4gfuQzKX3HOS5zGpeHSrkx6Eyul4sw7/Kj0GXJLbTQzhbwzpiiGBNGkpGHMsZA8mpAYy7bbiWt9PU7sFWbSBpgrhDSwl7V7VQf6iQRGYCFIHLKk+zY38lmjMBfhnDoETpnCZ6PyhC1cGY285w/rVRXPzQYeoCGZyqLqZ4c4L++8JEojqsey0oRgG9V6N/WpKyT5L4ioxI1/YC4D2UScXcBlo+KwUx1X8Wcwow1Ufh3QOce6IAISZi6hYx9WnY2pO03ZTE+YORhF0gY0kb9XXZuI5LDI1TmT/5FD/srkSMg6qHgh1xmpbJiHEB1ahhyQ0g73ISdUPMpeI6IxAohGWLD/LdW9NJWAXytw7Q8LiBRK8JTdbI+UEgcIOfPIcPX8yI7hU3Lcvgtmk/8t6JGdiqjQBk/xTAW2FhygPHqLt9GL1THVg7FIJrvNgMcVpa0xEiEsVfJQk87Cf2vYdIpoYcEpBDEHNraBKsW7qNPx2Yj6lVh7lbI7HYi7Tdia9CQ8qKoGkChlPm1HNzKWQeEEiaBGJpKZ1Y3AHuMwpD5RIxl4riUHBXyShLhvB22zB26kiaNZKeBDk5g4xzd3L47fH4SyDziIoU14isHUL5Jp3EZV5yH9d45KuNPPqHOxmYpIBBIXeLzOAIiaRZwzFmgPj36chhjdJVtQzFzLTvzUeKQ8KmUfi3MAmHnrZVCYwnzUTGRJBkBfdmM71TwZgfwP6FDf+KAIlaO2nnIXBFgGiHFUeNiL9MZdivT1Dz6lhka4K6OR8A8GhPJc9lVjO26jqyl51nR2c1Z+MRHiqaTv2L0/6XzoLOP75MrK3tP7SjMhTma9m/uv8/col/g5Z1vzymadqkv+tF/w/FzxOi/wBKN0XYEjayMKeSbaOcuOQQFx7KR5g4ivVPXUXxljtwTOmFpIjyaQaucwqxx7xUPrsOz75OMir6sPQkePq2D1OWc6+OlQ98T+KsHUSwtqbIqvOkDluTgKhLZfWEm+30TtWQdQqaBvZftfHCXW8zGDJzzco9iD+kUd2Uj7FVj05SiDsE9LkhTDUGWm9WGNqfReixHIoubsHcp6Lvl0iaIeo34NhnpGr8RjJsQcQrBjjaUYCuzoS/WCStVqHzYoHsxxtw7TXQ/oKRuEtBF9IY8bwXMS6gGEDvFwjky+haDUQjeq783S/57V//R6v3+r3zGGhMo/wDP2Ub7uJkTw7rpv9AzvsGAByWCJ4MP+bhXhIJiVCOhrlFJpqhoW3wML28kasyjzI4K4Z3ZJJ9myag98Kly4+QO78VMQEJu0ZsQpD4gJELt68nWJhasfmKddw6fzdrP16LtUWk4Uwu6VUSX3RPYGC0RMKm0TdVQblukIkX12AYFJDiAva2JOEhEzXdGei9InWrDMxfdJwvJrzN3vZSkmYNqzNMz5Z8ju2vYE9/Bc4a6A1ayXrLyMVpdSgGMPZKCHEB0aCgBWXaLtWRyLDSvm4crrMajtM6Gle8iTQigORIELnJy+hX1hHOEok5NWS/ROl9vShGMPcm6Whx4/OZ0QU0TB0yikmgbU8BJX+4QOdLZRgGNWakNRAuTLLm1XvRtesRNAhtzUKOgLVOR2hRkBPrxlHy5Rpe3LuQ6n3l9I2TUDWB4CEPE647jQAk1VRiubHeQGR8BN0hGxnHg9yavx9DY+rZJRypqZ79hg7qujJQC6J8dcmfKfmrSnfExqjxzRgbDOBIoPNDy1Aa8+7fz+A4lVJzH74ykU3HJtEXsfDy5R+y6+N3MXztRO0wY+lWCRQrRJb6cZ7SMTDSSDhbIBTT4/y1AU3SOHewhMicQCqHZ7yC3qdh/cLGuccLyC/uA1HD3KMRvdxP80oNgzHO4FiNB9dsQv19BkZ3hKGxKppJ4dBfxyNoqZRmTdIIPOzHkh9AjAvofCKXFtTinxahcHYL18/ezwu/ehNNguP3VhIshFCRwsAEF5KkYs3384e5G4m6RMINDrr8drp6nIx+8hSSJcl7u+dgrDUihzWChSqBIjMxp8iRvgL6JziwX9VJ0iwS35dO++kscnZICM44vhI9/X12Ih4Nc6cA4/yEs7WUecCVZNvqi3Ge1BEfHsHg0/B32njv0ZcpGN2FbY8Zuy2MJsKiJVVI7hi+UhGDX0Wb6SX7pwD3X70Zb2kq/bxwawJLvQ5VguK0AYxdOhxTexEUIC7S2eliz+YJWFZ0o1hVOi5TEO/roa/diXdmFONXThJuM3dV3UhssY+0k6k1XMwuEneoaAKMcHcTGB8l/4ZGTnXmcEveAS5deoxfr/4MpThC/V0SgyN0KH49oeIkcrMRpduMrTWKGBVIJCSsrVGyHAHuWLqTnc+8iGGPHc2o4B2dJO2cQP+qCUwfVU/p8wke7alkwVU381xmNQtzKrmnYi9SWTELzy/5d5Ghn/F/Bn4mRP+baLj2Depu1rOpfzL1H40HYOs9c5BywvgqbAy79xxNi98mtDuDOZXnufi+w+TeX0+O1UfZylo0ORUwp/PH+c3JZWS9dQzNovD++WkUbo8QKYwTTReQdUn0fg3xsn7UkA6tLJwSPccFol0WzNUmlmce59EX7yDcZeWTb2eTnOND1KlIEaipKsLYr6HUWwmXxcn5XEfW4ThNywxkm/34VvlRSyNYpvZTtFEgYRUo/XQtDe0eQofTiQyYEGNgGISM+xsp3RjlxHcjiLkFXG9aQU51RF29eR+CCkzxESpUcFzRSTwjiemwBVt7nEVLqhj/9DomHL2Gko0JpPQYA79PoEkaapWTvf3lRNIl0AS0DR4Chz0IP6QR7zVT/qemlP23NETvvAQXpdXzzOcraZz3Hvk7QDGBLpiqPaltzMZfCqotyUVFTQjWJLPX3ImpW8BRp+KtjPPBuanEi6IEi1TM+QFEBc435AAw7qI6rpt+iLvL9hBMGAiWJ0CFtkVgqdezrPwU+glDmDpkdjZUsOibh7AY4iRdCZLH0rB2qGQc1TjfnYkqg78ujUi6zMb2iXiqQ8QdGtiTiB1GDH0ySnaM5sVGjP0aggpiUuPy+dcQ7bQwo6QBX0MaebsC+EckyPkxib7UT+PaUuJpCt57AggxkS9mvoF3RoyyeY1IlwwQTVf5/sQovGUSnuoQmzvHcdnEU5i7VYo3B5m/4HiqUqNEITIugskQp22hhU8ufx1jt4ySHyWWG0c94sQ0pZ+jX4zBdVoj3GJHDglEiuKkfW/E0aRQu0bH41VX8OWtLxCcGCHhVLB2xOnfkofRFKc4q5/HW6+keamO6qZ8GnaU8OyqD8jeoic0PkKO3c/GnRdh6pTwJU1Y2jUcp3T0nszkCksYgEChwLDfnaJ9gYaxRyLSYgMt1acnRSESNtA71UEsQ2H0jHpmF9UTzlbx5A/hq9Aw3dJF+hGJ2IYsSj+LMXRplBL3APpuHeIhB3nDe6gKlNAxR0/Ub6BkUwJdjw5BASU3iqUz9cdE+SYd6+d2knYVQYFjT0/EeM5EbWsWT2ecZvV3t1N783rCOUbEmIAmasSvGuLxMVuQtzt55vwikmYhtco75MJ50MCBriIsR0wI7hgI4JsVBSD/gTqiLki7ycvgWI3Qhhw656rofRpyUKRzkYIa1CEv7yPnW5lFC46iypDpCGBtE9CEVMdh62U25q4+RP5HMsaBBC/M+5Tlf7uPhCKReW0LPr8Fy0V97P54CunfGimZ18RPr7yJ/a92+sZbWW6rJWmGuFNj5kuHiGSqrL53K6cPlhHNSqJs9DDyokb0aVHkXh1Ji0ZPVRa/mH6EqSMaaT+eQ94OkWyPD++iML3jjUjnLQR7rAxNSGLwwsB4NdUTl5bk4L5RuPcYqPmhlGSrhVefWskPX0/kiW+uRo1LuPcYyN3Sg2SPoxuSUqv/Yi/eMhNKTgznNguNawXGpHXy9rcLmPD5gzjr4ohhCSksYu1KcvTJ9Rxty6ftMdj61xk89fF7FH+Tion4evlFKPVN1DZn0fiv0/8rj5T/gZ81RP8w/EyI/jdQ9tAhHusdQ+E3cH4wk/LnI9h+TEfac5zia08xMFbgUEsRF9+9hikrTrGvroyIqufJvG+5yNnAsTMl/HbHJuJfZ1C5/hSaJqBt9SB5ZdR6K413Q3lxN+knY7i+tBDOSk1hxbBIzsd6TF0Shn4RzaSgzfDx2gsrsPQqGHolsvcnkfal3ByhoiRJd4Kj/7IeTQR9p46hYTItqxRMPSI9S41MzW4l6dejfuOmY46MHNYoG9eOwxkmPjyC/ZwOADmqcbo9l/pbZBSjhmFQI5wh8/SsL+mdqOP3m1eg88OyklNIrhjhT7JxZgaIzQgQc8qcfWAMFTdcIPGTm6Yr9GjAwJAVLT2OJsHZphzSb2/Bvd3I0AiBhVdWkZjtw1YvUf9yBvHJQdyOELp2PdfYLlA2qznV5RTXsDVrRJb6sdeLLK48RcZRFaM9xoHvR3PfxB9o/YVKLA0W/tOP6Pp0SOetmC4YyR3RQ6jfzNBwcJzSI8bg/NZygoqBj9qn0f9OITn5A8xacgLXCYncZw+wtXkkum+dJEcFEWstpJcM0ttvJydvkEhJjJx76ulfEca5xUJ0XoAZM85x82PfYpQT9P82iqHEDwEZOSxgm9iPlhSpuu6PHPn9eg6+8AZDYxVqb01j2H2Hqd40Gnu9yPbNH6EbkOmcJaOedpA0aby9+B18fjPDx7TxaPFU3LsNxBUJf8CMYVBECkiEhsXpmmFllLObbcfG0j0/SfuvVbbvmcAta7fy/MK/YjDGCZ5xoeo01rxyLzGPgr7ehCvTT3hYjFHp3YTHRkgaBSZPrqXm1vVcNeEY/ZMVeqaIlN9WTUa6n2urb2NueS05P4j03xMmlKMhf+/Epoty+mwBUkRgZnk9UhTu330Dwx86g6YKJJ/JxD5ygDtu2MqX+6Ziu6YTR3MSOSTwaSCN4i13EM1J0Hp/JUVfq8TTVFzDBslY0crg3CiBUgXrT2YSFgEhKVCzbRh7dlRiGBTp63QyflI9vogR3zAILfPTvNRE1ld6Mo0Brl2yD11Qw78lm6o3xnPh9vUUfwatC1Lp476RSSqeCdI/RcVRK6BKAt1zFFAE4k4VS3OQcEkcky1K8bd3IFoSzLnjDp545l0SpRFkvwQ7XPy2ahneERqhJgdRl0bz+nLC+Ulm3HYcjyVEoFQh+2s90WyF9K0GHBcEesI25PFean47DHOhn/5LYojWBJFMgZgnSe5WCfsFmeDeDDqvTLC3vYy8HQO0VefgG5lMObOKJHRBOPrEJPrG6egbZ2BD91Q0a5KOVjdNu4swnDFhfDsNZg8RdYm0f1nM2KrriFsF9AGNDMmCJmpcuH09n38xG3O3yIuH55N2HkxtMt4REL/LQd6bOu5eug3b6AF0o/x8daaS8xuHc+OivXTOEvBHDZgPWnAu7EI1aOTtFND1yalYAcDcp9K09G3EBNhb4kjjfEyaXkswX8Tc/f+W5f6kR0zC+QfcSE0mxNIgz9/8HqWufsy9SbIzvMTSBMR2IwdenQzFITzHoWuGHnu9iGdUH/2jdVyy+nYSQ0ZiUT2jrrzAaH2C8rVVND0zHd9oFwBNl72DKv8D2ITGzy6zfyB+JkT/TqSdFej/tpwTSwpY+fx2HJfXU7PGQeDifupfnIbtx3QWzTtK7vt6hLW9tD1Uijak5/XcQ1z34iPYpAhXTjnOjTvXIkVTawj9ARtDH+YjhQX0w/3kuH3UtWegPdpP1xyVUIFCsNqNtUWk44YEkWyFP972LkJI5qK8RtzvHERMaggKhDNllq/eg6najL1Gxn1Ax5zb7mDYlBbK3mghVKSQtteI+0yC888Wsv+bcQgJgcHpcX6x8CBDkxO0f1+AtsuF0GYkmqHhrFcwXN2D+YgZS70e5wXwVkAwT+CtR1akrMn5EQCmWBpQhgx4F4ZJKBJrR/2I674WWhcaqdkwnGhlGNWkkr3JgOWEicZ57xEpjVHxSpSBiJmEBeKeJJuPTCDSbkOd7UVrsnDTiCoSmzIwjPaSLln4W/k2/uncCnSPdBNYHET/nR1bW5KTz1RiO9uPZbsVgGGGbtIz/DhrVb7vqkDnF0jYVOJjwgx9l01hUR9qUQRNhJhbZcySC+z+bDJj0jqxre6grzqT429UMjgtTttvZxButhPOElAVicwqheAhD1NKm7EuaqS8oAejlCQZlwhcESA3zcdPx0bw3OFFDH1QgK/WRY7TT+5uyN0XRfzczYhnh7iq5louemAtlc+uw9wqo/eKBK+eRqgyQsIK449cS9ahVA5NNCOJdfgQr7bPw2SO0/FNEXV/mcDvfvMXgnEDSkQiVhwDwOoOk7DD/s5iLpt4ipycQaxf2EDU2Hr7bB7ZdzWRgJFJsy+QyEtZz387bzNokNiTDgmRTEOAhkvfZ9ra47T40/giaOfUUC7GXpmylxpQv8shGtfhH7Dw07ZxdE8HnaSQu0/BsLiXrpAdW4NMIjvOsW9Hk/95CyP+OET1e2PQtRnonGnA67Pwp/3zGTO+CX/UwJI/7EJMwpOnlqAbkEk/LBMdFSH+4CBkx/jzqA20DTkxnzJha5Twl6vkLGlBsyT565oXMQ4KRAoS2DxB7srZjb/GxVWL9hMJ61FzowyMkqjzeagLZqBJApbLuzGs7GHh+SWM/UM1zlrAoCJaEyg2I5n7U+W7UQ+U33UcR42IOS/I9i0bICny0rjPyS/qx73LyMAIHWt+uAWh24CtGTJeP4DYYUTQIKMKrJUDyFENISGyfd94mg/mk35MJHbLELr0CJGVPrwjNMQ/pBMOGsjZp+F610rj/PdI22Nk5IJaBHNqda7JoEkw4tfdBLxmzt/jQLEpoFMp3RQmWKwQHBPD1B1Gmj7EqUde563iryEqMWJYB5qUmrK1zxMINjqIuSBhhZvKqlhy/156L49R8tUaFJPGrNO/IFocIzQyhq5HR9Uz6zFN60cqCtK0Mp3uqQa+ve9ShM1uIq02XD8aiGRofP7pHFSTgsUQRw5rDOzJRtHDxb87iFoYpWBSB5I7xsAYgRkPriWerqDJAnMLajl8soxwvkIoT+DV1W8yZc2JFFFzR7GOHUBttPLw8ZVMdrbQflOCPq8V//AkhnI/sWVeil4VGRwtwIgA3rEJulvcqJP9eIfp0fdLXFxaT/tLw7ji1nuofXMyUkygd6LIjs7qnwXU/5fiZ1H1v+f7AyKR4jjltx/lttom3r3xCtwvtXPu0xFMvOEUP1SPpHxtFaa9mXw9bAeXrL6duF3C0h5lcJSZLx9/nrmf/ZJ3lr/JnZ+vQRPBNmIQbbsbYdEA4aieeJ+Z9CMiwVwB+0W9OJ4yI8QUmpc5eOjqr9neN4r+PxbTN05GTEIkV8GQESYW0uP6Uc+ie39iR/sIklvT/x/23jtKjupc9/7tquocJueclXPOQhIiB4EAkYMNksjB4MSxwQYbY0wWIhgwSWAJEYSQhBCgnBAKozAz0sxoco6du6tqf380x4fjzz7H372+Pvdbi3etXt2r+tnV1RV2vfWG50FIiaLDQAlk7TFwnRpkxBvVrNk3EW+1hjFzAFUxkUDKSjfClDTNs2BaJZZBhaRJHQxJ6mRPUyGuz9zIC3qYmV3HZ2snobslFAUo+oOkb5ibqEfg6DHxXerDttFL7wQdTIHqUxE5ofiDT7sdZ7uCs02iXNlJnqef+v4UBgN2tENusuc30fteLr1TYihWA/d+B4NDdewpIVwb3Zx1+w7ePzmGE9PfBGDSwcV0NSaRmtePqpjcU7qZn350JZl7TdonK9h6BalHdfpu8hE7mIRpkeiFYZK/sDP91v18vHs8jhYVMXGAswuPs+bgeJLSfASOJCO1eA1GXkE34VWZeE9HqL1Mw9YVF8Jlej+xygRkRQDTUND7rWQW9dDekIK7TiOSKCmY3EzTzlwiqQbCpaO22XB0CpQY6K74ORUeGWJCYQMnujLw++w8OHE9X/QN4UBrHt41HqJX9uKyxjD+mIEWNOkZoRHI10nfrdIzWiIzItwwejc/T63CkCYXTL+I2htycY3rxn8oBd0lydol6Rmuklhj4mkMU38reD1BwrtSSTsc46UXnuLan97LnsdXsjBnLD+rPciKtjPYe7iUNWc/x3ib9S/XQNG6H1L0vqT5DAtLzt5GghripbUL+ebGp5n1iztJWNJC4zc5mFZJ8ZowaijGxnVv80DHGD76eBpJkzoIRKwEfHZ+MXEdzzy+mGiCQMzqw++zY/otqN4YGR9aCaUojL/hCNtOl3DL8B1sGuGlcJ+Dz44OJ/FrKzEPqFP7ODxpFcWf3wgDFly5PqLHE3CN7KW/PonS90LUXezEzIhgrbejl4Zwfe3AmDlAxnN2wikaXWMVYmk6Wo+GdVBw3uJdrH93GuFRIQyfBU+NhhqG/glRUnZa6Jmsk7tBoXt0PO2ddkindYZC8tAeek6mYNpMbF0qpga2oQOkuIKov01BDekMljjwL/IR9NlQLCa5b1poukoHCdNLa9l+eAioktRdGr4isHcLfIUmZqJO3kcKltvbiT2XSW+Fhve0SdeFYfRBK670AMFmN6jgOq0iVQilm9h6FXS3JPWQpHuMiEd3uxXU8f34O9xogypGRoT8rF4amlJx1salQURWGCOqonRZMbw6aBIUCSGVxGMaWZ+1o7wcourrAhIqemFdCoPFYNok9k4Fc5wPy24Pji6Jvd9A3N5J7OUM+oaoRMpDVNzVQO3dFcQSTMreDHLyNguZ6610ThCYTpOi93W6xtiIecFzWtJfDrH0GAmHrfgLTNSQIGd7jJbZFjz1kHwiRM+PgwQPpGLrhfxL6ji+vxBPnYJvRpDZxaeo++VQ7nx2FXd/fiUlf9bxZ1vpGxrnWdLCoNvhV7e8wb3rruZ/xf4pRdX5eTL7/n9u48/pmjuczgAAIABJREFU2+/7vqj6H7TvI0T/oD187mpyzmjigWkbeLdpF5e5B6j5gZ3dlWWYFvhj/g48NRo1r0wgelGU82rOpq/Cys6nX0TsPoy7VWdPOIfSdwf58b/djLsR3r/8Sfq6PYy85ihLy7aT9q6Tu+dspGdeGG+DiW9HOqGHfDRckEDKUYNnX7uIw425tF4eJe2ITrAwRu5miR5TcVTbiLkFb++fgrE+hWgC9I3VGSwBBDSdLWmbm8yhO0ZjSw5hnddNgjOEKQW+bhcdEyzUXyLIGdtGQo3Ac1rS2ePlqxPlRFtcDJZAQUIfW1ZNIumkSe6WKI49bgynheRrG4kmwECRgtiTQO/kGA/M+BQ1oGDxxeUoSp+IMXFyDYHhYfrODZBgC/N1bQEjUttI/shJaEiYro/y6B1ncPbIoyiKJOncVoRTJ9xvR7uki3WnR+D6zM2esMHM226hu8dD/YUvEdibyp9HvM6D712JVuCn5VydvC06mXsjtF0Z4a0xr6FXBMme3IrLHUZ3CDZsnEjaPoXQ0DAJ73q4MXknIqjitMYwHJL0MR2cMeoEzdXp+HMFkWQLnlMa9p54J1u4JoGll2xg3/QXmVpUR9FagxkZdQinTnBUiKLJTTTszcU6qh/Np+KosqOFBZa53YRm+AkU6oRHBbEdddD0VBn63iRKrznIb96/hN3bhjOvoIbOcyPoX6SSbA+ScVstCx/dSvHZdZS/HkIxJI52BTOm8sb6uVx9eg4TDyzh9BMeqn64At+xFGJ5UQrXx2hZYKI7JfMe2EnPj4O8PPkNepsTeePmp2idobFw/d10nKFT8sUNnHxtHK91zuTH2RtIyh1gRccZDNv1HzeIjLw+5I+6sHULzvYcYfUjZ+IZ18P4l+/Cc0UrHZ/nMnLKKdSgoHWWk7ZZCZR9dT0fbJgaJ3Q8kYZzTQJ2Z5RENRjvqJwxQLAmkR2znwWrScJWO8//7hmkIvj6zdFEfVbW/nIBwUWTqfvRELI2aQxODREojZH0optJP1mG1mqjbFgLmU/ZSP/aZERaGxl7oXu0k5If7SY/s5e0wwalWZ1Y53UTrvMwUGKldYGJdVAgLCZ6ooEw4YOq0WTuC2MMWkg4puE9sx1/ocR93ErfnDBKQCWUohArC2HrhSsfXY+zVaGnOgUlLHC0aiRO6OLCs/bgXu3lqxEfEr6/D4CrfrIB+XUC7oQQymkHkUSVlM12ZJ+V2ieHkfyNijqoouhgGRQEsyUJJwUObxglKok9l8m2FS9x9M4VBJYMkPClA1uHhqKYKDFB/noTf4nOxIsqcTUrMGaQtG8kEa9AxCBnaAcxr8Tf4+T8CQfRAgL3YTvt/R6EX8MY40MNCwxdwVFtQ8kLUFjcibfSSkbGANhNlJikelk6pzpTUWKCqK4xWAK28kHK3gkQyjS5uPRw/HhPNwgt7cMXsXLjwx8hdJADVu7YuxPTAhm7BO3TPTiP2wmlKWhBQd4GSd0VCvYeScxt4isQxJINsjZrRGcPIkyBZfgg9RcrxBJN+itgsNAO61Kwd0EgR+K1hLH2K5QsqWHFpLc5/PJIQikq979/DWn5ffQX2xgsVkg+LkmsNXB2GFh9kpHWdmqvWPk/K/D6fQ3R/5h97xD9A1Z7xUreqMjD/0oOHwxPZ/yXtzHs+eXkblSov+Alii+sZcgryxiyqJqCDwRtVw2l/a1CHrnjVX7YNJ2aFyYRTNNwKRHUHl9czb1B58INdyAUScOjQ3h83YV0jVG42HOMJSO/pmOOjjHGh/ZECk9f8zKGNd5unPSVHSOiEkxVKVll0DVKw3HASTQhXpRr6dEIneEntVIn+3MVT318YrX0agyWGdz0x494fcLrzMo6RWtdKuHqBFAlCCgs7mTgw2y087oJpwi003YsbVaunbsN92mofb8Mf3kMcWMnFn+MUIak9kqV6lPZhHJ18s5oRAsCEYU//v4Cxk0+ybzzD6BWuahb7KXzF0XYau0oNS461hZgP2ln15YReG9qpiinm5SjEUREYePxYVhtMZ4pfxdFlVi6NcRbqQy2e+idEWHJZ8vY/tyLSEMw6onlpFbqXHToJhCgKJK8D1W4t5O6xSqOvS7uqLmCpIQAc9JPMj6zGX+BpPqGF1AjEunXaJ8G5350D442lbauBF656CVGpbSy+5NR/Pasd3G1Spy3tpC5N8ioJUfJfeUoycfgo3vmM/aLW9m9bTgdy8Icnekk5wML4wsbOb0rDz0nwsTMOMnmsdtXIAzIuM/EvttN+dJ9IAWJtSYxl4JhB60wH90hsfgE674eixlRcSzo5Ed5G1hbupnzvIc5sbeIk8ut9IwUeOe3k7rDgq1HcGbyUQJHkrFZdIa+tJzfXfImos9C2sP1eKot6B6DvT2FBMM2Xu2cgYgqXPfiXXFR2xaNgrxuXAcdKF1Wqp4ZzrXP3k2KK8gXR4Zi3eb9y7WwZ8waWvdlk3IixpLPlmG5vgObphPzmATfyEYKON6eibcOGD9A5pO7EKcdONsFSjSecu6YoxPy23jgretRFZNlQ7ah+QUJipUXZ/+JodefoNNw48+Pz+YioNE5XmHKz/eR8OsmOs6PFyAPe7ST1lkaXTNj2HoFSfYgWl+I1tmC468Mx9Gt47iwg7p3xtDQmErzfEn/q3kMHEmhZGwzoVSBvcWCxQcypmBNCmPYwLHPRTjFEhflzZG0VaVjZoXRglA77zWGjm1AXNSDopjYe0yerJyHZW43S8/cTFIVGFZJT7+bNQfHM1CsUL7tWtpq0hj5XCVvPn4OUgG5IwndbTJQrKBGJcmVCnl31tA7QSehWtA5J4Y5dQDdITEW9pO4xo0vT0O3/8eUfUfFl0QSBbEEE3N3EgWfxmi83KB0VYzdm0biHxUhHLSSd2cNad/4yP88THN7EtY+QcnbJl80lpNYY2Kf14Vx2o2nTkGPaYgyP0P/rZtwhonbGaG5K4nzr99O6LN0sjeo+Aqh7B0/lv0eXC2CCVlNyNwwTluU5KeaUUOCTSunU3FJNZkFPfh3p9Hb5WXN4jnoY/xgCJZtvxo9JUbMLfCVGky+6Aj+PIlhkzSdJbA3W7Bf2Y6jU8G0SoZWNKNFTJ4duwotIFhccpCUAypSlaQelnROkfRO0Llx+XrKfn+K3bVFpFbqHDhazO3v/YDeURLDKij++X72jV3N4IIAMbekr0Kh56oAHZMFvkK44MX7/7J/8z8z/s/fXP4vMiHEBUKI/UKIbUKInUKI/zKqJITwCiFe/3bMN0KIx4QQ2l9hsoQQHwkhdn+Lue9vrGeIEOILIcR2IcQBIcQ1fwMz9dt1bBNC7BNCnPW//4//xn/6PmX2X5vpMbC4o3i3OAnkCA7d/DQXz72CT796/y+Yd31J1EXSqQpkcPjdEdyxdC0v1s0k9U6dX21+jy7Dw6P3XUfzIh18FiaMPsXXh0tJPKpgG5AELhtA25iIPLsP3VTIeyDCyV96kG12iteE6BnlJOqJt7Qfv3UF048sovV0Kin7VXpHSX4w90te23AGY6fXcGhHObrLxNGmYlpBiUDGgSiXPLWJJ/YvwO6KEm51ceOcrby5bi5GUYjsP1sp+/Fxdn06Ci0EuhOUKEgFzNE+xDEPRWt6OL0oBVsfDFQYZJR2k2QPETE06hrjE2XrgrjswKLxB9j4/hTOumQPG9dMwd4rCczzE/HZEBaT7I8stF0UxX7CgX1qN+amVPIW13G0PgdbvQ3TGp8Yz5x9iO3vjUMLg6PLxDZg0DRPY9zUGvbXFKFYDaYW13OwLQfrlgRCc3zYdnsYHBJDcemYMQWhSmy1dhKndKC9lEo4SaF7io6n2oKzw8Q2YNJ4oUneOoUxDx5ka3MpvoYEnK0KCfUmPcMVUo8Y9IxQUaLxFKVIjOI46sA6IOkbHedScdTa4jVc5RG8h2xULK4mqFvpez6f8LV99Pe7yPzYSufFEbQqJ+4mySM/e4W7//hDliz5gi33zySQoWHYIZwiMDVIP6jzsydf40xnnGyu0wjgMyW/aDmPfduGklANfcMlp678j6fZsy64mjc+eJHp78Tnncw9Jm2XRnF97cCfZyJToqgdNgybhIQYKJKKZdX0XziSpA3VnPrREMzCELQ4cLUK7l/6Hi+enoWqmDS0pVA3/1VKttxAVtoAO0atZdiK5YSzDDwnVQKTgni3O4gkCUJZBt6TKoPlBtJhkP+BguOeFmoq85A2k9xNgsgPeol8nsZF12/l9uR9PNY1gyN9Odg0nY/LNnJF/Rnsqy1kXFEjhxrzUJrsJFZB9g11XJG5j4dWLaHoz910zExhoBQcpQPY1yXws5+8yT07Lyd/tULrdI0Js6to9CXheCSBwSI73eMkpsMAKZg5uorWn5RSe2U8MuNqil+TyZWDtD1oUpDUR8crRfScFUZRDX44YicvbFlA3eKVFG26CYDC3G56Ak58HW5ScwboaUgCXeDM8+FZ7cG0CJzXtNK+M4dwZpzWWugKE0afoubPFQg9zvcTHhbC1BUsdh3Z5KT4/t20/mgatn5JyhVN9IUd8OdU9v3mBQCKP78RGVXwHrUSTYSY26R8fCMnTuYw9Ml+Bp40aK9Kh5QIyUkB+o+mIC2S/E06pkUQu72HjmPpJFTH091t0wW2fD8hnw1PpQ13i0n7LIm9VSVUFEWxGZhBjTmjq6h6Zji+S33kJ/XRsaqAzPUNVD2WgTloiYu7DmlGU0yOHSzE2qegRME2tQef34FW5SRlWjtua4SeN/OJuQWhDEks0ST5kELqlY20bCggkGcgbSaueguRZMkZcw7RNN9C3ucxDq4YQyhdcP6SHWx5cjqBbEHBqiY6zswlkCUIZ+q46zTyPmyjZllGXE6nrB/vy146JmqkH9RpvljHddRO9uO76Fw+DXebQcs8qHh5kJPXJKCGBYU/3/2v5SHKz5PZP/onp8zu+O9TZkKI8cBWYJKU8rgQ4jzgT8BwKWX73xmzFghIKa8RQliBr4CvpJQ//fZ7BdgHbJBSPiiESAC+AR6TUr70LcYNHAceklL+UQiRCxwBlkgpN32LyQMqgUullJ8LIcYC24GpUsrK/51989f2fYTov7DaK1birLOQmTxIzCOIlYeYeehKGPAx+nfLOWf4XABeunURq1fOY9euYSgxeLFuJpHP0nh2y5ss/vgOHnj+Jkp+fILMT604s/zU/7GcpPw+DLug54IgzjUJ9A8z6W/1Mi2nnsGRqViOOUnfDx1TXPRNjaK7IDosxENdwwhGrGCC7hS4mhRe/WwuycfAqhi4GyChSkWd1IcaBlerRPPFePmF8/EesJO4xk3dpS/y+uY52Ib3U3LlIXquCXDq18NQo/Eiy6zdUYIFOuZoH3qTi7TDOoNPxEg4ZRJKl0inQeLFLbSuLaT+WDZEFUyLIOkbDcWhs/c3EwllGXy4dRKxBEnGFQ3EWlwoAxoJe+30l6okJwVImtVOeFcqLOjFrUW4ZuweTtyyguobXvjLTV6f7GOgzKR9GnHRz1HtnOxJoyC3G+8uBw2/ryDc6KF/YoS0VU4GK+IsvAsrTpCy00rCHjvKqAGiH6TTM0Kle4LB0Cf6SDkWI5CpYOuNYmu1EElQ2ffkeGwfJOLKH0ROGqDvMj8xj4kvV8Xig+z5TYwY2UDSl3aO3rGC/mEm9jaNzM0WrIOgGCBEvD7o13kfc6ozlc7xCsnOEJOLT9M7VMW120kkzSDt42o+6R9DKMfg3XfOoPWGCN2TTPqGSbQQnHfRbqJuhS2+4cy/8sa/nJPtRlxpvubaF+ieGcPWozD+oWUUr7mFEXuuouUML7Nf+RGWQUFKpaTzihAed4hQusQ6oCCDGtY+gbQbaPYYP5v4KebwYpDQsHQoD1z8AZnv2zBSYiDhFx9fhv/DTHo25FD+dITpdy3FcczBjlFrKVr3Q4QBUjMZHBqj9EmDtEubOG/xLuztKoMTwlw3czuJqX58uSrNmwpwNSm46jU6JyicnXsc//gQb341k1TVxeOZB9k09BM+LtvIvkiM/ogDBiwc31iOvdKBzA9hanC4Op81neNRRw0w988HSFzcQtYuE6um018BT992BReMPEzDReBqFuzbNYTQ2gxmPLeXwWKBu17hJ7PWU1rWxt7GQhrOsYGAzBGdOM/qIJAlCBS48TV7KXN30n1mmOTNdmpmvcGLR2YyZGUPw3ZdzeTyelSbQcdXOTjWJpJ4xEL/sRQsqSFsPQqRk17UqKR7rKR1XzauJom9VcPRaEGqkkPbytEdMFhmYo7zIQ2FpH1WTFPB3iko3OcgnCrxzQ9QvzcP9a0U+s+Oa2rFpIHFppP/kUI0EZytEtMq8VrDlJa0MzAimTRHAHuHgjQUoptTKX++CcNp0n5zmMm/3o/2fCqmXeLsMvHnqJheHa8zzG+mrSWULhksUsCAMeecQO2zkP+WiuLS6TjPSswlCA7akXclYNgE1Xfno9Xb0Xwqmk/lVHsajWuKcRYMsvyy9cS8ksGaJPLTezGtkt7tmfEmABsMjItQfeMLuHMHGSiDiKFRefcKssu7cNVb0AKQNbqdXWvGUvNvwzjYlUvXNJ2YCw705tM5XSdQoHPysSQG5obQXRI1IYa/PMbE92uwDCgYSToDjQmYVkHR+70oMYn7iJ1Jlxxh9pEQ2jndnP3QVygRQdekRBztCtU3vsCSqtZ/veL9/4z9BNgkpTwOIKX8BOgAbv1bYCHECOBi4Hff4qPAU8Bd3zo5AOcAY4AnvsUMAC8CPxdC/LvjeD3gAF77FtMMvAv8/Ds/dydQK6X8/FvMQWAbcD//ZPs+QvQ3rPSePWxqPQTAlPuXEshSqLx7BWcVTGLy1wHWPzeLnkk6iek+/FVJ6B4De1qIBFeIyCfpBLMkqRM7cJ9Vx6m3xvLk5Pf4aeXFBDpdeE5qSBGvQzHsEPVKEmph0s0H+XzzWHSXxNanUDG3luotJeRtCdI21cmzy1by+PzzMT0OGn+hom1LYHCoDibMHFvF9iMVKE4d2W/lxtlb2TUrg/q7hpN0wsSwCebcvZvKs9LJ+ijI7nWj0D0STCifdppPyjcw6eBiYrpKf6cHS7eGuxEC2eBujkeLhAldM2N4K60oBvjzJbrb4J7Zm1jfMYLAc7m0T1JQIwLTIhk18ySH9pSRdBzOuWsbD6Udo+iTH/LFwic5e+8y9JjGmPwm1pR8/jePwcr+HJ745ALyN0bpvD3EouLDvL9qNlKBYK7O0KHNnDiVgyMpxNXl+1ldPxanLUprczKKzSD5CzvdM2MQVtD8cbXzyJAQ0hQk7rbhnxPAZtMJBmwk7LATSRY4p3XTfywFe49g7mX7OfTrsXRMVInlRch/T6V1poaeGWVGxUn2bhmO7oxrYFkHBc72eMqyvxxMLV4MG8wxKV0VoPUnBkF/nLjQs8/BYIWBtVdBLw2R9LkDLRwP5+feeIrukJvOPVm4WiT+gngXke4yKRnWysTkBg7153KiOhfh1HEet+NuMhkoU7B3Q0J9DPu9rUxIbmSEo5nHqhYityQTmeljfE4Te7cPRRQE0DSTxNVuLn5wM593DKX943z8E0OUPG8S/uUgjrMbwTRo+vk0Di97lmFv3YbMC2GrdGL1Qf/oOPN04lGNkstrOLSnDGe7IJwiSRvXwbT0enY8PhndIci7/hQ1n5YRTo0XwsYKItjq7FjG9BE7lEQ4XSenuJufl67nLGfkL8d/ecsUDCmo7s+gd2M2qUci9N0eQFNN7H9Mon2yAvkhkjY5sC7poKU+lZzPBaHr+4jqGoFGL7bsAOEeB+Wvhqhb5MbRJUg5q4Vuv4vKye/wxmAqr/x4EU0LwV2rEUmW3HfxRzy28QJsvXGyQItPQa8IMjqvmfanS2g9P4brmJ3QmCCjcls4sq+EjBGddB7KwFLqI9Tt5M6Zn/HBA2cy5Vf7+HD9VGJ5EWRUJXm/hqdZx5er0TteJ32nRswFkfmDzM6v5ctPxqG7JM7yfgKn4qLJgQIdHAbeIzYCuXG+HmHCQLlB+b0H0TancKwqj6yvFAaKFEJZBkpqhKuG7+e9j2ZhamBa4/Vm1tndRHakEhgSweGJ4NjsIe/KOrqfLYwX3Z80GShSsA5+WyeXAtmzmvG9lcPwpUf56tBQCoo76duQTSRJIgyBdWwfwaCNpya9y+FgAVtHO2n88wiizS5Ii2BGVVw1VnRnHB/OiWHp1tA9JlKVWJLDnFlSzaf7xpBV0oX6QirNl+iIHivu0wqmBdzz4uz0lg4rUgE9QcfapVHyZDUnHinlzHGV7G4tJHIskWhmDJsnAlVuxsyr5t2iL5h6+BK6+jwkb3IQvHAQz2oPu59YybR7ljLu3oPEpMJnB+LabPcvWMcbD53PrifjDPpVrw4l9bCf6uVWlD7Lf3nf+GdFiHLu++dGiOrv/IciRP3EIze/+c6yPwEVUsr/V4hMCHEX8BvAKb91IoQQRUAdcLaUcqMQ4qlvP1d8Z9xc4AtgqJSySgjxIeCQUi78DuYG4BXAI6UMCiEOAXullLd8B/MQcIuUMvP/+x75+/Z9hOivrPaKlTxav485N/2QBUtuYM/vVlJ59wqKPr6Z5vfKePPQZJyXtlP+w/1k3xUXVB32SDN/mvAq2qsp/OSut7H1CbTnU+m/ZioVOR3cuWMJBcu7cTZoeBa0ExobIpIsSa4yUAy4+q4N1P+gCFu/IHubBAlHjhbGt+cSO4EREbb6h9D6lINAsZfcxxR848IIh86w37Wzc+8wPln4DGZMRToN3vh0LlVPl5BYbdI5CbrHSZK1AG2LS9nTUoAyfoBYapzor/eFAopXL0V7MwXrmiTsLRYyx7Vj2ASehnj6Ztevn0N3QlrWAHkX1+PoNsGEB+as57mjs+kPO2i5MJ7WUaJQ8k4fVevLQZEEswTbOksZ96tlIOHl3mnML6rhqhH7/q4zBLA0sQWZE6ZxoRXrhgRSNT/JVfFCaSySE7XZ5OV3Y9np5eU9sxisS6TzUAbeSisVDw8wsDBA6g4LyYdVnl30KkLC6IJmLI3xaID1kBt/mxu10U7MLQinmsgPUrB3CeSUAT7ZO455D+3A1CBhj53OcRbcDZC83UrDo0PQggI1O4ijS2CbHtdHC6YLkDBxZhXazF7UiODkrRYCPjv24w4cxxzoLnA1qjhG9+E84GSwGDqmSoKZgro1ZXTsz+SNa5/mijs+o2JGPXpWBDWo4P9jDqsOTqL1g0KEQydpR1wmY/6PdsJwH/3DdDrHWqg+lc07W6fzxG+voL/LjffcNuYWnmRPfREnr3kBy2E3ti+8IODFjQuoOZnN/Gv3kLrJTt2lDjr6PZx8ZgLtHw4lkmqy4AdL0T0Gaq2DWIJE6BJ7k4XMbQqLb95C768KsRb5iHrBPaKXnr2ZbH59KjGnoG+YpNWfQPLctriOWIWf84ZVkrc5iHVdIiduXoE9LUTXgQxeaZv5n47/ipw9hAwL1+TtwbBDy1wr/V1u9E9Tab5AR0/WyXnbgueaFjp6vWA16R2m0n8qmRvK9uBoV8hMHKRojYH4bS+61yCQY3L6dDrZ3kFK31nK848sxjJokFHQi39olKQT8NsvzsfdoOA5LfHWCtyNEscBJ3ZVp2W+pOIPIfzlUU7NeZ3KPaWcOfsQKY4g+656gmi9ByWs8NyGs2i82GTdmmmoQ3xcMOIIWq9G6iE/3geaUC/sxtKrkVAbYnBmCHVnApuqh8IIHxafwDfg4KzZB1EjgEVSnt9BzB13tA0rjLv6CCnFfdT9ajxtqwpJy+sjnBTvlHI1q9gqnaxePRtnK1gqBpGZESwByYDPSeLcdh6YvJFIs5uYW9Dyp2J6h6ggwVQF7maJrySeSk6oM+n0uRk8x8/Rl0ag+hUa25IxLWAd1U/aYZ3QiURy37CQp/Xz9qp5TDsUwfWZm6mTq1BVE63LgrNN8svL3yWaYHLZxP042wV4dIRLRwjY/8w4VL+Cb3MmwoQJxQ3UXr6SwaE6igHmO+m4j9uIpcbQU2MMfaofqULVk4WgSPb9aSyh6kSiqTopeyyYp9wIU9B3dw4THlyG76sMSp80iC3qIxK20L8owLYweN7dw2efj6PY0c2Fk74hoUbh1fppeOoDjHxyOf6YjZRXdiP3V+LwRP6F4q7/dB6iVCHE19953fzdnxNCJAMJQNtfbUk7UPx3trIY6JD/OaLS/p3v/v39b63zH8EoQOF/g8kQQjj/zvb9L9n3DtF3rOyBbyjefCM/fOxOmq+JsXnVawBM+MUyypfuI+U1F0N+68N9K9S8OoGTt2TzwEUf0HBNIdf96U46xyv8eNelHLlvBVtffIm9j72AeVYP7hM26p5NR3dK5GvpmN02rP2C1lkCJSJYd/s8qm5z42k06boshBg5SM4WsIzro/SePSTtsvLpE7OJ7k2mr0zDtGncOHYXSqeNhityKVkd5hdN5zO+7DT2Riulb/Qgwyoxl0ANCtwNCuseOoOBcoNgv4PYMS/eSiv+HI1x9x3E0aHQc3GQUJrAWye5PO9rDt+/goFSSK3UGf+H2wlmQXRjGtWtGfivGMC0S1bUzCIasJKyPIoMq5S90o6zQ3Ld+5vwNJgYHoOShXWcrkuPa0oZgiMDOXz5wXgeSjv23x6PpIQA3mE93HL3R6x8+1xaLonR9AMdzRXD1hp/WvvZsrepeDGItEoy95mMW1LJqYc9nF92lO7JOs4uk+Xrb0CZ2M/ohBYy9xkMlElCGSZFH5rkfhElaUEb9i6Fnsk6oQyJxxFGSwnx7to5eGvB6pc4JnVj8UPv7AgtsxXCQ0NMKThNIMdEVST+XIltQOLoFOzfOYT+Vi81177A3Ioa3IfsRBMkgfIoulPirTdYO+YVfMOiKLpAWk1SK2OYZ/QRzYly+RfLeGflQiqr83BU2zGyInSeHcHWbMFfYCKjKr3jDGJekw9XzSQa1rB3akgFUvZqXDDja7qn6jjqrXTuzaR2MBVVUvpjAAAgAElEQVSt1k7xmlsQEpKrInSeH8FVOsDvz3iXtUfGsvexFxgxoZ6cly2UvBdlXGYzSnoYV1UnxWsMnl7yKt5aiLkF+Rt96Ff2sv3a8Zy+2sRm0UmolfT3u1B0yHn/NChQ+EmU2IdpDGzMYuJVhzHq3FQ427nw5S34F/oZf+AyIh1OKA3wg6ztzDiyCABDmjzVV8iuuhIe2XcOwVydsXOqUfsshNNg/rAqMARSETR2JiNb7VQUtWHtBzMxxp8bxxEb7aflQDb1l6h0rs5n5tgqiteGQZXUVGczbspJpAqxB3pJ+KUDdEH/eQEsfQq+YoOeMRLdEdcZDOSafNOaS+GHkvqfaeSuVyn6+GZ0j8GBZ8fi1iJ8E/Ugs8J46hTmzTqMYjUofOUUOU9b+GztJNwNgpPXOjnWmIXxcSpaUFB/oYMLK44QnhggZbMdvd4Nowex19hJtARBwpNzVnGqMjcujyEkb976JDGpMFCZgu6JR3+7WhLRgnG29puu+5SYRxIqidA3Tmd2fi2qZpC5rZdRuS1078vg2TcvxLSbRJIl02/dz4Rzj6IFIZAtyL3pFEmVgmCGQu95QfIS+7Fv9+DPF5xY8jwXjDhCMMskw+On+QyB7ja5+sl1XLr7FkK5BtvumYo/D3YeLkc56SR7u8Gqhx/nd9Vn4q1VWL1jMt4GnZK8Trxf2zENhe5xEj0lhv2MLnSHQl/ESfFnN2Hr0BA6lCytwuKTiIiKCKo0XpgWTwl/ZSP/Y0E4FVJHdVL2pyjhswcpeeQIQoe6RW7s/Sb5r5/i1F0ayvokUpN8OL5w80j9eQQumUzNdS/wcuV0Nq6fSPrzu/DtT4N9lWy4/Xd8WLYJpoxiU+shjk976//P3ETdUsoJ33m99Ffff0sCQuSvlkeAv+dwuP4Onu+M+Vdi/in2vUP0HTv52DhEj5XSq2uQhmDK/UuZdHAxiScjdC2biryti95xKZy+IguHN0ws0eDRDReROKedEzevIHl0F0fmP/+f1ilHlxMcGcJi0Sl5qwt3cxjp0hEz+jATYkhNYu0N8cD0T0na24be5iTki+v7GHuT8F0+BcMm6B4j+eFVn+Ka28nb7zzHG5/OxUiKESjUaZrvpPvRIg415iFVSSTTg/uUhWE3HcPdDOasfu77zds4c/1YOixEM3Qy9gUovaGaLevHE8zXSfrYiaNLol3ZwdMfncfM227BcJqsfPYpguNCGHYIZXwrYbA3iYzdkPNvEkedlRO/TMPSp1H1YBIpr+zm4TeXcPdDq3ClBqnZWoTqU7FO6WXXeX/gRFMmtr5/LE07NKWdvj43vz+0gLkXH+DWsVuxVLqwWHVyprXQvSOLR1ZcRe/DUTKKu2k+U1L58ghsB1x8/sYUFKdOywU6JSNaCJ9MYPPDMxn+b0eYNe0YZmIMf7aF0+dbMGQ85aUEVUrHNzKwM4OaWW8QSTXoGy7pGwLDU9sJ5Ajy31VRwwLNapBsDXLZvF0YpsDIjuBqM0isizF8ch2OFo1JBxezbetIgtkmxWv9qL3xBozWswzOf+F+UCTmUD9JWYM0LdEJVSdy2ZgDeI9aue3WtaiDKmLCAE53BGutA2crGF4DS5eGtUsl6ZjgB9d+itpsJ/WIQeJJk6SaMJtPV5B8QGP6BYf54NonaNyej71bMGr0aaSAhh8YzC2tYfP4V7h//ZWcMbSaoo9vZl5qFZbPDxBJtrCnqZBlo7ZRe0M2gWwLyz+9npE3HSXlzFaa53lwP5dAw7mJAOg7k0m78TTuQ3bCmQbHH8xFC0oal+pY/ZKyRTXs2DCa3PGt/H73Qp5/+3yinU7sFp3czZLogI17D1/Kgqwqpty/lHf9aTy9cwFarR3PITtTR59kqLudjxb/gRO3rOCrr0aRsVPB0RrAetSJzIhwujsZ9cxu0rZa6d+fjnbMzaULd+LO9HP1rZuofHMEwpDYG6y4MgMcX1dB1xSD9m8yqbnRgbAbWA66iSWZpO8VnDP7AO4Wk5hbUjqqmUjYQtO1MaxWnfYpChdO/Iby8la6JpnUrhzCj399M5cPP0AwS7L9o7HIASsnfl1A81wHQkI0EaQrrr3luLgDiw+K79/NbG8VqmYScwmuO+tLohELofwYqz+ZQSDPZF3PGNwNCjGv5JIZ+7j22bvZ8+VwYmkxREIU68IunjvjTQbOCeBu1vno3gW4R/TiPhZXht/1xjhkg4sr3t9C8J4M9JIwwdIoaCbRdJ0vm8rYtXMY9l5JOM2k/s9lDC4IwMw+Cp9VOLmvgGCOJFoS4vGeYazbNgElJjhVn0HG7nhE9JmXFuHe6iS/rIPBfCuGFbw1GoYVmi41uOzRHxE8lIzlnC5yt0h6h2qc3p+LxS9x7nXy4NlrqT/nFYLb03C1hJmfXkXSbitJVSaDFQYNviT6Rhmk7ldw16m8t/QJwmkmC+/YQfN8BW+dxP5UEg3nOAnXe/AvHME7Nz1J/uYobTMEhZ8MYgQ1yq6rZnxqc7zJI2bFW9nN0JXLeXXK62Tujivcf3LD71DLijlj13IW5o5H7QvSbQT+dc7QP7vl/h+bagPfvtv+arkNCP4XY/4Wnu+M+Vdi/in2vUP0N+zAvjJKs7voGSFIdgSpvUwjnCqwP5zAnsdXYu+W6LoSP9nSIwyGbTzbV4DthWTciv0/rav2chdClZimwozVlZw+1wFS4LFHyNpkwSgN4Ximi2JrJ+t3foRMiuI5amPxfZ9hjPXRPkuiu8Be4OPpbWcysCed8x68j2FT63CnBLln5iauWPQV7VMs2I46yNoZo3WWjWuv28SOk6X0jTRxWGM8/IdreGzU+/zm0rcZXt5M50QXe2uKiBSFsbdp9A0TBDMFbR2JONoF7VMUkor6WHbrnRhRhYy9JoouuGHEbmJeidVn0HxWMloIyq4/gJkXxlZrp//aqahR+NnXFyOExNEel1RI+52dNb7hyEErA2WSmxpn/N39fygSYcGJ86nszIY+K8nrHRx+dAwrP15I4qx2OOSlfUsumBCd7iPVGaC9IQVXeoC+uWGSq3UiKZKUZD8yqNLhc5O+36R9qqBuaQldYTfCr2HvM5AWE8dvEzEtkDmkk8YvC7ju8s2cU30OSPjt2auIpehsP1pBoDxK0xIdZ7sg9X0ni5P2sbllCMNT20ndYsPWE6F1hkZVRzquKd2EohbuOO9Tyt4aZNnbH3DqypUUrekj/8N4Hciwh7qIdToYkdaG+4ADPcHg/c+nknwiRrfuwdajEA5ZCfQ6iBRGsA1IPCcspIztJGl8F8GzfOzqK2H63KOEExQ6z40Q82iEWtwMzg5x5PlRXHfsOqx98S6mw8cKECaYvTa27B/B2b++D2tugP3teWRuVfjD9oX0rS9joFBDr3Xz/OYziSabdJ8ThsQYe5sL6NmczexF39ByfQTdLSnJ6cJfEqP2yyJSj0RwZftYPmMLvkU+VFUSvGyAE5+W89CVb9MbcJKU5iNUECPxmILrV146rwrhSg3yyKiPeH3vdDqmmfx86yIwBDGPZPVdj3O4PZvXd85guNXBFfVnoHsNusdC8/wErANw+YgDuBwRetsT6B5ncua5+zGskveOjUfuSWTFkVlkLT5N+TNVhHNiLCw4gWGHguJOlBhYu1XK8zoIZcQJFf2X+OiNuii+swp3s6CmJhu12U5RRg/D0joQBQHWHR9J+MlspCK54+d/pnuywbpVM8jeoRNNMnG0qNhbLDi64rVlUa9E7bGQdtjkwtzDJJ3Tyqk3x/LoL68l2uLCXyB586O5aLV2rN0a4+dWYetR2LpjBMKA0skNfLh5CtYBiVLqB13B5Q7TO+BiU/9IvK4wzfNU/NkagcPJBAoNEBJ7r0nZpAaeWHkZg6VuphTVI0IqjgYr1k4Njz1C0pBeHJe342xVGBhiMCSrk0hlIm1Tnfxx8QvomVFq573GT1Orqb18JXpalNLX451qml/BmDlAbOEAPJ2GbdAkdXQnGfuCpIzqQmuLX+/Gt9PivId2IBUQhQH65ofRnfDukgXc0hzXDTt1s8LKXXOw+CXG1T2ofgVTCtSgQtLru1F0WHLoRqRN8s7hiWTugu7xkobzVOaceYj8TTobnnmaSz68E0tvGG+twrY147A3W+gJu6i6fzgrfvs03tskPZPTyZzZwvIXl2O5L57Juepn9/Hp1rUUXXGEpp9M5uSNaUxefe//idvM/zUmpewF+oG/rsfJBGr/zrA6IP07xdH/juc7Y+r+zjr/EYwJnP5vMO1Syu8don+FnTqcy7YrH6f77XyGPHAizq4c0Zn0k2Vk7Oij9M52tME4qVq2d5DXaqfQdKlOpxF3th/oGEPZV9ejRARqnQN1WwImAj07yhdnPknw40wiV/VidNk5uqP0L63V5c9G8S5o5+Wj04l0O3DXqbhndaKpJtlF3ehOSdI1TVQeLML+cQKvvHwuH54eRSRNp+Lsk5y+SGH8mcd5Z+VCEvbaGTXyNJfkH8KwCX5+7CLu23YZFZ4OrAu7cJ60kfa5DSUGpgWyt/qQQY3UygjeU9Db7cF2TxuqzaBzfPxUybX2IAz46pWX4+KXVui+ZSoHZq8gUhLm1p+uxt4tsRxzUjn5Ha5ZvhEE3PLaWt5tGk/dxS8iLZKDnTkMe345Je8tpWjTTQzZcQ1tup8j0TAHwgWcqsvEsSoRLT1E11kRms82sfULdo5ai+6UxEb7cU3tRq9zU3UsD21QxfJFAqbPQvf1AZSoQDcULEkR/C1egukqjnaFzl/o1OwuJPmIQtMFJijQOc5OzlX19PpchIojvP7+Aqoq80ioVnlw9ZVkbFfRejQ0u45QJZEk6F0c4Cd3L6W3Jpm6gVR8BYK6SxxYBgWfT3mBiuROzik8zpoHFlJzrZe32qdQsuUGOqYl0TNMwzqzm6pfpaCGFDpvz2dweBQlqCBVSdJPT7O9pxRznA/abKArpH4ZZ9UVErJcg/hCNn48YiOVbdls/XoYU2/9GrpsNFwgcLaoiNMOzMt6iOkqKSeihNMMkvP6UaOQtU0gnQb+PLB/5WFyViOdk8Cd4ad3wEVs5iC541sRusDZFBfczfnQgvsTD9det4kVOXuI9dsx7JLNQ9chHAbmMD/dtwUJnUqgPpRGuNGDWeOmLKWLiy/fzgNbL8PcmYSUAndaAN0pqFsuiHQ7ME3Bveuv5q35L4LNRPWpnD/xIGpmiLM/uBfzSALOjPh1tb+uAEdakLkz4t22mTv7eb9mDANVKTgaLKTvFXyydxyZ+wwy1tkwJw4ys7iWrjcLeC5nL0PureLj6lGEc2K0fp1FNMnE3iVo/7CA2stXokTAutnLroMV7KwsI5wM2A1i6TFCMQsNz5ejVLtRW+xEPQpokqs8PdRf8BIxF5z5m21UPN+GYoBzfDd9owwC5VG0kGDUpFpa5kvebxxL274sNKtBKF1BDcUjLaLCD0Dx6gF2V5YRyjRJOCkIpUmqKvMQ+UH6RkqocqN5o2R4/Bj9VvZ0FBLclYq1X6H3jDDOUX0MWdFHtCBC73DBqb0FLLrhK/w5Cnvqi0g+rBBzSaQA9flUsjyDNLclE8g3cZ1WaRn0kj2llWCuyc1vLaPu/2HvvcPkKM92z1+FzmGme3LURM0ojiQklHNAQiAQyYhgMkgCjE2yjY3BGAcMBoxBIgmRMwhkUEASSiinkWYkTdDkHDtOxwrnj9bxer3fnvV3lt3v7F7c19V/TNUzb1XXVV31vM/73Pe9aN3/7tnoSgmy5OU9RJME4hlxQh12xqR3YRyMcfFjuwhvyqBjlpXIlnQsPQLOhoTRre1vybx5dBoLrzxM6UODqH4DaZVx+ic42XZoLJFUjeSDJhy1BqLJIuIHKYya2oj1mWTc1QINf5mCpV8j7W9WbE0JsdNgtogYFdiy7FmWuU9g/EUX1TED2Xt1pEE/xT+qY+89z3Dl8r20HM6lc7qZB25bTcczZg49tZY56fXkb+yjN2Bna2cl7uODzD+zjF81VpKzJ4xm+K/yM/seP/8etgP/2ng98fz2/wjbSLDDRv1LfBjY908xJYIgJP9LTJuu67X/FDPhPEX/n2P2/1Oys+0/eW7/0/ghIfofYManD5Ky7gDxC0oZedtp6u4xYfap6DXnaFpZwq8u+wzjih4ae1M4PvEjinP7aIybKX1nFRkGP0UZ/dTdvJb0YxqONpX1p6ZiO21i3qb7sQxoaFtSseQGQICizxMN9LW3WQl/noHloJ3S92IEh6loH6WRZIkQ+TwDOSTQ3O/G6BWJpApEUnXiqoS7UuLk8WIMPon9J8p47L63GcrR6VpXxCcvLCCWDP6GZNyHDfREnQgfpTJmaQ2eixP3XM5uhcar7Jh6ZRqvkjB7NCqK2mmozMV8ykr6cY2CDR4+uGYBlyw5RMl7q4ik6cSdOr5SnQv2rMJij/Lk51fTP1UhqUlj5P4bODuUhajAU/UXEfk8g11hkcYrX+H4xI84c/caxkxowuEK8cjYzWTJdl7vn8m2wZEYHFEufmQXeosVsyWGaFWYdtUJft4zDkuvgKpIyO+7Ue0apfccQs8NE8zTMffIXFVSiWkQ1B0pWPfbyNwroIswlK8SU2QQwDsngskR5d1FLxPK0QgrBoSTDsr/MoStTSftiIig6qRWaigWAUGDpWXViEKin0g87sBbIpM3upsRrh4i+THsLSImj86SFx9m/5Fydj8zBe/tAXZe9QwnjpQA8MDPPkaKQtJfHYiSzqjJjdTebmVpRRW6USf9KFTvK+Gr4ZuJ9FlwNIvMG3eGvjkx4nadUJZO19piwkMmCoz9RLxmdFHnhewj3LtoC4JZZahQSbCE9qQSjcs0L5MoHtWJtjWVpMVdJK1q5cnpGxA0iLpgsauKOxfsQFEkil4C6xYHA0NWzMV+Rl5aS9rXJmxNQZwtUbbeO5spD6/E1iyju+IUbbuVmyoOonRasXyejLVb4M7U3egGnViKSqFtgM82zOSZ2R9h8uqoe9yMz2yHmR7UgAFb5hCmb524Twr8+MvVGHoNqDaNTTsnIog6Gy5/Hnmcl/RXLZS+uwpDk5lwv5Vdu8aSNLebhmuTMB61M3FaLcroIEPZIoZBEX++TN94AaXOwa6qcgbH6JS8v5Kav5Rj/85KWo4Xc79ASqEHBMi9oomS91cSdeuoJgHdkDAbNfeD5ZyJpsWvM7Qxk57pie9VMrkFs0el8GONiqdXc1f7VIw++Or3c+mZn80Dt35KfGcqq2fuYPeC54lkKZzZWYrsl3A85QAdYgEjwQlh0k7oCDoYDzm47NID1D9konxtgPTSfsweHUcL6LKO+YgNKSQQzVQwnLHS2JVK3lYI7k/D0quTdlLhlrEH8AcszP34GAiJiU751CZWJB8hlK1hO2zBVRdB0EGx6djub0e92UTpzccw5QYJDo+T9oQJ/yfZpJf1kXFY5b1Ayj+eiXsiYDIovHJ6Bval3Ri7DBh8Isfa8xgYa2XTb+fw+E/eJpSvEMzXCJSq6KJA2ZpuCh6vQYhIVA7kUr86D4BYkkTJrbW4T4mYBkXiDgFHu4ZvhIphSOdkQx7NlxjwFSdsQboXxglmGxF0yN/speyqWuxtAvcOm85b3dPp/SSfW968l66roqjrNRRNYtZfH8QhRYinxbG36cjfHuOW0gOMeGU1+yuM1N6ZQvb1rVyUPY4N295DXtDK74vG0bj8X1dp/t+BoH+/n38TfwIuEgRhBIAgCBcDWcBL5/9+UhCEakEQzAC6rp8GNgAPnd9vIEGPf17X9eD5MTcDlcDPzsc4gTuBJ//puG8BEeCm8zE5wLX/EvMCicRq3vmYCmAW5yn/3yd+oN3/Gyj6PELbQivzLj7ON/XlWCwx8h4K0/e8AeNbLvorRFJPasjhxLXsq5CJuTTuWryNV79ehCEg8KsbP+I3m69GM2pg0hhf2kLlySKGj2qn/mQeziIvvtYkxo5tRhZVqnaV4miB5HNRWpaYeO+aF3i7fwbHnh2PoEPyXa3UdWYgtplx1kPq9a30BBwEa12IMVAKIuiqgKHdBCVDxLwmZI9MxlEN5856wpOK6ZpqwOgDS59O/OpBPN1OkqoMhDN09OIQmR+aEGM6gqYTTpPxlYioZp3CSW10bs7H2qPTtzCKPiSTckzC1q0SuNMHm9yEMwXSjyoEciUsAxq9E0VUu0Z+aQ+7Rn/xj2tb+MWdrJy5k4+bx+Pz27AftHDsFy8ycv3dpJ3QCKWJeKdGcZwwc/LhNUx9YCW+K4OkvWlFNQsIKgxcP4RS78DWJjDxppPsqC0DnwFHgwQC+MdGQRUw2OKI9VbEeML8U4gk5gO6pINRQwhLDH8ziGo3su2D9Vxw7BrUbanIQzopbxyk4c9TyP5OpedC6R/U+swDIc7dlvC0KvhIxDPcwFCuTtZ+lb6KhAdW3K5jDAhEXTp1N61l+J4fI5yzEUtRkYISU6af5fCuERRuHKL+OgtGr4i1EzwXxkERQBMQIyI5I3rw7MjC4Nf57JdPc8nRu1BPO5GHBNIqY0R+6kFRJVJ/JeFa082xb8upvW0tw99axf2XbeSFM3ORZZVQQxKmAZG8b/w89NEHrPrkTuzNMJQDziaILvNi2pjMiDtO0x+xc7Ypm6ytMj2TIanEg6fHSep+mXCGQN6mQS58p4q3903HlBpmWIoHzzt5CCok39xG/0d5jL21ml0ny0k9KKOaQBcFxCX9CF+koNgE1Llehrk8nK0chqlfJO7QUTOjGFpNyCP9OD53MDhKoOTNXtovzSDjUIjb1n3Jr49ejt5nQgoJmLwCJo+OahAIzAjh2mYhnC6gWCHu0Cj/WxcjP2vlm3emkn5JG0szq3jr3BRCp1zoUoJmnnpKp79C4PKLD3DyzjE0PiAg1tqJZijIzhjuzRZ+9POt3O9upPDLO3FXSnDJANrmVAIzQpjNcW4dfoD+uJ0PDk5BTopRP+dNAIo/Xom9WeTkw2uo+PNq/KNjGHoN2NtAVGBgkoLslRGHDSFV24mURjCYFJQOK8MfO03rvWPQxgeIBI3IvQn6ednaLpp/lE0kXcPgFxFGBlBVAfsuG0lXdtLcnor9jInwuDCOfRZMfo3BpWGUXgvppf14TqSRVAfStb30nEtFN+iUDe+g95N80o4FqbvZAiJYW2SeuOVdrrT7ARj/h9XEbaBf6CMaNWCxxMh6xkjzJRbimYklf6NPxzsvjN5tRooKLF54lL8fHY+pR04oa1vVBNOuFvqnKpSVdBJYm4t1ZSdNPSkIgHTOQjRdRYiKuKsEYpd4yXzKyLnrzCTVSHgnxCga1kuGJUDVlyMYyleRAyLl05qoasglb6PInjWvUvj1HdhSQgxP7WVoVh916y9g+C3HGLhjKum7elBSHQgHThJfNBFB0emYbSLu0P7T74nvhXafl6fn/uxn/3eG+D+g8YEH/i0vM0EQlgGPkqjySMBPdV0/cn7fM8AVJIQaw+e3OYG/ASPPx28HHtF1XfmnMbOBl4E0wAy8r+v60/9y3HJgLSCTaKB+Xtf1t/8lZhoJPaM4icrUo7qub/lPXor/S/xQIfo30HiFmZo71nD41fE4vrPy9QWv8odtH+KvTKFrNnx0/fN0TxVQzQJDWRLpx+OoboVNXaNRs6J8ePuzFBj6MOUGER1xCvL7aH27BDEqUNeewfB1XhRNZMK4Buq2FzPZ1UTGUZVQlkDvT8PIYYFnOy/ixZxDmG/tQri5l3MHhzH8iQDOepCv7MP3Wh6xg27UzCjykEDSPjOjCzpxNEI8KoMuoKTFkaI6HTeU0zFHJpoVR5cgcFnC0BRVwDc2jq0dNF2g85o48fsGaF4u4miJkrs9hFgUZEHGWUI5Gt5y0EMyOdsFhnIFei6U+MPILzAu68M8fpCuGRLGoM7gNSEEDWS/SLIpTPHHK7mmcT6FW25HDkrs7BvOYGcSu2b+jeisAEsv+zH5WyN0TxXIXdGErogECjVm3n0XAxUCerWTq57aymC5hNGnoGkC2rAwGQd87GwoxWqPkvNtopE1Ni0AUYmMXTJqtwXKg0QyVGRbHINX5OtLn0v8CkTQjRpjXjtD60IzxTtuYVRqN+EMnXC6QPOTU9CsGmWPnKb2lrVoJp3UqjjBfDOix8CS0adpnycTc4K5T6BnRYS0ykQvk2LTSV/Uzux5p5h0/BpKHgvx1I/ewdQnYSgIcqCxEF2GtGdbsTdLRNNVwhkCV1QcB0VA9kkICgh/SyN1YQeeCpVb664n5LNQc/taqu9bQyDXgCToeKtT6H1S5UBNMWuue5UpD63k4kVH+HT1RVi+cRCPy0ycUkf23jAtPxd4sPpqZs6pwlOhYu0SiDkE2ONCMQtMS2qgqT+Fkb/rw7WvjRXz9uFtTqb8xSDO1hjVP1lDzX12vmodRfkrAazf2pFvE0n9rpuLH9xNy4ALT4VG55QAUkAi/ZsWFJuAZoTwwVTCGQKZBwLkPRLnbHsm8pBAOC+O4lDRVRHVoiMeSCJ0tQ8lN0r4JYW0Je1s+/hNRpq6UGMiaInKwd03f8ngzCihHB1TtRXPggjV963B2q2jG3Vc7/kYZ2vFX6rgfS+XrRfmoOx1kzShn3iqgnPcAP1jBWbMr+Krj6cRyrEgn7Zj7dJZNKGKMbkd9C+K8NrHiyn84k4waomqY9jE1FuOI8sqsVon7z93ER/tnoatWcZw1sotrTOZe/oy0MBfrjByzWrC6TqyVSGeFcNTofKTBz5B9iQa7p1bbJj7dOROE/mpHqSYQMMjo1HNOvFmO0Ql1l29Fi0thtLYTMyVMGIWtMTyp9ptxTslRlufCyEkoxng5jEHGPfjKmJ2gXjAiKBBsjmMwS/QN13BG7TirhQxd8rUnssGoH+cnecWvo+pN0GIeGLtDcw4dQWzVt2Z0CCz6ainnYzM6SbYb0OxyRj8AmZ7DGnWIN4RemJCljtEPElly/aJyD4JzaBjHebHMChj6RG49+efYOQhSZcAACAASURBVO4wUNuSSSBPwv9moiKtA+7TOsga48Y34J8fIusPMvV3yhi8Ir4RKtYGI3FVourLEVT9dA1SSKT4Aw/R2d0My+2nd3zimg6/4wjS7iTuzN5D54aRDL/lGADBhUHO/txNw2qRxj9NxT/MgPztsf+pZOh7xX+Rl5mu6xt1XZ+k6/osXden//dk6Py+B3VdL/rvydD5bX5d1286/z8TdF1/+J+TofMxnbquL9N1faqu6+P/NRk6H1Oj6/pcXddnnh/n7f8gZv/5MWadP973ngzBDwnRv43bWmdw3X1byd7cQb5s5+7aFbgv6AUNbnj5ZziaRdJ/0kjGtg5yH60n5YCB/m05vDb9LZbvXcWvG5Yz7KYmUtxBenfmMJQjJGamgs7geBdXFJ3k7KbhhHMUvmiv4Io/fIOlRydanUy8NEz7c6UUfbKSoZiROwq+Q4oK1NydwmCFhvsnKkl3tGGe1o/BpDD7iuPIl/Rz5kgB2nmj8rzCPtyHDZh/2klwUhjFomNKjvD0XeswyCoIgABCTERYOoDRqJDzgYFAxASSTjTFwLJXdhLrtvLaxkWklfWji2BvlHH/pIVIpkLO7hirdt/IQHUaxs9cTJxZgy4IRHutfLriOeJuhYVpZ3CdFjhSW4hoUEkeOUB9Rzo520SWP/4QEb8JMRQjnG7EUuzn7IFCbPVG7MN8mO/pRLFpjJpfR55xgLztQTpXxch52UhxRj+ND8uofiM5V5ymYy5IETB+5yCvsA9vqUjqcYFYxIAYFknaaSGWG+P1gRk0LXuVC0qbsTYb+O6ZycTzorh2mxll70IqCwAJgUQhLtD4SDklu27G1CfSf9cQggopZQOceWwMWlYksQx1QZDamW/TOUMif0IHedtUomuzOLChAsfzDuzrPPzy3R8zbHob+hkHuteIoMCBQ+UEyuO4c7zoks7BP00iuVom/aiGaUCk7+YQno05mLtlZFEjc5vMba0zKH99FZGlfjx7Mhk1tRFfwIrjtJHbd95Cweo66q/MRtp1HO/MCBG/ieN7yoglG3B9ZsPbb2dvUzHSkIi9W0UzJhqwr121jbefuJRbR+ynd042XZfms+mVGbgKPTQvd9F6kZF1vkyM9hieJhcty1wgQO3dObz67TvsuX8qggBph0SaPxpL2qg+hPd0UqpjqLN8uOo1BA3aFjpoXJFK8Vod57mEeWxJWReSWcEQFBHjcHPJIeaX1TLG1UnTmSxuaJ7Dg9fdhbPShLlPRIwLvFQ7G2OTGbUgQihPQWg3U/zxSoquqyflmEjdKyP489lFXD/tAAPjNKJTy9l539N4fDYMAzK+gAX3Bb3srCnD6IeuKRLieB+ZewfZvWUc4XtTMdVYAJDCIqJP5vijCQuNM78ZgxKXkaIC/lLI+0ZFm+Sn8L1O9jYW09yaxuTJtdiaZMIFMXImdaL4jcgmFYNX4i+vXEPWfhUpklA7v3TVHjSDTveWPEomt2AaELCOSyg8mztlbjt4M9YaE933TaNgYxjD2MT9EkmB/M0KjYvWoUYlGpe/Qsyp8+H78ziwdQyhbMjITSwRtgy4CeUlqjSRQTMpp4dIOa1i7JVhySDWfo0/PnEjcbuOryUJdaaPviMZaLLAmCnnyDiskjW1kzNHCzAnR7A+0kHGkSgpH1nx+y1k7dMRJZ3RWV2IEZF4soqtQ4DCEPrBZAx+Af+YGI/tugIxBhk7DARGx4glJQosajix5GmwxWl9uwRB1Ol6REH0Goi7NLJ3gTjJS3u3i6gr8dYfM+Uc7b8VESaNwbSomZTTGmVvrOLu+joMQzqLrVHybmon6bsUJKcTzjpwHzFQcuMJJs2soeK2qn/bpuP/UfwXJUQ/4IeE6N9Cw7Uv0z4lyEvfLKLtihzGH7mW6KcZ+ENmSIoz7L0Wiq6qp7Ixn4ZbcqnszkFQwDmvm5Wf3Yko6VjvEmj8xVhS71MIZ6u4ajVM/SLpm030zY/xzt4ZpM/rwN4g01Odzgu7LsLap2HtEjBVW+icfV4tuieJJ7Yvx9qloxt05PQw4bUajYfzGZXajdJlZf87Exj6Lg13tUDJ9XXoIRnt1XR8s8OEXsohyRnC3C8S7bdw3we3Eo3JmAYE0CC9cADhixSuLj1BeJUHSdBZOWUX3ddGWPvhUlKPi0gxAdNaN3U3rUUOQuAPuaTvkxj35xM8Nn0jqZU64opeju4uJ5AvkHpE5JHm5Vwx8Rjvt05iYHIcMZCYvR2Z8DGmOgtht4jJr1H0ns7geDe6CMEuO7k7E83m+n4XwXU5NF75CqHlOi+vuJz6lQakYw567w3TviMfiyWG86xM8++noht1LL06+lwPBc4B4kkawoo+xB4TuRVdPPnzN7A0mtiwbxJFn6zk5L5SxAu9+K8MUP6nAIMzo6z7+wKMe5wkNWlcs/g7LF0ibQuNqD4j9nadIZ8F8ZZevCdTufEvfyd7g5FwhobhpJ17OiaTW9FFU0cqrStUrnpiK+EMja4ZJjqCSbindjPwQR5aWZD8zRoTZteS+60GmsBgVxKaDNLtvXgnxvAXSpQtqac8vQdBAzEORY5+lOsHqVw/hnlLTnBdyVFytwc4faAIQ40V8/w+BINGx19KGVhjpOyogdLnYwgRiTsu/QZvSUK3qPgdDaMpMakrefgMcXti/NuSKxFv7uXAYBGeBRG8ozU8ExTc1jBGLxRd0MaOwRGo7VYEd4zf3fgunjEqelaEBQdX0TPJRMRrZnA0SKfs9J1Oo/+1Yexc/zrGbU66Lo4zVBpDnORFKQnTuNxMKEvAvy+dc7VZCK0WpLE+Mpe1UhvKYN/GCna9P4mHF3zFq/nf4C+ykHN5M2I8oeSd/LaDrGkdJO0x4871YukRIDVK1e5S+qcqhDMEzJ8lc3xFOY5GidiDHma9/BAZbj8pVTolvwkwcCIdudNIeHYAOSLwyrh36P4jxFI0YqlWVKtO3KlxbsXL2Ip8AOSleOmYIyO2mJmyuIrkGsj9dT3hgIn6O7NQvUbkfgNVvVlEUnUEWWdqahOphyWMVVakiIBpQR8dV8eJZimElvv4qnUUGaN6UY3wdsknpJyOU+QaYP6UKgwTPGR+YiLm1LH2anTMsRIOmc4v4cLgSCOzVt+JsdPI3FvvYNacKoZKY8QKI+RM66CnJ4n0w6BpiUZuMSrw1NyPaZ9rp2u6QCxNwV/nomMOGIc0SIuiG3WKUgYwjvbhL5DoebEY7+0BujxOzq14GbdjiJ/lfcPAKBOeFUHem/46oTQRY5WVVr+Lxy7+FGetjHlQQ9cFImka4SwVc4sRRJ0RF9fRuyjGzBF1KPO9xCqGyMkZpGxKM67NVgKFoDfZkLckIw8JLJt6jKf+vJa8X6tkZXhZvPAoRRvu4tXCL4jUJWF/touee6cRuM6HtVvgvl3XY+9QGP/71Wyq2UPza8Oxfm1k2GP7ST0RRJl3Ad6bXezaP/q/5P3yA/7XwQ8J0b+B4g9X8pvG4wCcemAN5g+TidsExmR20bjwDXSnjdq+dBwnTBT83c/I9G4s13bzXNlHWDsFnLst9M3MQg4JtDxlRQqJpK9qwlWvMbAsxOjCDgwekb4dOYy/opqkOgEhLhC5zYPRr3PhZVWU3nsIOSiQ87VExj4Bz/QozloZTZUQBZ3yqU20BV2kVAoI8wcJj4jQP17n5K7hCGaVvgkie2e8yN4XX0H+zI1rejf2JpmS9V04rFGSmlUenLMZ75F0/AuHeGf3DPpaXWQ9GOO1LQtI+dKKatbxFwqknlJJe7iRsX9ZTaBYo/UGld7Zcaov0Hj6ravoXqDQ3Z2MpVfA5IVQpkDv+gK+2HMh/UcTatLukkFMljgj167G2q0TKAT1tn56LjQRv3aQznk6STUybQsMiHGQh6B7lkb5a6up+2Upzcsc6CEp0W9yLBk5DP5BG44l3cRcKvYGmXCawK2lBzj1/mjSjkI4ZiBjTA9RVeJn795GOEfB0iXhOiNQ9IsDyNuSiTY5qPmlg2Hvi+TuiiMt7CfqFHj/5CRyL2ohnhrn57O/pn9eFD0u0lOZgckjsObcLCK3egCwdelsrhlJS0sa2RsNmCxxNv5sAYacIXK3h/B8l0mfx8HsVYcQ6m2MfvwUx/aV8bu/vorslTD2yugFYTzbspD7DbjPKpyoG0bb+hIUC7hqVbbvr8D+YhITb6vEbRjitUOz8DwWYc7cU1h6dfo6kym96TiCquM5nkZAMdM5OwkxJPLuaxcRSdHJW11Pww0ikqghxgQWuM4gj/GRv9nHlL1309GcSs3WUmxHLWCPk1Rl4Fx9Fv6xMVr25XH2vREUP3AQfdDIo2/fQOEXGvdP2E6804a7RsFZbUAKC4QL4lg7RfyFIiNeXk1qZQjbaROWJiPyjmRyPjAi5w+RdjLO0VXPY0oNoySrLBhWi/fNPPZuHI8yJsh9d3yOisglN69CFwS6PikgnKUxfFQ7qlEgsj6L4Lwh4rtSGXvVGcp/PYB5QCCp2kBKdZz+8TqTPzzNlvv/TK/XjqBBZ28yvRdHqbk3DbEkyPR51cS6bDAqwKoX7yEYMpG5R6DxSol4soYUEZhw9EcMNSdxV/tUzp3L5M2rX8LaJbCvuRDPojD7qktJdg8RdykIdoXiSa0olcmoKXFKXlXY3DoSe0ecxVcdRBzjo7/BjeOQhfFlzZg3JjHYkczAoUySGzQu/PZeWpYJVLblcnXKYWIxmY//+hdsnQI9U0BQYfg9zQzbGiE0O4ghoNN7XaIC7PhlG8feG8vSsVXI7SZMqw0U5PYTShcxHbKzau520k8o/GL/VZQubsDcJ4IqoNo05k+uJm4VyP3YgBATqDqbT9BrwdKr0zVLJ1KVTPHDPgq33E5nawp/bV/IFbfuIl7n5PZX78UzVkOa7OGe4p183nMBP7plBzGHQMEaMA2KzJ50hmWX7yczb5Djx0uwnjbTFUpiqNOB81srHV0uprkbUSwgBwXWX/MSlgGN1Ik9nL1A4ZbDN1N/o5v4hxn8vbICW4vE0kceJGtcN4GZ/WRv7yPr8rMo8738esbfMfpieCfEuKt9Kof/uJbjDcOILp1Ez2QHO95dR+2q9P/CN8z/hu+7ofo/0VT9A/ghIfofouHal0k5KVBy/0FWrr2HhmtfpvjDlQSGiXxy/9O0/m04hVtup2aVixWlx4jb4d0Nr3C2L4PO6gweLZxExuEhjH6dqEtAM0A8llBfPXlmGOrN/SR9Y6Pz7UIeuOpLRl9aw9vD9lD84zrQwSgr9M5S2LdzNB2/mEY8WUO5bQDTLQnNDH+5Qu57Mt0+B+FfZaI/l4565QDiJhc5XxgwBASMXgHRqFJ7y1ou+f1DFH+8kqhboLMplcuu38vZ36RwZMLH9I4X+esXlxDNi1Ge1Ysj3489M0jXwkykGPiKRZLqIJas0XNNmOMni/ng3r/gPiWQmeYj6ZSR+pcmk7uglW/nP4+1zoTJqxMcphMeFcZbnngQOhshWKjhq0rBsNeJ0Qf+Yqi7aS0HKj5Dk0HZkYrBKyFFdNJO6GTvCWC+pAfBqmBv1zH3iYlG5UGJrO8UTIPAbA+2OiNdNek4zskYAjrhMWFee/dikpoVAssDBNqcuMxhvEErkdwYGXtFYmMS/U11ay/ENyWCpdiPrgqkPdpE11Qj2pZUXvrFi5TndxN6IYcRxZ28/twyrGfMlL0cwlASIOrSuXLYSQRBZ/S4ZkKZAqY6C7YGA/4b/YzK7KLnrjBCtYPOn8WJloVx7LLy5c4L0YrDVA7koGVH+GXdFSguBWcjaL1mwhNCmHsFbPe3c0FZM9HLvOS9VU/qT5r56vJniTkl9n0+nqOD+aQclOnrSCaqSQzNHcKUFKF39TQ6Z0jIIYFVGd8SGBXD6BdIOxnBfVrnyJkirp94iNgJF9YugT+v/RGWr5w0Xu1EllWMfRJLrzyQqBr1GfEXazQte5XU7wwMm96GyaextbMSg08kmqLh/FUbx/3DsHSJ5D9ch61bY/Elh3GeMRAsVJl+6UmmXFyFYpMJjwszenEt4TRIerAV1wYbRk+MnREnNkuU7IJ+znoz6VsYJXVmF5YDdv6w5XL++sUlNF8m0z9eZ/otx1CdKsmmMN1L4gyMFkj/0AIzPdR50mi9JhfLwl7cNTEGRhpIrRTY1lXOzI8e4v6xOxi1tBb3bhNCjwk5PYxpv4Pv9owmt7wH8YQDXQbjCTvTf34Ie7OMMS1E+jENWdLQLCrfHB9DUXEPt799D8mXdhIPmBif34a500DkhBshKiKIOvzcjdEHxi4D5j/2JJh/l4kceWIiIa8FKSyizPVR93UppbfXIEZEJi48g+OOdi4ZVUXSWZlh6ySmmQOsKDvG9I0PYLukG0GFqFunefUIoi4D8T4Lc1cfRD7u4Ib5e+kL2fBVxNi6YwLl05voWJqB9tcMhnJ0gkUqa47MoXuyhK4JNH1eTNypc83UwzhrZCr7cuidqtM+V0RMjTJ9bB3ERLzDYc3iN3E2wpnfpDOqqCOh/r27kA3r5pBcA5kHIxgHRCJnk9nuGcmplhwq/bkcf3QtjXclbHIOfTWGvU9Nwbc/A92sYvImJByKyzvxzwvRdNE63j83kcEKlUiGxgcDU7F/cojOdjedG0aixBIEEfOKbn4+bTOaDIeeWkvfgSxc+9z8YdO7iBUjEHcnc6m9gaFcC8NvO8o3p0bxkjcP93dGtr66hoy/7f//sgL1D/ie8UNC9D9A0ScrOfzHtWztrMQyp48LT1xNyf0HyVjQzorfP8jAKIHhtx6l8cpX+LJ1LJF0jUENrig6iZqksLWzkrjTSCRFJOYEBEj5u4WcPQoFX+oEd6czMCeKFIOnjl2EZ/ogAC8O24jBL+LZl0nhRxqus/DiHS8jhURiisTqgp00LniDpDMyXdNkLJudtFxiYdrvDyF+noIxoNM5Q2DV8s0YgjqGGiuTjl+Db1aEtOH9KDN8uE9IfNk0hktGn2LUi6sp+syHrVMgf4NITVc6sqRiN0dRLRBzqZi88Ppvn8PRLPLCxI9IOiMxymihf36Uzg43zqVdSEGR9h35XLz+YUoXNxBYGsTeLKD7jCTVJdhWsWVezL0iGYc1dBnQYeLMGhauuIVZVctZsvwgSU0KORM7ybqumYFlIUI5FqR1qRjaTESW+MnaF8Z1RuCXV30GQCQdrKYELV1QIJas4x2jMX5YGzOWn8BXKGM1xVmz5E3CT2STbA9xQVkzjpYIX09/iWAeFHyhYbFFyUv2IvcZ8fwyHzEGcTtc/8XdnDswjFCqxKayTYy//RSaESKZVky7nGgG+Pp3czC+4eZMRybhDA3NkDiXYEsSN2QeRDuVROahGKMyEsnsscfW4ij1IogaAwEbWkSCN9IRIhIpVUE0s8ZlZacSBrGOPk53Z/HuuPXUPJ1Px1tFfOybyLWPbSY0IkL/e/ks/ckepo+u53DbMHQd9Fo7ihXsbQKzlx/nzmfvw3rOiGrW8Qw3EU4VmTmmlh1PTadsXgO6DOEpQUbfWY29BUxGBcWms3HzFMLZCvKQiNEr8mR/Oe7r26hryuT+xz5gysMribs0LD0iMU3i26oR6BKcW1OOv0BkT0cxv139NrbcAMffGMueA6PoWRXBZovQ/mIpSmmIswcK0W/sp+lyK0eHivAMOOhsS6G+OhfbKTP3FH6Ls1Ulb7tKUj3oBp3kGoE9H16A1R1iIGLDUmsinhlnYLREii2EpyoVTYL+02k0X6czVKhiur6bXo+Dc9ev5ZC/iNpPyxisSJy7waAiRXTyJ3TQWZXB6XvWIE7zkNSo8tUXU4nbINZtxV8g4a1KpXztEOsXvk5TZyqR/BitnSnMGVNDRzCJaIqKpUdHDorIBpXG5XYci7uJ23Wis7uJ9NhIOSbROUNC9MkoSSraiSTidh1ZVBHSohxpy+dcVS4nHx/PnJsOo/2inzHb7uatXbMQXTEcV/YgxgRKHzuJ6cJBOq6Mo9sUPj04CUerxoa3ZtPncWBuT1DUq07nY/DrdM6QMPcLTBx3jrRdxoTVlQbhKUNkHNH4Zt00Ysng+qMVOSCSP6YLx3dWBu/IQHLEiefF+O2jt+AZCXKfgZZNhaAKSOEEy+/JR1+nd4IZQ0AgacwAdjlG+lYjlW25xHUVg0kh/UjCcHbeL/ZhmjhI6ZsxwqkCxR+vxBu2kO4K8JPOSYSCJsw9MpYukWRDiNbfTKPwY4jWJHH1mONEnSJ7xmzgy2tmkr/JS9GGu8if2cqZvgxW/vo+9DMNhLJ1Fv/+Qb597m9MORmn9I04X03K5+gTa+lRo/9r9Az9K75/L7Mf8G/iB9r9/wlK7j8I8A/X+4uyxyGVFNJ4YxY1d6zhouxx/4jtfGgaOTv9ND4gIVfb2LXyaV71XMCOh2fSukjCXuQjUpVM6cxmOv1O9K0p+Eaq2JolbHN7UT5PI5wmJExHJw8SP+DG5NETfTTDwFLuJftxgdZHBSIhI+mpfrwHM3DVaHhLRMI5CnJAwjW6n6Hv0jBNHSCmyMTPOMn9Nsbc5/fxzt/nopp1xl14jhPHSsjbppL8cCunmnOQO0ykH9Pomi6QUjZApj1A3e5CLD0C3nFxhm0A1SQiaDrW+zqo70yn7N5mGu4vR7HpiPFEmX32xDOcWTOaQL5AOFtBUAVMfRL66AD2b+z4i8Fd0cdQ1EjKqzb6xhuIOXV+cukm3jg3lZgiYTIo+OtdGL0i0eHhf7jDO+olfrbqU2529nLRFT+md6IdX2mi2Vix6eTsUfDny0hR8C0aQonJmGvNSDGQZwwSrnTzyI8+5oVnrsYyqKGYBOSITtd0AfeIAVy/s1B3h5HU7wwMzI6hawLWc0bkEMy98TBfVo4j+bgR7xgFRB1knaRKI+bFvTj+YKfhRyYQwTgg4qrR6Zmio5s1hhd3kWYJcqChEC1oYOvFz7Fkzz2ML2ij8lAJapLK9ZMO8tGWGeRP7KC5O4XLR5zki70Xoss6KcdF3G8coO3X0wgXxkAVGFbQh/SnFBpXiAghCVu+nyG/mYwtRnqm6uTsAtd9LUQVmbrGLKyNBsLZKqvmbueNs9MYkdHNw7lbuHX9vQmBvrwIY/Pa8T+WR+tFJlynQdB0BscImPoFDEGd0TedpvrtUfimRWiYv54xz67GMrePaFwm0OpkwvgGjp0uwtwtc+MVO3ht32xyvxFQLALdszTsjTJGn07yuRhNV8g46iR+d8+bfNg7mZNfj0AOJejn5kGNiEsk6oLbr93CQW8h1VvLcDZpeC4Nke32YVzYQtMHFegtVkQF4k6NkWNbMYoKTZ4UhsJGDCftyCGIpOtIkcRLQRk5hNZh4dx1L1MXH2LxFw+Qv0UlmCsTzBXInNpJ8KMsBqYoNC19DYDSd1dx99LNBFUzb22ei3m4D1UVyXjdQvM1OqWvx2mfZ0MOgWKFM6vXUPj1HZiTI2SvNdJ4g0BezgC9PjvCGQeOJp2BCh1EMOUGiQwZIWhATgnj/rsV3+VDxNpsZO/V6bo6CrrAry/4mudfugrfCAVHdoBQfTLkhLEftBLM0xNLcetVABqXm0iqF4i4BcwDOsNvrKXE1seT6VUUbr4dySNjGhSxTe/DP2RGFHW0005ihRFko8KwNA+tB3Mpea2DhltzmDC/hsF7shl8MkZfmytBmy8Oc0FBK8HL4dyLOSj9FvI3a4RSZTyjQA4JpFSreEskTB6dcbdU0XZ/MY13i+S+L2N/qJ36njQc220MTItBXEQwalidEXJ/D41XOUkb34N3yEKey0vdmVySq0UUm0DWc4foeGgyUgQ0E8QuCLJt6hpWLr6VoeJk1Lv7kUWNXaO/oOLwCrTvXFTdv4b5Z5bRfiAHk1fg1ANrEr5+yvebLHwftHtzbp6ee+/939cpAdDwi/v/Ldr9D/ihQvQfouT+g2ztrMR/3RRG/3X1P7ar55qI5kdZXDiZrZ2VnHt3PM1PTk3I8nd7iPuNjL24hilf3M+bW+dy6TM7aLj2ZU5e+AGaCU7X5+J4LYnYPB+6SUWd7Ce8LR1fSaL0HUnVyHwsocMScQuM/vFplOwooZpk6m52oigS5Q91MhQ1EkvS6J6lk3UwStJZGUNAwGUOY5wySOZNvZSn9RDPi9I+38j+gSLsYwewt4hU7yzFdVrAf5efk7X5CB4jyRWJvgLNquE7kcpj+RtJrtXxjlGwpw7huyuA/yY/UadE18ZhZKb6qP1bAe6zOporjr1NIH+TRnV/Fk8/vpbyRfUIakIxOf2EwuMVX+FsjWNrExAEnYsLzrBr3WuYB3SSa+H54/MYOuMi/8YmPAP2BP06WyH3YwNxO+hmlUiazicLJlG49Ta6pttxtiiYc4K4azTsrQJtKxREBWbdfYj0Ty04jpmxTe2n6mdrCNa6kCLw5OdXE1ocoK9ConuOSs9kkeSzAv3nUlCsBjJ2yngXhsn5UuaXUzehVAQZujDEtk8vJG2vAXSQnDEkexz3IQMxJ4S/SadjtpVl046hm1WS63TMt3Qhp0WQvTK5Ni8nvhrJnyd/htEdYbjBRsP89XxavJ17lmyhpKibDQ1jsXYJdO3MBR2+2jSZy2ceRjdo9E9UCV9+Ie4aFUSdayYdIfJ2Jm13KEwdeY7kGoFoxIDLHaR/vIC1U6J9iUZ30MG01EasjQbGX3qGxitf4SF3A86Nduq+LmXFzruIJWvY2wSK/6ZSeaqIpkuNyMHE0m74ah/mXgFXvYLzii6uSzuEYhUwn7Uw6sXVjL+yGu3zVBzmKKMrWmhdV0rqIYkR8+p56+/zAEj5aTOj76ui/KensMzuI2tFMy2LjaQdFPGPiTHK2Mvhg2WE8hX8Y2IIik7/JRH8U8MUvt/Ji7sXMtbZQcFn/QwuDWM8YieyPgvfphKy3zWi2jXuWb4J04CEpgv0/6WQ+H430hk7oYI4oayEGrMuJibKalfCNqfom9u4aPt9SGkRuqbJzF11kNlLTtC7O5uU50ODOwAAIABJREFUdQcQYiKX1C1h/O9XYyz288KOi1i/Yw6lb/ajHk8m2mbHf7cPa32CwhnOVojbweiH6T9dSea3EvluD8EcI/kbRLo9DtI+tBIvDRPMFyh7uQ/nOZFoxIDYa0IMiUhn7XgvG+KZ8Z9AepT2hTozixoQJZUXn74SW7eKoAu8MOYjCie0c+voA/gnRCnYFMV90MDAQyHO3SKRUjaAp0IlXBxj0q2VtPhdbHplBsPfXEXKfgNqShyjH4IH04h1W1FrHMTcKnpMxHTcztBrOcTSFM4+kEXut1Fanh9O/f1Gsu1+TD0yj1/7AQXrRU7uGk7XNWUoAxYERcA3zIBy5SBqegxN1um6LIa9XcNbrrNv61jOXW/GlTRE50yZJenV1M58m5888AlNF63D3G7AdtpE3uMqrUuSsI8apKcqg3DATNeXw7B0SqRd3Ubul11MOh5jwuXVWHs1zP06WW+auOaRB/E9q9B+TZyku+KsHf4+AGnPW4hMCHFR9jjE36WAkHjOluy6+XtPhn7A/z/wQ0L0H+A3jceZevJKrD1xcp7az54INDwzhR/XttF00Tr0aJQli66lYd56RsxsRFQFui/Op+zlEIePlyKFRIa/OcjX3aMp2XUzhVtuRxd1Ug7LBHIkjDuTKH5Pg5NOxlx9htL1fTgbE0sBHfOTiTs0QiUxjn85Gutpc0JMsdBLRU4HLTcVszCvlpSyAUrfCeMfZiTtRAhrt45JVtB0AVKSOVZTiOO4Gb0oRNPuAmyvJyNFdMz9At5yneMTP6LpktcQ4gJGScU3PYJgVsk6oDDOKPPob99EsCiEw0a8gzYCbU76F0XIfOEQ+vp0DI1mFLOAo9qEr1Rj0pNHMa138dOnV1HzTSm6RSWcodExW+Txj66l6SqBQJFGdGM6W1pGsHTG5QRzwdqvkpw8RHIt1Dw7GiEoE01XkQMSbVcrhPIVkk8YUew6Z36dg/20iUiKTutyjeRP7XTO07hu1VaM9RYqbqvi88MT6ZouIC/oxxdMGCjtuPZpjP6Eam/YayZ3ZwQhKqLJOmafhiEzxFCWAe9wAa3XTPcUEY9iQz5px73VQiRdo292jKFZQTLcfv4wcQNpP2olUhJN0NTdGpu+mcSIZwP0TYSLs6qZmN9K/Q1rudh9CjEGTz9xHbEBM7/tG/mP+2xLzyhmp9WjnUoikqqjjg3i2m0me0on29vKsNcbcGQHMPhVui6PUfgenLi7gqFsEdNxG4cOlVFyfR1Kr4XU35lQUuKEsjSsjQb6Bxx801VOOEul+oOR/1BCD+YJCV8pAUreDSCoOueutWIckCgY20m4KIb7+jYsxjjBYhUxptN6NpOn77wBYaYH1azz9h3Pc7ClAMUm0Hc0g9rvCtEMkHJjK5WN+cRcKk2XvUrDYApHu/Oo+9M4hA9TOV2bC6LOFQ9vx1llZOlbD2HpFrGfk0nfZcDepZL9vhGpzUzNPZkM+7vGuqMzKHirFZqtDOVp9FcIxBSJlqs0HPUSf92yhLN3reH5ok/omCsyNEwh+ZyGYNIgP4y5XyD/t/sp+NKDnBMib3ucJNcQliYjYoOFWHac610H+e6L8TjadFp/Mw1TeojudwrwF2vE65xYuiQmX1hL/MUIqklHjAl4WlxYu3UiaYllndzZbehiwtKle4HC1hFf0TsvhmIRSfnSSvsiHanRTLQsTNeiDKw9KhcUtJJcCxfPPUo0XUXXBdbMXYD9iAXREaf2+YQrgr1ToXd5lIx9Ar954HaaulN58+t5DH8xiqfUxOAElaGTbiSTyn9j772j66rOvd1nrbV7kfbe6pLVm4tky90S7tjYpplqwIBpptihhRISAickEAiBkNAMxvROsI2xwdgGG3e5yd1Wl6xe99bude217h/i5uae+31j5OTjnpxzBr8xNIY05xpz7bnG1prvmO87f09gbwoABTn9bDtajvBuCqogUFJ1Dud5MYSABs/kMIlNCtZWiQULj5BySEQyyRhnDeAcI7Bs6n6ydkDnHANlD5+kcJXC2f0FRJPjPPrtEnqn6sn5Noy3WCHzeyj+0I82qBIM61CDEvZaFccOA2GHiMYvYBrvxH5CxHsiiSUL9/L2ixcDsCxhGK5697Wb8BfIlL1bT8ShIG1wULv0FUbm9uAdGyU21k+Pz4p7UhqHnbnsPTQad4lIKEVAuX+Qd556gdH2Ph6ZtJX7tm9hlM7EpH9bwXcfvw0dRvxXT+XbT96h/tbXqL/lNdTe/zdv8r+afiqq/tfpp4Dof6FlG1ZSPW4d2z94i1Vte5lpgMKHDvB+aTalb69ga/dxUtZ0U/rWCjYUb0UKwcqff4F67AzF9xxElVQGpjro3zKCwzNW0brwTcZOaSb5unYidgFPsULLVVry1g5wcPcoah+2kXw8iOOkSN6lLQiOKLYaHYktCtYOBXehhGfITM25HHI/asMrGxE/TqJ7hoXrH/yG9oVGnNNibCzeQuYTAnX3p2JyBIklgCgpROwK0Ttd6PwqmqBK03Wv/22ucUucfreF3Pcl7py4m45rZc6/404uMoW5vOw4ap8BQVJJLnCRuV5H/xfFyMucvHXDqyy4dy/STBdFn4b4YsdUuhbFST3sI1QYQdAqiGlhtF6R9Gk9WOu0pIwZIGIXyPq1Qt296eRv9NN+EaT8VociAaKKoV/C0K2BESFyP5bILewn/5pGxk9swpbhZcQrR5HNKqOecbH419sRDHEedjRz55LNPJaxBWO3hsQGAVenDfMeC+v8CdzbeiURBzx16acU5ffROVePGBUw9ot0LVCId5pwlQtEMmOothjLL/yO1w/PwtSr4hqrolhlkEX0Ryz4v0nnl1uvpW1nLtouHdZ2hZK3XMgmlZbf6Eg9CFvunc2lycd5eSiXX396PYIK5fecwtKi4a9N4wEYjAeob8jiseQ6tD6IG1Xs1iDhZIHOATvJL5kovrSR2DE7g+V69HVGeqfq6fi5QuLcXhy1MtllvdSvKyX5mMDgeAuti94kpUYg7XCEtBQP+8aup+Wq1axcuQGtR6TimZVULT6BdvwQekuEzgWJeItVVEkloUWle88IEo/rCL6SxaPFmzG3SvRUDRvutSyRCNbZGDu3gYearsZ40AJARrVMao1Cyc11tFbnkLJDh+O4RP7m5TxZtpHfjP6KovJOBheEGTeyndzNMT7/yzxCaSq2BnXYXiAKvlyB6J0uumdKxBLjqMlR+m8PUvpyiG+3j0cKCWSU9nPTou8ZclrIWScRN4AYE7iscQEXf/wQSoKMxivRV6lirNdjtYRQdCCfP5GemXZiYQ1tl0iY9VHKFg6jlIrflrm/8Rruuv5rsm9vxNqmEuk2E1joR5VACgtEHCpH2nOQfmmj/tbXMAwKFI/qonB5Pd0zRCIpcbQrdKBC2KfHdlTH+WcvxX5IR99UCF7jISV7CFUC1aXD4FIYuCLM4doCpAhsXz8ZVatQmDpI659tFF/VQNpXenqnq2jOWGi7UmV5+T6GRgp4czQkbTWg6FTqf6bHl88wPHlyH5ImjjDZAzqFyOsZWFo0+HJE5i2vxvlGLsSGa4c0PXoEFWIm+GbHJJwVKsajJmy/NxFLVDjjzaBnukDS1F6+PTyWjgVmps0+g7FTYt+lfyKcrCAbJLJ2DTvQ98xIwF0CFVldaLwSybe3kXVzC4IC4+fW42my454RRgoKSILCUFWE0ndW8KyzGO2EIf6040JaF7/Blk8q0XlEpIjKqI/vpr5z+ESqJKnE4yJSVMW/egQjdihcvLh6GMb7ZCKLP3mQ0650ZhibuMAUw6+EAZh9++2oI8KElw2xsmsazTH/f48C6p98iP5l+ikg+t+o8NO7WJBZwfzv72Ph4huHa4mmjSVzr0zBujvZU1dMLDfCnFtvxzRzgNsSe9nafZxl9R2kHYKUA04ULVSteYh7uydz/FQBK7O/J3Vu17D3yIBE76xkBBXsNVqalxj5/LHnCD2WQeZ6HUaXwj1PfoYnX0SKQNo2LZnrdTTcm0NEkTDf0s1dt2zixT0XkPNtGFOLjrHPr6RhWQKFo7qJRTWEU+OUZfSg9Qs4T6bQf2EEVYJH+8YCUPbiSlKrJSy7zKCofHvvDHQGmaESLUtazscrG/lo8avg1uKqSyKYIhI+5sBfncLTVYtY21BBrm2IrodkSia2Y27S0XifFqIihW+qGIxRwnlR2mvTST4d5aKsM9iaFBgcwtQpUr/cgKlNw5aNHxJOFkg4q0XnHkaJiM1GXKVaXEEjLWuLOX6wiK0T3qLzkwJSiweZ8Hkja1+Yh6iLc1njAl79ahFzt9+PYbKToQkygkkm6WyYCfpuGr4tRD/JxWNHLuPGrGpyz+sgbpOJ2FUErYIu18+I76JY6nRkfqXh4zfnQ1RENgloAgIJZ3QkntSS+fx+/JND2POGsFX2oXcKXPXYNlqvSkbnEYgOmOidHWfg3iBPnr6Qv3y7CN1YN7ZmmSMfjKPwkmZSE/zsDsMNjUvIzh+g7KWV2BvkYZzItlQMgyr3jvuezjk6Ot8qwl6vEE5Ryf1qGMwa6Tdh/YUOTSiO/EYaGXt9hBZ7GBo9nArQLO2jY76OZ0vX/e27fEdiN0VvdBLMVGn83WiCzYlotfHhNJJNJq1oEH+OQDgzRsQGQ6USawcn4S8dLqLV+QSSjkhcvWAfsiLSdjYD+6Ju/LkKXTM0BJNFDhwpRRVg6MJhs8qMbzWYhAi/e+FGWg9nk/aVnrq+VDSP9uGcHEcVwXZrx/DCbIbY6CCuEyk8ePFGSkd2kfKtnkfGbMX5ZAxbmRODC3oaUjjpzcJWo+eW5zcQs6qQF0BWRYre6sWe7ENb5AOrTGxsAF+dg3B5iP67QwSyVfBpMXVKWH+hoz9oZeLcOoZ+FSTZ6OeFXQuQFQlfnkDGPqjKbiW9GkyTBhFlSF1rIOHP3RTvvJnYZB9t+7LpC1opmtCBfkCi9pd2AiOG68pQIcXox58NQlxA2WvHX51Cao3CkwvWor+5l/xUJ9k5gwjXDSBM8pCzSaDzqzxiUQ2tbge+HJGXF77HFVfuIXmfltf3zkEtDeCdFCZ4uQedSwRZJLHciaFXIvZRGtJpC1Mz2xACEoPl0jC2ZLKPbe9VYumKIBjiNH9QgpoTondhDL0HUo+oxM0KcT00LtMjJkc4u60ErVfE9FQCC6ecIHdmG6cHMtB54Mm+eZgLPPRO0zHq0VPEDSrBNBV7HehEGTkxztlzmTR9U8i8W6s5059O5h4V83EjBqfKuuYKDI0GpLDA6l1zcZhCWFolSt5fwamfryKcLiPd2E/cGufRSd+Qe3kLxh0WNk5ajWZFL8EUkWCKxH3Je8ga00fTjdq/nVC9aNPPWZBZgUU0kPRmNdc+t5nCG45TM/GvNE8Oc8G6h/5Vy8lP+m+in4qq/wEVPXCAyEWTSX+0mU/zdwBQ8YeVpL20H4DItjx2lm1g0QXX0nCbjeZrXueCK28imGkgeJObo5M+Y0nL+TR9VIJsFNAGVPw58PPLN/Js9SK0A1qkoEDu7DZ8UT3u3cMnlQwDIpGRIUzmCEZdDPHjJAYWRhjxmZb2S1S+W/hnltUuI/RlGp5SlfETmzi1q5jUowp9U0Rkk4LeJSFGwdin4s8Z3gUhLiAFRIwFXrJtbnrW5uGdFqLo1Tilr9Ry9A8TCNtEYhe7iUQ1RPtMJNeIROwCJx5eBcDoVSuhwkuiOYTrSCoZU3tob0jDMsKL1RCh/3gaxn4Bf2F8+F69w78Dw+nDGolAloCtUWFg4nCdR84Wmf4JOqKJKqZegagNtF4I5CiYukT8+TKGlBDhQSMY4gh+DamFTjItHpo2FBPIHmYkiaFhtlhCg0j8giHCYS3XjaphpqWOu9+7k/w55zi3PQ9hvIf5efVs/3QKgZw4qkYFUSVrm8jel1dTuP0WhD49ik6l5arV5G+6nfx1Kv13hfjVmC38/oNrkE0qmqDAossPcObO0dz40Te89uur6LowTkluLw3n0mld9Cb5X97Bs+d/xi+3XovjhIj1mm7aupJJ3qkj7BDwlchk7BTpXRjDWK8nPCbEeYXNHFtXhuX8PgwamfNSWviytRxxtw1/roIqqkghkYJ1fvonW5FNoFZ6SHrHTP+yELXnffC373D+5uVcMPYMR1+vIOWgi8E/KsMuzbUmUmtiuEZpkSu9hAM6pL5hqKW1BXQ+Ff+VPk5P+wiAOWcW07s3C6HcS84zKq2XJRDLjaDGBQxteq69fCdbn56J71ovVNuQf0jPhTNjoAhgiJNwUs+Ea06xp7mIzM+1dF0VQ9dkJGZRiFsUrI0aTOf3s7bsXeZ8/DD2WlA0EE0UiNpUkk4pxHUCA5MhId9NoNaOrQ4CGcIwJLXKSTiqJfVtI+0XiuSU9NFbnYmgwvj5tTS5k/EfSEEc7yEc1lLwssq5S00YR7nxdlsx9GqIm1S0xV5sn1rouywKwvBpTSEOabO76N+RRdSuEjcpFL8fpPF6M1JEIG5U/lZcf9eV33DInU91fSH2w1qi871odyQSToaIXUHrE8j/wov/9yF6XQnE3TqmjmviUFMeGr2M7Rszznlhzitq5tC2MopmtRKb3UPjS1PJ3qbSP0GDoEAoN0bORoH+CRqsbSquMVD4cDWdj1aR92k39g/d7G8qoOTFCM1XJyAnxDH0a5ANKmpOiLKsHs5WF6B3ChgHVIZGgrFfIHOPj/aFVrRe8JXEUQUVS4sG2Qy2aX1MTW1jT3cBrt5ERKOMKovYHH70n9qZ+uARdnUWEdvvQNH9MF+/gG3SAP5dqcQm+En8xkzSMTdd8+zIJjAMqqTvclJ7XyLvzHuLW3bcRuoeDWG7wMKb93PoV5OJJkioAgQyRQyDKut+/xyXPPsLgukqmXtjhO5zk3hhE0+2HmaKXsuCzAqUWeNpWfz/P6j1RymqzspWc3724xZVN/76p6Lqf1Q/7RD9A2p6YRqxe5wc3VNK0cfDO0cJbcPH6p9tPYj+gnMsyKyg5XEdzde8zsSaJQSzDLz4/MsMdSUyavVKErVhvIUq/lyFoTEKepfAF6NT0AxqsZUPAjDJ0Y47YCSaqPL0ws+wT+8lc72OnPt9uGqTcJYLKAEt8bsHMbVpWbD3HnrrUvFOD1M+vpU2j4Noeoy+ySKxlBjWHC+x4hApJ2JYrulBFVUmj2qhoKgXx0gnkbCWzaWbkSIqYpeB1vvhwCuT6J4F3nkB7KstiKKKNdvLQKWMbPx/nsnZlavQ7ktAVQXkvDBPFG5EkxxC3GHHvSsdckL4imUcx4Z5U/5xYfK+kNF4RHSOMK6xCuHMGM6xAilHYN55J4jYNJRfXIeaHca6sJfEJgVluocb5uwh5USUpKMSkX4TgiKQukOHqlNwH07Frgtx8qFVWPI9FL8fAxUs50TCqeDrtqLXy7x/qIqjoTyST8dpOJJLzhYvz4z9gm/PlbLwumr06UGkhBhCTEQVYcqxq8lKdaOmRxhR2k/BF3eidWnonKslFNTxWe9kYgkq5jFDoML64xNouEfHEzWXwPIBJJeGhsbM4R0DIHejyht3Xom1ReLI716jpzqTpRWHuPLB77j4xr0A9F0coeWCtwjmyujrjfSHrIQnBOltdxBdk8Hm9tFkJnjxFcQx9oiIUYGEZuiebUV/UT/BTAWhOhHXrX7CTiNveDLpkf0Uv7+CySNbOfjBeBJbI7QtTuI3pZswHjGhSvD+63/GVyijqU5A32xg7uzjFL/nIjAnQFwvEG61clnjAh7smUCB1YkUActXVgK5FqLJcVK+06Pr1OE4q7Dt9zMBSHnNRMa+IFG7wmNLP8PYoSWhTgNxAVOvwkhzL2lJHnoqJRL3GYjkh7l4zhGkgIg/N04gomPmhgeRIgKDE1RMS3rx5ypE0mV+88zbJO3qYNmc3QiCyoQZ9fizhoMhTRA8zXYAnGOGSeztvQ5ytgQYN7+Otr+UcEFWHTlz2oidSYBuA+2LTCgaiB2xgwiRJAUlO8zNpQforRT48rxVJH1jJJIWp+LCWnr3ZXHmnlWkju1jdFk7DbcZ/naSLbXQyYVTjqNK8OKeCzi5aRTLJlbzwkOridYlMOuWQ0QKwqhmGUUHfZWJdHU6SF8/vGB770ihLK8bRRFxLwqghDU0vzAKe72CTpRpeG0KJR8E6ZwrEjephEbI/Gr61wwVaxh3QR1xHcStcdw3VpKz2U39k3YOnMvjwJyXqV9u5uoF+8guHMDaqvKHKz5Cf8pE6/pCEhohYlfx5QmYegX0cwcJ/95HuDBCYIRKy+WrERSBWdfUgAB9fTZqJ8q4mx0IIRElJjIytwd3h42sFU0EZD2hE3aCY8JESkNoAgKzFh7HtyeV3PeacXxpQlDAX5CAb3QUIQ7pOwZQ9Fqemb2W73xjaL1oDaKs4h0doyNo54o/bePRp98lcJ0Hb6nMwCSVxSdu5divV5FX1UHvVB2SOMwhe+Dhu2mN+XHeVvmfEgz9pP8Zkp544ol/9Wf4T9NTf/7zE9aqf853wj9gpvDhA1z1p3paX9Ohre/iJdflHM1Ppdsyjo/XfkJZage3PnIrNJio/tPr/KLzfIxXNNO/JJNHx25k37pJJJ9WCGRK5M5up3VRPn+Z9xGn5ycyND2N+pO5JOzXIUZFtnaXIVVbCF7lIRZKJWYRsY0fxJQQwem2EJVEFFUgo3CQ1BcNCO/LdFYlYKw3UDCrjag4HOvmJA3RKdiRUiMETQLhtel06y3o9poJ2AVeOjMZRRQxDArYDw3T4Ue8fIpwWg7BTIn0MQO8MuoTHGlBqoPZfB7K47lNs3lv3WRkk4BphwFPEez4eioRUQOCQDg1juDWolri6JwaQoUxTI16nONEXrziXQalBILbUpBCEjMvOs7NF37H2ucXIMbh8kX72N04kviJREJpAvKQngbJjv6snt/87l2EZAWnxsCAQ4exXUtSVS+1Bwq4Z9wR/tQwGc9UFWODntfufoW/OivQDGmovegd7i85yjZ/Jl1jtJQWdHM8J419Wyai9huobR+B+ZiOGMOeJ+ELfXhbbUTrE2hYsoant85BscZRDApZO1Ukpw7jG+C6EBYXnaSpJgdDr4SiSKR+LyHXWnl4+VqqD5bzb7d9xK/a5jBUodJvSECKCERGDdBktnHo0EjOVhdRfyaH7G9j1Dz4Nre1T6fVmUTB5yE6xhvQHbZQXNVOi82KeYcZT56K9rSJYJaCYpMJJwmkH1DoTjVS+FmEuF6DO03C5Aiy89xI3j04E21QwHcgiRcffI0t1VO5fOX3/PHLy1E0AqpW5bGJJ9kmpJFQ4iYh14t/qYnsj3tp35eL46Ju0p5TOJuVxqLiU3yxfgbhVIVAYRxZoyFzV5zuRXEyd0H/1SEMUz0IJ0wYNx7C9kkMS0qQlnAKxjta8T5sIjPVTZfBQv+aPMxfqAxO1hLKUrDU66jtyUTKDVKY109vYwqmHomcuW04Un34P8oikKcgRES+ahxPoMCOnB3D+3kWN83bwfa2MVgnOjFUG4hMCWI1RTAXeQn2WNCmh+kfrWMobsCdqKchloTwUTLhJBFTr4AgC1xwyRFqVQeWpCBSvQlNv5ZTNUUYxw/xadck4k4DKCLngjYcJ0TWZafhXZtFh5iAblDCcUbF0gXusIXgJyk4ru7GmBDBukHPvsx00hx+PqvaykJrNx+8M425F5+gS2tCOGcgc6fMky+tYcu+qXhKTXQHE1gw9gwNvWncNmkfx1oLMV7RR9OJbKZMbKQxnsXkubXEP3UQmBjlyHfl2JvitPpSiVX6iSOgneAj6rJRPOccvu/TefvkeTAiTPvGQoZSBQL5CtWeAiLJcVSfjnA6KFqVETuiuEdK+DQaosfszJp+Bvm9JG6+9Cgvnp5M8MN0xIjAiOnd9F6RRcSnR++UMHZoyCzrI/HnIU6WjaDjdCb3XvEVh46MZlpFI30GA429aRSuGqTzJSuOLQLOMi3jbjvNkGogGNLjGmdG1en57lw5p1tymDq6mv35Wdxb/h0TrO2s+v1VnFgeJnhdMvm/8+MttRD0Gwmk+3gjt5pNl8bwWEayd+NXPNc+lcemnODZ/v+8jRFf9QEef+CB3/6fjPHUC39+InFy5Y/1kQBw7dja88QTT7zxow76P1Q/pcz+gxJlAfsZsL9X/be2/9urCOCq5nn4ZgzS/YsqzLP76W9Oovieg7iXVRJMF7jnlg28Wj+LWSOaqJ8Uw3P9NIZGClTNP82Z18vwXeQn92mFYK6ZnmkS5ISozG+h8eXRw2mCKhn7MQ3mXgVPvoTh/AE8x5ORSn0kf2hiYKwGBDBMcOF40Yzx37o5U5dN0uHhNJLn/BCyV0f5qHZOnc1h1MhO6s5kU/JugN7KBAwuhVCKSNwAoXQFS6tIOFUlZlNIPCPhHiuTu1HFt8JDVoIXZ8hE4Ot0wjN8CKetxA0qmXtkuuZoULQqEyc3MvTLbIIZemImkaFRYBrpJhjUI4c1CKKK7YAOzSWDJJsCCMtEmm/PJpImk1CnRZ7uoTSln9P7i5CCAinn9TCwPwPrOZXBuRFw67AXuFiUXcvBeyYx69VqNj89m55ZCnlfqgzeGeDBUd/xXkcl5zpSSE71kmH1UrcvH0sH+PIgoRkylp6jbXM+QtUQgdZESivaqW3ORPRqUEwK+j7N8HHlsWHs+/XoLuvHfSANodyLxRhhcCABQ7MepdyH+TsLgRGg6CCeFSZlq56+WXFS92j42aOf81HXNPi1g8alRqzNEvaGGF03xlAUAdWtQ+MViRtUECB15ADRDamkHvRQf5sV/YCEEB8uwpbCAuG0OIZeCYNLJZz0Q8rjvAjZf9UwtNxH/KCdx2/+hKdrF2L+ayLyUhfBfcn8+bY1nAjn8E5dJZFOCzljerhhxEFWNc3E4zVT+KKMr8CMIsHAZEgf2Y8AxD5Ow/7ZUbo/K8A3ZELfoSP9oMztL6znqY+vIZwhM/qP/ZziESg8AAAgAElEQVR9PAl9hw7ZOMzgy9op075QwtIuIs104fUZkSSF/OtO0LBqCoIsYBiQiOtUjAMCmoDKnJ8dYMvHlfiLhr2fNC4NxpFutJtthBd4CbqNaJxaLO0CUevwM1FFkHPDLBp5lm3fTsDUI3DxbXv46EAlhZ/JNF8r4ajRkHwyQMMtehAgJcuNy20hff0wCb57poDGLxCzK2jdIlFHHGO3hhuWbMeh8bO6cQaejkRsOW7i3ydRemU99YOp3FR0kDW15/Hi+E/ZFyhhV18xWilOc1cK9eevQSsM/w8eisR4vmshHT4bBo2Mqgq4tmTiz4sjhUVkm0xu3gC/L1rPXa/fjWFQ5a1/+zM3P/tz9Jf249+RhjzFR9ivQ/BrEOICKSWD6KQ4u8u/oOrnd+EuFqlYVEvTGyOZd/8+1m+cjhCH2rtWMf2eO9He1Uu/18KZyo+Yfs+d9MwQEMMCmqCAYRAQwV0e4/CFf2Hy5vsxt2hJmdfF92O+ZPzTKzFf3EvPQCJGc5RAv3n42V4nIQYlmpe8TsUzK1n30B+58MOHuXPxVjY+fD79E7TIVpVps85Q3VpA3KVn+sRa3FETve/kMzhRoeXK1SzIrKD9N1WEM2QQVNKyhzhQsZb8rbch+DQU33uQ7l9UkdAax/L5QXzXTmNggkDREydoXFOK1RLC22z7P3rP/zP60VJmK37klNnjP6XM/lH9lDL7D0rR/GCsBohWKz0bRlHw3a3Mv+4WAE7tLkYsG4lh5iC+PalMm9DA1u7jHPzDa2RvGuCZ6gtJtgTYdHIc3b+oQhtUUSXYeWwUa3/3HA5rgOaHtbgLNaTWKIitRmo2lRFc4iFqFRCiIgaXyuA4CX+ugsMYpP6W1wg5jfROkzBMdBHJjxCrdtC+PE7nunwEWSCUKuCaGUEOaGm99A1On87FlBqg45s8NMkhBiuspF3WztBIAV++ghSGq2YdIJCtkHJUQQyKXHD7frRuCW+uBld3Iuc2FuDymQmlq6gNluEC2WSZgQotgjy8mJ/6tpQJLx8nYhOJ66F02jlmZLWgP2lCcmrRdOsQLnIifJpM8Pksah/OQogLpO7V4BsXIdqUgPupXGIOGdmiMjm5jZhVQdGCGtSApBKJaVnfNI72BQY+/GIuyjA3Fn+GBvO6BF6oOx/P+uH0ledUEu3rCtB5h+0HLB0Cjzz0MQWWQeJaUPcNm0I2HM0hITlA1i4F3YCEUhIgkK3g2KsnkAXhqJZYSQi5yYrrbDJl+V2oEiwuPkVkoZdIcpy0gwr6eiPRJUPYazRoQgqtkVRCspaO+RawxYic52PE442kf67HttuAxi8Sy4iiOmIoCTL9tSkMjVYZejrKxZVHGbErhKlPJZoZgwovQlwglCXjzwZtAFxTYyQcMTDuiWPED9qZf9Uh3vjZFQQDBlb821rCO5ORK/zc9c2tfPzaAq4oOkH21jjya+k8/c1lhPcnM6eogYmrT7Dt+b+gDamY20WSjEHi76XiudjPltaDPDTyW7QGmUhqnK6ZGp575RpMvSqV5Y3Ur8gg4YQenWcYHXPpvIMM3h7EXuBCP2eQMSm9GE1RYh49ztsqESwyibkecjcOEU2XydjtZtztp9jcOhrL+X04stwYurUkNkOgNRFvEdg/sXDV+Bp0hV5Qhv1lEiYOMnFOHa9Wfsz2ryYSSx6G1n7TMYqV07djeaILgyNMyrXt6J/tp/WSNWgHNQR3p1D6ywG650D/JBH7WQFRFpAcEeIFIcSwyC3XbmXdq3N55w+XkvSXYR7h/Ox6olU+jrbloP3axst75yGcsPKztcs5eGsFzm2ZGDUxBEH9WzAUVxWu/+xeWt8qoa8pmXJ7N/17MokkqegHJYTs4HCAtiWT29++m8VL9yAosPS1B3CXKThPpGIcUJEkBUe1DmuzhMYvYPs3A66AiYL1d9IzW2HsojoOteViGpD5pHoasUSF9BldlLy7Ak+BhOvrLDZOWk3pOytwF0mM+E4hnh5lweJD6PwqsglaL1nDlC8ewNihRTareNZnUrDtNjI3D3uhzSutI1qXgBAV6Lk7Qu6XkFfWzWP95chzPMz7+gFks8rrp6fTW6lBlaDoD2c5+9YYNHUmpo1v4P3c3YwwuVE0cO/creR/dTutf6gkkhyndHWA0U/1MCGlk/wNd/DYtK+pmlRP4t4kghkKHz7/JwJXTcU1WqD46bMowSCF1x/7lwRDP+l/hn5Kmf2Tci0YwaHV69l002jsb5ym+7oCnveMpfDeQ2w5/h1vrJpJWk2IDXdsAGDyYyv45p01bH56Mq5BO6mHVIbOj6Dp1RK1Q+EDB3nXew2Vs2sZ+DSXcIpAKEVE0UFCC/hEA+F0haT8IfwBM1Gbwl3zv2Xb/vH8pbOC26bu5aScOuxDdM7Eize+SePt+fRPN6N3isTGBkhMDBLpM7N61xTyNobwa60ECmPYdxtwj1YRbDGun7qfWKJAd4KBjWVbuGzUTl6SK3GcFrl6zh4uGXOEjZqRZH2pIewQiYV0yCMiCCENYlwg83sVIS4w87qjNA8lE1C1nHWnIZYFCCgGfEaR/rAVr6In+7s46Vt6CAfS8F7qRzvdS2JyAK9JQla0yCYVQ6/EwGwZU4uOpCl9nPl6JJZ2cE6Oc8nkY4TeSsd4SoelRkPW0jYM75sw3NuLK2AimAZxvUT6X0WMzjiRpZ5hOnscblq2lWP+LAzNGr5tHUur3kpqqZMBjIg5QZK/1+LDiBgVuePWzXT+qYRAhkRgTJSsbSr9aUbycvtxDVnRuUX6B2zE9SqjCzpp/qqY9AMKcYNI3nVNDG0cQSAH/DnQvT6XaK0VrR80Ti25azy4azKI2CTyljcib3QQKYtiPmIke3OEETd0MFTvwOc1wR8SabrGyO3Xb6FxfSmPXPYlO72FZG3S4BmlEs2LUlncQnSTHXlKGDE7wrFdpQRGaEg4paE130LaqAG6BxwknNHiK1IYNBjwehJRNAKa4PBu4OCudNr2ZCNNHmCbPQ97jYRj2gA9nSlMn36Wz1xFfLR9Jo2XvEUk2cthXxZfLv0zb8YnIn+YRsQmEklWqL3zNQ6mGTn7+7EEYyaM+w3EOswEtjuIeY2sue51vmiaQsM1b/BsTRXKZSEyUjycG2+mZSAF8z4L8TozHtFISkUf7oAFKT+ApkOPohFpSTbz6Ogt7Kwt5/Fr/8rOdVMYPJVC8fgO9raVYu6UCE4KEYtpqDk8Es8ZB+EESH0gTAe5/HGoAlUaPglmb4rjrDBi7BMxDqpwwRAxWSLuNGDqkdgfysFxUsA9UqTqnqPoUiN8f6QM+vQY0oP4slQMzXoEBazt4Ck1IMagQ04g6xuJvxak87szVbxYU4mSKDNxUR3zis/y1SfTUSb4iJpUZleeoWtvNraNOqJWCf/IGN1f5JF0Qzu/nrueg+sm4C+JEcyPo3aaCJeH0XdoST8cZbDCjLZBj1DhQ3fGROQbO7FxERLmuBBMMrrDZgYiFtIPqEQtIsYF/bx2ZC720yIpJ8J03R7DbgtyZk8R3rIYOV9FmX/5Xt4/Mp1IuoyYGSZolhCtMVxzDYQ7LKivOBDiGuITAtg/M9N5oUpEFDk9kEmkz8yoV91IspF4SI/WJ3DpdXvZOb6cM8vf4O5Jh/lsqJBN3mxq3h+Hf06Amj1jyNgH+kWDcNyKfoWTgflm3C/nYprnpKkyRr93HPUJNvZd+gIzt/ycYKpI/q8OsKXlIK/Pr6J9QumP8n7/Z/RjpcxskyoR4Ef7cX7/U8rsH9WPukMkCEKuIAifCYKwQxCEU4Ig1AiCMOfv+pf/0LZXEIRvBUEo/F+M8aggCEcFQTggCMI6QRBS/12/VhCEF34Y57AgCGsEQTD/mPP4R1X46V3UL3fAlHJGfOel+OYa3DdWMunxFRz83at0zjWxILOCBZkVuMpVlrctwHSiA8OifjouUUj+Rs8V9+1g+eJtMG0sc352gAZvKpqwSiQlzivLVmPqEsi8tQXZGkeMCES3JxPMkUnfr7K5p4y8TTFaF77JutYKYgNGzF9bUfJCvNk7k9pH7EilPgQFaDPhPZuEolfIefYIjcu1XLt0Bxq3hqFylbGTm9lX8SnLEo+xoXgrjbPfBSBHY6H1sjewX9fJY0cXc5EpjHWPka4rh00AdSO9aLr1LF24m3C6jHrPACPvOsO+9ybiHbAgxEREScXyeQKpR1UKk524Tycxtvwccb1Iw11Z9FfFqcpuJbg1DYcxiPmwkbxnj/LJ/NeJJCvDOxFjQghAoDSKepmTh2Z+Q/Wrk9Dc14tsEOkfr+X0qVxaloj4ozoKPoCEeg1SGFqWCrRcJ+I/4yBqU7j/tvWsPjETod2IbBKIpMQ5UfkeO8s2oPFKCPVmXKMFcqZ1ovOrbOgZx8Tf1aBKoOnV0T1Dg6qPM+g3Y+4adhbP2RplSmU9X2ypJDw5gCYUJ3TDEN1vFOItUCiZdo64WSH99lZ01/ThnCoTzFLpnZtC+wIdvdMVjjTl4RsxXBs244Yamm7RcKomn6QpfSSVOumeacVxSuDjPy4ilCKw6vGrMR82MfKXp7GdEUk8YMCqDdM/UcOZXUVIokLKpD6eWvohUatA08FcEnVhVkzaiW9KiKTjApdmnUIKq8hGgeRTERStiq9QIXCplz9+fzFSqwFPvsjhU4X4RsW4JXkvx74ajWOUk/IXVrK6Ziajc3tYsO1+DE0GznvkIInNKnFrnNJ3VtDzcAGFv6olWBjFftqLtzJExr3NGPtVZhrA4FQpWHsntgwv2k8daASF1LUGNNo47jIZVRJIOwA9vXZSJvcRcRsIZclYuuN4fUYe23gtiS0KLz6zhMzn9xPKUPnL2kuxtAuE0hSslhCW3SYsnQIxq0pZfhdl69sQY5CV4yS10Ek0SWFgqoO0A7Dyti/xLvERPeBAHjQg2KMoOsjZMPyqNAxClbWRUx2ZaPwicmKcRFMIx3YD9hm9BAtiOCcohFJU/HkKgizgKpVwv52NqdrMnMlnKMjrZ6S5j3e2z2bZsq08U/EFhkYDu7aP5ZqrdzJYrsU9PkrmNpHPH3+Ogc9yeGTNrSSd8GNs00FIwlTiRm+IIkVVtn/wFvI8N6E0Ac3eROIG6K0SMG+20n40i9ChZIS4iqnIgz9DIjbDy4DTygUTT+GcHaF3mpGErWZcTQ6iKXFWTvue/glGZn34MNY2QBWYkttG/qgeUAUs+0zklPfQfK2G5K+bMOqjuEZLZOcOMj27heR1JgpHddO8NIlQskgoTcHeGGP3U5VoenUsbR1eEjwPZFKzsQxVgkRLmLrbV+HLkujrtREcodB2NoNjkz/FfE8nZY5erq/rZOlvN1Nyx2G2BgsoufMwhQ8doPmj8SzIrKD37H8NYv1P+u+rHy0gEgQhGdgBvKaq6lxgLNACjPmhfzHwNHCRqqrTgS+BbYIgGP5ujHuBG4GZqqpOA1qBL/7drZ4FxgNTgSmADVjzY83jn1HTtWYyXmlja/dxfvH4R1StOMKFWRPIfnI/ffdUce7JSpKPCjjPG8K+PsKVOccoWX6EmEng6+4xbOgcR+MNRja3jqa2JZPB8QJvL1rDNm85BVc3ErvFSGKthvxNEZg1hKVFQ0LtEIYHjSi/GqR0zzLcgxYqxrbgGqtiqjHR7U/EmhRAU52AooPEepBtMgWfx2l9YiK3TdrLWzXnkXRCReMTWF/0LVpBIkNj+f/Mr2T3Mjr3ZPPHievJ/+p2Mrb2/ABsFVAPJ6J3CRy6aRyGpBDpZi/HPy8jYgNHugdVr2DbaWBgvIB/qYe2DQWMO6+RBF2IpPvOkbk3Tt4XCkc+HYuvKE5jfwr5lzfTc8cElm5eiblwmB0VD0sMes2MerAZ75kkXjwxF08JdFZn0TNPJmduG5JfpOTOw/Q7E+ifpCe1JgRTPKTs0XLJ+OM8f9V7mPM8PLX/Ykru70QTELjopr20XLGah3unctF5i2m88TVks0o0PYYzYMK+r5PW7mR2vzoV2SFjHOmmYnoDxESC9TZ07mGLAOWXg/SFrGi9AjGPno7lMkNOC/4sEVu9wGDQjOOYxO9zN5BkDGJq1ZJSoxJKEYjrVczpAd6e8Q4Rh0o8rOG4MwtbjQ7FpPCbok0MtNs5+eAqJt5xHGeFSig3xsDlIbxjonx3dAxxnUAwQ+W7veNIaFWRLSqa+R1oXkri4W+vQ9VA8jGVd3L2sK1vNMbTRgamx3gkqRFPVRhPWYzWm1Syx/Si6FTSE31MGteEqoHYOD/Jh4cLtZ/vWoAqgCioaAMquVmDNFTnYa3T8odl77KlbRRJt7QxtayZwqo2Up5tY1fNaO6ZtoO6e004bAEanSkMTlRZ2jqH0Fw/uiGJyBEHcx/aT6/PylCJhKoKJB+USLy0G61fIeGoHu+2dFL2a0BUEeIqRlOU9PI+Bq8M4i4FYVIZZVVNRO0KvikhFINK5LADXz7I57uZN/8YXR/lc2KanrADulqT8YX0lK7xEbEJ9MxR+OyBCwm3WAmnKKQWOlFdOuQxfoxbjjI4VkIKqTzx3vUk7jWgAq2L36DPmUgwXcDzfTqSSaZibAsVMxvI3Rwjc7fKwqsOcOCPr/PVQ39kd0sRExwdfLD2fMZMOMcXT87n90/fSM6cNsSYwCdfziJqUxECGqI3DTH/84cIOwS0fnCVWdB5AElFI8VRTiUyVK5Q/P4Kgq0J2BoV/NkKkeQ4V80+QMT2gw3AOC+eEpVQnY1QGoz4k4TUYWD/2vFML25C0cDYO06h2qPkFfax6sAcBHXY2d09SkVnD3No9yiCb2eiqzWCAK6vsigf1U7V9g7i3ycx/5LDdHQk8X1LMf2TBJIMAfLXedH6VfLGduMu1CIbBQwDAo1vjmTccyuRPCGCxVHS93gwr0qk8sG7uPi2PWRtllAdUbJ2qhTvvJktI79mTfY+ntxwNV/3lTN4RyW/3bMYgO6Hqkje/NMpsp/04+jH3CH6BXBQVdWdAOpwtfaDwFc/9D8OfKCqau8Pf68GkoHrAQRBEIFHgVWqqvp/uOY5oEoQhPN/uMYO3A28oKqq/MM9ngOuEwSh6Eecy39Yu/ePYWHdRTRF0tjaPOpv7TfftRlE8C320fxcJT3BBB52NAMQTRDo6bNhWdhC8ftBzJsSSDipQ1vgY5O7gm/aRtHkSqb55gzefeDP+Ebo0W6yoQpw7opkBibb6XUnkJM8RFqGm3Bci+qIoh9ScfrMmP+aSDBd5ZIr9jP33mpyN0LXbB0jtkf5umsM+g4d8etcqEXB/+28Zt15B0mJAeKlAV54aCnp2S6+3ruBhHoJbyEsuXYn/pIY3bNsaA5ZOXyykOg0H1GbQnh/Mo/M/BqdT0XrF4ictOEdHePYkSJO9Wdydl8Bg7cEaFuqEE5SkRwR8p6McaI2F3GeE/spEa/TjNBjQNunJeLTU/dEKfHMCIoiIhQEiNkUJpaeI/p0OuOnN9DzQBVip4FQRZCmZRrUI4kMVMbZ3lbC008s4+SUT0g4raP5lQyiiQqf7azizs5K6s/TUPtQOhXPrMR+Zhi0m/B6AvnrB9C1GjA64xi6tMRkiURtGDQqilZFlMFXEqPc3k3PrhFopg4x+qkuVFVgcvE5Uk7EcI9S4eNkUq5t5+Kt93KqaQS6aS5iN7qQRwZBo6LblsC9p65BFUFjkPlt8Ze4x8oIhjhbPeUYejWU7LqJM64M9C4RISxiNEbR9WjRJEbJ/N6FY1I/ikFhYOqwx03v/ZX0VGrI+0JmyhUnuerxbYzceyPn+pIou6SOUY+0UPnQXRTdeIzE01oSDxrwhAz8ct4mBv1mTvVkoinyoT9iYej8MHNymqg5U0D64SiRzan4c8Cqi2DqEZAN8OibN+N4y4I3YkBWROq70ji8ZyT2nCF2OksQtXFcTQ7SrT6al7zO2U9HoSgCsUQFc6fKkZ+NR1EFEmb2IQc1iHFor0vDNVpDzDyMIZCvcNF68Rp8t3rRbk+k93Qq8biIefQQb6x/nZMdI1A1KqpbhyXLizTRjblDINCewO51E0i5tp3618oxuKC0tIuJmR20XpGIvyKMaI7hv9tD9ndxRr7UjfVZK6oljuMrE/VvVGCscDFr5UHsjQreGWHiGRGaY36a5rxD1KaiTvEwJe8c4Vst+O9IomuWDkUDjf5Uxj23kkufeZjUDQa++WslkaIwA6/n4VjRhvHaXuK/TaVsbgOaoEDu5jBzpp4muj2Z0ZPPEUlSEGMq+bc2YLmwl3kVZxHXJ5E4pR+dSxw2rhwUCSWLFI3rJK1okLW7puEfF2bKhEaiES0wDM8VVBDiCvkbA4gx2L93DMGSCCfWlCP16YnFJRKSA0iRYcyPFBTQauO8ueQ1YkYB87RBPJMixKxwqi6bt2rOwzs6xhX2GuypPlRFZMGcoxxqzqN3RiJD88KcO52Jr0BBNgr4CmWcs6IkLuxhyZe7kfRxtnz9ER3zJQQFjiwfR9f5KqJG4cnn15C43ciYl4d5kiMmdtO8P5fkN6r/Bt9ds/JlBibyX5Na/8/qJ6fqf5l+zIDoSmD33zeoqtququq5HwKZicCRv+uLAceB+T80jQXS/t01fUD7310zC9D+/TXAMSAOzPsR5/JPqfF4No8kNZLxrh5NRjrS6BJe3HMBRavayL9nkGnn1XLuZCZ3dlbS/FwlV924k5JXYjiXVzL3rQMMzo4ScUDIY+DQU5MRd9ixfJKIrUHlyi/uo292HG1wGHegSuAap8AZK63Hs4h8nYpNF0JrkCm74zRyi4X+KSCFYUvbKLatqcKTp0VQBFz3BejtsmPsEwiEdew471VORsNc1rgAgNH7b6D4wxXMXn47JY+doXrcOtZWrqbyiUM4T6Ry/tlL8YyLYm2FDatng0bB4FKRIuDIdvN/sffe0XKU17bvr0LntHPOUdrKOSIkUAAJCRCIaEQOEiYbMAYbY3xMsI0MNggQIgsDQggEEggjCeWcs3bSzrH37pyrvvdHy+f5+r73rs+7jGOfc5n/VHd9q3p09+jx9aq15prTutHOLTM3kVKv8/Lp8wlmy4yadYKEVeA6akC3aQSPp5K7Q8O61olrj5nyD/swHbbSMjsNKSahf5tO3wiNjGwf981eh1IZoGCtQtZeMDWYeX/CG9g32BCK4Gh7Hn01JvY3FhEcEcbZAHpCpnHOMi5ZsIP8kl4SCYVwlszIpxfhq4nzi+Fr0Q1J24M9746g6+YRWFsUghNCyT8DZ5TWaSo7lo/E1AcdExUiOQlkWXBP9gZs9QaKhnTgHqHjPGngjC+LrANxwqdSMKxIIEkC70/y6b0ziLVNpn9g0jPO2K1ibjESDJnwHkvHdNgKAvKuOcuMwtNIQEFmPxdaNB47/0tUU4JPd49GDSfNSe8o2YKQofyjKL4+G3JCQj1txTcgha7aDC4ddwA5IqFXhvANjRJP1bn15c/Y3lxKWzSFeJuNfVOW4rkvj/jHNnrmRDnz6liC+QJvtU4wZGLpy5eR+7SMYZeDWEwlY1YbcrOZDV+MwtCnID3SjbdaI39zgp7XSkg7FSOSoxGsjNE+WaE/aOFQSwFCk3A2QuBIOnXflJGzxohIjfFe1UeUrbwL36gojwz9hooPQ+jz3dQvknlo4Lcob2RQUtyDyadh7lYIDIgRydZRIoL+dhfDnl+Mvy6FUI5g5vmHcH1nYWR2Kxd++DBGU5xbJ21h8NAm0l6zE25w4h8XpuLDMMGSBL0rilB7DMQccFP+Do6+N5i4Q2A/ZKbsddC+TKd9kkrtXfl0jrFgbDcgXd+DFFZ4oPpbNr86DltrmNxVRuyHzZQbkhVVQ7kfVdEptPbTMTOX8z86SPnSRgy3d9H2QSma6VxCcks3weoo6BJdF8U4eaCYwGc5xB/vJxA34WjSiT7hoeWBcvLXdtHyURl3XvQXfFPDHGgqIs0SYsu3QxEKdLWkIkcl3E2pMMqLtVunyZ1K79EsjH0yGRtN1PZl4NxmJm+rwFUHKeO6qL3eRvt5dnyDYqSeBEdqCNOCLuxnJaTXMonGVBDJPSZrZBfPDV1FXCi88sRLDM1oJzenn5pZZyhZDcY2I9n5/dzy1e30t7rITvOxbscIRFQhZ7sXvd/Iry5eyZJL3sVfBJZ2lYplGj8pW8+v1s+n6sE2Ljo1h+KvElh64tReb+eq8XuQmyzcsvVm+obpFDyzgxkn53K2MYsXr36T5c3b2B+Nsb79ENd/dvc/b8P/Af/t8L0kROc4PGWAIknSCkmStp/jCF11LqT03LHj7y7tPHcdf3P8X8WIc+eAf0+s3H8T809F+Yd30TJDwb3cDvEEeRskGm4vwTuxmL4b0xg1tpazY8OUP7yTTxuHIfYeJf2Nnbz16QxERKH4yR0UfyLRdmmCcJYgcK2XhEVi1Nharhy1j8g1HqydAnW4B2EUZO1PIBTwl+rsqi+lOqebPWuGkL85gbPUQ6IkgvyXVDxDEzibEiDAYY4i+1V+9+BruGxhbq69lsvW3keB1UPpl7cT7rEiFwe56YXP2bp+KENeWMyl6+9h7UcTyd6tE3s5F0kVeMbH8A7UaJy1nJq7j+EdHOf3NZ8gCUhTAyTMEomDKQgV9rcVMnvqfgLjwqBLJGyClUuX0Dc0OSrdeGU6seFB7K06NUObCU0IYq9XSbcGef/Z2SQa7bRdCMaATqwizLXf3sX0RTtJLfBi32jDNrsTISSUNjPeakHeGgNln9zJsdk5tDWnI52y89zdy7Fc2kXGTpVfrF2AISChZcQ476a9/PSBDwhn6/x0xNek3daMrkuY3RKeiVFK59fjrHGDQWdB+UFufP5BXrhtGT8u3kT5JzFSz8S5KX8HoQwVXRWc7MymJq+T7pF2oqddHH3gFVJPCI6dzUNOSFxx2Va0LguDJ9aRv8lP2iGZ4DP5rN4xhvTDgt6/5PO6N487XO1I9Vbyy3ox9QkkHd667zKGzDhN0491JFVHHoE4JxwAACAASURBVOrF3iww39FOw5WvcfjxEaQNcDO36ihqlxFzl8Kbiy5jaF47W/40DtdpCZdsofZ6O009qVT9yk/ppzpqSCLlpETuChNzbttK24UuDFN7ceyw0rUlH1EcxtYmsHRJuIxhcrdKdI434B4sYfDFMHcrNM5+g/KxzUTPOjAftqJ0mAhnSpi7JaIDw2x/8TVM1jhZio2irzUqX43zzNrLaH1YI7o9A2O9hY+umY6nQiHX6qNtigzDfWRvTPqpBfPA4IoSKNbR7BqXzN7NntdH4Bko2LpxCDk7dcJ9Fvb0l9ATshFNVdDNgsKsfprmWDH0K4RyJEyVPkbNO8Zzf7yW8NQkx87RotF8kZlQrsSUmUeIpycwnt9Lwfg23F4b5naFJ7deTiRN4uw90HZpnD/9OKnaflXDhcj7nAzM7GLN5xOJuWD58QnEqnJpP5BLdJaPSIZO3CoRX5FN6h4jqTuMGMwJ9JQE/aPi9PhtdK8somtWnJazGdRda+bkQ2lkHAmx8nczUWqtmI9YaFqd3ObidglrRoi0UxqWVoW8FB/t03QGZHdT+lmI9Imd9A0RxDZn4J0YoXWmwHVDK/7N2bjOyKiT+3h04lcYgoLIiRRKnX14BifoGSEjHXdgu6KTcI7GtqGfMsca4UKLxliTgbiQybd7ObKtkpJfnsLokehuSCflmIwkJPS3shD2BFJUxl9mRyiCXx2aw/3rFiJrEuG8BIECM788OZeRI+v48/7POdOajZQQ6I/1ItJjdESclD62k4aZy7GfTU7nzck5SuMly1hy3VVkKiau3XX7fw1fsv8ovmdj1x/MXf9j+L4qRH+dc/w18HshxCTgceAdSZKuA/5Keo7+3XVRwHru8T8aExf/s3jS38b8D5Ak6Q5JkvZJkrRPCwb/0c/zv42eUxms2/wp9pW7KV3ppr9aYd3mT/m4bENSt0hWSH3dTud9E7FszqZ0VR/mdpXal8dh+EknQpOoWNZK9HgKP7r/K06vrGb3r8bgP5OKboTqjG4w6jRfppO3Jfl1iJBK45dlGD3QNdqAcVUqzp0WIhlgb1DpvCGCbhC0dqXiaJR5un4ufQeycK8swNAns/elkdhrDRjTIkwsbuRX6+cTzdQIDomQs1nBEICuMTLq4k7UdiNXDD2AMOvMuOomvjs0kJRsP/f/4S68UyI8t302kfke5AQ8esdHaJpE/YI8sr4w4cjx42iUGb/6QVKPSVx40y5mzduDZY+N7ulxGteXMra4ieioAGdOFPDer3/HuMknydopEUlRyP7ChKXJwJcfTyS+NZ1AIcQ/zEbpNGHwSuTsFHBbD3JGlJZry7A0G1CHeHng3VuxGuJIV/Ry5bRdCAkyMv30Ru2MMbVR/FWCW12dNK8vIeY1cezeV0hPC9DkScV/NJ2q16IYJA1zv85LrdN5eMcC6q410HyVzmPfXUkoV8JV3Ue828LRljxCuYKaCQ0M2HYD3TNjyKogfUIna5edhxKROFhXTN11NqIpEu23xMjdKuEtk5l21V6+6a1hzBOLiKfotHemknFdMw/MXkv3SCPHvq4m/XMri0ZuJtRjQ1nQQ/OBfF72FNJ+noq2JgO7GgUJDAHoHG/iRHcO7qlRJB1qli5GGAQOW4RYrpOJz+9GyOAZHyN0l4eHM/YQytOJb8qAGX2E8xIUZPbjnhxDM8Oho2XEb3SjhpJVj+ZZdpQwzD49m9On81ELgxTMaqLspzuJVodJqU8kNYe+vP3f7USa5kPrBXb0zBihXitChrhNIC3xEKqJ0BFyYi72MzKvFfcQiWh2goIJbSTcZtSQBIpg13NjGXDzSYweGUmHjokyw6qbOVpfQE+fk65xIIdkIu/kIGQ4c+NSTtz9CqGAiQZvBpNu3E/apzamTT7Kz595m2su2YJtZC+tdxYzsKKNyI4M1BktCF1G1sDaYCA2PIhosmI7bubWTxZR9e4i9u+pJP1YggZPOgzyM23+fiqeDBHONFI2thl1kwvdLJJk5vE6cYdEaEaAleNeJ32ngZyNKvphFzPu2Mk1w/ay+eIl2It85Be7cT3TSjRFIuWMjr1Vx1+pYeqT8A1IoB1zMe5ne0k/mSCaUJGsCQ6fKKbhcitOUwR7k0ygKo7luIXism7cnxQw7+ptaCYwfJrKyvZR9FfLlI1v5ujHNWDSMXolbG2C7l25lH6WoHzDzbzry/j3fe2WrG1MTz+JoxG2NpSTszeMo8BH1lXNGDNDSYL0ISOOWoX4TX3MH7sPrdWKMOpcfuk2DB6FxEI3B0Z/ROgWJ+cteQjXbjPti2IYZzSxa9ofOdaTS94uB7PyhjPmqiMAfDOpmNE/X8T6z97DJBnQOiz/86b73wU/tMz+afi+EiLt3PFLIcQBACHEHpKE6AeBv2Yif89+MwF/JbD8ozEGSZL+Xvzqb2P+BwghXhdCjBZCjFZs/7nDaH+9g9GOn0YN8u8TZ8+5K2n8t7EIRWLDT35L+Pwu2i9MgyF+SlYnuLVgG2ZnlIbfpZCwCJauvpjDj75Cy+xkW2rb4y/S8VI5Va9GKSxw0zFRwuiVuWnCNiIZAv+kMNGspHmnpAs0kyBYoJGb6iPtuGDgI21EU6GpKRN0EBf34zgL4ppeAlVxyh/3s3X7ICQNbE0qVkcUf5GMGhLE0xNEEirF41r57NvxZG5TaX8gzj2Tv8Xb7CJrX5DC91SqyjqwfuTi9uvX8eT+ucQ9ZpquzkdIoK5PYdEdnyOlxQhnSnzz5/F8tW4M6gW9iIhCwipICBnXNzYMmWFmf/IQda8MoG+QhL9IomOqzuCLTyNrIE3qJ/WkoHeUjm4U6CP99FcptNdlkpfupeaKU8RSdYzfOImWR/C/lU9Ck1m3YiJqSMK5xEFX2MHVx24mnKFS+e4iElZBwVcy5Rtu5qaynXg6HeTu0HC90M7KxhH0DpOo/66UtG1Gbp24hYrXE1w/bleShFqXRspxGT1gIGePRuinOTi+toMEJnMMt9+GISjIOCiQDRoFG3QiWQL7Bhv+AhlltIdKSzcHjpTjLwHsCYr/LNH/VhEX2U4ia2DpFniqZF7/ciYl5V0YX0tjyPg6Vt07k9M3L2X/k0tZVT+cvG0aL/z4NYzj+nBYIgwo6iScKREpj6KkRXG96KBnuJkVe8cjFMGAh5vp7XEwfO29LJy2Bf+AOGZjnNJVGu6glUGl7YRqImTtlPDty0wmRFGJybMPE3dA7Z5inKdUjLsdnDlaSGzWaIQuEcxRyP3ARH6Rm6rNNzJn5CwsTQYGXnyGlNQglhYDjuak11XrFyVkfmsi2+on1GWj7uUBJBwat07YyoaaNRg8MtbB/dhPG+mvltl+qgJzb1KMse76pRxpykc2alQ+H6FgY9LG4fmnXyXzkGDUU4u48MQ8Zg88jvZOFuuOD8ZfILNpy1DW9I9gw9OTCe/OoP4aF4296aSc30n40jHcP3wDcZsg7ZQGZ604G5KtaM2q46yF5y/5AE+5ilHRSDTY2bF8JCcfTiGYo1B7pBDL7C4s7Qol6X00XP4achRkWefmZx5AzHPTPRqiZVE+2TyeE75czv/L/Syu2kL/5hwO7q/ANzBOaIEXX4nMexcvxTGjExSBpQvW1g+ifbJM+7Hs5HfZpuJokqjdXYykAapAG+VneHorcgwCmgl/ZYLMb5sYktpOtCJCS38KMRfIPpUjD72CtwriDp2W2xPoCZmIMFL2yZ2868vglu038cyWOfSNTqC7TShPdmP+NIX6PUWI03Y6L0zgrUkQLNTJsfvZ2FpF9ZIWZHucHT1lHFi4hJ6WVKbefjunF2URytPxjIxh2OUgsGAc49bfx4HRH7Fv1RDWtx9igL2D+t9OoPG+Qdi6NKreXfTfszL0A/4l8H0lRD0kqzStf3e+iWS7rPHc85y/W88B6s89bvgHYySSXCMAJElSgfS/ifmXwl/Jfvnf9DDqoM769kOs+v10zty4FNuZPm4cewVqQT6+ygRFC44CcI2jH/2MnUifGUNAouyZI5SuvZ2cwr6k/5EWo/PSGLWLjXTvysXRIKNEYOt949HzImheA2WfxMm5pZGYS0INS6SX9dOzKY/OCxN0zivDEABjd1LV2roiBffoZE47qLKVxh/lUvmulwG/b8boFYQCJkI1EaKpEo1zl7Gk+mN6VxeiGwU9U+LcWr2TP+6+ACUoE8o303a+SvPWIgD+9NVFpH9twZgaQUgQyJfxTozw5nPzKMlxIycgf7Ofkp/vJNMWpPBryNmtUftONUa/jqbJOBpk+mok4jkxtIFBSj4XBG52oSsQaHYSKJApXC8o/EZD2ecgVB5Djkj0bstlb2MxqcclAsVw8MKXsXYnGJzZQWBgjJGzT9A+yUTg3Xzi6zIRP+olkaLhqk1WwpDgtTOTsTQbaL5M5/DmKlKWOij7NIAcBVunxscNIyj9wxmuTdmDHIchIxvJubqJuaMP0jtERQnFEDLcO2ojIa+FR4esZ89vlhLMkbHao4TTFfK/S5CztgmjX5BIKLz8yRwMfTImt4Rs0OkeacR5Uys3PvgQObNaCOZLRHITlHwZpuVQHk/8/i26XyzDc29yFqHy/UVcV7GP7hEqi/b8CE+Xg94jWbjfLKb4y37MdSYMx600zzISyRRk7FSRExInnykmfYeRlCMGNnVVMW5QPfqfs2i8QiGxL5VmTwqyKvBeGiSWpjHk+mOY+uHb4wNxjOoFAeknY6hBgTDptE01IBt0xt95gK4xCvrbWcQDRkR6CnIcDu6vIONZM86zOv3VEjmT24imC0Lzvew+WsGA1/z0D5Rw1Ku8e3wcw59djKhI3jdpRrCN68V51Mikm/ZTf+FbAFhOmdEDBk4tctB5TRShCha/vpiekRLeSkFDXQ4bmyoRMhjajKhhSD8CQ20t9A6VGXfJUeKpiWQl680s3DUqLxyYjqseOiZLnFz4MrMWbcdbk8DUo7Dv6aU8ufxHIEP402wSLg0lCo4TRjwjYkhxCZOiYRzXR9NfSij96jYOP/oK0bMO+idGcf7JyYTxp0jfnORXHT5VxDvT3mDJ6nnICcjcB8hwZOyfAVj42WK8m3IoXg2e4XEyPrJi7pHRzYKvR7wJw33IcYFm1zn4xCuo3QYiPRa+2DyamEvi2M+GIcUlvMvNrDk2FFODGV2XiZZH0E06U2+7nclTjqE7NJSTNmyuCO83j0OOyjiVCGmbTSCDYksgrBodnxfTc0GM3F0amhnMziiKM07B0E4a1pXh6XQQGpwHbhPesJnpj95P4dfQM9RA1l4oXhcnY7uBIw+9gr9Q4a5xm5my6A5+f2eSNP1wWj2layKMmnWCvoEqmvH/gJLHDxWifxq+N+sOSZI2AC1CiJv+5tybwAghxAhJkvYCm4UQPzm3ZiCZSD0khFh+bsqsHfi1EOJP52KygC5guhBiwzlydidwhRDiy3MxY4A9QKUQou7/6z1+H9Yd/ztIPywRzJPYtvh3pCpJjaIPW3bweOcFZBt9rGocRk1mFy5DmE3fDqfivV6aLs8knKth6lVgsJ94s42qX5+mb0U6xtfTMPg0omkq7VPB0p60dYgND1Ka5YYLWwleMY7ZT37HG1ungiOOqd6MrVUQyZTIOBJHiepE0lT8RQr6JC8hn5mMTUZ6R+lYOhS0EX5KMvo4u7WYZ697l1+8uhB/dRxjjwoVQbSEQkaqH/eRTIQh+Vsy9cqk1GlIOtgbAny9dgU3N5/Hti2DMVX4KHgaztxiR/UnvaRcDQl6h6hoFkHFlLNJO5CokeBZFyItRvp3JtzDdSydCmoIlAvcKKvTECq4J8Q5OvNPjP/jgwgZQmVxFFscLargOGriybve55E9V6C0mBk6uZae58rov82P6x0nnoV+gj4zFnuU4idiNP7KjFZnJ+HQKa7uRJV1On0OzIYEk3Ib2LZ0DOr8HqT3MgjmyoSzBMuvWspN225B6jXiqk3yXEomtqA9nYX9qTZCj+XQNNtK9j6NjAcbafikEm9NAgw6BV8qmHuiTH55DyvrRxCNGJAkgdRg5YPrXuTK7xaR95UBhCD17masaozD31UhZLj84p0cuHcEAJ0TLAQq4liaDVg7BLoRLG5BX41MxmGNjgVRrNYooZAJzW8AVUBCwtitYm8FMaePcNSILOtsG7uMG+qv4OzXpVi6BVc+8C0AHy6dQeahIPWLZVSDRpozSGBTNg/f8jFP7Z2LetbM5BlHOfDuUCbcfICvDg0hr9DN6wNWcPWrD2Gc6Ka/0wmAwa0Sz4xjbjUSt+s4q/oJhEzIskA+7MDaLbC3JfDd5WNUdiv7uwoACB1Kw9wrEbdDypRO2jtSMdtjqLsd2C/swu21kfNnM12jFW64dBNfPjeVSLpMoFCQejI5Wr7346HEJ/iJxxVy0nwkdBnp/QwG3Hucem8Gbd0pDCrq4NihEkiJ4dpjRp/Rj+HLFPoHCbIG9BCImHCtcBBa6MFliTA39yjvLb2I6utOsa+pCOt+K7oBhs87wfH3azDM6aHI2c/+U6U4TxgQEkQyBQmnTkqhh/C+dLIOJOi9OYhrpYOOGQkkVae8oIe/DPyC0b9YRKBQ4tTtr9CRCDB5yz1IXSaUghCpjhBdHSmUFPXQuS0fZZiXmqxOfA/k0TLDQXhABEOTCdMQD/quVIJlcQqLe1FknbZ9eaQdE/TNTRbXtU4r1a+5qb8uA+vQfvSNaQRGh0nZaiY804+600n27hB1Cw3MGXGEdScGUZTbR0e/E+WIHZMHUk9Hab3QiJCT3mg5OwXtFwqqq9oQF7Rx5vUxXDtmN6vWTSJ/c5ym2Sq6TaNojYTRE6fxLtACKga3SunPdnL23yZw7KY/MXLJPYTy9P/U/fo/iu/DusOSVyhKbvt+rTtOPf2Ddcc/iu9zyuw54FJJkkohKdIIXA68dG7918ANkiT9tbpzO0ky9AoAIYROUqdo8d8ILf4E2EFS3wghRD/wMvCAJEnqudbZT4A//6+SoX8FuIcJ7G2CwzE7cyZdyplXxnL9kIvZ+NUIXGqI3MtOUmbtZdtnIzhz01K0k7UU/GYHuVslzD3AcQe2Zhn33AF016XTOVbh7CUGOidIVD10COdZPek/tsXK2d40ziwfTecVUZZtPx9sCUaWNmNrF7jHJlBD0HSloGOiCd0gYe4VZLxhZeDjHViv78BRr6CP9OO0RfC/WohQBc8/8SOUKX3YzyQ3desWO0IHz75M5s/aiRSXSBvgxjyxl47zwHeDj87zXFw093o2n6lk6vlHyHd5OX130sMr4dIJ5Qnapyhcec1mElbB6T0lpJrDDM3qQHcmoN9IJE1CDchE03UimQKnOYp7hE4wV8JkjzL9p/cTztax9Agem7wWpz1M9gYDQ688wWP7L6dwhYpeFMEdsdFykUTa63Z6rg1hMcaR+oyE2+2cfDAFISBnl8a8Cfvp6HfS93EBjg+dJL7K4JQ3m1CuRHRtFj0jJfw1MWxtEjd9dSeVf4pja5NJPxHhkStXc0fhFmIpKqc3lNP44yQ5e+avtnDoaBnmPh17gwq6RNuFUH+1iU/fmko8ruDaYEE9ZsPSJXHl+h9jajHSMQk8FQpFtn6Or63G6JFI5EdZuX0cbffFqb/ayO/vWAYGnXB5FOlyN75yQf9VATIndNB+noQel4mcSCH/AyMXDDuJpOqoXpVl1y8l5pSI7Ugn0mtBHHAx4a2H6FxRQrA6ypS7d7Pyxem8/84MHG0JSv9wBmOtBdMeO109Lo7e/woLnb1UF3Rh9Evs+XQoA68/yRF3HsZOld692dy96F7kGIzKbkXtU0k9oGLplDC7ohh8YPDLBA+nobdZiXVayd4fR47Dz156G3VVGqeeHYy3LpUDoz9i9cLfE5viQ46B/Rc2BpZ0YNzuIJohsP4uBS2uYG0L8fiClXyw8gLCWTLBgiQZ3TszyLa/DEHIUJzeR8omC12HsunudvHYU+9y4KMhXJR7AkOjmWNHiql6LwBeAwcffwW+SyV8sQ/druHbkk2wxUHbnATKZ2m4v87nywcvwD8pxP7t1WhhlWCBTjRNsP1UBaFcwfjssxxuKcCZGcA+qxPdCPFUDaEIhEhW+wyB5ASje7CEMz3IolGbqT+Rx+0tk5DjSXLssN8uZtaLjyDJyRsPLaEQihkYWNaO++t88rZF0XWJ/p8WUne/StopjYkVDcQyNaYW1MFYL1JYobUrlcAHeTx62WrS9vYwurCFeJ8ZkRbj5L2pxFN1ZDnZys1aZyJ7ex+6LhEo02idbgVZ8Kf83ZxfXUtTWzpOWwSjDwyze4i5VOSYhKsu6fvYPl0n/1uJMy3ZXHzcQ9o+la+WT0aOSzQukKh8x0/V7Xu58tn1qHtPMrv6GMWfwejzT9H5wEScw9wYJOVfPhn6Af898L0lREKIb4C7gVWSJG0DPgIeFkK8dW79c5JE66/Orc8HZgkhIn/zGi+RTJC2SZK0C6gALv87EvVPgSPAbmAv4CeZXP2XQO9wwa2f30HNqmY2znmBxmVFFP9iJ6t+PgtlUx6ffTKZUEGCqrcXsb79EOvbD9ExWRCcEsDWJlCn9+IeKkAGg19iwIttGIqC1D81EktvAt2iE8oFyzY7kqqTstGCuUPFdNZE4N5s3KMTqB6VYJ4gY5uB2MAQ7iESvRdEiaQpeMcXElueQzhHcMvAHUQ2ZtI5AYadf4aOaTpiQxrlc+tRwxJX3LkRPaoQK4ny8e6xWAZ40D7PQF+fgaRLjMppZcrCvbRNczG0uI1Drw6lcW8hZSXdpNQnMPYq6IUR7Gcl3js8DikhIUehY2UJ75d8h6nFCIpg2jV7eemqNxk8upGKpc30r8vDVewlUhol1mXl9sc/QxLgHq3x0tuXIUmC7jHQuKSaX41cQ9MCgdxipvloLs/O/JD4vW5iYQNzCo5jbZO5fOJecjYpGAwa/kKVO9O3IMuCYD4Ec2S84yKEXswnMShAoFhwzYxtlK/QESqofpmBS0+iBgV1CxWe//RyHv3iOnyFKpbRbgB8Q6O8s24ahoww8x7ZhHlKL2gSltwAwqbhq4mzctzrqBGBIQByXJByTMXaAaNH1ZJ2SmPHeyNJWAWGKW5MjWaEKpD2O1HTIzx0dAFEFKqLO5OCfWZBuM9C79bcpK5NnpuSL0I0zxMc6skDIZF6An516y0svHE9weIEk4adQaiQO6Gdq+/7BuchE5uWjcMzMJlQDH7iCMefH0reea0YAgK56/+m+k3LPE2gJEFsVICdJ8tpa01DEiDHk5NoQoajfxwCgGaWyJrXgm29HfOFPURKosTy4hRs0FDCMhVPnaB3tM5DRxfQMzlB62wdo1em5uXFXPbeQ7g+tRMcHKXtpzq1e4p56u53sTVLNN2abPm2T3Ey3tJEzKUTTYHqsWepvWEpotmGZobAwBjKQggUSBgq/IiYzM+W30T1FafpjDmxt4CtWcFfasNaEKB660IiEwKEPBZsdQaeueVtjsx/EdWk4amC7NktqP44UqsFg19CCimUrolR9HUMKaSQdUDnSF8+sqIzKb+R9rMZyDFwnUgKTIaOpiJHZM7OMWH5zEXCLgjWprDsi5lk7pX5bvNQQrkSWnWQQJFOaFQIwykrBd9qXFR1glBtCqeOFzLg8tM0XCclf7cFZlRDgjUvLqHemw46rN06ilCrndzqbpz7ktN0v/twPqfvyuTQ1wMxZ4axHjdj7FVQgjKBI+lcctFu5j/+F+qvSSPqNWPL8yNq/ORsUHm4cwRbaiuw1poIRox4R8TorUvHfU0IvTpAf42g9kdLUQIK3lKFqpeivLVsNmknwngHathH9zKupp6v167Au66CFc9cjB6JUD83ne+WL6PrF2XkLNmBf3/G/3mcoR9aZv80fK/WHUKID4QQI4UQk4UQ44UQb/zd+ht/sz79/6mqI4T4NyHEiHPXzxdCdP/dekwI8YAQYtQ5svRtQoj/vPGx7wmffjeOjaEKihYcpfWxiRg9CZq+KaH0gzbev+hVEvlR5kycR/mGm1EiMmemvEv6sp3ENmdQsFEje7vEmruep/bOAiIeM2p5AG+JgfS9CrtveQHfAI17R29ET2qyEclKUHedE9cxA/YmiVXXLEG6ohdZSv7hTaxswFsmY1u1G3+RjLMBlm6ejr8mhu5IcGnmIQYNaMHcp9P0cTnScC+rXr2ArC0GJLeR+WP2EQ6ZSNgkAkUCV1k/jb506q/KZ/yCwxxpyqfvwgjxFI3JmfWE0xVsI3sZWdyMd4BGZX43Wk4MeytY3DqT7rsTSxdYcgNs7yjl7p3XcbQ5j7q7igiODRGNqwwra0Uogl9vmwu6ROYuhWBJgtDeDIRR59+eXcYbd17O4rGbsA3sZ8Dvm3l7wiiiq7Ipyu3jg1OjkRNwan4+tvYYNlMM75gIszfdg2G7k2h+HNPMHopy+mi5TMNlj/DTuavZ8Nwkwo96mHvjVi6avo9KSxf9Y+LYUsOUv9VJ2lGJmAv8p9JI+A0QVnA2QqLDSrbBy9MDPsfSqpJqC5O7XkUyaVy5804SVgl/hUY4WyIwMYQyx83B7VX0DlWwdulQFcTTlPRMs7SqqGHITPWj6xIF30j0v11EV30GSloUa6OBSLaGEpHoDdiou86M6lHpa0lB8hhwDxNc8OJ2Vv5uJpItQccT5cgx6N6Uz9INM8jZ6Sdhk7CVewkMj3DgxeFc+stvOduRTv/EKE/N+5jf9pVT+uXtLP1uOo56ldxUH0q/gZQDRuSoxIx5e9GNgmiaoG9OGMsAD9YunS6/HUufjn9vJvbjJs6rOUMgX0Wz6mz/YhiSK8aw7HYm1NSh9qtEyyPY2wSaReCdH8C1z0SkwcGsCw7wh8bp2Lo09D4Tos+EboTZnzyElprg1O2vcHp3CRUrFmH0SBj8EsWfSpx4OododoJY1IC9zoDRI5iYWs+29jKqbjpF8ZxGfKUK8RNOzDvsOG0RUg4ayX9uBy80zmTYRK04hgAAIABJREFUx/fzo0F7MAQkHIYIZ+8FNSSRdlKjdHWCULaRvgEmLhp3mFCWTFNTJgNzujj4wnBSjqlYegRqONm2lGMShd9qFI9oI+vms9gb5eRwgAr2hW0IRWDqF1Q+7gVAabDgatBpulSiIZDO8PG1ZO2U2HusHGOnAdtqJ12XRFEPOLj85HXk2nyYM8McWLAESZNob8gglCdQxvaDSFboLpu/jaLfSwTL4uRtT2DwSdib4bMTw1i6YQZaWQTncQPxoy5i3Va650T57C/jyVljRB/hJ/sNC7bTRqS0KFqdHfmMDUmHjwMu0ga4CRZrOF/sBAlcz7Qyelgdbred/kl9hPQYaTf0Iceh7dGJrN3/NaXrbsO0t5b17YeIO36oDP2A/zz84Hb/T8RvvrqM9e2HsE7qpf/+AMfveYWz1+Tz1A23YHeFKfi4h4obDlL2yE7GPbqIlp9P5I+LXqVvoIFQjsz1P/sJlAUZ+Lwb/YwdcUkf4jI35+27BQS8uXw2sRk+5DhYOlT0zBhxO3gHatx2/Ab01RlU/szD9oW/Y/uhKqLpOt51FaQfi+NsipO3Cc6rOUP6TgPXO9z0LS0mlC2TeSCIdtxJOBt6xuog4PON46id+jZyFKQE9Hc4aarLonN6Lns+GAY+A449FgwpEQZZWvFUgfWtVHrCdoRB4H2zACSBPrufjhkJbnjqS3wVOolTTvSvM7hrxBayMnzYh7mpm/o2JT+P8nzJp0gJKcmH8UqYvDqXjj2ApEPWDoXbdt1I11gzD6fVE9udxolf5tN+/QC808I0NWWSaLOSUp+gbV4hHRPM6B9lIuIy5iYT5ZfXUl3aQU+Hi/D7ucwcdAKjmuC9B+fy4m/+SChqZNXq8/hi7wheOTkFpc/AyNwWPKOy6Z0WJZaqUzKqlYEv+SheKyj9US1Gj8yvN8/jvj/fQmJwgEFpHdjaIkj9RmryOnGPTjBj3BHSjmukOEO4G1MpWxWgaEoz1/7iK1z2pPN6PDsOEvgGxenoSiHktiLd1U33eI0Br/STneYDwJofQCgQ6LEhR2TsZyWUoELheo36q19lTesQekcIZEXQPdJEpCZMwiIw5wdQAlH8NTECDS5sx8x0j4Hlq2ZhaLBgO2bmiY1X8P7yWWTmezD0y/gHRxma1sbQcXVkXNFCuCROWDOStyWBpVvCdMBGwSMxCu+uxfxZCoEcheIvfbx39xKOvTsIIcHauUuYcMkRHh29nhp7B+7JHq6euQ2DKcEtj6wh9biEdMCJr0zHUBzkm29HMj3nFO3TQCiC1BMSlm7B1gW/I3ujytiDC0hkxZALQkgaRIpjhBZ7kFUdS0YIqdUMOlh7dLa4K/EHLBxoK6A7aCeSJlCiEv6RETzH0wlOCnLFyW4i7+Rgcsu8u2kKZregK+TAsemcuOb9dXSPNNF5URxm9VHry8RbKcjM83D4dBEpJ7xo0/sJ5UiEMyXyN0qYPJDzs3ratxRgN0QJ5QvkOBhL/XRsLkBKSJgv76L+WQeZ1b2oA330DZZQvQqnjhRxqKWAGx5fi+pRkKsC+Eol7hmxCSFBbHkOp76pJHOFhctv+jH3zPwac5dK1j6d8KkULN0CoQg+/3gylue6mD96P5veXMb9P/qMSbfvQ2gywhlHC6r4BsUx9Uk0zH+NwhUqibQEvcNk5IMO+isNhAZFMFtjmAZ4SVgFdde+yr+9fD2+oBk1I0zDW1Xkr+1iYmoD+/dUYmg1sb79EH/sH4QeCKLEBNOv2kP5x3dRdds+KMz9P68yBD/oEP2T8b2Rqv8r4J9Nqv5/Q8WDu4hePAZLR5Dmi1PI2RMl5lSxrt7Nc427WbDyfn556ce8W11I4OsyUu6M0zG7AGm2m+CBDGJpGo9c+CUvnz6fgqeg4LUmbGqUdevHoIYlEhZB4dg2tCXZNF8kI6waxk4DsawEpnYDcYeOs07GV6GTvQdcJzzU3pCK5tBAB6NbQU5I6KpAN4FmEjiKvQzM7GJx7kaeHT6FutdLSF1nw9oVp+v2ZBfUscZB6HIvicMpJKpCzKo8SeP8TPxvGLDfr+IfmIbRm6BzrCnp4ZURw9RsQhnsJdRlw3laJW4He5tA0sBTBYvmf8XnD06n6bqk6atlmx2LW8c9WEIokLtTY8srrzOv9iKONuTTOGs582ovIn53Cu0z0rB26phv6aA3YMPxoZOOaTpp+R76G9KQYxLWDomCuWc5eSYfyaSTutNIxlUtuP9cSChHImtKO13b8rj9qq/54MVZ+EuSvCPHQy3UdWcQ67ZSsSJC4z0Sxctkmi4yoZYFsJhi+GpT0aw6mDUsjijKbidqEIYvPEr9MwNpmyYzfeJhNjVUkvOhmet+s5Y/fDIPRyOEciVCRQnsdSrRDIFrsJvy1F6OfjmAWIpAjkHhxigN8w1Y2xSC5XFSDhoYe9NBdnwyAoNfkLk/QOODElK9DWO/RKA6TmFRL/6ICU+fDecRE/5SDXOXgiEEvsoEakBBLg6Sl+Yl/HYuzoYw9ufa6flDGa3TBSMGN3J8SwXxVB1HnYJvaBS1x0giPU76DgPuSXFkg4boN9Jw5WtUbLqZO4dt4c2Vs8jZFcNTaSSSAZpZkHDoGDwyC+ZsY8uTE8h7pI7jnw3AcJ4b35lUUo9L9I5PYGk1EEvRsZV7Cda5WHzRN6z++Qw6FkSh3czESSfYtncg5atinJ1jRtLBPCBJKhYSxFyCnN06rRfrGJxREm4Lhn6ZWFYCZ1YAvktl0IKT7Dlbwp/GfcC+UCkfrrgAo0+gzHET3ZgBU/qJnEghb1uC5lkKls6kVpG5R+CeFsVUZyaWqmPplAkOiCKrOoUfqsTvceP8iZHaxyy4nEG8PhtSmxndKMjdKui+MkLBcpXmm3VuGLybFWvPR4lK6AaBa0QvCU0mFDESb7ORelzC2qsRTlO49P5NfN1ew8iMFk56c2j7rpDSZfWsPbCe4c8sJlggMHokJl9+kP3LhlN24xmOflMNQ/wcmfg2BklhyO7r2DnmTXZEHMy0xqn44C6mnXeULX8ZSsKWJJTzbiaRdJmcLX2cvsOFsU8mlqaj+pJ6Rr7XCumcIjBlhTDscqBEIHaBl3R7iAlZjTyXfYhZecNpeG4CZaNbCMSNvFT9IaNMRoY/u5jsl3YQnT2GlunKP3dT/v+B74VUnVsoSm/5fknVJ3/zA6n6H8UPFaJ/AdS9MJ6WGQr6oROsuGMJhi1HCafJtK4axHVvPEDZIzt5se4Ceu+YgP5OFj3TCsjZ2IPxz2lYOwU44xwJFlLwpE7Oqy00jwuyo7OUhEvH0STQS8I078+n9boE9kIfWd8ZQIfUAyrR7AS6WUeJCsiMYmuL0jE1jdQTYOpSSTmmotb4iFeHuGT2bhIZcRoWvEr4RArNL1Vx4zd34Ps4g7jPhGYC22NthP0mbOYYtvYYoaCJp65bgd5lZuuKUfROK6R7dw51N2TQPUqma4wJR7OOqU9iQlUDcYdONGIkb5NEYHQYe5vAWwHdk5JjvUtXX0zcoVCwWsVoTBDKFfQOT3qP5ezWidlkxj62iFM7S1FNSV7JmsqvOX2XC3+5RueMBH1BK5FaF4agjvOUisdrw9QnU/nkYSQdet8upnHuMkRcRjdIeN8swD0+jnmsG1/ERDRd53QoG/cojYGTG2ierzEr8wSpjhD5G6B+sYzeZSZhUUhkxXB+Ycd3JhVdBSkmY2oxMjy3Dd0A8eleTrwymM4JCpPGnWB7axmV93fhK1T50zuXIhTI3NrJMze/TUaBh+CQCCPPO43vcDoD7Z3IcUg5BcIAC5euofAbQdwm2DFrCYVXN/DNgSGknUwQzIOWGQ5M++ykHxHEXIKLhx+l/Wg2gZOpjKk8i39YFGEUGP2gTHNj6lXQsmLEQgbEC1l0na9Re4dK58vltF0Wx9SrcOhgOZWTz5Ja1I8SEUgBFS0rhuO4EUkHZ1oQek0UfaUz4+Rccj8xsrZ9CEoUusYY8U8I42gUOOsg7YBM9h6Nb387CXeNyvHPBmDyCPo7nZw3+TjWHg0UgWYWFGzQ+G7UcjS7zoe/nUXFoyfIWmNGc2p0hx2klHjoHmnB5JaQEjA8u42YS2Ca4MY5yI2yqIuBlW0YDtvJ2wTDpp5B8SnJhCNPsK+lENrNPPDurSzfM5nY8CAHfr4UbW06tg6d4FkX8RQN68NtkBJDiYISTt6Rb5n6EnGnTumaGLoBZIOOHjAQyFOJJVSafqlCp4nQvgwshy1QEAYdlv1+CTV5nXSOT04DrvjqfCrf6EDIEMuL09OeQrotRMrnNmZNPoR7tEbHBIWUM2HePDgR570SXxwYjntlASMvPkH7/HIGv7iYqTftYdjEWmwdgi3rRuCeEOeCtFNIOiQa7TzRPQqAo+M+wC6bmWmN80JfGeUjWtn+xTCmzjzE6/OWsXPYKp54+m1S6uL0jkqlZI3G2BnHmT3uEPlj2jl9Kp/I9f1UP3IM6bCDgi+7yV/QiMmQYMuQ1TyXfYhdEY3fnt1F7Q1LWT/wSzo6U3my6VJm5Q0n+6UdrG8/9F8yGfoB/z3wQ0L0L4S6F8ZTZZDovGs0Zo/O8QkrSAwJELtoDD0tqXimRNBVidS3d1K/MJOoU8IzIUrVHcdZv3U4p+90YVGS3mj66gyEWSN2RT9lfxSUPboTLagyPLuN8BUe1BofMZeEZE0gJSRMC7ownbEQzjKiRASBQomR008STQdlu4vclSY+3TuazdP/wI/OTiX9mCDn7noa571Oe1cKvz7/U+JWibrNpUgBlcxHoHW6EbXJzPsdEzAWBAnlCsx9GsahHmxtErHcONZOQShHJutgnA9KNyEU0PwG/Nf7yEz3M2zREXQjyBGZ8od3YmuD9rlxWi7VMHzjImHXQYeMw4KeYQpCkQjlSpxZuJTaqW9zWe0sBixbTGpxP2p6hIp3NBzmKKIgzMLfrsE3IEF+poeTd75C+PxBTLr+AJHLPTznrgQdQucF6J4VA03C/EEqAMKks+P/Yu89w6wsz77f391Wn1Wm9wYzwwBDGTpIERQQVBQRu2ANoNiiJibmSdGY2FsCWAKWiB1BFBFBeu8MML33utaa1dt97w8rb95j7/0ex372e+R48yQPv4/zda513v/rOs/z/28vQN8jUb91CFOKG/joD1cxIaWF2b85QEqSB2udSMcdYUw1evrGaCSfgfxvo6xZuIFHlnzNua2lSCGInbUxMD9A6phuTn85EnmPje5rCvFN8aN3xmdJ2l4y8MjBWwjvTMZy1kChqQ9Dj8DGqvEUXNNAaJGLSFqYN5+/kY5pEruWvcj0zx+ndnchd045iDdT4sr5p5D+tq1kP9WDvRa+OzcSJc+HYUDgxNmhFK2LYGqRUTwamdZBhJjAJzPXYTutJ2YQyf1awH5cz/gnTqJFRFJPxZADApMTG1ldtAc5AKlHBAqye/FnqvRNiOH1GBBiEEmQ6NiZQ9uVGkW2XvzDQoTtKoZzRvpHa6CB/yoPeleEwXyRiFVj5tKTzFxxlLRsJwd3j6R1rkDR+ghSiYfmRfCppwhLg0z63Y3sOVXKoVfWkbVDJPJsOinP6lAlkEKABhOszSQ0grY9ibQ7uvF/lIHnzRwC6SrdE0TO7i5GTQ3j7zWjuAVinSYKytso+KKP5PRBYp1GCj9fgTcXBJW46aFBpbImi9SUQWJ6iJrAcEsXNz3+ODFrjNbZeoSxbob8WUWICgyMVpE+TiQcUoglxCj4UxXGHo2kbQZUo8q1h1Yx3NqJqVtD3y1TMKGVxlszyZ/eTNY2CXOdQuPpLEJ2kb1flcfboK0CXT8NY6w2UHd3Knn5vbzzs9c515OBJoOhT2PHpolU7C1ClcFY3k/RuxFerZhD0oUYslfgq++n/L9q0WOJDdTUZmJp1dj/9Vh+/vv7KdhyP0+uv5vBXBnzLZ3Yn26m/1YHDXP09P2QxT2X7cO61krvLaMIZEeJVddxoTablGd1/NmVQ2XYz68Lx3EkUMjYZ1dR9uoqTNYgrVsKeK3pEHJB3n/PNtn/k0tD1f80LrXM/guSfEbA/sFhmFjG95s/ZMHYuSzff5TX/uMWEj45Eo/++Btzbr8HV5GOwSGgFHpQDlgx9Gv0zIyQ/qNM0CESmOmBygQQNGKGuMOuuVnC1hAjkCxi7o5h6A3T/liEtLUGwjaZqFGgZ5KG4hKR/QL+4hByj453lrzF3QeW03DlegBmVFyP6/sMPGUhRhR08E3xdxTvXUY0LIFXYdgb/XS+IHN6wifc3DgbVROo+KGEYH4Iiz2ATo4y0GZH3yuhDAqoCoy95iIXPyxlcIiGvUog845GvM9k0f+gH1+9jT8t2kC6NMgD1bfQ0ekg/1ORjrtCzBtSiSdqoM1np6k7iUn5TbR4HHQfT2fC7ErOfD0cMRL/SJj6onTcESLi1iP6JDKG99BzJo3CiS0EowrNzSmIxihJOw0YbunC6Tfib7Si2qM0zn+X4WtWERoWwHQuHiEQMUPhjCaaf8gnYtWouXMtI95chb8ozD3jD7D1xcsJ3uAieMFOJCtM0fKTAEh2G6W7PFjlIB+cm0TKdj1HXlxHwXf3IigqY/JbOV2Zj7VKgZlOpO12nGNiGLpkhCgYezU8+ZC7I8Ts1w/yzqGZSAkREg4aQYDgLA/axQTUEi95f5ZoXKRH1WuUj6mn882hdE+Ekjfb6ViYQ9gGeqeGJgn87KGPea1+DtrHKQxe60WSVKIVNiwtGqFEATEC0247RdNcIx23lzL85kqqB1Iw6yL4wgrC5iT0Ho2QTeA/nnyfX61ZjmdsEC0qopjCGI9YCKRqUOBHEDWUUxZSzoQJJsp0zVDR9UskVWj0jBe4evZxjr48HneBiBSBR+7axFsN0zk29nMKvrmPxJMyqiRw74NbeXnnAqSgiM4pEMiIUT62ntMnhiJl+En7zIi0ohvv5xmk7uul+yURT0US9mqwNodovFbH14tf5ZpvH8Ge40LelEjv9Ahyn8JN8w6wcd9USt52U3+bA9krEEpWydyv0b4owrDnvbT+XibZ4sMdMOBqsSMGBfQDIkkzOylPbuXh5D0sf+Ax2i8XEUMCmeM7aT+TQTQhRuOitxl24I6/t8JcswN/99LyZatxM8R0PxGngaQTEqGr3XgHTIhuGdUapTC/h8b2ZJ6ZvIVXXl6Kc6SKGBbQ0kJIrQZGXlbH6YsFSB4JU5ELAfC4jYwf0sxnhbs4FoowUa/8va44Y34+9RTxwq6rUQZFItkhzBcMfPfgC9w/cgFafiZVj8bdUZQuHTq3ABPdpL+hJ2qU6J6kkP+Vk/4xdo79YS0Tnl7J8WfXsnDCAlxTc5DCGnpXhIhZxnSymW2nd/zLi6F/VMuscPk/tmV28Y+XWmb/WS4Jov/C1N+8DoBhB+5gyEM9ODaFOLavlIKfH6ZrcymGL+24rvahO24hmKyhZgVJ/sHAZQ8fZduWycgBeGflm9z9l9WYp/Shbk1i8G8xu46yPryHUpADELaBccwAru4E7OcUImaQpjjxtFvR90gYygeIxCTMX1nxL3bj8xjQ/DKKPYjWYiZqi2JoV9CVO9F/ZcfaHKLpPg3zcSPlN1dwYO9Ibp2/j7/umo61QcQzJUDMo1Ba0kbPX/NwlmlojjCZWxWkkMbAMBnv8BC2U3oMV/XQdzEZXb6XUKuFotGt1LSloUVEHMcV5tx/hC9OjcNaoUN3ZR9ur4ENE9/jp1VLSXxYpfLJJDJ/kOheFMJyxIi5W8XUEaT+PpHMrQruQongGD856yU6putRRrkwbLUhqHDdYz/yzsGZfDxvLSN1EZbOW4Z3qI0Nb7zCA/U3kWN2cfyj0WRtbSf8rkqGyc2Zz0fy7urXmahXKH5/JaoMc2ae4Ycjo9B0Gn+du47XO66k+a1iPIs8cMZKwbv1VP4yn8JNERa8uYdNbWOIqiJ9LguOH4wMzA6iqzUSGebHaglg+sBO93iR/Alt3JBxiuf3LSTva43mxRrWCzr86RqWYU5cXQkIQQlTp0jIoRFNiVCwUePP777J/Y8+SvvlIIYFlEERMQK2hngA6fHfr2XoxhXU3Ro/fyNfX8XQBfU8kv0Dr7VdydnqXMpKWqk+UEA4OUruNwK9Y2SCqTEUl8iQFy9g+EZP/aYijj3xOmPXPkzK2SiqItA2V2PciAYaPioiOt9FJCIT6jSRtQdcQyR8I4PYHT4s+jDiG8l0TVRwTOzG90Ma9oYoruUexB8dWFuidI+XyNvmJ/Y7J773MulfEMRu9XO8/DMmnr6RAbcZTRVIT3ajbUilZ7yAblDA2qAiBzWSH2oirEr0fpBHKFEgpoPUU2F67g8Q8BjIzeyn7WwGikcglKRSuDlM09U60CB5eB/9FSm8dsMGHtx5J0JIpLisFXFpkNa7SvBlq9jyXbj6LaSkuentsGOpVgiW+0l2eOjps5K004B7gQ/dSQu33L6Len8KxzaNIpCmErNFsSb7CF6wE06MYc3w8HrZp/zil/dT+sh5Ou9Mo/KRJExpPoJ+HUmJXnKtTqq+LSaQpqIpGilHRY4+v5aPPEm8sO4mAikahn4B35gAcouBZ27cyK9OL2JqXiMbcvcDsM6VxQp7OwB3Ns+g1eugqTkFcVDGXiUQWeDC05WAqVnG1qjSOUvFdlFGiIF8VR99PVYa57/L5DNL2Dv6Y6ojMUbpDDzXV8LeUca4U7/Xyg2W+OD/hXCAey7eQW9VMv/qXBJE//pcapn9F+Z/3JiqLvuQb099z4vZ3/xdDG0vf4fETefIv+kc3qFRoo4otkMGfvr0Rs65srBN7iGSoOFSTQRyopj/bOPkr9diq4eE4QP8ovg7gmkxxtx4HkM/uAbMJJ6UiRrA1hBj0Gki9bBAeGiA8JFEtBM23EMFpB/taDEBVLB/Z8bQJyB5JXQeCJ63EzXBlW/uZ+aQWoLJGoe/LyNmUqlwZzLslVYCKfDKxE/Jyu9DJ8YIJcVnPHDpKHn8AnvfeptgkkZK6iC+KX56emxIOX5+OmIno8fVU5jQj+aV0XUpGFwaX/04CdEjIwc03BVJqK1mln/xAJ6Antp709B1yYRsArGAhLs8RM91QSI2BS0i0jFdIGt+M7QYaZ+pQ/aDrz2BwNWDJH50kg82z8aU6uPOzx6k7OuHqL0jETmgsvDoSsLPpfPjwTIGi2N0LMyibW8OTYNJBMb7+dw5kYLv7yGpQmPYhCb2bh9D2hEBa6XME0+t4rPCXXizBeRDVkIlAaofL0AMCww86uPdT+fj3plOd7uDmEuHpSOCGpARNCh5ogfv+USid/cjAN3bcvimZxQVV79B221RlF6F65bvJZoaRtrqQOmXGTO2Hn92jNLJjegtIdov13Htpz9FEyHtSFwMxUp9lC6ooWthmITWMKXrVrHmur8A4FWDBFNVGpyJPPCXFTQ6ExGNUZq+LsReDY6MQSJmkVBJgInltWRO7qDyhWGcPl9A+c0VjF33MIGCMB3TJVxDJYzJfvz32HCVqcSOOkj5q5HXrvqQwTyJqBHMFQZSrq1GejWZlrkSql6jqzWRSALoV3fi6TejzXbSXypz1VXHqb1PJvKndLpnRcn4VIdRiVB+4iYS9CFiQZlJBU10dDpIXdUImUFMnRp9C4PYV7dQ159MmtFDIE3ANzZA+pw2esfouDDlI3QtOialNHHDFYfJm9WMnOanZZ4eyS+gpoZJ0Iew1sPq75aRUCNT+noXTQdyyfo2iByAzP0arnYrCYk+ejttCEGR0uuqWTlqH4P70khK9BKxCKhNZsLjvCyynuHHyhICaSp530ZQzBEc71h486Z3MXTL+OpszDKq9I0SCKkyjTenkbdVY2JWC/Tq6a9J4uSFQnK2uyAlRO63KvblrVxZeQ3PbryJmA5ipvjwva7eiCbAf3x8G7oTFs73ZTD2uVUUfHcvB1xFjHxjFWP+uIrTXdkkGXwsGnsGc7uIe2YQw1d2it4PcWH1GmI6AVOzzJmfr8FVFmXAacZxTGHoxyuwLaij7IOHeCJ/MguGz2TvKCM3VPYwL3MM7y5ewNKGOczLHMNj+VP+LcTQP5RLLbN/GpcE0X9xhnyygkk/X8m8zDFkyBbavhxB+JSDGQce5Lvag9T+eRIbrnyX5CMyKWsP88ZTN1PfnsLvircgjxzkybX3oO+RmPv8Pn7WPQYA4ZtEnv3jHRRuitA7Jxr3RImIRK9yMXHJOTpnQdpOhe5ZUQQRDP0aUaOGFITBcSEMjXoQiHulXN6NtU5gsCxM9d1rOfWrtezuLeb0e2UUXdaEzg3ZxT3UbyqicVkeMaPG418ug3dTaH+vEE2IG/g5Cgeo/eNwxh6/mfxvAvT1JWDdayQj3Uk0IrHhV4s4XZfHhd+NovG6t4lYVXwZIsnD+0g8JxBMFojYYoyc2IC+X8DvNmJuE5D9Av2TI8waWU1K6iCJ2400Xy3w28s2k/ddjIGACX3xIAWbBtFESDolkvG6juXna8neHcL+aULclyXDS8yi0j5TQaux0HSNQv6oDhSXyJjbKxAjwJoUSrO6ON6Xh6lWT9/VQao60ogaoWdBiNKlVfSWCxRsvxdbo4pnWAR9lZH0IxrGbpHoniTSjoWpeGwNxARyhvTScnuMkqEdhG0qTXfmY22A7pZEyPMTSYCq9nS6Y1GUaiNDN/TwVeMohq5X8WULRG0q51qzMTdL1PaksKT4DCOm16EbOojzVi9oYK9TocnE+R+L+fCyd2m8WUBQ4bF372Po7rv4eedMpk+7wKjUTqQQeFqsaG4dYZtGz4wIgRNJOPY0Yjpn5PSeEqJr00EfA51K1/1ZBHIi2M7qEP82dCwet9IxPw1ju4QcgM6bwjy1YTnXLNtPKCWGd0SIxuem0HZnBP2AyG1X7QUgf7OTuuoMkg4rBKrsfHT/q1TfNZSM7Qrzf7cXNIF9a99mX9lXpD8RQ3nKRtoumcMVRcg9Oqp7Ukl2eNAW9ZNw0ETXh/mqvn/aAAAgAElEQVScn/wRey6WoIkwOreN5tNZiFEo2X8nxm6BE4+P45vPp9JwJBeh3owYhnBSDM0n0747h+gCF5l7wVsYo+t1HVqJjxMfjKb89nN0XB0FScP6oZXigi40vcqZfcW8UzmNQFYMf0iHfmEPckBAvGDhmq8fIfdLCXuVQE+5HoMxzIpXvuC1a69HGDmInOuj/MRNTJxVydH9pSBC8yKBvceGk3YMNEXjvql7qb3NhhqWaLlaJPxCOp1uK6qsEUyJb5IGUzRstSo1y9cC4M9UmZbRQMYNTRTnd3HoRAkAj636DPOXVnRijDsSD6HKYN9jIGV/J1e+c4DivctQ/CpZe3w83VOGqUUmZbueYIrAo1d9i6DoSGgE35JJJH+nUffXsXzTM4p7ahrp+j2cqMsH/mfW4yUu8V+BS4LoX4C+MRp1r0xm/K9WkveTLnJntlB46xlec+bTcP1bPF17HXJIIzarHPOXR5E69OTIbsybrXhKIuTsCPDln2dzrC+P1M11eApg/a9epafcQM1zo0i+pYXlkw7ir7Lz47lSzE0SgwUiWd9JFKzVSLypjdTTKubLejEmBAkWhOJ+Jpuq6Oqyo7+2h+QDCpPPLGHyEyvo/zAX5+gYtQfy8Y8LYH4iLqDyXjtL0TvdRC0qk546HvdjSVOJpEVwVyXRO0YmcjiREa+fx35Yz8CECKsLdnNNaQXtc1VEXYy+sngifWZRL1JQQ9UEBmYHkYJQsClGzc4hJM7uROpXMPWq+IeGWTVpN8e2lOH2xgd4C4u6+OOHS+m+N4BxrYNwSEYaGCQ4IoD7Sj91t8r8rmIhOz9ajyYQT3XfZ0czR4npIHN/hD9e9TEDm7ORhnk4tHMk4kQX/cNlIvNcJBp8LLrxANY9RoRWI6pRReg0IKKh6jQMzTrSV9VjSfGRPKOT9is0lEENeVY/bZcrjD1+M8ZkP96QDs2po64zFSkokHw+iqk3hmQN8/HkdxgxpwahzcAdTz7O6HlVVD2YzNPDt1G/RA8apB0QqLt8A77cGJlrdXyxZTq5ZifaaRt5T4cxOGP0l8XNARUPvNU9i9zNItapPYQcGrqLRuo8yew7PII3c7ax/9GXsVVJCLYw0ohBdN0KwYwoF5/LYf7Nh4kZNNx3DpK+Q2Hh6Ap8hVYWjzuJP1Mje0o7mYubMHdqRMxg7NFIrgghCBqqAlvfm85P52xjUnEjkZQoUZ+C4oGPKsdjalIYGGVn2pgaxOv7iGUFWfLlw/Q9p+IsFfnk/Tn8evoWSv6ykqI9yxHf8jEwwoKqCKCLG/ulv6snuimFb0avxzAQF4wA1rM6ElpUutYMwdAXn2ETqiyIEY3tH7zNgiWHUTwC0dwgapEfU5uM7JGwTOlFPeDAvLIdxS0SPJQcz2KLaPx4cRgZ6U5um3SEzuvCNB3KQXbJmDsExFMJmJskwiGZxCckrA0a39/7AsmnRExtXvzzPIxfUgH7Hfx24y3U3ZGEst9KrNGCs93GkcYCNJH4BcUrMmydEzmgou+R+OT9OcQcEUSXzDvz38WbpXB+8kdEEzTU1DBRE4RSY7iv8xHSIkTNKqo9wpazY6g9loc7ZMDQK5G538+v9yzGc72Hii2l3FdxB/b6GPbaEBd/lsLnzeUUZ/TQMQsaF5n4/JvL8OdG6Z6ukvPMId6tnUbrE/EwWTGi8VzWNkbkdHKhNYOfb7+ZjGWdCAPKJTH0v+If/Tp06YXo/xeXBNG/EM4RGg8dOUDL/lzkwny+G2FnXuYYBvanY996AWnPKSSrlcQLUKozMTA/gGSO0j/SiP8KL80XM2hYPZS0YyqLtj2EYUBjxNgmFmecZnPTKGImlbdmv4fepREZ7ieh2k3jdQa6vsth4DYvihRDOmhD8MtEDRr5OwLMHXGRyKZUPPnQ02elf7SAcziICRFsY/vQVRrpnuogpoOqV4bTMyMN3YDI9i8nExoZQJM1jNYgqk4jYUIvviERdjQOI5QocMu4Y/zHlzdzdiALa6XCwmHn8edHKPqgH+e+dHzZIP01iUSHD8OcXron6IkkaHi+zkBNCTOYJ6L0Krz13VxSzkaItpuQAgKNbSlELRqWrQm0XK+S+oWB+hftPFG+A/0pMwnVCqGAQvEHK5GDKoOjQ2R/0oD9uJ6MQ/FXot98eBv+6V7CjQlYx/Rj2mRDjEHD02Px/jyTj49PQry2H03WUNzxDblzW0sx5Q8SsWhU7SjCagziDemQByWkMCzKq2DKzAsIgkasMgFnbwJJp0V0lUZUBfQDYdpvjJDxhZ4l3z9I9dZioskRuqdA47oSTNlefvbjTWTtjVfBUY+e5cH2STTc8BYpzzQy5ooqXss4QaAwTO2yZEY8c46aZWsJZYfxFkU4dGAE7TMlku7zoy8aRIyCZ10OtmqBZQ2LWXTfQyRVBhHbDWTb3TgqNUrfcJL7lci58SJiWCBQbcdzo4fvDo7FuczL19WjUHMDuAIG6g/m0Tc3SKAkxLJHtxFIVogOGChYV0fEDJ8/MZ9jp4poXPgOcr9M9vf9GI5bCCWqSBGNQydKcJ1NZt6wSlS9RiCsoJZ6CU7wstzag6UVpDojJjlM/yiN3skx8r4QiKaG8WQr6Jd0c/355fSOA2ujRtEHKxkcGaH4gYt0TdOwX95FbJwHMQKT7jvNNdXXstB2lkB2jLtGHyZhtym+tZYdILA7hZyFTcjLQSz2IgfBYgzhmhGEqMDYpHb2PDMVy0kjCS1gH96PL1vjmxUvECj3Y9lrou72RHqmxtgfyMN9lY+qVWY0TaDxt8OQfRoJTRqZ4zrxDI1v8wnmKNkpTmK2KKOn1rJ0ziG6ZiQxMEwGILkijOiRSR/ew33f3cvAqPg50PeK4Fbwloa4YuwFUj8wMuq9h3h0wbeYrEEkp4x1eD9sTCZq0ugfYcTYLuPvNRNM1PCdTWT9y6/QM85A2bBWbH+0kG8e4MqJ58jaG+WVmzcwurSZ4hXHqHlnApb3bVind5P54iFeeHUt2bKF0MwuDBeMZO2GbRf3/nMK6L8Il4wZ/3lcEkT/Yjzw9V2knoiidnT9/W8XH1hDx71lMLGM6t8O5+jzayl+fyWCoKHTR9C7NSqnfUh5eR3pR6OU/eIspb9twlkKtbsK2dE3HFlSST4l8qvf3ktg/iC6ChOBnASMXSJRM5i3WglsTkOb7sLSIBFNjbBnczk7zoxEWdyDqoDmlSE7QOFXfiQ5RmhHCgjEt6L6NeSECP1TIvxi6edMXXQWh82Hplcpz2wja4+G61wypgaF/BWdKB74+ORElJJBXH4jniExtlaMAgFarklGUKHwYyfLfrUVf0jBfSoZf34EKSDA3AE0VcA3NEIkOcLkaZV0TVLQ9Brn7/oTSpuOvO1Brv3pbhCg87owS4rPsOEP12LuVJFn9WO4GN8e65ghkuDw03JHIaoC7XNACgqYujRirSYmT6vEbgzQV67hy40RzogwUGqisLAb97kkitd2M+TVGroencr7P3kNr8uIzi3wyG2bOThqE4aPHPx58btEjQLf/yE+xOpst6FJ8PHlb+HLFAimqOjcAhs/+TNKvZHuG4PYLsp4C6MUbtS4ZvoJAje4mJLVhK5PwpMtEUqJ0XZTKjt2lfNcXwk5Ridndg6jeO8yTLYAYy+r4eIvy7i5cTYJST6MbQqJ5+OOz42vJxKqs6J4oevaEAltMS4ez6d7ksJgnp6kCg2zEmLpz3fQcUUKQbtE48YRaBIkVWiY9GHSD8aDS2tnvUfdrPcIR2UUj4DlhBFHkodXjl6Ju1DEXiFS+9gQArkR2mbJ6JwiV1Zeg94lULXKTkwHYlaAnmtCDCtrJeKIsf3waFZevhN5tw35jIVoj5Hi91fiHKmiCXC+K4PCLSEkn4i7QIGwSP7yWvrcFoba+1CN8XafrmgQyRyh4q8jSTkuon/Rgc0cQArDj9+NJRSTuefr+7E0SHxSNw5NBEu7yukZ6wglaXjeyKHv8hzCbWZGLz2PL6hDqTei65bZ117IqJ+fRZNAWNRP8tMyGnDjM0+QmeTGXaIRSYpibJd5/Q9LocZMWUkrNxSdoW+0wuAQCKQK3Jx9HGOHRMQS3zjcM3IzSp9MoyuRLZ9dRsQqkPd5B44alb5VPqx1Ir3OBMaPqUPVqRR9uBKh3I2YGGbDjA2cfnsUijdKJCfEa99czarSfZjaRQYrk3j5d2uImlXs9WFMHRpCVCD5nIbOLbDk9L0kV4Sp6U7hh083cO73o9n37VhcQxVeeOROLnbGM7urFqzhtZffxHpVPQA/e3glANGduShe6Jgu/Mtvk13i35dLguhfkLbZIlpZEXJBHkwsY+HEhSRVhJj57jFUW5QRh28ja2+UpK1GTN8lMOnxE8y5/R4q9hTRvFgjEFNoeGAIhmI3mTPaaHYnMng2CW+2QMghIAjgHxqma4qMqoPcGS0Er3MR0wsEq23YGmLYk7zcuHQvyBrdtclEbHHTvKJHO+iaYkausGDuUsnd7gFgYKxK1KtQ8Bk8e3ohIyzt9LXZkbwSFZ8Ox10oEUmKggBt76biHh7FcVIhfY2BoqReJL9IwUdgTfFi7InnrzlfiPLS6bn4XXHxYq1SCGVESHzdjK5dQemVeX3mRgbuSubO63cxeUwNl/38QRzlvbif9JKhuEjer0C/nk2fT2egDFzX+nB2WgmmqAgxiJlVvM02QnYNLneSNqQPW73K8LsvYOgVOVRXSJ/XTOI5gZJ1biyJfkJJAp4PsxCG+Oifmk7nzSVIQY2f/P5hHp64i4ur1jBE183CcfOJ3DHAT/Yso/yOc4x//CRNdWkY22UiGWGea12IGIM917+EKsOVp+4hXBDEsc2Ee0wYKSFCx4owB/80AfWwg11HRyL7hLjfTLvExSfTMPQK/GXn5ex/eRIFm5yYD5rhpI3Ol4fSPF/m6IUheJwmpCDovCrTplxkZHon111xBHdJjCSHl5YFYOoQ4947wLTHjjEnuYqvn5zD4IQgwSSBxG9M5OyM0D0rinGdg5888yVPjdj+9zNbMWkj+Vc14imKoduYiCCpSJOcJJ/1c/q2VzG2KdxwxWEEVaCuMf5xJSqglnuIOPWoAzrU+f0IMYFhf+rjw/XzUBUIjQyw9drX0A8I/PLKLcgBAdP3CUR+6cRxQcCfoWFP93Dm6FCGPNhJvTsJISyQu0Uk/4E+xCYjvhwN5dZumu+J0d1lZ/qNp5h45QUCEQXVEsNTFCXQnEDYJtBfJjDt5ccQYhA1CGTdW0fp2GaOfz+S24uPkzypC3VIAPWwg/oJQfyZKob1Drqm2dFygrhKoK06FS0pjBAUKb6inpgexJiAqglcbT0Tb9upkHVVMyc9+QyZ30DMEUUWVS576CckNMFAhw1Tl0buB/W0vWKkf6TA0MQ+fNkaZlOIk8eLSKiTUQYFIrVW1JjAXXvuJpQkYP1tK1pURDdkkNe3XI3i0yjY7Oeuo3eBNUr3BD3OERrmZomu6Sp3Lv+eUEghYpEIDRhZcOVNzPjNYRKaNTL2uzB8cwz5vIXw/AmMP7aMT5yT6L8n7m1k2Hos7kxdn4an4FIu2X+KSy2zfxqXBNG/KLW3mKlanQHHKkCWmP7yEQ7OK6D47hOsGHYAf4pM8CYX/ZeF2XJ6DFm/r0WMCfx4xWt0rcgh42AE7Yidhrp0+hsc6JwCaHDfiq0Ih20YmnU4JvSQWBlDfNiCpyuBwbIw1809QsdMgYT1VvRClLRdMkmnRSRbBHOjQvutQ9G5tPh20AyNgREWkrNdyIMiBkeQtrvDaC0mXt83l+wdAl8sfp2wA2JT3dw3eR/ps9vI/lmE6WOqCFsFfBkKJ5tysVdB22wdtvcSyFjeiL5fw3k6BanRABqYugSyNtZhTfUSSFVIPxoj8aLGL99ZjuUvTj7cNIcjJ4vpmRZDXp+E/t1EPl65gP4xGqklvRj7NNJGdaM/nEDyERlN0TC3AbKK3hn3YooeddDVkoi7SMQbiQeJrp/2HtZ3rEx78DhVj1gIX7RR8cgaBso01CYz8q3dJFaGCNsETvxuLY84mgCYY4xR+VQe0W3JIECVK5VtNSPAEMMypReTNciFk/n48qLc+thP2XT3S4SPJGJOCNI3L8g1o84iK1GC/UbUxf1/3xrTDYI8qx9zh4ahS8bYo5F+JH5mWq9yoCoQKA7hzZDI2RVD36kguhW8JWHu/f1XHDwynJPHiviqajQNS97C7TUieyX8GRq33vgjx55by9bakby8ewE9dwfQYgKDo0MEHSKNi0VM9Tpal0T5Q8VV/Gbz0v/bmXWuyUNJDvDU7z7AXGFA/7Wd5qtNvNg/jsI5jZjEMMH0KJJbJpSokdAgEmkx89jM7zl53au0PTKOhsVvsW3Plxjm9GKd20XaZh13VixDE+HFzxajc8HQ5dU4t2VivbWdz29/FVdXAsYekbIdvQyx9WPM83DL899S/XgBMYPGyKl1uH9MJ3GHkeK8Ln78bix1rw9H1QREXzz93VLg5sFlW8j5IcRgaQTZL7Dy11/Q+t5Qwk+lInth05uzEQUNsd5IxKoxcNeUeLjuLX4Gh6pMH1JHyikNQ4/EsJwuika0c+58PtrCASrvX8PFilzuOX0ngdwImqSxfdi3nFo/iorqHOQ+hdMtOXgzJAYLIb+wB4NLpfW2Ifhr7Mg+gSZXIuIQL2m/kUg+JaDzaEQS4pEdml9G6YlbalR2p5P3pUDAp0cTAA1CiXoS9piwVOgZtrAG1RzDfHkP6ftENnwyD0WJMXCbl5TDEtt++JSNZydw1aP7SPhTN+7bJ5N6KoIUjOEdMHGuXCP5nJea9eOZXeEjujMX0XfJffo/y6WW2T+PS4LoX5y6Vybz7aGvmZNwgW9Pbqfj8al8M8JB//wgqYuqaJz3F4rvPcGFDSNQFY0rNj1Oww02WpdFUScMgqyCLYKqgKqDlw7OJ2KBYH4Y15E0DA92UL3Chj1jkPQfZQ71FCCmBpn7zD42t46if7RA1vIG0jfp8A2J4MlX0fn+9is0R4nd0E/k+2Q0AeYVVpL5gR6pwIu1WqbsqbMse+NRokaNSESiK2yl40A2VauT8Ed16AY13IUishJDvb4fnVugdYFGz9v5eHMF1IIA1nqQ3DKWzhjRwgwGeyzYK5zEDALhpU70/RqnW3Kw1akYuySKizoILHMir+qi/YEIsk+guzaZgXFRFmWfxdCvMTgERo1qYuSdFxACEi/dvp5QokpwRIDEkzJo8Y+KpUXjvqN3oioCF1wZSKYoYkig7NVV6As81N6xliSjH+WXXfgzY3//n/XFfCypv4KERonkG1oRJI2Bw+kYT5mwntHTV5eEbpcN1RHh0MJXsDR6ORQoJJii4m+2kvOxTKU7HU0TmDWmkiuya0g9FaZ7qoYY1ZA2J5K4rIWCL/rpvTJE71iB7tlRcra7eGrFx9w8+jiuCSHceQqWNg1blUDGTpnnvroBW42A4hX4dMrbDPl0BQUp/SiDAnlj29nRWUrx3mXYLEGGDW9lWk4jkj6G9awey8Iu9N0ymfv95H4mkfCtBYD7Wqcx6ecrKfrrSsY9eRJNFXjk4C0gQvhaF5oAX781k4a+JD79bBbmJhlNjKfH+3JUUAXWfLqQ8h0PESgLMOzAHUz8xUrcZ5KxPiyid0ZJes5I9q5BpCB4czUqvxjG2SfX0OOxcN3Wh0k+IiP7oXowjeO7SvF1mfn62onErDGSzwgMPJ9PyKExeuU5amozSZrYTf/1fgaqktANiLx//RrUAw7e+tMimq7RIVsiyBOdHPEM5fizawkn6kisjhI1CnScT8M8eoCMCZ3oPSqyV4DzCRh6RbqXpTFkdRWmTo2ej/KQRBVDmg9nu40hu+7C3CKhVthwnJT/7gHlnBZCGZDRFw1SmNaHOsdJQhO4v8zEmyGheOKxJ5oIQxP7iEZktn/7EZ58AX+aQNQWY+Q1VUg+kezdEV6+bT3hoMy4353EcsZA1B5j0n2n6ZoiMTA+StQI1X2pWOoVDHIUS2sQ8+Q+fE4jq0v30HtZlKtrrmLNtI94OvkcZ/cUM/OnR9BvO079HSLF95xA2p1J7S1mGue/y1++uYKmc5n/p8viJS7xv8UlY8Z/I/6HkePlFxYxLaWB42Pit7Lu1VNxj4hSvOIY2Ucs7D46EjEKYkgg/WiMXWvWMurt1SRdiNE5VWDcxFqq+1LR9jlIPRWke7yBsusrOd+bTvSkAyEK/rwo1koZxafhLgL9gEDIobF43mE+PzsOPDKZe6C3XCScEcGQECJtg4GQXcKXLhKxEB9sbhHwp2mE06NkbRfxZkkULqmlaeNQfNkQTo6Rld9H/5F0Us5EGRgmM3vJcXZ/MoGoBRIvxJDu7WHXyC+YVXEj0Y1paAIc+8NaFsxcTDDPQcwoMXC3l8S/WGi5GtDg6NWvsvDsXQT3JhMY68eW4OfkuM+Yef/9GB9rp/vLPFxlUYzJflRVQDqTgBSCwmvrOXc+n4RaiSGLa+n0WYmpIr2dNqzndQRSNWImjWHPN9H1thVFUum7mMyQL/w0PixQM/N9Cr/8CbJPxDDMRcCvR2w0EjVpWArcSNvtROa5EQ7Y8ebFMHZKKF4YLI6RUdSL8kYSLTfHaLhiPYU770Zu15M3sY2m7iS0DgMxexTbOR2Z1zcReD6L21/dyh+2XYeaFCHlRx39ozXEiIAqga0Oblv9Pe+tn0/honoqThXwuwWf8/SB6zG06gg5VMQIxBKjrJ74IxXeLI58V8aIK2o4eaGQrB8EesaLRK1x53N/ukrJ2h76J6fhKoFwShRdj0w0QSN7l0rrfAExIBAzqZha/9aO/e0hojtzaTuahRQUyJjZRse+bHQeYKYTj8tE9maJqF6k/PHTHF1TjuvKAEW/D9D9Rw1nswPJK3L/1Tuo8aWzq7oEoV9Hzs4YvaMV0OK2ETm3NXDhaCFRR5Rhf/LieT5ER00KiRUi3myB/K/dDJRZ0d3cTUdNCppRxVahEDPEYzp0Lg3HrW20O23YNlnomhXD1KgQKAkhOBUMuR4CXj12h4/Q4SSCIwLUz95Aweb7sV+UKbvtPA0vldJ5XRgE0Fw6fnPFl9QEMzi2qpy6m4w4CgcYvJBELCPEO5e9zxxjjFsbL6fDZ6O5NZmxRc0YpCiHK4cg6mOkJg/SV5HKE9dsYf2z1xJIEVFl8OWoJJ4XGBipYauJ33m9ORpyUCCYFUH0SWiOMFpIwlKnECr3ojttIffLTuruTifjUIzBPBnbog6ib6fRPVEEDRKaIGoUyDjgQappoenBEeQ8e4gXm46w9Nh9iOcSWHf3GmYYoHjvMop+4Yq/Yv834R9izJieow297R9rzHj+lUvGjP9ZLr0Q/Rsx5JMV3NUyndBfMv4uhr7vOEPGPieN174NwN6GoTTcuI7ySbXEcoME73cycv2DAMR0AqeWvkptfwrCLgc6j0b9bRKxyYMEYzImXSS+zn08xPTRVQyOjODLjPvWBFNVcse3c2ogh4xtCs/M/YL2q2KYOgTsp3S8Ne5D2uZIdM1QefWBtwhmRnFcBNe4UNz2PyLw+ktv8trD67jQmYEcgKgp3vaZltbAjAWnYWU8A2vXlxNIPx4gVuKjY46GN6Sj+NsVeL9Pp2hFJSGHwKr2yVQ+mkznT8L0jJUJVdvonihhrZRBgBtXPsqYlHbQQKkx4bmQRMn6lfSMl6luzMA/w8vc8grCIQVNi8/l6Ac0GpyJFJe2oSpwpqKQFJMPUdB4fOr3aBIYh7vI+SFG6y2FuAfNRDelIPsFau7SY7f6KXtlFcPLWuLRDdZBsj5SUDwCmqwROmcnkCaQ8mcjmhj3lsndOoB7bJjckm4chgA95QpSp55Rx27hlxO2EdNrtO3PoehZP1JQQOlT8BSqVFXkEDWKPLv/Ggq+DqFv0uPNEojZohiGuZg8rRJvDgRVBUGD4Jx+Cke18x/f3ojOEmbIrEbSintRPCLmah2PJTawIXc/eTObaVs3lIen76BjJkQzwtjPiyy+bS/2aoGq1SnxPLrCAHJCJF5hNDDXDFD6bDOqUSP5hERsrIdgZoSGP07BFTD8fd4h/OcMQoUhPKNDnJ34MZnfyPhTJLqvCbGvvRBBBdtuI5E3/ASOJmNpkLBXw8Z189hdV8yovHZSj8Gv3liPfUYXQ+Y3IEag650Cvr3pJWRzBM9QK6WObhS3yGABhPODzNhwgmCiwOfDP8CSO4jjpIwmgW9kEE9RlMErfbg+ykarsOLJERF9Eulz2tBUAZJDWE1BzBcMCN8kEnZoJBwzMuHplRSvOobi0WjxJKKJ8LuJX2M5YSS7qIc/frSU7W9eRuMiE2IUhM1JOKog4xuFlcdu56GOCZz5rpTmphSkAYUzZwo5cqoYY5OOVyd/ijeoJ/2wytsvLaJ7ukr6NS1ErBo6l4gvSwBbhDtWbsebo5E8ugdV1tB1xyNfzNYgjvRBTjz8OtGggqFPo/0lA8ZuASkYQw5odB7PIHF1M1l7ogi5foyLutny8AtkvtGEc0EpAF+2HWGUzkDVZR8Sdqj8vnAMVxVOJhKU/1uJoUv8e3BJEP2bse/QCHw3u4G4GPpJ2xT6xtmZe8Mykg460HoMFH2wkgvfliC2GwjsTiGSEyZlaifmzjA20Uj4aCKeQpWIWaCsuJX8pAGqdhQR2ZyCqtOY+tIx2n5dTOkL/UxdcC6+mVIl0OG0Ud+aSsfcKM98upSMH2T8GRo/W/0x921cSeKwfiz1Ms81LgRFZfrqo5Tmd/LeA69hvyhz4/4V3LX7bsy7zfROiZFb1okYgu0fTOWHw6PxbM1AC0oEU1U6JxvJTHIzbM0gJl0EwRBD59Y4crAU30Q/hzvykb0S5u0W0o9FyBrXgW6UC+/4APKgRP9IGXfEgDx9ACkAS+cfIJwYI1IUwNigQz5j4fSaMdBhIPsthYwjYWJ6yL6rC1HQyHmvFlOrRH1fEt2ddj596u9lX28AACAASURBVCruu/tb0p7X0bI0Rniyh5wPZYz9Kgnl/SyZcAI+T0bv1ojEJFRFo7omi+arBQIjAgzdGEQTwdSp0bxcZeIN50jLG6B6hQ3BJ8FrKVyoy0LvhJhRJXrcwSsfLEb2CcQMGq1XJ5MxsZPCTweZMKGG0hfbiRoFhLBI3XIRa73GhdVryMnpRxZVDp0swdoA7xybTtQAkRmj6dyRg6lLJDvJRfvnBYQi8faVrfF/DsMm6v04SwUecTRh7JSQDRHEMFR703AXacyfcpZgosCo7HZifXomzY4Pn7f+UYfnPSOWeom+yVGiLWZ0PTKlUxoZ6LQhlnjRRnvwLndjc/iQDRGW1F9B/0iJZY9sQ9FF8XQl4CkQCCQLNJzL4uKqNRRfX4O6uJ/BSQGUaiPNnw5Bu6OPZ1bfTa8zgej9Fjz5Au5FPlbnTSPJ7iVkEzn+19H8duknSGEB00UD3/z+ckYtucgVa59EVUVcU0PEdKC06kk5LJH5oR5fpsDouVX4c2KQHKKhJRXBK5HyvZ7QllSiEzw4y1SSRvaiznaSs7yOutcm0z9GYzCox3aml6d330AwWSO4MR21zIMUIW7qqQpYm8MMFgj0jhN5e+KH/NhSzNzrj6HrlRGz/QAYOyQME/t5eP+tZF5/kcFlgxhu7GZSWR019RmkT+4keVIXZVdUk3BGz9pzM5DCENqSSiQ7jKDC0jmHCFXZSHlOz5i3HsZxSIftlnb0X9kxze9GfbIfg0sl+azGwJt5uAsUjMfMDBxJ59GmG9h3ZATeTBFro0pQi/G2O5N5mWNwlAzwfccZhIQERKfyvypPl/j/4pIP0T+VS4Lo3xBvgw2Agi3382rmXu57YgvC4bP0Bi3UL12HatB49Z532HHzixh7NfKy+ug+kY4/TceEp1di6NOQggLaLCdNWwqpupAT36oxC1ha4Kv6UfSU66h6MJUfzwzHlucmYhII9hqRDRFS9ynIIweJmARMnQK/2LmUc3e9QW+njagZ3B9nISgqW2vKcL6Ty41bVxNOgIYr1mNs0uHNBUSNmCqi/q2umttEbl3xPfZzMvU3reP0Q2/S3JBK1Qor7Z2O/4u9946Wo7r2db9VVZ3D7t45R+2knHMmCRBBBImcg5BNPtgYBzA+HGyTMQgRRRIIGQmQBJJQREjaymkr7Jxz3t27c1fV+6O59/H8TvB499xrP1vfGDV6dPVcFcboXj1rzTl/k7iDZoayBGp8BKc9gO+Mm5SDGr4sQc8oA2FVJt3pIX6HmYxdUTIuaMLzYCoDPXb8WSpfvzULw6CMa4eZQH4Y07ReuieriIwAwZ/1409SGHdbObX/UkJ1ezJn/zWXQKpG9KyTnM8FvSMUlq+5lJ5fBKm74D3cDj+tcxTarw2hrU9gy8fTCLkEnnwIRA3EVQOyjikhgOZXGPP6SSj0MVikow0ZONiejeNZB+k7AQG9IwxY64wE53qJq5IR4wcx+CB3ejNFUxswDuo0d7mpfsTIoYNFuD8bouz5FThqZLLXyXTNUCn88D4sSgR1RwJ5pe04b2jFmegj6XiEhoUGfAURzLN7aO2LI21XL+dlVmFv0mmbq3HBktupivhQJJVwaoQLltyOu0ol4jHROyXKwWOFmPO9HOvJQJ81gCR0zJ0yjc+U4M+OIoTOstxdpL24j7QdMul7NEz9gop9edgS/UTCCsEeC/6gEf9JN4uKT3KqPY1/veFj/rThEsIBA8VFrYTygoRH+UGPhWW6/A7CuxMRnSZCBUGyltSRsCzMrnfeJs4RoOKXTm5ZtB2HNUj7I9PpOZ2EEtCZe9tB3l52FaH8IL7cKF0TBUMRE+Y+HaMSu17VEuv11jdax9LoJezSqFhdQsnrfYhOE8sm70Q36nSPB28ehLqsJOb30dkZh3Gji1N7hsV+b44o/mMJnH00EXuNQjhRZfASHwDBeIHBG1tlbb5DJWrVsXQKfv7bewn4jXz/1iRUI0S7LOR8raKO8zJQ70byKLT8Yjrexjg6T6TgvcEGqqCxLhmbIUzPU7kkLmxB7Tfxwg0rCSQLxICBUFaYb96bSTQ9hO+pIc4uXY69XaXhZDqhRQM4fmdnQkITQ2kyckjDf8sAiVc1M++mg4y74CwNawtAg1tu20LC7hbmHb6blz+5kvZHppP0s9hvtfIXBf/H57x/FMT/hu0cfz3nHKJ/UGpenIoUkJjw9kOsLU1GTknm13kbmPnAvVw04zgvDBvB0iXL6B2r0/NtBgavwLF6P1GrIGIXGDyCULkLb1EUW5MMAoqvriRqFdxSdJCcT5t5/ML1OCsVpK/dePM1cgs7ifiMRK/tZXhyB33nBTEM6Zg6ZUo2/AQkyNwZQDUJErebWFRygpBTIDRBMEUj//N7CWRGMQ4KrAl+tLeTYyX20VibkJWfXkTqR+X8vHMsj3dMwpXmAYNO2mYD/llDuCs0XMeMDNa4efqq1Xhy5Fi7iAi0tbvpWZWNuqiPgt+cpWF/FqEkK+MKG8nYAQZ/7Dx9s8J8Pn85ic+aGVbcjhqRsBoi9I4SPJuxBeOgIOoxkrRPwZLtZfn1b9F4uUAKQ1pZGP/xePLW38PQ9hTypzRhqLQScQgQ4Jsce8pXdYEug2yLoNbbsTYa+PzIRPKe1xGqAEnH7zPRfIGZniV+pGCsu7w2zsuRae+SvrGFBblnefWh5VTVpDE1vp6QS5C0yURWSj8A7X4nx0MhUi9romOqTMkbXpSAoPpMBr/7yfu07sugudsNwK5338bWKpGUMYD6dQKOTXY8pS4+Pz6BhBuaiTur0HCZmSV/fIyjX41kxvAaQgkGWi9RcZ42kLZdRh6SUE/GEVmbTChooHJDESlzW+krVRj2cRjHKifPnLmYqhWT6b/aR+tsCX+GxuKL9+Drt6BFJOqveItoqxXjoOBoXxbJqyyxJqA6KK0mqppSid9tImWtiScv+Zwbhx+i9UwKtjYNY7+E45iZqs4kwjkJ5K+9l6Sbu3l80mY+qxtPcHsScggswwbx5Eps+H4iK1e+grnazBWTj2LpEpzdnY8vA6anNRDv9lE4r468Ve1IYUHdr40YvBJhJ1T/2saLV3zI6voJGHtk4ssFQoW9l73AZyNXIhs1otZYcvqY55ZhcoYIZYbhh6qfMaWNZLxrJBJW8BaopO31kz2hFZfTz90LtuEp/SEU3WlioFRHTw1iHJDoHmck0ekjrlJCjw9j6dIxeAXOWjj7SAaWVgVzh0J9VwKWyk7qT6ejyzqPrrqDgvn1aGYNqzOI5ZJOhue04/8qhYfaJ6IpgvwvQviqXbQ8orJx8xR8M30IFRxvxzEnqZrxtgY+ydtJyqv7GDGhgc//cCHR5hbKp3xC1u/2kbm5B+H1n9MYOsf/rznnEP2DE47T2NJ2HLWzi/v/tIw9r77J9y35VL86heYL7MSXCzJ2DBJx6AxdOwVvnkb8gjYCmSrhJBWEzqxrjxIsDdD3qxzCDlixdx5VyzJ5r346nuIog3OCLJh+nO7tGQzL7WTf+FUc31NEwlYzCcf6MfdCaXELphYDSn+AUTefIpgoiGoSEbvA3Cmhu8MYU/0YBmT86RrRM076C2X86ZBywIs+wYO1Q6f27Tx2vDqN8fZGBhtcSOZoTBSy3kbvKIFpQEOKCp7883WcfHQ5+uRB0i5vJD+rG//FXr4c+y6n+1JBB80gONmSwZqXX+DT3z5HJCuM6DNwxysPUbvYQl1bIroq4V+RTvIRjbvqrsY3PASKRmjRAH8YvZanH7qT1NxexMx+6hfDttueQ3aGYcYArYNxxJ/VGMpVCSbpJH5tRuT7sC+oY859B5DrLRgHYqXy+XmddE6OJW4bu2N9oaztcH5eFZYCD/ZWlbMzPuKiB+6n9o4Mvls+hXs+WkbiAYUvXp/HtKtO0DlLYzBgJnuLyuBnGdz8+sMoN0aIq4L+Z2P5Xzgj/OzTW0nfHcZgjGL42sXwN5ahGqH/VCLJ7xwh6/Yach+pJCujl+qKDAbHhzAMCi6+aw9pZQFqlpfQPVohMcVDKEFnxCPlTJpVARr0zQ6R+JUFX06squ6am3ZRc4fC3lfexP2+HckvoRxwYPBKTJ1WwarDUxB+mZ9O3AnApqtfwNynIwuN5gVwJBTG0iGYOKcCpcNI7wSVjqkSny2Yzvrlc1h6wVYMAR3njC7SXjuI2mCn9lojhuQAVb8sZnP3CN4e8yHekWEiDjAqMa2rovcGuO2uh7A36Ww4O4qkE2GcdWAaPcDmquHEL6yicncelU+6OP/8YyiKSjAnzHt3/gmp1sJvn7uVgep4Js0/S9K+btL3RJn9yWMsLr8D0WrGP30IZXI/il8n784GzM4Qmd8KvEUR+p/PoXuskZQNJqSgRO09EnVVqfSfTeDPL5/PiKIWQvE6mlkntaQLudWMY0IP4Tid9opkojYwWSIEL/WQdFwjkCyIq5Iw9+q4KjX0RiudF2XhaJBwpnoJpUY4XZ/OhBF1cDgOPkrizKlsvDnwdeVInGUNtM20oPgF8n4n8eO7+OPEtbEcqlSZz9+Zzyc3LeBrv5ktbcc5dSqHA394g6Ynp8c0hj4Zy9KvNlLxcObfbqL7R+JcyOxvxjmH6J+AgtVLkZOSCMbrzLv9LoZ6rRxZ9BKORp2FD37H5g2rkIqGiFglLpx9nL7N6ZjbZJRBmazcHhxKEPMZC20zzAQyo0j2CLYWweTkRtDAHefjZG86pj6dpn2ZPNA6m7wvfQRSBH1j3aRd00CaxUPEodOwKJ6y3SNieROAZoiJQDqPmJmfV03EHcVRL1HwQQdJ89pIPqJSu9iGXu5EV8B80E7f/CBv/vxqANLXGfHkSRiLPJhHDgAgsvykT2njorMLyYnvp/pEFl2bMxFCZ+4X/0LvgJ1wvIovWSZ9jZHZqx/j/B0PYq0woSfGqoCyR7aT/I2J4xf+iZyHqugrlan7Jp/zSisoXnqcxD9Zefn26+maoBBZm4y3w4Ghx8BdNUvQNYHlyzhClXEMZUiYemUi7ihhp0A5bqf61Sms3zqFcLyKLsFgkYr5do3B0ph20qNXfUXPJUECSYKJjnocq530L/Ex6uVlyEGNUEaEgfMCSFEYKIasG+o48NkYcjboyBvctMxT6J8TRJ3qoeLxPNxnh5A+TOTyRfsQfUZ0GTonmfj5iC0IFSydOv7CMKOm1dBz6wR8s7s5uLuUgYAZU6eMyRZm9CUVfPHFTIbSTQxlSph7YNBrIZQSZfvhkZxcX0ooUUVpMWHui+I+IXFBSgXbnpqFNKBwNuwn/Wc1JJX0kHA2Qs7XXk59OpzFEw5Td/WbfPz6RQC4JPDkwYKU0yhxYW5/7SE8Y0PUDSYQtWpg1jB6BJ0XZBBfEeSzly9kzm/24dmXTM2HI7EX9/PAvC2EB02YeiX6/y2Xn5y+EcIS5h6dQa+VqEWn9WlBx2QjwUSB9YQFT7YB45CGd8CK5ZiF6tenELXpaEGFsrZcXJ85KH1ukAee/mms8vHGevTEMLck76XnRcGkfz0MQGhnIkKFSL+Z6AE3/aM1xu0exLzLgedWD/ZaA123BPCnaeQ/WEHxay1k/Vkhv6iD7K0RvLnQ5nHiHtHDeRNP0dnnxFEPPZ1Ozj//GBMmVnPedQcJdViJswbovCqEatHxDIsJiQYSJS4+/zDWazsIxusom1yY2w043H6OnM7nxuu3M++xfcQfk1BtGrlvClqWFKArEHarBMf7UVcn8+jXN7H7jbfwX+RFDur4M6z8+swVABT+5AB5X97DG7es4LmG/Uh1Fh7ecMv/2UntH5hzOkR/O845RP8kVP6igKhNx/zdKYruOsyk1Y/i+rCM3d3DAEhyDvHBb19gmKWL3CvqCKaqCB0Gvk1j93NTCWSqTLmsnKQDMubTFrwFGt+/PwnFK5Ni99LaEo+jNUokO8SOmmIGh1n58id/pPv8EJXHs6n64wic+QOEklQMBV4CyToXxZWjmaDgI41gAmzZOxaTO4g0r4/ntn1Cz5CNjqkyik/C2q4zUKwzlKdiO25BUwR1167gqeffQQpD0psWnB866RspKErrom9jBtVnM2jYkUvhmGZ+d++HTEhvBkkn0e2l/sq30BUYzFWovukNCMj48iJkfyqjz+mn7UA6XZNg7u8fpayiAHMPZG4bpDPoQJ06EsMvO4jaFILpEaRFPbjKFdxnoPZgNqlfGVn2i7UknNDxjAhja9ZxpAzx1CMf4GzQKFgTQgnEwmiqRcfglai6PwvXKYn22Rovrb4Src+EpUvn3768mrBDEK12EEzQ6RlloPRFL3qXGTkA9kbByYYMojaQAxoht0DxC5I2mzDsdYIOVfeauOAX3/PlN9PI+E4jYtcJpKu8/YurefTnq3E0R6lf8A4t7w5jzF3laNuzsLYKokfcsV5zuxzUvFeMFAHzHe34s1QGxkSIj/ORmDGIbo3iy4+AgIhLpWOykahNsHLzfFou0bDmeXi2fQHlG0uIrk2icZFO0wIHCWdCbGspomD77fSPizLpl/cx9atHOG/BMT5Yfgn2fVaGsjWET6GzMZ6Uwh7EkEwgO8JdD69n66cr6R2vsu7PswjH6WR8aoQt8Yw1N2GtNxAe5af55ij21+J47fwP6R+pYztkJRKnMTOjnmCySmjyENKMfjzn+RnKlFk06hiZb5bjzu5Hc6hkrxeEDsXTPgNqbk1EU8Beq7C+cDPx35v46Z/viulBOc9iaxNEpnrJ+ToAZhXVDDPHn+XzzTPIW1LN3okrWbXsRVLivBgyfGi64MxvUgjEywTeTqdlvgFdgeiuBMxvuTmycjR6p5mQW0BE4v6kHRyqymP76snoNpXuYymoEQkpIhDJIbRL+5l+y1E2fzuR9hOp/HbxatBj1Zqh0y7stQpv75vD55Vj6Z0R4fHzN9AzwoKuQDBJJa2omwk5TRgCOrOnnuai9LGcmf4xiW+V4U+O5eABrG89hPukzLMFoxltNBO1nvvXPcc/Bud0iP4Jqb1uBcUr7yP3l2VUrZxA0ZsR6hbZyP95GR0PTSdiByUAGdsHqLrFieaOUPRmhKrbzLhOKnhnBEhNGKR70I5ro41gvEAJ6oQXDHJhdgVn7iim7olYNnTyZxa6xsWUnl21Gp1XhDCdsnDF4j0c7c+i7kA2uoC0MpVAgoxxSOeGp77m+e8XMGlEHRU9ybwz5kPe7JzH9Lga1l4xg1Gf1fLd81OxdkfxpRjonqIhXGEspywERgSQ2szoWQGyPlCI2GVaF2iIgER6UTdtXS6sp80ILdYewTCkYxrQGfbQGb4/VczFY8v57ovx2Jt1ohZQTQJPoUrRB0M0XOEkatUperYK53o4eKQQc6dM1KGT+5WfmhvMJB2S6J6kseGyl7n64D1QYcfeDM7GCPVXSWBSydygoCmCUJwgGC+49Lp9HOjOpakiBWurjC6BPKWf8Ak3jkYddOiZGeG1OR/zs3fvYMzCsxzaW4Jq07h15vcc6s+hfkseUatOwcxG6r/LxTWlE++uFILJGqZcL6ZtTgLJsRTLs/cuZ+Qrywgl6OiSjqNwgOieePwZGill0LPIj95oI+GkTiBRwlMaRfZJmLsl/Okq7rx+7i7Yyyun53N2xkeMeG0ZyfNaaT2Uzp6bn2fanx9Fs8d0eoIjAphPx5rT3nzNdjrCTva25zExpZlh1i42Pn4ei/+4iS/bx1JXngGJIdLXGgm6JcJxAk9pBPcxhf7JYWaVVNP0dDHWml5aL0kl7AJbq04wUeDLjbJ4ykE+Pz0O+2ELI649y/7DxWSXdtDS48JyzErIrRN1xjrDn126nNI3lxEt9rN62lv8sn4RLQMuImecqBYdZ7Ug/8ZqBp/Iov3hMH6vCQScnL+csasfouaGFYw+eD2GTS60S/q5PLecNV/MweCJ6R5Z+mJaPuY+DcOtnYQ/SqFvoR+13YpIDpH4jQnb7W3Un0kDVwRZ0XBttTD63nJ2VhTjLjPinefDus+OpVujZ4zA1ibwFKnoFhWrK4B+LI5bl2zlnW/Ox94ssHRreHIlDD5Iu7oBTRd0+2wMViRQuGqQwWIn7XM0Eg/J9I7VUZIDRAZNlCz30HJRPEOFEWR7hMI7Kui4azxpqytpfzeJ+4p2c09cbMW1qj4VQ7eBvF+UMbQ5n44zyX/Dmezvi/8OHSJrSpZeeN1/rw7RyVfP6RD9tZxbIfonpGD1UsLpYQC2zX8F9p9kyUV7aH9kOt6xIbTRXpKOh6i+yUnWNhVji5H6h8B9Qsa1qBXFEMX0nJtIUMGTJ3Bc3MGRJ99AP+DCroSouM+B2mJFOW5nMFfmlevfQ+gQuKmf4ZntqOO8NPgTqN+bTdK4TtTUML67Bgg7BT2jBJu6RiJ7Zc5sLGZx/jE6onH8IWMLW/uGU31HMjvbCok/3MPCF3fQfUGIORPPILX+4OR0mSidWk/8txZGPXOC+AcaKfhEZeH0o3SUp5C1JlZKPjQswrArqhma7yP7gSqimoylyUDlYDKJJ6P0TNAxeXSCSTq2Jpma6x3kz2pEtWv4pg2jbSgO3RHFNq0HS6dA/9c+AAJJAmeVzHWvPYpprwPr2L6YYrBV4vxxpwF4/6UX6BsukXh9Ew/fto7vfz+V9n4n5i6ZxPIICWej2NbEEbVrJO3vZShHoPQYeLz8KnQFDu8uIe+rAEQFa1bPZWPRJtL2B0k9oGJVwiy8vAzP7hR8uVHQQT0VhycfQvEan9/+AgXbbyfi0Ek8pqP4BaY1LqIWkIKC3AcqKUrtxtQn6BkPwWSdlD0S2VuimHt1fjp/K30tLq5yVBHyGbm7eQb+gjChd9PQcgMsqbgBxSdIOKTgqtWonb+SsEsn5UiEz+rGU7mshCMT1lD55EhWfnoR3WMVPnrm0piwpKyj95roHiehy2A8r4fE/QrhOMH80kq+P11E260hGn9v4do7d2Duhr5ZIcy9Okn7ZTasm47cbEaKwqG9JRSObKHtaBomU4RwnI5QwZAUQB47yJPdIwimRKHJwh0nbqX2cDahaifmkQOx5r1XdvJ5wTZqbjTg67WStN1E9qcyFz38IMPGNwMw1OTkyJNvMNDl4Mv60QRTonhHhjF5NIIuGX+KTtxdzXSXpeHNklBO2cn4TuMPk9aSeHcjjV3xSGGBpGhIdRb65gfZVTaSzC9kXNe2orVaUc0gh3XGzaxCDulgj4ImSHvNRNSms2LfXCzFA8y67RD9JRJx8zqIWqH+u1xatuQgf56ApUvQOTUOx2f7GTG8GWuPyrA1Qai3IgUlGq+IRzOAuc2AxRJGCwaJ2KB+WTGGdW6+mjcKgC2lG7HFB4gf082WtuPnnKFz/MNxziH6J0XqN1Dz4lR+2XI5CXvdLI0vw5+mU5jVSc7icpTtR8gb10rTAgkpKqDORiBZ0FCXjF5tJ+npejJT+0k4o2J4KZ7hbywjY9cQn301m7x1GkVvdCKHwV0dZdnXtxNy64SjCuWnswl5TYxxtqAEBJOSmhib30R0eyLxFRFUq07FkRwUn4QuwCBULrf5SZRtJBp9PHr5ero74qhfksy27hLszgC7jpcSTYiQ8eJBNLPOqcZ0pv70MJu2TWTgxWz0X/Zw7N/GY+4VsTykPh1LQoCarwpRTtg5uamEppeKGH5hFY2dCbTPkLFkewnFCURUoBkga0w7jVtzMXUq9N/lpXtvGtYqEz0tLvypsdwpe4PM8KsqUII66Rc3YevQODrxM6J2neZLdbYfGkn9Re9y66OPEi4IEHwunTWlqXRdEaQwpZu8d2ppvALsp7pj4n8hQfMzCsYftIfkHS5Uo07ErVJ7n2Dk2AaKL64mf929KL/upOlqlbNdKWz8chq6DNdMOYStRUIM96LaVZTkAB/2T0NuMVMwoxH3PU2EUqL0jhZYunSiiRHKThTSH7Rw6sHlaAqE0yJElvTx4Tsvgw6fNkyktKSFqese5Y7xe9nz9RjEkIInR4I2Mw11yYRToyS8XYb93hbyvroHW4uge7SBuPcdtD8RZdRLy2ifqRDIjmD0QPek2Hcyu7gTER8inBsi7BAEwgYO/esb2Fp1JKEjDBp2a4hgs4PP357P4OQQcQfMRC2C/hIYfmEVtuH9PPngh0gRqKpNI+KOMiKlg4hLw1UF6fGDfDzuPfY9MBljn4ylZACjolL0ahOqWSfxVSu5G3zsGb0OAFN8gOyvBOGr+2m8Mnad3rCJiK6SdFji8uoFCL+MIqskHZAZU9BMxzSJnvE6Vy/cS2V1OpHCQOy3MLODqElin3cYI+LacX5vwdQrYai04mgE89nY++ZLdNr2ZOIs7GeoJEzH1SGO1GejGgVKhxFLowHVJKP4BLYkPzcPO8j3708iFK/Rtz+V9IuaMPaDrzhM4MpBolbwDNMZewzONKXhT5RRatog34dICpE2pwVTn072U/sI1DrRp43B3qKRtc2H+4My9GiUJzpHk7flTk5NXUXC3eeqyf63ci6p+m/GuZDZORj2yH76bp/GoWfeYNyh65C+cTNYqFNzwwqK37uPgk/7aD0/Ac+YMM6TRrzjgjiPmJHO7+XPY97luvI7WJh1irUfzEWokLWonorWVMbnNFHRk4y/2oUcBHsLhNyxEmXFD6F4CJcE0LtN5K6PUHeTwNxoJGVGG4/nb+KFxgsxyVE2Fm0C4MrqizhxOgdzh4ISgCU37+D9U1NJ2mimr1QQzgyTm9lDe1k64TgdU59EMCXKjdPL2Pz6TAKJgtAoP8bTVqQJg5yauoqi9+9DisbaWJz38F4++246R695ieurr6YvYMVTlkzEoaMmhxEDBsZNqOHoyQJsDTLP3PM+//L5reRNbqb6dAbGVD8hn5Gk74wMZQjCLh3VoSLCAnRBdmkHTe3x6BEJU6sRzahj6RBEreBs1HDf3cTZqgysjQZcszvoOpmCpujEDevnloIDWKUwH/7qMgofO8Ou08XYzxpJv7iJLaUbKfzoPqLJYeL3G/Glx5Sv0lLxngAAIABJREFU0SH1gErEKjFw7RDRiAxCJ+I3QljCVq8QSojZ6VkBrEesBJN0Iimx44horL9d6uJGqg/lYG+IhUZt17UjhE7TqTQ0s4aICpL3C7qm6yTtl+iaF0EaVMjZFEVEdcz1PXTPTse7cIhIow1nrSDsjH0PfHkqrqwBorsS8BZFkYIS86eVs6NsFJojypMz1vPMl1djKvLwQOlONneP4NipPADc5TIhl2DEpZW0DsUhvZVEME7CmwfqsAA5b0m03x8i0Gon6ZDEwEIfer0NNSNI4lYzB37/BiP+tIzImCEctiBpDi+N3+RRcGktXxZuIX/dvZi6ZHRFRw4IHE0awXgJJaCjGgX+NB05LAi5NTSbimyPkvmJQs+dfiwbnfSN1Pn1Jes47c9g3++n4E+KPX9eec8u1nw2l0BxkJLsDobCJtp749DbzAwb30x9WTaFMxoI/C4dw0CQUW+fodKbQvnZbCzNCrmvn6bljhF4iyNg0DF0GtBlkEIQyQ7hcvtwmkNkO/roudJM+7XDCM31YDOH8RxP4DfXruG95hn0rcsk/ZsW1u/9kuLv7qDwDyEMr/ZT9V0e+R92EMqJR1MExm+PsKX1GAvyplD97Li/1TT1d89/S8gsOUsvWvLfGzI78dq5kNlfy7kVonNQ8+JU4leWMeyTpURUmTVPPIc538t5N99J1KZTfUs8hiEdc5ORGTce5b2Z7zOUrXF9/mGuOHIvPc0ueiM24isiSFFoXZNH0kYTp78uJnrYTdHztdibYeGy3fgKIgwVRPHN9GFt03FvM5N8AOqul7A4g1gn9tDYkMRTT92OQVJp/iqPJzpHc2fTTBoH3FjaFOQwGD06ftWI1GCh47woeb89hGTQGPgiA0cDOOoljOP6QdY51JtD/wid8Cg/osmC0QvW9U5uapiLuUeQOLkTb5bg84px2BskJn78CN8Uf4P4OJFQcQBXBThOmFB8gt5n8tBlneSLWninbTaqSadrXTbDRzdhs4Sw1JhQr+7F0qMzb/ZJZK+McUCicGQLu0Z+Sd0F75H+rYIcjDUwTb68mdP3L2f4A6do/joXdEHykTBtbfHIQYgv6iNUlsCevmH86b0rMSztYO+ukdjcAYJJOlenH+XelmkYCrxYak2YBnVyZjZxx+Xb0BRoWRLTUAp22Hhj8sfMy69G6TFg7lDwZ6gYBgWWToHlmBVvQZSiqQ2YmkxELYLBYhia46OqLYWMnVECc734MgRtR9MYCpkoWjlAam4vsk9ioFigeCTkiI4rYQh7k0TEJlN/hYGET/pJvaMeccLBhXOOU3JTBRGnji9XRXKHkL6KtbVI3idjzx1k97ejKVo5yGVjTvDbPZeTtl/FucbBVEsdtWsLuX/2NowJQfonRcj4zsea/O14v02l5VKVnlkRrGP7iHoNNJ9nYnpmA7YsL4EkgXuDlZyNAeg2EVfjZ/jyZaTtD2LdayfpaRNN6/Mw9ep0+u2U7r2ZwtJWcuY2EnHo6Ap48iUiDsi8uY5jv1pO5Z1vxJz7gMBVboAOEz2jDETOOlGCOppNZUXdbNbum4wnW2LiLScYGB1h1ZlJPHnbKhxHzVQez6Z31w8tLgTUHM2i8vY3MMsR6q9QsL/cyeeHJ+L5QxaWVoVQaYDBC0vxp+mIoIzzhJHMnRFGTa3hlevfw2iJIIROY10yZd+PoPtdFylXNaKdcTBwOoG8r4b4zaZrkZ6K5+OfvYC+Msrc++8j/zWd6ItDRC72oA4LUHtrKs++s4KdK99haFMeQ1oQPRT6W05T5/g7QQjxhBDiqBBivxBirRDiv4ydCiFKhBA7hBDfCyGOCCFu/ndspgkhyoQQu4UQB4UQC/4dm8uFEId+sNkrhJj4F58/JYQ4LoTY9aNt0197b+cconMAMaeo5oYVxH9oY1nOTPytdprvjkBSiOTDGv5UwXWLdrG7uYA79t5G4jEYb2kgWOvE2COzb8VEfMsGkYM6SlDn9t+sp3BBLWMvPkv1IwWMvuMUm5+fjeyRSd4rQ6OV4KUehrIFfVf6iT+kwHEnntMJyB6ZnrHQsCOXiBU+PTCVnRXFDLbEIcYN4s9UESoc6svB2iFI3abQuXYYuiaQVD2WPKyD12Nh+NOtiH+JQxdg229l5txTDI6K4E8VWOQICWciDOxMRTPpJLu9fPDgSySc0inafQtx1T7c35npG6XjKY6y9ebnaLxEwtqo0NQVT+O6fHJHt2Fe2MmZk9kE9idy5ifLmZVeRyBJsL8th/TRHaAL1pd8wbBPY2EG7w2D+AvCZH8zSFOPm+kPL2V33TCEDgmHZNJ+U4vSZeCmK3cyJaWRYIrG0RMFXHbDHlqOpiMHBP42O5pB59mdCznz+9HMyq7F3qLTfmGUwfcyWbFnHkrBEFKbmc03P4frtMSDb99Lb8hG/CnQSodYMO0EgYIw8ZURQvE6w//YQe+bOVi6wTs+yLRZp4l2W8hM6seTq6BX2wlkRLE1C3rr3LRcGE/3mSSGrR4kf1YjUZuGN1si4SUrckCnZ7SMbtQ4sn4knxR8SbAoyNad4+h4Mh9XJbjLJdSQTOYtdQyNCDH6Jyexr3ISidcIpto49fhoDF0GBgoUpj12kJuP346lR+PL1jFkvaEgvApXvbON2eWL8I4KIQwaqdsUShK6sLQYKFjdx6mXRzHUYSfiAG+WRNPFFqwdErWLLQSTVTqmmPFODlD3qCCQrONoiWA1RNA0QVV1OpemnMLeJJE2p4VwqZ/0vQFO1mcycv+N5G+7A/fsDoQK/jSd5IMQcutY2wT+FAkEuH5pQreqqCbYdqoUc7uBhA0WVvz0WoKJsaa3/uwo1gPWWA+4+CjFK++jcWUh9kaZI+X5DP99N00LYrIB9kMWOqaDHARbtofMdU00LJSp7k3iFy/eSSSsEN2WSFKZzOzZ5SiySv2+bAyeWFVj9Q02dFmn819C3PPYQ3R9koPBE0XsO4F0XjO+C0ai9prI/XUZv86bxLhnluFaGsUumal5cerfeJb6J+HvOGQmhHgAuBmYrev6VKAe+OK/GGMHvgVW6bo+C7gCeEUIcdGPbLKATcCvdV2fDdwLfC6EGPUjmwnAJ8CtP9g8C2wRQqT+xSkf0nV97o+2i//a+zvnEJ3jf1KweimtsyWEyYStUSbvupOYT1vIeLAGd5XKhztnY9gRR/pXBronwAvTz8PUJ4irBXdlEKMSJWoV9JfCs99fypn2FAYWQsKobrqWuLjisR38dMFm8pZWgg6haifOOp2wJ7YiYW/RST6ikXQUEk4KTH3w8u1v8+y8P1N34bvULXqTpHetmDtlrNd2kGQeQhfQO0ZgWOfGftxM35QIwdF+AlN9iD4j3RfkkLWiAVtLrA3I7rIRFP/0JIEUjT3fjGEg38BFi/eTtT2IbUEdS1Y9hKM5RPxGK8EkM6E4gblbQoQFF77/GMVvDZJ1QSMmc4S8q2pRNYnommTiyyWiNp0RZTeyYfdEctd0EimPo+1UClqRj0VjL2HD1S9yU8NcvL02Lh5zis7fqFTO+hDPEi9jslpQjaBe3s+liScpfL6KkZYWvt02nuxNUa6bUcaX62ZiKvQghQFVsPTCrchxEVrPg60HR3PXz78ieacBb7YEio55p4PEUV3s8A8j5BIYpvVR/l0hnjxBJGDgxB/HIJtUmhar/PGajxiYlIYuwcDoCC6Xj+6700neL7A+aMCTD1JIgEFjcLjK8NFNTL7mJEmHoXZxHFVHs7E1yegCPI958RTAJZfvx5nmJWleG6fCBn46fhdqcpjucSa8WQJfBlirTFRvLgBd0DTFx5VPbsPSKiM91kX//T5MAwL/+AA7355CynOxyqzZKTX0PuxH8QmaQgmxFZawhHuPiYz7aqj8oIRIqZ/6axLwXONl2KowwdQoUgQK3u9Emt6PpV2i7po3iVp15hZWQ4ONCbMq6Ssx4gmaqZz1ISIieLt6OiceW059ZRq0mQk7DaRvVDDIKmiCrmMpGAcEhlIP3eMEV16wn4HxYXzpOi/N+5TKe22kblVwNmikbldImd5Gx/wo/ocGePH6lfTdOQRmlcgsD3JKgBklNWh5AQw+HXOvjtAFDUvSScjvJ/FkgPjKCIpP4K7QGeqykbm2FyUpiL/RiW+WDyFAioLRp7HjTAkdrW7C8SpDw6LkbArhOisQEUHmL6L0XBPAd+EQU587ROPT0xhaPJUpTx6i8P4DbGk7DkDy6/uoeCD9XM7Q/0H+XnWIhBAS8ASwXNf1oR92PwdMF0Kc958MvQ2wACsBdF1vAVYDv/qRzYNAra7r236wOQbsBn72I5tfAFt0XT/zg81GoBP4yf/anf3fnHOIzvH/ovrZcaQ/v4+al6Zy+v7lTIhrQo7oFL/WycCIKEID92mB2tlFwtx2fJd56Hw0SFt9Ir5pflL3ayDriFobc3a3YDOGiaa5ef+b+axrGcfAfB/GAYG1TaCaQLZGcVzcwae/fY64nTV4syWct7agy/DAqrv5wyvXk7fxbt4dTGXXO2/z6A3rSLV5OLuqFCWgE3HH1Kpt7RpJ3xmwHLUS8RuwtEvk313J3g1jmHz1ScJTvAz71Eflq6MhKYS5G7wz/Zy+p5Sm8814bphK1KrTM8pCz1jonGiAGQMsvHYfulXF1C94fuNK9CcS8LfbOVGXiSxpRByCwQJwnwX1tBNzt0TLH40YRg0iVEHhI50Mzcjjiv1L6QtZUXoNDIQtJP3BxISn7mNuVg2N7xXy81vX4Kl3caOjl7PP5vPYFzfjqIe22Qa+rB2NrVVHlMVh7tPJ/lblq5YxKLVm3Nn9mHpkPmmeTNecCNLkAYgKTB4NVZN4/ZVFFC6oxXfGjSgaonB+HXGHTaTcX8e8YVW43D6e+OAW+ktk+ocL7NUG1J0JVP7MSt8oQcNVSUhZPnLmNpK020D6TqjZncuRT0bTfVkQBCQfiskY7L3/BSLfxOy/Wz6FeJuflhNpPP6Tpfy5OZZ/4stRsXbpaIV+5lx1lECmCmGJng1FvF0+E22cl8hrqSRfWUkgReO6EYfRZUHtNWZ6hmysOjCVoUo3QoNv3ptJWlkQU6fCwHCdio1FGBd18f7U91BLfGjlccghlfnjzmCc00PjNan4K13Y2nTyvrwH4+gB3s3eg54dYCBkoXRxBT1NLkrfWoa9Qca40UVTdIjS59ox9UqkPV5D64Uagw0uMr+USTyh4y2KwsE4VIfG5sZSjG0Gti5+jj/UXIS9VqF/kY/+4QJu7qa934mhy0BnYzzP/vxW/A1OUrYZ0E86EbVWKt4rJTpkwJMnEUwQFJS0YRgCuylEzQ0mbI+3YG+CgWJB2k6ZbVUl0GhFGRKkxnuw77eQtrOH3pEyxlYD44oaSTogY0n0M5hnQrq8l8KPvPRNSCDe6ePsjI9YvWcaIioweFVOjtfxXzUFgKoVk8+tCp3jx4wGUoDD/2OHruudQBNwwX8y7nzgqK7r2o/2HSLmSFl/ZHP4L8Yd+ovj/jU2/0ucc4jO8e9S8+JUapes4KL0sTwSX0H3aIVvdn9BxnZBf7HMwHCd5l9PJ6zKaJrEkMdCSk4fVmsIoUNudjfheJW3tp9H255MWufaKfi4n9CqVIYuG4u9RcOXqYOAkowOCl3dvNR1HmO3dpN0LEzfZ5loBkgrixC1gTCpvFY1lye7R/BxyxSCqgHPtADBBIGxW6F3vMqCx3cDYG/VsMQFSToeRhEa9mndnHp9FHEb7YSf8WDoV7CesOCd6eebGa9TfaMDZx14rvYiRQXeXA1bi0RcnYavycmGtdOx1hrRFLjriYepv9JG2i4wdBjp3JqJa2Ebqk3D4NfIndmEEgBffRzlUz5BTQ7T/mYcgzkKFxZUMvBmNnpmgGNtmQSTTfgyBDOdVRTcVcknN1+Ms0Zi2KdLuXvKbqQQOK9tI2LXUSsc+FMF4Tgdy6JOGq/SuSLzBOHsMH3NLhjhpaXbTdxJI8mOIQz9MgOX++hudqNf3E9NbyKmkkG0WjvP567F1qHS9NEwjqwcTX+3A2ViPxGHjuwXFCysxVMUxXbKjOssuGo1HLYg+vxW/KkCz00x1XHXwjZUj5GEkzrK7Z3cfefXjN30AL4MHfsuG72TVJrK09AVnY5pCv2HYqkGo0c1cN6yMlLWmNm2dRwpewSXTTxGVJXRm63IskZfiULt81PI3Rjmk5OTGBgTwdwtoWoSU0bWkjCym4hTI+F0iHCcgr0J8ke1kr2uA+cldawfHE/BHyJosk7rXAeHV4/GtCoe07ReTH2xRHYpLOHtsTHhyGJcWy3Ul2VzYmsJjnQvxsFYD7yBeUGuKb+dwRUKUavOtUmHyc3rYvLEKgZzFX73zNsgdEourcJ9UiLpdQtx43uYv/0hhnam4MvUMBmjuM/odPU4ibZbyZnSApKOL00GHfpGCoLZYRSfQDULlk7bRe6CejI39VBXnsGwxVWEVRljkp/TFVl48iFtWhuexV70PiOqWSeSGab3uzTCTjj7cBxpe0KYewX1awrpnh4l0GUl49Y6At8n0v10hLhVB+hod7PZb6Lu6jcJ5YcwbTrEnVX1WNcd4KH2iUj+c38PfxP+fkNm+T+8tv/F/o4fffYfjfv3xkhA7n9hkyKEsAoh4oG4v/Lcd/yQO7RXCPGREKLoP7m2/wfnvvHn+A8pWL2UmhenYhAy2U/v41ddo+geJxEYEUC1acgB6KpORDrqQI9IyCsTUDa5SHukhs7vMpACEqZeiWB2GIMH+se66BsBjopBescIjIMCS69Gj99G0xNF7F49gTU7pnPfn/6M0auTsKCVpgtlHM0aSdtMhA7Fs7Z2LB3fZ3DmcC4lj3cStetEskK4Mjys2jSH/uEQdgoCvRbG/NtxTFKUdLuHoUyBvTWM/8N0bM2x7uJat5mf1FyHruj0XxAg98EB5CDoEhRcVU33RNANGmG3hj8niqlfx3V6kMRjOrokSD2g4suL0v91rIlm+6Iw9WXZGLw61lwPs8sXEb/HiLIuHm+hyvet+QzmS5hOWZGOOvBmKoSSovz6q+s4uruYxosdsQagEcHquvHEn9FpOZZO0hFBxu4IugTpU9rwbE1F6TWwoXU0hlYj9rQhEj6z4tphBh0GP81AzQqS86LA2qTw7IgvEEJHkTR0CRateIzOyRKKX2f23YeQjCrsdhO1x1ZuAnM6sbQqpD9XRvfcMIN5Ejlx/XT9ZDr2Fo2hQQumPomWHhcAoThBWJX50/pLuHPK97EVxGtacaV5MPVJkBjiisvK0GUQskZFZzKbPppO8M5+Rs6sYShD4tuvJ+GvcIEGFmMEU7+O6lRpuFvDvdeEpcmANs5LoMnBgfJhzEmrwZblpeXeCPf+cS1hp6B9axZdc1MoPqSwbss0Oma62HDL8yixxvJIUZ3BQSsRh453no+4/H5+NWMjs9LrkFR4eNF6zt6znGDAyKW37GGwNIrUbGbAa6HzRAqRnBBPfHYjkRWpXJt0mKEJAZ677gacFQbKdxXSP1yn/hoZbV0iN44/wNCwCM5aCZMhSubSGrSAAjp0bcjinfPfw+jRKflTOwnlOuZmIwsX72NwUpC6QCLtn+RSf20iukGn/U/DcC2NYt7jwNQlY+4WNNSkEC2Pi6lLT6ym4F2N9O8DZO7wIXtk6hdLBKYO4ZsRUxNP3yk4U5aPs0FD/zqB9kem8ebsD+hV7YT0CNOLaum5dxrvFsUq+DZ8d64o6R+IRCHE4R9t9/x/PI7th9e/zK4PAVb+Y2z/wRh+NO6/svlrz90EnCC2mjQTOAMcEULk/SfX9z855xCd47/kf+QPrDowFVOfQPMZENYopkGd3A0Rctb3krlRpu08HW8OHK7NIfVgGN2kx5SWwxIF11fRNzKW2NkxNx50eO629xj3xFFmpNTRMt+ILoGtSeLJj2+kc0bMGZCDgr7hsSf6hDMqqirhmtaJsU/ihh0HMPUKXPtNDLQ7MQwKps0+jdGjY241UH1pAnsa8yk/lkd8pUrHZBOBJIHRq/PVA39k+JhGapuTcdbIiBYLbZfnIDSBvVmipi+RlZevwNylYG+UMPTL9E5UaVjkxnZPK5oCfcUK9Ze/RSBVx3VagS4T0Zwg7utb8DU7iGgS/SN0iu45CwKSXrYw6tIK/i/23jvaivre+3/NzO7l7L1P771RD51DR5Cu2BBU7B2MSTSaaJrJ1SSWxBrBXiNYUCMgCigg0uHAoZ7ee9+9zp55/tjGa+79PTfe32PKc5/zWmvWOmuf7+zvzJr5zn7Pp+rLBwiN8eMqUtANSGjzPUhBgdUrduGb5cXWAP56G71TQLZGMQxG0f6km1BJgM7D6QRSVMQcHwNfpJH1WQifx0DXDAHvIi/u4ijZ19Vj32dgYJQZUYZ7Tq1Au8NG6Gg8UYuCoofM8Z34U0WOPjIJ2wEDactaGTGinaHZQQqOGoiO8dK/uYj4gzoSqmSqe1NwFyk8+uBzqF/9sKu9BkRzhMk3VtJ/Nona69bzVs1kNCPciIJK6Eg8l674El29kVPXlBJOlilM7yPk05G2rBXeT+Dk8QLyLmwklCIjJ0aou3Y9rtMJ+NIFnj/vNe4ZtxN/mkDYriBUWlEMClmfwvufl5PjGGJ50WkeqLgQw6CKqUvFmyngjBgRZdj5k8e46qF7CCaCuyxE0C6ihCTCiVFM+y2Up7Xwm88v4sm0YwQvdfLRVXP4ac9Ykj400uxPIH137PFYntNM3p/9lPzej2FAIOgQuWfvSspy2ln46n48eVHCGRHECNjOafAt9rLl1Vno7UEEBZwnEzlxtJARRR2MmdBE2AYvd8+mb6JK/U3phC0CoYQo296aTnKSm92fj2NoRghTt4o+2U/wmkFqfxsfK8xY6sWXFyWhQiJxajeeoJ6jZwrwZOtpW2Cgb7wZMcOP1hom74pTRJx64i+o5Y3f/wFzq0DXIplVd3zG6EurWGiKsNo6wMWLrmFotY2KB9azvbNy2E32T+bvEEPUr6rqpG9sL/zVfILwkCAI6t/Y5gJfvVqg/w+HrAf8/8Up+f43+/CN/f7WmG81t6qqr6iq+gdVVWU1VlPoYWCQWIzS32RYEA3zrah/vJzi246S9U4LB5Y+TuH6KL4MgZ7JenrL47HWOr+KQxBQoyLNF0mYWiXm3nUQQRY4ub8Ic5uA1iXgLIsQTQtx3+lL2f3OZD7aMwVTl4DOpWLpjBIp9ZN0WETrFog/qyKGBQILPGT8qI7498y496YQyJI56s1DE1A5/PM/YujWEDWqfHmuGG+mSMaXQeqeTEFTaUE3JDK42ksgLYo3P0r/giAXnbiF5sF4Uj/RkrC8HTECvtlegqkynnFBLsk7xfXbbiNa4iNiBjlOIXO7wMSF5+jdkkXkikEicSoFu24gdUI3+mW9aDL85L4i0vlZFqpRoac2CSkocOhgKYYUH403wpG6PEIRDekJLozdInJuEKM+TN6zNby5+TwsX5hxzg2iGFSKX3cTf0LCnaOh89Nskrfpyf/tSQQFdDqZqEGlabkOkyWElOZHf9CKlBDCf3cKKXv6cBXDlEtP4W+3oAlA6cI6coq7+cHKj/BuTMfaGmX6fYdxjlJo2ZdN/aEcsjZq+Oyz8Vg/MyNtSkCIQs9ECcsWK9YGke+duhKASJyKJs3PWzNeYqylnfrV6ynceDtyg4VIXRzNFbHO5x/Ul6GKUHObjazsfraP2Iq9Qs/AxixUCRpWPsdgwITkkTA261g6+xLiy/rI39DL3adW8vwfL0IVVew1AufuWIdgkukfpSFvfAed7jg2f1yO2RRC61NwloBsUagdSib/3SGWPHAPg2UKqUciaLt0XHfXNjK3SYh+Ec/UANv3j+Pm2XvI33kj/lo76c+1csqVATf2UvN6KaE4Admi0PToCOqvNNC2NB7/FD8Jp7ykfq7hzIFC1p2cg7VJomnxS2TujhKKh7yfB7B0KYS8emQjRFIimPNd9Lydw8mqHOQSP3ZdAENvLBBfNgqIYQHNrEF6muMBMFUZCMYL6Pda8VYmoBKzfKa+bkAMCgyOVeisS0LY48DcpKFvEoybX4OrPIjREMFwzMyttY2UlHTQtmk0z/TP5cTP17Fy/DE8UQMb8nZzW/s0AD7Z8TZKTx/AcPD0P5vv2l327VxmjwJZf2M7CDR+Nf4/ZnWlAg3/xfc3/m/2UYDmvzGmW1VVv6qqg4Dzvzv3V6KoCSj8L47va4YF0TDfmvrHy6m+O4vrrvwe0sk6UCDz/FYSXj5I/dXxFI9sJ+GcDGERY6qX3CVN7OooRpMQJG+zH4SYlSdpn4biJ0IEa2zIZkjbr+LNjomhgcv9WA6aEK/qxV6n0DcRTN0q0Rorp7aV0j9GIBSvYkgIsPP9KXhyYMZPv0cwO0w4QSFjm0RksgdnoZ5IQEsoUaHqtnWcnfYWjSueJ+mIiKXCCHsdGLfE4coXaTuSgbFHIPs5DcVFnRhrDGw8NwnbOQnTQTP+wjCqpGL9fhsNfyxFPG+QwS4bhj4BocvAoM+EZ18yYY+OwVI95pl9LBt3irgGEcMIJ0JakORXjSTEe7Ha/YxI7qH/yzT8WTKrRlfgr0ik+vEcCl/pYvx1pzGZg6galRUbdzH2xjOwcBBvvkzpD84SOG8Upk4Bca+d0zf9kWMrHmdZ3lkifh0JZ8PIQQ2eh/wocUYK3naz77MxqHoFU69M88ZCBj7N4LnnLmLymhP0TBX4YM9U8kq7CGWGUbQq9z79Jncu34alU2ZwUYDBcQrhVBlzdwTj0h6CAR2GhACXLjhIzaw3WHvmKh7fsYy5Zy5GjAgYBgUi8TKyTSZUGiAU0CJEAQG8IR0P9I3CP8vL4HgF0+XdFL2xBt+WVGx1EEyJEv/GED2t8Qw9qRKttJGyooWoAY79ej0TK1ZSlNELgPDLBIY6bUyeX4Vus507H3qXSEqECVPrMGhk6lc7KLqpGkEWcGdrKHz4LOs3LGPlg5/H0LVCAAAgAElEQVRi6BfJ2qhBn+llU/M4EEAMw/7tY2kejEf/ewfO2UEGxypoEgO0L1bQD0r4CsNI9Ubcv/bTsySM1i1Q/JAXd5FM3uZbMfyoE9mgUnVXPP1jBYxxQWQzIAucmrKRoRkhLp18jD9MfpfPdo5HtqrYCobwFEaJWhSiX8QjBkXEcKw3myYI4TluwhmRWC+2BBVzVR+GPpHkIwKI4Jvi57c3v8aSmScYCpqwHTHgbbRxybVf8EJpEbVnMhmd2kXlQCbjHl7Le19O5aHk0wDsPDz267X9Sf2BYTH0/yiqqrpVVW3/G1sIOEUsq+trf+pXNYiygc/+iyl2AhO+ylL7C5OAA6qq+r8x5j/6aSf9h+/97G+NEQThqf+P+TOIudL+JsOCaJj/Ng2XGVh1vJ5QYRDNtQreleUUvjlATW0G7fMFdP0S/j4z5fFNDFUlYDxipmGtSDAe2pcoOEdA9ywbADeu2I4YVlEygsjf6yd5g5HUS1oIhrUMjhLQegUGZ4dQNBDIiZD3kQdVUtEctrJi1Rckj+vhyO/WY2zUoYoqnbMFzMYQvnQBe4WeS847TP5nN3JL2wxecyczOBKievCWBRFlKFjcSOFTDSgaaFugx/VqJmIE/m3CZpzjI2h8KqYGHVpbiIFXcnDlifjOObDWaNF6VOLPgHDIRs5HAyCpBBPB9rCFjyvK8EwJUJrYi6oKtM/V4D6ZgPltGwMP5SJEISV3kK6QjdRDEQy1BpzrRA58MpbgOTtNy1/g2bo56EQZZ2ccSdlDnHxtNK2LRVxjIiRf0EbZ+juZ9/A9KKqANKBFjCiYa3VoRQVniZmaW8zU3LiehGMauq8PIi0bQJDBNULmk+NjMOR7SB7Rx8BHmUwsbgbg+5uv5/G9i1j+yOeovQZ0QyI/mL6TluUSclSCZhNJbxm5M2EfTw7lot/oIKFoAHdQj5IVJGRXsST5SMp0oq8xYj1iROOHkntO4mxy8NaO2RQk96Nxibi3pmHqEUi5rIXAQg+WZokDFSU4KiXCskT6vhCR36RiKnYy/9xyvH4D/X4TskWl7notqHDgaCm2xhD3778MsyOAWROm63AaqqTS8XARqiOMqwgaXswlkBfm8QMLcczupnWJSDisIfXabn42eRuKBpQSLxy2seTJPawYdQKtW+TOMV9QetdZpADk5/aSt2mIInsfuiYDgeIQfeWJpO8RMbVq6H8rG0WnonHHGvTOy6kjb04zolmONYJt01O1NJG7Dq8iEh/F3Crgr4xH6xIxtmuYfWUFUlhALfUiRAS8EwJEGqxoTWHS/3CY1ENRXBNSCCYq+Fe4sJ2VmJNfz6/+cB2ffDme7SO2MuHqUyiOCBs+mU3Hj6dy+azDnN5ZwpDfiDpviLFlzV+v48bLngdgwrFVw2LoX4l/0aDqr7LEfgusFQThLzE99wAHgF1/GScIwheCILz2jV1fB4LAdV/9PwO4AnjoG2OeBgoFQZj31ZgyYDYx69VfeBhYJAjCiK/GLAXSgGe/MWa5IAjLv3EsVxML3H7x25zjsCAa5v8XD267lJlF9Xx8dBu6W7qwvDBAwdsy2Z8qZJZ3MPJ3Xbz36jyszTFXmH2vAUu7ir5LS+m0JjQBFUR4/pOF9I/REHfQiPvzVDpnSFyRfpQTk98mbVonlHpx7NcjRiD3A2i81AoCKFp4s3Iqrj0xC+rVKz9HNyDxy0UfMNhhJ1QQROtR2fnaNPJeE9jbXMATz69AINaOQvVrGBwNdp2f6scy8RTJqBIMjooVdWyLxIOoErnQScihIjSYkE0QNahMmF1D5if9JB0eIuQQkILgHmEneZcOU7fK4I99iOYI2W9K/C7rIzRamV9e+B7mdhj3o0qaLxFJP7+N3l4bh7eOoWtazIqiqAKKTkU/JFD41hrCBxJI0Pr4wewd2AxBnNNCXDL9KKJRZsBnIpApY+6J8mF1GckVKkNFerIXNTM2vgPz9Z2IcRHKHlmLNwuS3zLi35+IPNuF1imRsxk+mPgCTq+JzMua8Eb0WFpELK0ioiXCi1UzsNUIpO8N83LtdEpe9OA9mkjScYWOuSILXvoxGx5bgjtXxHsoiaF+K8qQDkWvohyx4w/puOOqLYTjQArBJ42HGFnWguMcKHc7sNWBMn+IspVnaDiWjc0cAODmOXswOFWEzQmEHBoeePFlxiR30VqRgXGfhcDhRGSrQsoeCX1CgOIxbTgL9BjjggT8OlYkHsXQL2DqEumcJZGQ4KX4N+ew7jAjejXMHVNN9I1kLig/jthqZNu5L7jJ1k3dtespTeslbFd58fRMPn57OuEkmdefWMrAyjLkiR6c72ZQf5+BE92ZGHti68BdCP7VTgwDKvFVftJG9pI/sQ29U2Db2dHE6/2kJrrQfWyn5ob1VD2QC2qsAKd7WgDZFPvFMvap7Ng5AWsjGA5YsJ8T0TQbkIICSpsZ6xcO2paodMxXETP9yLLEzd/bwsmXxjA0QWbJrBPk77yRUksXI3/Zw1VL9iKG4cySFBzVCienbOTklI20vFvwn9ayq97xD3hiDPNtEPjXrUMEoKrq08BbwD5BEA4Rc0Vdov51DzATsbpDf9nHCywErhUE4UtgC3CXqqrbvzGmDVgK/EYQhL3AS8Dlqqqe/saYCmA18MZXY34GLFJVtfsbc/8M+OFXWWYHgFuBhV/VNfqbDPcyG+b/CG2mj+qZbzLtnttxfFJD3FaBzscKcV7vIXLCgcYP9oYo3utcGDfa6ZqjYEzyIx2JI5igIihQ9GIXcqKV2psMSG4NOrfAuTXrmHbyMpyHUtD4wDCgklDppuGKOKIWhfxNEZou1KE4IjQtepm8j2/B1KglMtZH3Ocmlt25l41VE5H7jZQ+M4D/GZn2PgffG7ebl19bSmiil7ykQdp2ZROxqaQdiNJ2gYL9uA7DhT30D1nJXQcNt4roDBEiHWZ0QyKCApY2lf55IbRterQegWCSQvxpAeNAFGehBl+mwvQZ56j940j6JkHe5jBt8/SoWjh//gk+OTYWU6sGa1usN5a7KMqa8z7j7WcW4pvvJdxnonRkG7XHckg4BQNLAljNQQInHagiSCEh1k5CgGicjBARsdZJBJNUjL0CWq9K/6wwjiM6QnaBrAUttO3MQdHCqAW1dHht+HamoJk7QPhAAqjw4xve5Q/rV3LyJ+sofWkNEYvKuVXPsLz6EkbbO/myu4Apya3sf3Ui3myVumvX/9V9MOXE5QhvJ2K5roOms+lklvagfSSexqsFbCdiAfP+DBVr6SCBYwlIITB3qlx07y7efH8+EYuKFIr1XlM0Ah3zwdghEUpUuH7BHl7bORdbrcC82w6x6dgkLHUxERlIUZHyvMzKaeTFrP1MOLaK0oReTvem4euwYuyQ0A+pOMcoLJ1SyZlfjSXxviYAup8sIKoTGCoVGXleHXWbi2KuvaIISQc0DC4MgCqgrTMSygtirDYw5oJqevxWIlEJ77ZUfvG9P3HvJ1eR/2EYZ4EeQQXdyh6U15Ph2j5mpjRS5U6lqj0VfY2RkENh3ozTfF5dQs4GEdtP26jam49hjBOP24j5jIHMbYPUrLGxvLyCow9PxHR7J863MpHCKpqASn+ZiFwQpOQnvTTenEPErGIocCOftnH1JbtYHlfJWJ2Bsb9fy6l71rEofRz9t05jcEKU4teCbP/gja+v27Bl6Lvju+hlZk7KUksv+W57mR1/cbiX2bdl2EI0zP8RkXYzBW/fTnCVk+4rSnk7bxeWXdWEztixNSjo3CpxFZ2MS+mgpxx08UE0B+KQJ3uYOecM+VNbabgundobjRQWdqMkhQkmKow6uBpvUE/6rHYQIXSRk665NiytApk7VRIfaoakEGPyOyh493ZEj0Tq/HZMB8yY+qK8cWwaSrsJ4iI0r0gm+lwK8XYvBiHWroMmM227s7nuip2kHFF49PH1aEwygqLS02cjwe5l3NMnUWWRsF+HuV3kgdUbSZ7diS9DwGAOk3Ba5ZFbXsFeIxAxC9z6h/fxTgygHxSRFYnIFYPYqwVab42CAKOn1zPfdi7WRyxHpm8SoELuRzJvvLYITx7ILRZyirtp2JeDolOJmGNvjcHj8VibQLaoKGM8CNk+pGwfQlACJVadWCzxoh9SCcYLGBv1OKeGiRog+If0WB+3KJz9rJiDZe/HWllUxWOd04MvP8JvNl2OpzBK/nu3I4YFFs2qRC9o2T5iKx8emowvqOPQ8xNwlQf/Sgy97IpZ6I6Mf4++SSpWbYj0L1Xaexy0n6eDiIBsBq1fZfKMavwnEhizsAZfrownW+DFipmE8kKcN/sUhtFOImaRqE5g9JgWNAEQogIbN87D0iIStgls2VqOxhJh1TW7OHXPOqbNPkv1zDfxyToAjk96hw15u4lEJMSAiK58EF8m2M5KfFo3gpIHznD8ZAGyIuHJkuiZqWJpVznVnsH8K4+giqDr0yCu7ENTZ2L7zGcYc34NVlsAf0EY13w/s5PrUYFLbtrD+30TQRN7qXQVw8DsMP0VKfjSRQLbUvhoezkt2/IoedDDHVduQecUOfxOGcKAjpYrVRo3F6AW+gmfcDAyuwtvSZjqO63YzkrsfH8KXXOgbX8m7nzoXxrEnSuRdELBYfPRviInVlIgI0CwIY7kEzJvvzWPze5xsbU5zUN55Qo6752OpStK8e1HeHfT819fu2Ex9C/Kv6jL7P8Fhi1Ew3wnWFpFbrhlG+u2LCHvvoMELppCxxwRJU5G269F4xXIfaeb3rkphJc5mZjazrGuLEJBLYoqQKeBnE8iaH7aQ9uubKQQJJyN0D9Wi2FAZWB6BHuFDucYGdESgX49yUcg545ajh4upmxSAzopijtsoO5IDidXP8WYTd9Hn+klFNCSmuTCsyMVc7eC6cZOundlknBOxnhnJ9MTG3mnbgK6L+LwTfOjP2kibFNJPKliGIggRlUar4esTRq8aVIslbpUxVjgJhzWULC2nfpnM1GbzSScUYlqBfwXuFEUEbnJgpDtQ3/MQtihkjm1g5aeBLQ1Rgz9EEiBUJKMpVGDogO5zIt5j5nFt+5n96PTufxnO3j5rcX4i8JIA1pUSUWNj3DX5M94oXYG1xYdZt3+eVw4sZLj/Vm4Aga8PRbSc/vprkpmzIQmep7LwzAUxZumQZRBvaKfwXOJjJzSxJmTOYwua6FmXx5ydhBFFmla9PJfXdtW2cvc3d/HHBfk9NQNX38+49Sl9J1IQc0JEP+JEeM1XbSdS0W1R9C263BM6COwIxlljhPlqJ3E0zLq9/ro35dGsDBISrKLviErOp2MabuVwGI3c7Ib2P7lOHROEa0P8pc3ELw7kY7zbHgLZczJPoy6CP191q+Pc21HOXvfn8DjN7/IT564BdeIKKpOwdygZcmqgzyWeoKiPdcj+zWgCMRXaBgao2DokQiky0g+kXFT66k4k4+xXcO5O9axoOpCmroTUXv1PLBkE4+9tBJvvgwaFUeqm6E+K+Y6HdGJHtQqC1F9LDA7qVLh2oe2sLmnjO43cwE49uB6Hugbxc7fzqJ3CqTtU/EnxwReOE4l9+cHads0moDTgLFFx7jFVZzaPII1123hqc0XoGQFMZpCGLQygqAib03EVaJiyXXhaY1jwvgG6t8rpvK+dSxetpruWTYMi3sZdJtAFci78iTmvUn4ZvexvbOSRenjhlPr/w58Zxaii79jC9FLwxaib8uwhWiY7wRvtsIz2xeTd99BJLuNtgsVXr34OYSgxNIFR0mqlJm6qRpVhGnpzZzpT+P01A3Y4/zcNe5zjN0ibTfL+NZloJ08hLc4Qu9NAQz9KsmrW7h+4gFcU4OIlgjaRgPWBhHXpV6anitBPyByoi6Hu9O307shh2hKmPLHf8gXl/6ez6c8R9xBI2FZgyrC+LsqcW3MwNir0n5xlFuz9rKzqxS9NlZXybHDSPKJcCx2QwtNVwk0Xhern+RNk74+XzEsUJbSSbTVTO8lJTw68QMs7QI98yPovAqpzxoQKq0YewQiQwYMgyrh+CgAKQkuxDIXiKB3QuIxCdkEGh/oD1s4/ov1vPvZDMxdYd77zUL8WTIJ+7UUTmxF4xdQQyJP7FmMz21g/fE5oAhsOTIecX0i4TM27GluLNowz174Kqfb0tnw8O8Z9eApXMXQOz1Kvn2Aordc9PoslNx/htqeJFIPRvnd1A8gJP2na5utsYBHi9+rJ+/Tm5lx6lJWNs7n/LQa7DWgdhkIxwnoJZmGlc9x2djjSKUeQh8n41jaSbLVixSEvrEaflf0Plo3pGzX0dsfx+iMLkJtFgKJAsZP4/j0wDikdD/6oZggbthSwMxXK9B6VRIqJPxdFkKyREbaEFc3z6Xs0bU8kf4lZ76/joWmCHmr6hDsYUzNWgJpCh9+Xk7xG2soSe8h4bCW3LxeIkucWOslAhkxgZN4Ak40Z5HxuYB+CIr3Xkt9QyqWw0bS9qs8fGYxlvk9FL4T4aKJJ2BbPOnbNVhm92LebiFiUXnmslcIp8g4iyQef/di6nqSiMQJTF9zjFF/XMsb+2fgzRRRBfjy2edJOOXHmy9TNLMZ/RepaPfFoevWkvvBAKd704jEqfxh5zLI8eP4PCZy+3vj6K9PIGIV0A2KCLscIIH3hylEvgpx9f4ugM4Vq4qd+4yI0RjGe/lUfLNjafX5790+LIb+xRFU9Tvdhvn2DAuiYb5T6h8vZ9mBRrK2ivxuZDl5pV1MMDfjztWwed0c5CVOqh8cw+SUVsb/di2+fUm88fAF+PKiiLUx/5Dfr0dyabButqINqHR8lEu8xkfSZ3pGZXURzg6jzB9iQno7/jQBRQfp2zVM0WspvaEKjV5G51JZXnkTyypvwl2sMFQTT/rjh9lRO4LEPx1ncKwCLi2/eulqPEE97gY7CYd6Gb+2kr4yHRGbQm+5CmERsU+HscrA0MwQ/jQBIQpSWODsWyOZMK0W66pO7vnoalylUTS9OnxpEkPFOuJaFBw1MhqPRDhO4OgFT9A+YEd5I5n4P1mQAipiBKSQSjAzgrk75mKc9qPbMRU5aZuvx1UgovFIuEqgOK431l7DGkEMCuiMEdSARHFxJ00Xv0Dv6gARh0L802YaKjN5evlF2OL8/KTtIrYdHocUENDYwpzYXUL9lTb6nRYafj4W21YLbUsEfvfUatL2iFxQu+TrGjUAzzqz0PdLiJKK6NLQ2eXAt0rPxqqJKBLUX/kc3hyVhuNZTD6+kn2PTSXo1+GaGkQSFYY2ZSCbIJSkcOPGO/DmKVhbgmSnDtK6MR/FFEXRwcC0CDjCaLVRwnFgPFRLVA9LrKe44c5tBJK+evne60D7ZAIntoxElEEvaL8+1tr+ZEyVRkLxClk7oqhalUmzqnEGjZTddJr+HRkEqmMVtlEhN7c3FvdTb2Tf08/jGhVFHjCS8oWEJ1fBnySiKAIWXRhtl5tdG6YwNEFGd0sXxY5eAskCpjw39z11E9pBDTMuPYG9ViHs1uMpjOKMGNEEwNKsIZCiUjSujTm33crdb76NzhGkoS+R9o15lK06Q9ncWpovScDXGofOFatNJEckvNkCkk9EdGmwNovonCqaAEQsEFctEfydD83UIRZUXYh5cSNJu9oQBnV4Mw2EKx3sfyrmJnuy+UCsBMIw/7r8c+oQDfMVwy6zYf4u5H8Uwp+sw/LeYQa3FqOoMNTq4JLyo3y6qZynb36evd5S/FEdO1+dRvrn/bQ8qMU/YMJ+Uot3hp+s1zS0XBvFfMJIypEADSt1lDzvYqjMTuByF+FKB/Hl3fScTcZU4ML0no1gvIilM4q1wcP4V8+wu7uI4NYUUi5roeeDHISFA0T2JRCMVzGUuLC/YcWbJmHqi+LOkYjExbK8dE6VhGtaSTZ62Hd8BKqkgj7KiN8OUXtbMpYWEWtHlPYFKoJJRg1o+Gjx06z+492xuJ3iIGpIwpLoI+DXEw1IWM/pCCWoiMVepAor/swo2iERzUg3/j4zOfm9BP6UxuAoULNiWVeFT0TpmGPF2q7gSxUJ2yCUECX+tMjA+CjoFaQhDWKmH/1xC4oE/uIQybt19M6Qyd4i0H1NEMMhC/40lahBwdgtEY5TiZoVxJBA4sh+ejocjCpqxxvWM7g9nYgFIhYVY6ELf1Mc+WUd+F7IoHcK2KsFLJ0yHasjZPxJS99NfhzmAPcXbuMHh6+k+Dc+0l7pZHdNMWpUZEReJ3VHc1hw3gnODKbRczSV7E8DtC42csGSw3y8tZykk1HaL5S5ctxRAI4PZdG6OwdxvAu/2wAeDWPHNnOyJhtBH8VqCxCOaAj2G7l/7lZutXUCsRRyAOeQGastgLvLSumzbkRfgLYnTITO2MnYG+GyJ7ezoXUy3f02lLCEvk1H8dxGLkqu5JGPLiGSFEGQVFQVRK1CUrwHg0am43gaUaOK46yAP1VAHe3Bst3CwIwIui4txh4Bd2EUVRd7rmpcEkK2H3nQgMYtYuoUSDgbouk6ley0QVqaklgx+RhV7lTOVmeRk99L3550lPEeJEkhcjaOcFoEzYAWOSnCPeXbefmZCyi77gwdd+XT/qMopk+sHHtwPSWvrCH35wfZ2lGBVpCY8G9rGJocO4/GhS8Pxwz9nflOXGaJWeqIi+76rg4JgIpXfjTsMvuWDFuIhvm70HiRHk9WzP0Sf0EtruoEUvYLVK/Iwjqrl1v2XcebJ8rZvK0cBFj43hHGpHbhOKEhkAyOHUaaVgikbNUjBaDzh2HEhBAL3j5C/ziBk1M2oooqfU4LGq9IsMqO+YZOXGMjdM4Dd5GVP78/E+fBFLwz/DTvy8abpZJq9aDxQeJJldNTN9AxH6IG8GZIBFIVQukRbIu7ePGXT9K5NYd9FSMwpXnROYKYavXU3p4M6UHyL6/DnyiyZNIpxuZ2UPJygJsfuIvFVx0kbnoveen9GB0BXix7E02DASIiniKZcGqE4ICR0Fg/KQcEIjaFQKcFU4uGwU/T6ZuqEE0Jo0YF7LuN9P4iTCBNIZAgoneq6AfBfk7EVQSGbg2SU0PUpBBx6fGNDKHzgKRTMF7dRW5+L+5b3OQ/GEYVofAtJ6pVxtKmovUKaF0iWq+I60gyeluQxs/yiLyYQnS6C50TfrB0G78Z/WcarniOLlccvZMEpKDAwCSZ7ilaZLcOV76W87Lr8X6cypMtCxDaDdRfk8DxN8diqjKg6dVSVZOJsVtgrq2KjrMpiBGB1kVGLr9gHy3+eExdKgMjJeIcfjZ9OoOP35iJ99lMHDO68Q0aQYxVK297Kx+DI4ioUfHX2om0mtEOSTxSsYj9wVgjbV9Aj6oKXDbmBO5+MyvLj1B3vxHrnzxopZglSjaJVPnSmZnSyE1jD5D7DhTPbWRz0afcZOvmjZV/xFKtQ6OXMTboibf76D+XSGtdCtk7w6gWmaheoOr2dYinrSSubsVUp6NoRjPGfgUpMQSSimCUSa5Q0Z4yM2fiOWSrwkU3f0HjSonGBa9wd94OrCletv55Gl0eKyjgfS/mLou0mPH1mNG5BEyNOkxdAoJHwzv3L2FoYoSjH45hx/uvE3AZ8KcLTPrFGi6/YB+LzrgpeX8tAMmvHifutG5YDP1fxr9y2v3/dIYF0TB/N3yZCts7KwGIOiJEtQKqyUBPczzx+/SofgndKBfBmR4+6xtB1XuleLNBFVW8F3gwJfgZKhbRuxQ4ZsNYaeLFTYupX72epTVLKZzdTO4zsVT4+LI+Wk6mQ1Rg3JhG/CkSwbQoGj9E+/XY6sHSIlDVEOsFFr16gPwdN1H8mg9NQCWYqDKhvA5rlY4JiW3cVb8SVQNJR0TUozbCTj2ySWX1+V+iDOip+aQI56wgn1SMpfazAmqvN+G4to0tW8vpH7KiPprM2LRO3ndOQgoI2M5piKvWMG9UNeijyE4dgyMF5kw5R3ZJD0WLG9D6VKy1EvFf6tAbIwzOCuHxGjEXuBAUFXc+xF3QhXOMguyQqbp9HelfKiRmOxnxpJO0jzX401S0dUZ6v0inaygOtscz8EiUYJJKwyo75mo9nlwBeZQPKSCgG4LLLv4SWZZwzOzGeYUX85Y4TH0Kj3+5iEfvvwaAM+VvMXpKI9nl7ejsIZKPy6R8KeIqjdLqdxC2Q8uRTKbNPkvEHoWFg0TMKroiN4JJ5orrP+dnxy6haKOPYG4IocRLxbWjqKjPwTUjSHxVFHGnA02RB8+4IENXeXHtScWR7CFxlx7FoFLxq/UsLzxN1K3F2C2wdtEOpKDAjIIGvn/2CkpeXcOSwnMkW7y8f2oC5nodR/pzUJw6qt8uJbInkdyJ7XReHuECRyVb353Oh0/MI2oU2Vz06df3bblBwjCgEveZCdmk4vIasJUMokoqzgIdxbndeKYGKHpjDRGzijesJ2JTadqeR/d8Gd1pE5I5QkqyC19q7P7suTGN0vVORhg6mTG2lsXLVvOrqguZmdGEWOZisN1O7maVwXEKCSdVCie0kbZHxD8uQKA4hCrFgrG7rwihNUdI3+dnUdUFWM/pCKTJRJcPseHkZLaPjqNxxfOM/cNaUFS82cqwGBpmmG/JsCAa5u9Kwdu3U/f6BJqWvsScHxyic14CuR8pDM4McenUYwTrbMR9bKG6IoefrX2L1ENRwplhMp7SIh60ofNAz+xY2rovM4oQhVF/XIv72SzaPsyj654wP7lyE4OVSVhaRJqWv0CP34onV8F2VkKUIb24j/l37cc5RgZFQC4MENqRhL5JT8MKK4F5XkzdAlnGIeZeeZQCQx+dgzZO/3AdwQQBY59KfLqLiENh48ezeW3Z86y/ZR0N81+lafkLqAKUPlBPU28COmfMYt51Swh32MD2P00jkKrgyVMIpKjsqhyJ47AOjT1M2pQuzrNX0dKSRN2nBYTjBDwTg8xZe5icB6OkfqzDUGnCYgjhyYGM6R0M7krD0CVhqdMy4oW1uK73YNaF6ViYxOBICTknSCgpiq1JIaCSYlAAACAASURBVP79WKStf28SUoEXRzWEElR0TrDtMDF2aTWmxT1se24m0aAGRRUYk9pF/i01hOIE7p+zlX1P/3uadsPWAvKt/eQlDdC2GGSTELP6HcwjYlaJP6vyZWUplkYNgqBi6BcIh7RkpgwxydSI6bAJ74M+DJYwcouFzvnxxDn8CD16OhYrOMfIBHvMJCZ6EEWVH1//Ls4WO/4UgeKXvJQ9upbNf56OpUGDp0jm7vhGRpxfx77aWJuiZ1a9xJa9k7ghcz9Wux9jj4r6RDLWWglrR5SIFRpPZiC2GXi4cQnWWb04FwRIv7f+P923Rx9aj8GpkFSpkv97hcFuGwgq7gIQfmLHZA4RTQ8xdXYV6stJ2KtAP20AXZeWpZcfJDPRSXdrPJE5LqQgtC9OxJ8bx3N3reD4tpE0rIojqogsc1SS/aBKSUkHngwNKfn9TL/nCC0D8fRcEMJw1kjSLh3esiDedAnTATMRt47GSww09CTiKY2QvkfkyKS3KLruOAAlr67hnlvepe6RCX/v5T3M34PhGKJ/GsOCaJi/O8KAjqXzL+ex1BMI5w+iCUYpuu44u14tR4hC4sFeppTX8NL1FxNd088tE/fROdOIGAH9kIqpRUvcrB7MbRJhm0L2x0PINw7gLgvzs5Gf8MjGFWTuCmMYUBj91Fr6KlMQM/2E57sI26F7MI7dj00nrloDUQElLOEeIWOa1E/UpGAyhBBkle0tI9i2exIbWifx+uRXydt2M5ZF3QxMjGLWRUAFrVfg/vtuJ6j8eyDvqAW1dK8o5ncTP2TsinOYjxlRqyw09CZy6p51PLZkA4Z+ESUniP2UBneRiumIiZa2RB57eSULy84QSFXwFshkpgwRUHS0L3AQsglk7HLTXZ2MnBih7VgGggLm8n58GQqRYj/KPgeZFifBpFhgtvG0EU18kFCcQOodDTjHyvhKwuTe2MzU7x9DPyiQvaIRg0vh8KlC4n5tJpggQFiku9vO4TMF1L9SQtgu8OjHy5n0izX/fiGnO9l5eCwte3MoeEfm1rs/wnO5JxZknu9lYGkQBJBNYHrdzl1rNqF0G2hrT+B7R6+i5PIaek6loNPK6AcFFA1IosKGy55BMkcQzRFyirtJvAe8PRYe2H0pqqRimdNL3Y90hG2x1PbKu/6IsUPDFU3zONedyo3jDxDal8gvf3UThj4xVmV8t4P+yVEGRmvxFEbRrOlGU+ZE1cUKWg7sSKen284Now9yri/lr+7XG1pn8XlAouviMJbb2gmkmTC2aLGmeNEXuWldHIe4144oqQyGTPRNFJGNkGT2sXzpIT7dOI2BnemgCvhdRjz5Cp4REdpXR3Df7iaQFSGuEUJhDT/cdANCKEJTXwLrfv40gxXJfHh6POkOF2KHgYhNxZMtoEYFpJCKcP4gGTkDTJpWy0Ulp2ha9iJds1TGvngn3k/zEceNJP6Myq+3rfhHLe9hvmOGXWb/PIYF0TD/EGpuSWDKics5Mfltdm58FUGr48RP15F3/0Gq7k7gWFsWfePN7B/7AW98OJ9AaRB3icz5P9jP9ItO0t2cQNo+H2kHVDrPc2DWhTHV6SjVdWNuV+mboKd3QQRlspusnWGOznwe8aCNiFUh6tYhXduLzqPiOCmR+okW21kNgx12hPgwgx12rvzeDuQzcSCqdLckcPsTd5K+XaKrx4493Y0zYMBaL6GbOsjgSJF3+qcCUPDO7ZzaV4S9PsxT915BRUcWlg6Fgte6UBvNjHhhLS+0zUac7KTkZwO4CxXMbSI/XvMOoi6KLzdK5R/HgaRibtIQeT2FfRsnMOaSKoZGK9TcYkQVAFmMCY8gpFnd3Hzebq4ZdQRvSYSzfankfOwndbeEsU9FEFUsl3TzQeFOrHUaCIpUPzKCU78ch61RoX5HPp2zYplLLUssROJU4mo0OI7oGFPahmehDykEUlAgmCTww65YPKZy1I4qqYRtCj0/CPLSwxcRDkvkTWslKkvMzq8HjRKrSp0ksv43l3HT+bsZcV8L0S4jR2vykK1RHh/zLpnntxKxQGF8Py/3zSbOEkB162g5l0b1HfGsnHqE4uJOlkw+Rf/ZJJaVniGYKuPPizDi9Tsw9Ku0e+yEvHq+uHMaslmld0GYksV1vPHGIjhvCI1bIpCiYG6VaK9IJxzWcNXMA8hJEXb/4DGsZ3W8um0ep6ZsZLPPREPEyw+7JvFFxUh++Nxt0K+n8WgWpjYvwaIgodN2fE4jjpndpO1xEb/DQO+fcojERRl93VluzdrL8cEsrK0KWVv7mDy6AcIi+nQfQjCWAu+udWBp0JJyZQuCAJYRQzRekUTChyam6LWEE6M0LnyZqzKOkFyhEEkJk3FeG6gCzukh2BlPOCrhvlTDR7um8rHfQONlzzNx4TksixtRKs/RO/mfuNCHGeb/YoYF0TD/MAZqEr7++9OWI5Q9tpbej0oZ+VAHpWm9yCaY8OAaHNUKYr+OyuVP8UVPIYfeL6N4zRHaFpjpWRmAuUPcm7udQGmQ1S/eRcrHjdjmdaM3h4mENWh+2sP8X9+NoIAhz4OpRUPPyRQEBR6/93m6FsoknfCDVkFqNaBxS6w7ch5xjVD8Qj/2Mxrck4O4cyTi9+lxuUykPhqzUCiqQN5TVRx9dyx5m28lbb9K1KDSslSL8fudhEMaui8M0XpZOlJIQOOHoTeyCNbYaHncimKOokrw+6dWYThnRJVUPDkCGZ/DhIvP4M4VEWUYWpOCalSwpnjJHtENKugHBOJrwjiDRl7fMo8N1ZMQjTK+KgerXtmOKIO7AFRFIByVmPSLNVjbFHQDEtr4IP1jtfRcHCRqUtH3i1w2/xCySaXumvWEpnkI2wQa+hMQqi1kbmzAPHaQtH1+Nu+fyMgDV2PsU8nI76dkXCtXFx7F1CeT+GcTNXXpAPiiOjQD2tg5T5A5/Mh6Xto7l+7LClEcERxHtYx8pJs7/nQb/CyeqFHl2IlCzvy+jKE+KyNHtfLjBVuIrxQxiWGaD2Rx7NnxlKzrpOLRiRiT/Ih+CSkk4C4Aiy4EYRHNUIDUQzIbZr1IZWM23oIIkQoHtnqI2mRO37WO3I+DmHebeatiKuYaHbNevJfZV1Rgr4H8D27DoxjZ4h3NjqZSEnKGMM7tQ5fpI5oWonqNGUFUWb7sENoeLcvSz9K22IbBGcWyqgtDr4bjH43m/g9W80LRBt589Pc0r0ji3fzPcVRKaI5YSTkgoB6yoySHOf3DddQdzUEQVKaktRDJDxAxCRRuuJ3zx58F4CZbN30TRSR9lPqmFCzVOooze3BNCTLkMmN4T6X4N+d4urCUEc+tpemJErZ3Vg7XGPqfwLDL7J/GsCAa5h/KNwM8xQi4nCaqH06m+5U8FA1MvP4UCJB8FGYduxGdFEU/oKLMGo8og22HmagismbvNRirDUT10Hh7Ab4tqYTbzchBDa1fZBOMF/i3NW+Q9W8qwRSFolf6CFsFvn96FYREWpYZ0ZoiROxR3rv8SYgIeHIEqn5ix5OnoPo1BFIVnHODoELdjVqiRhV3v5mqx/NjYqtLQ+oPGlBsMgklAwwFjShhCeGrVVV16zrCVpX+yQrz5lUyLrWDjB0i3rIg9ks6GHNBNUIoVqxvcITE/WmfUr78FGOvPEPjSge7Fj6Bt8VGib2HjLx+Mt+qp2uaDk9Qj8YnoDSb0dUZ0Q8I/OmuC+m71A8KRIIaBk8mkfxFN658kUi8Qu5TAuZOFY0mSvqUTrQ++GDnNF669HkWXHkDI1O7EaLg2GhBFVVari9g94TXqL9Oi35QIvcXIVyF0NERj6IKvF4zlfbVMhGjQNpuibg9RjofK2TMtHpSFrST8yHkbbsZISLgywRjvZ70K5pR7BaEkR7cD/jQDQkYeiSksIK2T0vdvlyeqZqLbBJI1HqIZIXxpwr4i5Ow3t4GJ+IQIrEGuIIKQ69mY2rVUHefga4ZGq787HauGX+I88edI5gSRXNpH7oeLVNOXE7zMgNRg0DCIS3+jCjRET6O9Gaz+dePYW2QePzxlWy5cz5/nvw8g7XxeI4koVRbUH0a9N1aMpKdbDo2iU1XPsHuO6YjTR2i7SKFtnOpTF50hvwljVibYNnhNVz183sx9Knkv3c7Wj+UrziJq0BEnDaEpl0PgLUJcq6uZ8fJ0RhMsfMcObmZF7P2f70+hDwf+c+oOI5pmb/qCL3vZaPRRil4XKbHb2XbuS/4ccNpEqZ3I8rDrTj+R/Adu8uGXWb/PYYF0TD/cP7y4D7x03WM+FkPUrsB7ZU9vHvbH+gKxKFoBAYv9ZGw3gy/T6LwxhpSHmnCXhdlxV2fEa6OI/ddgXuv20Tm7iDXXPo5Oo/Ky8tfoPi5EJlz2jD1qPwv9t47So7q2vf/VHXOk3MeTdIo54CyUCJIAhlEziCJYLKzsTE4YEwUElmAyEgCCQkllFDOGo1Gk3PqST0907mrq+r9Mdj32vfd3/JbP94FP89nrVqrp3tXOKfrdO05Z+/vfmL1zVTeYSP2rEDVr23cv2oTfr+BzYteYsbs89BgRlAFblrzEPFpboLJEfTtOnT9IkmZPShGFdmjw2QNYbuoR4qW0ZgjbJi5FlECRa9StieflBQX7vNxdDVFY2zSYz9o5IN7n6PgrZXYGuC+mbs5sGMUlW8WEbi1l5x1IP4hjto3C3BUicSfjxCKVbj1V49g0YY4UpdLODXMnC8exV4j0uiNwbM9iYa7hxB7QcZhCpK+qx8xy0fB7FpKH1pD4yIRVRERZQFjo4GUse1U35lEdJWMJcVD3VVm3HlgOmCjOMpJIEll+vRSPuqZSOv9EhcODUEThNbLIqgaCEWpzD9/E1eMPodc4KN1XjyiBMgCdScGqtOLzUasbRKea/u5/+GNNF+uULspj/rKZHoLdVir9KhaFSkjhGRXCf8ikYYlUSS+bSTaGEATguQjIbpuCLBs/mFUDei1MsE4lY9/sYi6eW/hHxok57cV1HfFYmlXSTitMnpmJbHnVXov82Gf1oHtkJlwsoS5Xsf75ycw1VFN3VWv0XMhnpEzquhyOtDkeDF3KvTlgqNKAw1mvMfiuezcHcRcDNM7OUz91VpuffwRNMkBRGkgXgxAGOoh19FN0bMuVvz0QeruBs2uaAiKaPwCRw4Vk2XpQbIIiKKKL0Wgr1DF4BLpGaFyYM8IdON6iXndSiQtxJhT1yLZBFruG4PjvA7tQQeKFrpey2Lq+av+Nk6So/sRjpYQmOPlq6piAgkDZTjql9joPZTEkI9W8EzDQto6omib9v9L/maQQQZh0CEa5Hvir07RthPbWDzvGNYFdfyq6UoulmQiLO/Ctt1K000y4mOdnD5YwPGDRbRPg09XzyX+nEpftpa1T19N8xwDG1fPxnpzK6vWraDuQQ2NnTF0j1WQjaBzi3TPChMf0898Sw3aGhNuxciJ9gwSTypoY4JoAyB+FIspzo8YEpALfTh+bkTVqhg6NchldgIJKplfquRef45lO+5DssGdi3ehFHlR3k3g4cVbsNZrSZvWTCBJ4Nn2eUSygoSjBNZunc+MBecIxgn01sWQ9HQdxc+cx31pAPewCH1ZWkw5/XSNUznqzKZm1jpspQbM7SLmToVV6fvQ+VQcUzpQ7+qisSaBrjE27DstVO/IHQh81gzMjgfTwyRNbSWwPhnFqNJ6ZQRvhxWDSyCcIjH11tPE6r3Ya6Fs9TAONAyBs3ZSx7fRNzRC3hsRLMW9/Gj+YV4q+ojte8ehqAKpu7sxuATen/saUoJE14UEJIdMx3g94jdR/OXtZQh6hVCMyohhDVjaFCxtA9pAjmNGoi9C3RITSUfDuPN0GDXSQPbVLD0hp5kjXTlI0TLuLiu2enBOFsndcxuGGiPXxh3HeMTK7PuP4s4TiSginZNVjEaJdJubQJIKkkAgSUHxa3n1d1cx58Y7yPwqRP8cL2K/FtN+G6igGNUBcc6YAT0m+2t2FL2IrkVP/jof7TMVMteImKd0409RsNVo+cnwnRw4PRTns1p6C0VMZSaCszzk5DtJPKkgpvvYVj6MtCsa0GgUHHUytjqRYHKEwnGN5H7iZlJKA2JYoW7u28h7YpEsUPrwGgIJKtGL2sje0M1Pfvs+DkOQYS+tYtQfV9FxJIWsEybSX9Ag9RmIrhyQsai8fS06D2RukxAfsiL06L/HkTzId87gktn3xqBS9SDfO0MePoZufzJl5elcN+kYO1+dSu+UEFnrBVruiJBzcznua8bQWyhgaQVfOqgaiD+tIN3qwlUei2xV0PZpkK0K6TtUWqdr0GT4UBotqGkB0uLcdB5MQbINlMtIOiaj80aQLFqa5wskHBcQJegZLiBFKyRk99B9MQ7ZqiBaJIa8IlN3lQVro4D5CiexJj8RRaTGGY/k1aPr0hKxKthrNPzqvvdZ3TiLp3I/55aN92LoFRh7xQWOHB5KzfWvMuLZVSCAIIM/RcVeC9ZWmZZrJezHTCxfsZu9nQXUtMdjKjWhigPtRYXw0AAmc4gbh5xky5OzaVsUQePSEVUu0DNFwlSvh5H9BHpMJBzWknxHHdLNetR3ZHrezuT4n9ZSvHoVWh/0F0a4ecphzrrTKa1Ix1alQ9+voopgWtpB/64kZAPEn5fQeSOEonT0ZWsx9KrE3dqI680MPBkicaURJKtIzIpGKtsSSdxkQDKLeFMHyqqIEig6BqraF1ay7/gwdiz5C1ceX4FeH8HbamfEsAYq9udScdcacvfehhoRSd2spXUWpO9WsdT04hoTy3u//wvzv3qIgoJWXsr9lPk7HqToF3VU/jKPwmeb8a/T0tAUj65LR+rYNhpqEzm46Dn+2DGHbSXD0fTqiKoEx/JW6stSyH/fR91VVgqn1DM1ppbDrlxKq9KxxXvR7I5GExzQBtK5RRLOKPhv7yXGHCDL6mLPqWEsnXSSLbsn4qiG3mKVIQ8fo3PVFCJm8GXJxJ0S6ZoRpiDTiXivGcWsp+oWK1EXRbQBMN/QTmNLHFpjBDmsIfqIHnehijWnj8jxaEQJLE4FnU/l1edf4JrXHiHzzWpc7zpwLKoZjBn6AfGdKFXHpqvDFn23StUn3h9Uqv5nGZwhGuQHQcXJLPJXnuD0aBHXuAhavcze995CW2bhj1UHiS7rR1fcj2+Wj8rb1pJyIIK92kO/z4jeLbJ5wUsoqUFiT4u0TtdgrwPDCStKSpCU2D6y7T0IEUge6SRjUgu+JA0dDwTpvs1Hxg4F1+V+upcEQIWbLjmE+G4cStxAqr3VFqTqTj2oIM3pI7A5kbCsofZIJpZDFozNOiJ2BY1/IBNsY9dYxsU2MdUoUnPDWsruW4NGULHXCuQfuIVwlEpwvBdU0GZ76S1WaL5URO3X0zdUZqSpidqOOIQOA778MIbJPSDAiEUVaGuNmL5w8Pmf5tB+hUR8Qj+xJQIml0L6FhGDC8KNVjK3QPAqN9cnHyPS2EyUPkDXpSFG/nkV0ggvqR9WY63Vcrwni8bPc4g5oyV/SRXBWIHLV33D47k78I4KYupUabpGxp1rxJ+gwT/Bj7k7QsWFdDqmy4TiFLqHa9Hf7sSokVA6jHRMEOkZDmlzm0g6LmFrUjCN7cFcYWBvWSFiWOC17mkkvWskUBGFo0xDuTMRVasy9jcruXfUAQoz22m5MsKwkY24CrQ0LIund6jAU20LyRrSQednGVx77g70nVpqH8zH2ihy8ZepeD5NQdApGHoEFqeUMHvURdb0TGF16nFuGHccRw2YehSau6OYMr6Czl+FMRT2EfxpIh+/diktfQ501jAz02rw5CiEYgRS9w5UsfemaAiciKOhI5Y95YVoY4Lsf30g27B3mMoVM05R++fJmLsVfMODOC5qcBdCcXYbObYeGpfEI4RkVK2Kf7YX13CVpvIkdE4diRsNqJJIKEbAXisSPhONPy/Mn1e8RczpHixbz3LHxZswuFW61kUTe4d30Bn6fxCBwRii75PBGaJBfhAM+diHprYVucdF06+nkPHkEdq/KEI9HI2tScH26XGU6aOIGDU03qBQmO7EFTDjPZCAfmoPCVfXc1dZBY9vvZ68Dzy4htsRr+nCdS4eNStAwhdGHCubcAXMeA4mEDXDie7FWDom6hBDoA1AIElFipK5buIx6vxxDLO18c7OWWjCEDGr6HtFZKOKYlCRbTKZn0PjUrDE+Qk0DmSQieYIU3NreS/zm7+1bX755ews2soDbeM5sH48C249wob9k0g4CagQMQn0zglCuwE1KYShwkQwUcbSpCEYq5IzvpmqpkSIiBAR0Ngl/jR+I3+umUdHexRDsjqobYnnymHncWgDvHd6MvEHdHSPUzA6NQTSI0Ql96PVKFj0YRob4hEkEVWjUpDfSs/6DKKrAgQSDfTma/DlSui6tWRv9tF2iRVvURh9m474EgXnJIEhPz3FkKMa9mwdS8Ss4qiBuHNeJJue1hVhNGdtJB8LEojT4ZwqoPUJyHpAVDF1iCScCuGcaCCQLJPyDVg2HEedPJL+HBNX/2w3H788b6CQqzBQTy1xZAe9B5IIFAYxWUOcn7Sel905vHh8LvYYH3qtjEZUiDH5aelzcGXWBT6+OJaame/87Tt4zpXDy4fmYmnQomggFKfgqBIIRQtETCqJp2V6bvKhnnMgD/MyM7uar6sKidlj5OTTa8n9eAWKPULUGT0RC+indzMxqYmQrKUnZOHJzM3cUnIr4YgGf6eF/JUn6N+ei/nZKLru96MeisabE6F4aDO1XXGEOswkHRQIxggk72ij/DexGKuMZL/bSMUj6Rg7RWLLI3yz5nVydt3BLaOP8mHFOLLiXOws2joYQP0D5LuYIbLGpqvDFny3z6jjHz46OEP0TzI4QzTID4Ka5RYqf5EPQDhawXPtJJKXlONPUXBOVan/cAS6Ti/dIwdmasqrUom5rZ+oWU6mJDdQ84cxPPXCjeSNaka26jF1y/j2JSAO8eLYa+KyX+6j6lQmIUnLHTfsoOtMIuaTdcSUyTCuj0CCSsIpBSEs4JUNnKzP5O0zUwdib+IjKGaZ+KntaH0DVch1tjD9K/vJ+URBPu/A4BLJv/skmiYjpeuHcV/rRLK/vIvppUupvpDGhJ+tJEHvwT/RzxlXOqYOEcedzUx59ASma50sLTpH+XWvYD9qQrKqmFO9lD60huqb17KzaCvmagOmBh2xpzQkbjbwzFM34P86Acc5Pc6d6SQnuqnzxtErmUGFviEwY/xFYiplRL+Isj8GNsbSdjoZc4MOU5IXS4OWKEOAxFsaCD/hpu/mfhx1CoJBxtAr0LjIgi9DpmbB60gZIcxtQdTEEMrEYZg0YRSdSvSwblyjFPTPdtE0T09hYiembpXGBQb8N/YhyJA7pRHZJmNyisSWStQt0yDZVBLzukl7qBrnF0VMffUk4x48yzuVk1i48hDacb2MWlBOJFaivduBZFOZU1hJsMlG0cFbyTM4QVDxNjhQVJiRXIPyaDThkmi++Ggavxzz1d/dXw/H1CEGRLIW1GNtUUkvdjLs5jKuuXY/4ViZvkwtgTYrynAPRoPEnoMjSYjt5+TTa8nZeA/GLhFNn5a4sgCaILi67JzsyGBveQE1X+Vy1ZEVhE7FoNvvoOAtP66t+TgbY/El6ZCPR+MbEcSS5KN+VzahTjNxp0R8KSK9E8JU/z4Kg1nC1K3SvNpO4nGIL5FQNALX188i79bTHL11NFnLS6k9lzboDA0yyP8lBh2iQX5Q1Dw3icxtEin31VD/0UiKxzSQtVUma7VA3NsdJB/2o2syoHVrqXo0F7shyMmXx5CzMYBkg66PM7jkpeM0LVGYde1J4j8240sV+KRuDAgq/gvRvPvmAiJ2hdYbC0h+oBal1EHCGYW2WQMZSF+eHI25xITOJGGe2k3MKS1xqX00t8YSyAmTdExG6htIn26+M0LilDYiFpW2z4eyeMEx7rlvM9vOjQCdSntJEgaXSG8RbPvDTG4ZdoyGjlgiZpUWdxTl/Um0liXy5dZJ3NY4B1RIOSTz4siP/65fyu5bg6KH/iEg3dJDTIkbz9AwmrCKZAGbPkT/H9P5qqqYoU+0I5tUTn4+nPYlYdL2KQTjVCIWgYhFwdyuEv+WmeXX76WkLZX+59MZEdNK6EIUt//2C+L2GvBmRwglRrhn1l7yvliJGtbQMdGC9ZSJxoUmvnlhElq/QFdjNIYODWVnsjC4BWq25yIu6SbikLEZQ+gzvUi/TsQU50c2gj9Bi96lQYqL0FOSwIkzeXia7Jg1IXZWFwFwoS8Fj9NGurmX+kVvUpzejpIZ5ND2kXy+9AXkNjP37bqZ4uw2MnbIGNfHcLwri8r7jNx21W4KL6/inC/jv9xb5jaRyuNZmFwyEUUkrGj5oHwcgiIQVRchdR+Euk0oh6MZNaGGnrMJjPzzKt657DUiZpWEk+B8MER0VRiNS0t4Txw6YwT91B5oNZEwbaC4cOWdZjQfxYAC3NBNKE5Bq48gHnagjPagiwvQPTtEyWNrsDiCyO0mHh22G/e0IEuzz9NxeQjDVyfRrXByd9IBRKORlvkOUAdUtgf5f5vBJbPvj0GHaJAfHI2LdLh/ncGItFbKTmehc4fwphkpW1dMy0wzUrTCLQv3IdsjdL6fiSddQPv7LnzZEfxJAjZNkOSvtezcOQ7pzh5MY3tIsnlIPTCQ9eRPUsl/y0v/8DClbSmEskKIERVjmwb32BCGLg0LrjuK/rQVqyHEw498SrKtn4T9OhL3aenL1oJBJnQsFrXZTPDdJKLKwdtr5ssvJ7P5uulkbBFI3KtFm+3lvmu/JPVAhN6lPrY+M5O8FXWEkiPI5x1Unc5AkxzA0CtQ/0IhUUtbyfllOXNM8n/pl+QprSSNcdLVFE3rnGgSU9z4EwWGzKinfXMmmrBCzC4TbYszsTWI+DNkLGdMOCdqMHUI9OcqRGf3ovOr9Obr+PCT2QS7TLQslfny1GhGzqzi6ROLvb1pJwAAIABJREFU6B6nYE3ykprZw7pNlzJrbBmCTiHxmA9PnoxY4MV+Swv+3DBCREA/phetT8CfLjPz6tP4v4knc4uK9/Mkoj6zEv/HBgI9JnT9YGsOE12uYo71I4ZB6xHR9YlIihY5IhJtCdD0UQ6pWd3Ms5cCMDuugoQtBoJJEZYcWIWQFASjQv3ObLqH6+gcK9C/KRnBp+Wzl+bS6nWwo7bo7/puh9/AzbfsJHuzn45xWvr2JHHidB6RbhM/nfMlxgfbaL0yAloVBDh/OA9zuwAq3L5xJVJ6GNvHx/A6rTQsFhFTAxQtqyD1XT1DYrpJOKUSWpeE1iKBXuHYn1/F0KVF/SgeTZqfnL+onHz4RaQmC9nxPWjaDXzgiSXjcT+GXpGN04pBha2vTEcUVYTRxVh0YTa4xlP52lACCcpgzNC/A991htmgQ/R/xKBDNMgPkvor9Jw9PYS8n52l6m4DprvauOGBnRTOryZtt8q7O2bx8CW7kI0CwYIgPesyiUtzgwgn+rIIxAxoxHRVxqHuiCWiijQtVdD4BXK+8FFzg43okzp0Z6xcPeIM7ZO+jbWJ8RFKk9i0fyLRVRH6N6Tw9PpruXgyC+/ifrrmh/CND2CyhbBd0omiVQnEibgX+NG36ogMCdBf4KDpCpVAvEjiO0ZefecKJKtIwgcm5OUuUnfLxCX3EUwLk7ZHhnozvjSF5U9spyiqg2PNWf/bPtlXvBlZFZg5upyJ15bgPZCAOLIP/+9TWf/Qc/T+2IulXUKQVaytMhnbFJRL+rCMcNE3TEKb4sd7LpaYlY1EV0kkngpjatUiuHQIRpmKzwvQOA2ggHwqiq4+K5JNYe+FQkyVBhoXmYm6MPCTUVOZTP3CN9H1i/gropCzg6TtVtm1ZwyaIAz9bSnBuR46x0PvqiQMTh2CAvVXaem+LIh41MGyKw9x5KZneWTZZj5/YTbGKiPBDYkkHeimOKadh1ffQ8G6lazesYAjz7+KsU2LsdpI1D4T0XEeIqO8jLu6FGO3gBgBVaPiGqlwdORG7h76HwKHz7lyeKFpLpt/M5e2aRYSJrVjbVVQLTIav8jn181kVEwLdfPeAnWgiHDSMZmU7e30F0tEF/VQN+8tdrado37x68Se0WCzBgjLGlJ/Xc2nOXtw54l4fuRhyDNhNL1acj9eQShF4sgfX0HqMNE90krh5ntRE0MDzmtQ4JdHlpD4fjdnV7xIYFwOtXPW4Rolo79gRlAU1Js0VI8PoWsy/E8MuUEG+bdn0CEa5IeLAs57xhJ3WEd9axyvbZ5P6BqR1h9J6N0Cz5+egz9FReM08NpvX6C71UEwRaL63QJsi9sJpEdwVAlkXltL28E0jE16AjlhGi6zYGkV8UwPEIpR+bx8FFq/gGCK4PEZiT2qQ+sT6M/QIlkFEs5K5K/rxd9txnjRxNC0dn6Ud5aO1mjizgloAypilQXJriJ7tRx+4VWen/URvjSF3iE6Sh9cgyhDy9URgt/E8Ub6YWLNPp64ZAvuPB1Vt65FExR4efMivBE9el3kv+2SdJubw3uH0Rs24S8KYdBFaFykZcn+e8lwuBn5+3P4ZvlQdALtk7WYt9jJinIh6BVMh63csngvFSez6Lw9QNslOqKnOUEEsVuHZIH4Myqx50SCeSEqp71HzshWRuc3Ym9UECWB5M31COdtpOwTyNl1B3Pmn0XNCGAwhmmdISJIwKxe9m4fjXrejhgW6B4TReYlTcSf9WOv0qAvN2NvlPmscjQLS27jxYuzWPrgXkIxCn150Px7HYc+H00wTsXcLhBTOnCuvwpuCot78JXE8OWktcTpvTjmOAnGCWR/rmBp0ACw+tQsxv52JYVvrEIUFGRVxFWgYfzSUtp7HOj7FQSdgmxRaLgyii+2TuaXncNBp5KxQ6FjrAbFZiY6sR9pVxw5u+4AYMj+W0n4phNXSxRlB4dQ2pnM5JKrCWSFCYc1dI1zMGJ8LYYekeQ9Gm5omIupXcPIO0uxV2rJez6Mzq+iH9kLYZFD+4cx4tAdpP66mmNBmfolr3PNtftRSspBGdAdilgG/83/d0JQvtttkH+eQYdokB80niyFmHVHOTH7Zf50zXouPpWO45CRYLxC5Zw3GLK6nrPXP881x+7mZ9O3Yb+oQ5RgXeF6jLEBeofLrMv5nDdvWY0qqmhcOu5ZuhNVBEFQSRjZQWZiD1HVCr+ZuAW5T0/C9Y0IKvSNDmGZ20HXCB21y2MwtuoIDg0Q+EUS60smMjy/GXcBSDYBe/3AQ8vcoCN7213UhRKILRGIrpHI2XAPrbPBccKIL11mQcVl7Cjcxlfdw9GEVArWrUTfK6BqYaKjHvMnDjZ67f+lL3I/XkHp1kIiaSEqtuWjSiJpjj5stSKmSgMXWpPZ8/EElAYLps4wmrBA77wAPsmAscbAz+/7gPUb56D1CwS6zYiygP75GB68dDsAoXiZnB9X0D0xghoR+H13Ac7t6ZS1J3PPrzdhaVW5+EQG0VOd+JI0aJx6Tr0yGrHJhFRtJybPhWxWmZpaT8WdaylfsYbocuiZEaaqOgVdRz+qBoK5IdzLvTwy4msyHS6CjTYuepOpXf4qiScUdDsdxM5sZ8qcCxRcV0FfLiwfeZL2Pjtx1zXhLoslZVIb1//uUTbvmkTfniQO3vss9csEgvEqxS+vwnreQGShG1VUWVc9mZqqZMLRCkf2DCMtvpe2aRrSEnsxJ3sJpkn87toP2f/UFLI+gea5GhCg+mE9bpcVyQKmSgPjnlhJfnIn0towaFTEsIB+SxTSxgQcJXoiQR39c310/yUbfT+0z5c4cTaP667by8H6XOQZfaSvaaBnsoS3xY4xJsiNCw+Qde15uqa4+e1VNwFwYn46ABWPZAwGUA8yyP8ggw7RID94ap6bxJzTd7LE4qVg5XkS3yth45IXmXvPSurvyOHrQByyLLK6cgbKDDeu4SoLPnkMkyGMYJa5/JGHea97KinfhImqhNcvTkXWg+TX4byYQJbVxWU/28/aJ5eRtkugvDyN6HKVgkwnneXxpBwKYOwWSDoeRu3VU3unhltGHqNpQw66on6iaiL05YGjemA4DX26gy9aRxK2CzgnatEERPQ9GmQ9jBtdgztoInvz3XyaswdUSDit4BsWQpTg058spOvyEI8dXfZ3fZD95V1MnlSBcUo3eamdSFYVrUXC9XImnlwFebQHnU5GntjPbQv34pxkxD65E/0FM9UVqTimdvCzbdcNaCk5FBaPP0MwKULb7WG+unYyllYRVatyojGT+iveAEVgyzOzkCZ4SHtdx58+XMaC+w9hrdESlLRkLasl9ZsIfUPAMawHySHjPxyHEBHYfqH4b9fdOT2C4NZhdGoZ8lEzwTiVunlvkRbVx/MX5lByNI+Plr7M8YNFFL+8iqxHKlFFAevDek5+MZyTdZkkHZfJMnbjc5nQigqkB5gQ10jPhAiSXR6ogbfpYeovf4OiSfVkbHdz/tE1aL+KovKOtZgNYUxtWrR+gZgxnbScTsHcJuDelYzNFMRRquOnR6/Gnauh464gggoGl4CqCNii/Bgm9yCbVDxzfTTtzGJhYhkaq0Ti1Db8iQJ9+RB9ZSuqX4Mc0ZD2eDXBaR6Knukn9ozIFy/PwvqNmbQoNz0hM3lvSNirNBj1Evsfn0Lth6MA2P7VhyxceB3+URnsbDv3PzfABvlhMRhD9L0x6BAN8i+Bt84BQOXaEQg2K7edvwVBVQkmyrzvnMTRGasJhXTEWvwYMz3ocjwET8SiSiLOK0Ps2zuKpttlQtECcZ+YCSTLZKZ3owkJXFgznPc/n03YJtAxUSQrrwPnLJmmfZmY20T6co1ogyqmOhe6fpHc9E421o9k2V17ibH4abtGQhMUMHcqKHoo/008fduT6ctTeOvGVxg9tQptAFQRyr4qYFJiAxnbwK+EERQI2USy07rIm9pA00KQvVrEbj35761k/C9Xsqx2LhOH1XKiMRNPWSxVtcmEEyLEfmWkZ6gGa4OI5qyNzFgXiQ4P63bMRpSgoykGf5aE0anBuy8Rk1NEKfaSPtTJtt3j0fdo0OsjVD5mJunyJgydWoYkdZH95V1Yq3R0zI4QDupoukMmba+fz7ZeQmB0gDizn/7fpdG0SETI9+I9EQd6hcgoL1qfgKnOQO7e27itaRpEBEaPraF4ThVbD49FExAoem0VvrWp3FV0GGuDwK+yx6PzCAgKHDlVQF+hTN0TesJ2FcWnw52r5eV1SxANMu0eG9Nzatj66RSysjtJyu7hydveJ/E43Nw4ndLyDKpvcDD1/FX0FqsUHLyZrl4bCWckwhlh9GtisTYLqBrwZssYNDIP3ruBmIMGFC2cmPQm2aNaCSSq3DTqOOZPHXjKYxgxuwq5xYwmBKtPz2REWitRhgCXLjuBGIKuPamY2rQsKSqh+/FMcuJ7qPyFFdkAwRiB3rESy1NOUrk9j10b36VvmER2dA9do/Qo3QZGnBG4r3UiN366i0uf+WZwZujfmMEss++PQYdokH8ZRv1hFUU/rWf4jk7Ojv8Yw7aTjBtdQ+1H+ThlDZLTTHNnNKKoEmyzMHR+FTnZHQBovQK2w2YsszoJ3+7CnOalpSQZKS6CsLwLUydIFgFZr9LaHYWmT8uw+ZV4C8Mk3NqAa2yEzumJhBMj1JWkEvWOja0twxgf10h2UjdLrjrEXU9uIpgqYWgw4EtXsGX2cfcb99G0Ng9/qoxvWIiUwwHO/Wo0zskanuiciCcbTD0y9eXJ1O/MRrRJ6Fxaoi8IRFUMiDZe3JHPybpMNFoFRQMvzvqA6KR+POkioTiZuPMhgsUBuj/IoLEiiSVzjxFdGQFBRTRHuLhyDd6cCOLkXm4qOkGTMwZLq8CQl+swbHOgbTfw6+wtyCaV2iOZaHu1RNdEmFhQh67JgH2/CfHQOaS0MKoKBm2EjvEGhg5rggorqlZFa4rgsAbQ+UCyqSg+LRdfGYalUUvJkTxKD+Tx0NztxJbJxEx2ErKJfHX/TMIOaNlYzMVVazB1qmi9IjgkpBYLtuE96KKCIIDWD3Z7gA0j3ibN1EvaH46wLPUM3n2JXG3tpy9H5NS2YVgTvcQM7SbV2seYMTVUTnsPfbkZf5yWuH16QnYRX5qKooe89/z4P0xmU8cYPFkQLvYz+cWHqWsbKNny2/gy5Bt6kC0K9W/nY+gR6R8RwlpiJHh5iOAj8extzsPUJWBvUIiYVb44MIHpa48R/EMKiVsNBGMEgiMC1C96k2feX4Y82sNzrhwyv4CLziQu/HgN108/wpnHxnDgs7F81D6Bt76e9b2Os0EG+Xdl0CEa5F8GT7bCV+f38FnpGC6bciUAzmdzGXVzKbeV3owhxceQ5C4iEQ3mVg0tnih0j9mwnhyoB+ZLVeloiuHRvF0EA3rGTK7CflGHvCGevkIZc6eCtVFkRForWr9A+ZYCYk7oaN2QDToFo1th8ZizqHFhuodpsTzrYFPJGHo2pLFh+1SeX7sMIgLGzgHNm2hzAH9+iOEPlKLxi+ib9NRfbiTlFzWMnlnJxsMTsI7oIWISiC4V0XlB22gE4NIHDmNvDKFd2I2qA8WrQzhno2B1C4+e+hFen5HcBXUkHRFomaNH02LE8KMOxNgQAPE/qSMlswclrGFJ9XxQ4PyEj9jz+DRyX1eQZvZR+UgOhj4FMSSwYs19KHqV2PMqxoI+7vzjJvpvj8HWCPc8tJmWn08hJamXh8d8TWNvNMEEBc9f0gmlSvxVHKerw4Gig0isxIZ5r+Co9ePLjKCmBfn9NR/wWcsYtH6Zvv1JxJ7vp3uYkbFXXkAUVQoP3YRruErWVj+/m7SZB+dvp7fHhiJr8I4L0Fcgo34dw/31P2LT+hm0bCzm3qhmSh9cQ/HRGwjGKySclkiPcqPTyDS8mc+Z+gym/vgeEk+Ecc0PoGhB0QlI8RKyARovt9KXB6XlGZDnIyW2D6a4sZ004SjTsOCKG3C5rRiTfPSMVhBkyM/oYOw1pZT/pYCq+wzot0bRPzaIO1fEUQ2PLfiSdecns/T53biKRCIWlRHpLYz79Uoy/ngCrVbh4Zg62m8JEeo2MT9lFAvtJfz89Xdw1MmUn8n8fgbXID8MVEBVv9ttkH+aQYdokH8pcj9eQd7NZ2h8zgaAsSvMuoyD9NbFkHN/J1X1SXwy/g0u/HgN4+KbqbzbQiBRJZgZRjGozBtTyjN/up6YXUaaPVF4cmQ8WaBaI6AOFFE9XZGNUuBFsqp4ZvswdSnEH9BjWNmOBoUpQ+pQNdA8T0/0SR3uIgVLq0AwQQWDQsQCnrwIrp0pmKsN7Dk+DL1bIGd9B6IMVesKOV6egxgSmJlSQ1+OBkEGYa4LRzVICRKffj2V3oe9BA7FgQKZW1TmLj3JxV8mASBWWbhQmolwayfGLgEh20f/3iRkj46NhyfQ4omCt+JJ3qmltCmFscPrmPjTlbhWeKlfamR0citJwzvozdeg7wNtEBSjQvTxNkYntfCb3Vfj/LMGVYQPHruc+JlttDbH8vzZuSTYvIyfUIVwfycxJ3RINgVZEhmZ24ww0U1WRheP1vyI7hFm7FVaTs9Yw7O1l5Jh66XxMg06D1SuMuGZEGC4rRXlRBSaszbumLuPn6x/nxtsPVxnv0jdvLewHjNRnN5O/AmR6becpN1j46qbDiDL//HT9fm411HtEr0FOsJPJtFeFU/PCBXVq8XcGaZ1lg7Zrcczz0fXjDBiv5YpC88TMYGUHqJwdT+G41b6Aka0X0fRVxRBUFXE/gCGCyYikobCl7qQx3no+CKDA8eLsZXrmFFYRW+xis4YQbKpaCSVv2y5krwXI6z96DLCmSFks0qKqZ+omhCNHxURvd7K/JRRZMW50LsGMuKmGkXu2noX7VMGVRcHGVwy+z75zhwiQRAMgiA8LwjCOUEQDgiCcFwQhKX/YHOnIAinBUE4JAjCbkEQcv83x/m5IAhnBEE4JgjCRkEQEv7hc50gCM99e5yTgiC8IQiC5btqxyA/fGqem0TqVWVsbDnGpk9eZfRTq8j7wMf9h/ZTv+hNHssaELALyDpSs7vRFfQzfEgLsk3myGejSbu5jt6h0NltJ/qCiBgWEHt1dExTiJhA9GmQPAYSJjoRai1YWoMEFvfRfCKVTafHUv5uEcFUCU1AwHFVG8ZODZ5sBV1RP4RF4kolhMhA/a6Mbb0IEYGChdX0jU5gzLRKZqw8TtQ5PUM+8PBN+xAytnTRO1zF3W0lcrULY5MeQYV+jxlDr0rWJhctN0ns3DkONCqaCgs6HyQeFvDsSkKZ4WZ+bjnegjCZ2V08MHsnWlGhe6RIMFrAeNGEWRvGnQf97Tb0bpHjB4uIKCJ/vPUdzj+6hptW7MCa6KX8N/EcOjEUR7mGoXEdKDoBRSfg/zAZAP1FE3XnU7m4qZCh0U5OP7GW2uWvoms20P90Ot4eM03OGHp8Zrwz/Cy4+QgO0cTRkRu5MeEomdtkQlGgcWuxHzXxUd04VC2IEnzy9hxWnrgRgFmn7uLK6gW88OCrXGxNImwX2Ll9HF+PeoedrUUcmvzq3+6HfJ2FvbNfRLJB03wD5lYNSUM7MXRruGzNPpSMALZaLT8qOIv9nAFNip9bEg6h7xdQAxpiX3eSvqkFT5+J5St2M3ZYHaG5/XTOTMTaomIyhXH+RUek3oqiB7QqRpfKuY5UFKOC5oIVY48A13eTNb6Fpnk2AqkR5hRVoEnyc/TNMWx8/xUSHR5a5qvIM8ews2grkuM/cqKF/6rDOcggg/wP813OEP0SWAxMU1V1BrAC+FgQhJEAgiAsBn4PXKaq6iXAZmCXIAjGvx5AEIQHgJuA6aqqTgLqgc//4Tx/AkYDE4EJQBTwxnfYjkH+Bah5bhLTTt/K5ReX4x4RoX6JlZcmXULhG6uYXeoDYP+xYTjLEkiwe+nyWxD9GmLLJMq/yeGNZa+hqzMx4rYLGHtUHNUCyfsH9HeiLgrUX/YGbd1RSKlhPJlGJEmDJs+LoV1H2CGQf/dJZINKkzOGiFlFSAkiigoI0DVKR+wZkaXXHqR5QTT2XDfnm9NwDxG5sKWQr9dPwtSl0DXeTk+PlSWbDpP1pQQC9PZY0Xkg4oigKzPjmemnYpWd/Cf6yfmolyHvygSTIkgW6BwP+n6Vp4d9wYGWIeyd+wJd+1JYu2khJp2EtqgfFrnw50h0LxSYOLcMQ6eWYG4IbUBAEFR++t6tAKzedyn+ejszCqtAVOnPV9CJMraWCB0TRbqmSdgqdADoPAIJp4Ps3jOa7O13AqDm+on5VQP2Uj26OhO6r6JYWljCZwcn/u07u+/EdegfdzJqUTlar0hfkYx0KJZAagRvYZhfrPqAJ8dsYer5q/D1G/l8yFcc8eURvdtE3nWVSNEKYz9/iHBEw4RNjzDymVV/0wc6GMgi5qKMdogHe4NCaGMiFXeu5YNnF2L/xkQwVuX0rcNJurKJ7HgXd3y6klBBAFOTjtJPh+J+dUD08bOX5tL383TCIR2zVh5DskKwNArpYCzRZRA1r5344yJdE2QMn0VxzaQT5M+txZchIysi3d4BParETBdfnynGdNxK3/Qglz/wIJbbwpjjfejcQbb4zORuCNP8iymDAdSD/D2DWWbfG9+lQzQKOKmqqgdAVdWzQB8w+9vPfwWsV1XV+e3frwFxwA0AgiCIwM+BNaqqer+1+TMwRRCEOd/aRAP3Ac+pqhpRVVX91uY6QRCGfIdtGeRfgP7aKOS1ieSvOEHVrWuRc1KouGsNP4mtBqD2mlepue5VOvel4gkYsTSLdI7WkTGlhYcuXEPU+E72XywgY3kd7qEKvfkiUTFe3NODTC9dyqiMZpaPPIkvVUSRRcy7B5Sbww6V6nfHkHRCRpVF1G9HUVZ0L0XPduHLjDD13pN8dGEc4SgVX1k0NbPWEYpVGLPkAqoIHRPh4Uc+RauXeendJTTN02Ns1DP0yW48BRI/vmQ3a25/lUhYA0aF8sdj8Twb5i/vrsVeoUU2qGgDAtLlbjb3jOHxol1s9RYz46ozxJapLEwsY8PYNwgdi0UIiFStyab6paHIeX7i9+iRLCqqKhAzdWA4qmaZqEqBE60ZJOd1UbC2mwOVebhztORs9JKY7Oam23YSsaiEEmRinmok748XSdqjZXrpUrQ6mfLt+fz2vvdI2xdElGDL5inMmlDGq+5UAPZesprOTRmc215E3EQnGQUdDF9cjrlRi9Yc4Tfnr+Dpt64j3eam7tK30QgiD8SU0DMjTNm2AhwXNWRuk8lwuLHXiETMoOnUM/rpVVQFkzn08msA9IwQ6Jko8UDbeLonyNivakfRq/T+Pkzz/gycX2ZgaRUwlJsIFQW4++4v6T2YhLFbRdYLeH/ej9xr4Ou3JuOd4cNRMzCD5U0XeD7/EzqnyCR/IzLjkWN8VjKWsuM5ZBe146qJIcocwFbkoqsijqsmnGLC8hKivzay/KmviLS2EQrp2P7Vh7w2fTrFz5USih1UzhtkkB8K36VDtBGYJghCGoAgCPOBeKDjW0dmLHDqr8aqqkrAOeDSb98aAST+g00H0PSfbGYAuv9sA5wFZGDud9iWQf5FaJsm/E2zZecX6//L58tq5/L87W8QrrCjd6sEk2TqSgYe0K6SeOoXvMn1ycewpHkIJg8oREfvM/JM3gY25H5NvsnJ4usPInn1+OZ6cVQLhBMijMpuRpRU1ICG6pvXIggqpbVpNF2dDALs3jCBvJROxkyvZNyMCgreXgkqFFo60PepzJ92jhtsPej0EXzZEslHZBSdSu1tKcweWc7LuxZwz8f3kJjQh6ZHR3FeCz1Hk3i2fR7+ZBVVp1JwST0+n5Hz3Sms+eUy1r28iNr+OLpHCawpmUGR3swlS85iadJAu5GuMQLrJ72FvSmIowZc5+LpPJfIrLLFJO/S4ho1sG4T2JyIpziW+8buR9+nUr/YivJJPKWe1IGlIYPM2aZ0mlYUEzEImLQSjo1Wko6H+HnJEuqX6AnbBWSjSuWfi9naOYKcDfcw65v7Kbq+nEixjyy7C/WlBKL0AYwuleFprUR/YkEThCKrk7JwgOytdzFp9cMgqhjcKr5pPhquhfMNqWiDKvElErZCF/7pXlyShdy9txGpsZF4QsbQpuPI2nHcPOUwTc4YVC349iaQcEYidVc3Wr+KrUlFlQXerZ9EICdMZIEbc5dCR00c8cdEwnaI32TGmyFgmteJNNTPjw6uYO7oMjomwZn7RxFzWI+t0IXhXj1CQpCGljhMeglBhXOPjeaUM52YG5tZ3zCRnW3nUJ1Gcr6+ncZXYvnywLj/oVEyyL8KAoMxRN8n35lDpKrqO8BTwAVBEMqBr4DPvt2yvzVr/4fdnEDOt69z/kkb9dv3/npeCej5TzaD/Jvx/7Xk8Hjadi6GUkEATxYIUWHS9ioMT2hDTgvyVHchv1l/A6bNDhZPOMPIhDZO/W4tk4wDAa9P7lnKUwmloIJcZ+XMr9ai7dVSszkPySyi79Yw9JVVGE9asMf4MPSqXDX+FAa3SuM3mZzbU8D5L4vQuwUmTK7k3S2z6c+Dr2vzyd5+J1KlHSGowbk8hKKHcIzCKWc6Y8dXI0QEXKcTUDUq5c1J6N1Q8vEwrl/4DYIk0NpvR2g10lMfTX+mhmCsQMOxdGbMOo8cHLj+IxtGU/rwGvQuESU+zN2r76dllom5K44y5G0nKy7fSUNtIp7l/RjbtQD0TQnScU2ANz5dwIQVZ9EN7UcbUjl4MZ/8JVUYmgwIgoq1RWH2A0fp+TidYIxIz1AD4ik71kaR/qIIxk6BjqtDdHhtmNO80GmgzedAbTRT/0IhzpuCnOzIQBWhL2xC3y+TtKSRjzfN5OoTd2OMDuIbEsZaYkQVBMxHLJjq9UTHeAnECeT+qhxPRQymw1Z2HBuJKgvIaUHEsErEBH1zAqwvmYipwsjIMbWExvpom6rlgS2b6bkkTOyOGkxDbR4OAAAgAElEQVQVRlwX40BUUY9EY+oKY63X4E0TiJ7hxDl5QFk7yhjAeNaM7YyRU8508kY0E3miF08mPF6wC09xLObTZrQdejrPJ2LI8mB7ogV3h400ixvHohoK3lqJvVakbu7bBBps/9fHxSD/gnzXGWaDWWb/R3yXQdV3MrDkNU5V1SIG4nyOMTB789eg59A/7BYCzN++/mdtpG+Xyv47m3+8rrsFQTglCMIp2ef7P2jRIP9K/HdO0a1v/Ri/bEAVQc4JMnVILT974V3OfzQMsdXI3oenohvbS/dYhW9ac1mXcfDv9lcFlZF/WsXsEeUMn1LD0LWrMOW7UQXoujpAfIlCICfM+UfX4GmyE3dtM3vXTUKa14eUF0Ac6iFmVjsI0C8ZiVgHbl1NuRWTI4iUKDFubDWSV4+SHAQFCuI6aXo1j0ien8rb16IJC1TNfov+4WGiayTWH7wE21AXrm4b2mwvqkEhkKxw8d41RFVC9ZNDuWv8QDuSD/ooeHslykgPJlsQ7cweFI3Kp6fHUXtLEqtPzSL2pAZ/vR3ZpOL4zMb/Yu+946So8/z/56eqOvd0T845MIQhpwEkKygsIOacXZENZnfX1Q2u6966LuqqgBnTghEDoiACKpLDMGRmmJxzz/R0rvr8/hjOr+dv99a78867tZ+PRz26p+r96aqe+XTXaz7vJHUFvckOAj46NhTTZjfNs3SEZuC/ykHCEQP3Rjttc0K8XjYWS49Bz4gw1nPa6M+NYDmznRfOepb4Y2EqZqyi52gCCc87sGR72VryDmaPoKdAITOhh442F9YuSdvGTHpzTMRa/DjrJCkv2bhk0D6GFTbizTbwZUj8yZJAokFXYywJM5o5tGI4lq6BjvTWVhV6TdBhQdElSfslekhF+lV8RUHMqk44oKHbJT99/TpMrWaO3Z9H0dmnSN+m4zhu4dWly2iYYcU/zodhBv/bKVg6FX5V+j6N63MIO6F/go+eRhdWNYJNCzNkahV/WH4pzeeHcJzZSsSlo+T04+u14nkwi8r5T/FA+kdsaCrjxPUr+O2tq6IxQ1Gi/C/lWxFEQggBPAQ8I6WsBJBSlgMLGRBJ/6pEvt622QL4Tj//pjam0+f7ezb/Binl01LKcVLKcaojmoz2z8zfutEIHV5aO5v8n+3gV2Pf58TyYfzqgesYfdkhrF0C+71NpLl6MfUq7B/32r8Z+4lfxdKh4qrX2bx/KId2FjLn3N14a930Dw+gHnXSdn6A2P1mSnZezpoFT1DfHUsgHtJdvRSktuPvs9C6NxVLj6Tq4zwMZ4RwUphgkk6kIgbCCnv3F5KwSyPjLRPFz3roui8HxzVNmI7bmXTHEpJGtTJ62Y+ZOKSK2gUCYsIo78bz4vTnUPbHoHo0Ct7wcW3dVNom64hb23hmxzTy3r+RqvPsOEd2EvKZuajwAN4j8Vi6BUsmfsobVz5CaVEVnRMiqKEBYWFoA3E5whiIm1GbLXBWF/lrDGzHrRy/JY3uQQqhRT3EbrdQ8KLkvj+sIiu3g8jbSeS8L/lxwRZu2nsF/T/yMOilm5GKxHZ7I3FOHxMOXEgw0UDzQfPWTO6dtA5/gkLqDj++s7yUrx9Mx/gB191LhyZyvCkFJQJvXP4IkRhJ1YUrEbYI9Q0JtE8LY5rUhb3NwF1lkJDXTfHIOqovho7Rgqw3VVKyurFVWrCpYarnPoe9UeGVS/9C3FFQggqH9+bx2YqnUcJw4au3oYTB0AVKGLpKw1jGd/HYiYHK33Pm78FqDWNL9nFsWz7Hj2TRujKPvtFBjE4zLQ3xFD/vw7rdyeBH+jFv2Mu8jDEcDCVw1qXXAnDb+1f9938QovyfJuoy++74tlaIkoA4oOZr+6uB808/AqR+7XgqcOr086pvaCMYiDUCQAihAQlfsYnyPearomhlTwbZj+wnFK9z8oWxPHR0Lu1nBvEv8vBC9ucERvk4uT0XfhZPKC1MUIa/HPuBz8oNm68jYgNbcwBHrYazRrBu6zikK4ztuBXDIgl3W7B4DJxvxXDxhz/Cag4TP7mFnhezyHT0MCyviXCsQee4CBdftBXFqmOtN+OoUQfafdRqXDJ1B53jdOrnS+au3kn7KCu+59NBgdazQ7QcS8aXKjnanoLiCCO6zHSNMlhWP5dLL93MqNIK2sY4KXtpOK7jGv5VaSAF9hoT0iTpanJjdwV47Y0ZnH3WXvqKwxzozeLKsms5/NYQkAJbsyB1bAvWbh0jPYAeGyGpLEzErdNXEUsoVsNXEEJEIOKQZNzUSfJFdbSPtqJgUH8qicR9vVjvaOJ3716IeZ+TuN/bSDwgMXsUTjak0NwaS9/uJEx9Cu7agbpPf1y7GM8wnWuefQ8OxRB2SVCge5BG1mqNkoxm4g8J7qlZjC27j4UVZ5OT3glhhZ9N+pD+I3FEruqkP1WhszqO+vW5/GzShxhpAZqvCKA9n8DLNzzKF9X5jLn/ZkLjvNz685/QMUqSNbSF9M8Mhj65FF/awJ0j5JbcP+E9TF6I3WumrzKWvspY0rd00RJwIYTkg/ErSfsiwuQxJ+g614drv4XsDQYxx0189N4rJC+sp3t4LHLKQI+yz/qKUT49EF0ZivLNiGaZfWd8W4KogwG3VdrX9qcBPillNwOB0F9GEQohTMBIYNPpXeVA69dskoHsr9h8CoS+asOAa04FPvmW3kuU/+P8643nvdIC/nDsU4p+sovbJ2zicOmrxJRZiXvRyW/bh5LzjELsCTh5pR3FozF6x7VMKT+P5zypzLH1Y6szMWhcLb35NkYsPEYgEeIPCQiohGMkSQcMUr5QeOGBZYQdgphKjZ66WDr2pNCbL9iyZxgdz+YgVcnaOU/wVvVITFVWYqrlQEVnmySYYPB2xUgKi5qJ26+xrauQ3uIIHSMF7gltmOosaP0CdwX017gxVVuJL+oiplKlsjORj385jWPrB+EZphOY2UcoFmKua0ToAnuLJHtYM87kfuyWEJESL+tPlJC0Q6PyhWL6PDZS9vhJ2KMSckPopRRKflOO2+WDkMLWZ5/hj7NeY+nZGxn8s8NUz3sWS5eCrUVw6qcFNPW6CMXAT167gdVnr6A/z0lFWRZZH4eQpR5cDzUir2kn6UCEzJRupF8j/qhO2CFpuThIIFXH1i5IzO2iwNRG/HEDzSeIS/dg7ZT0J2s0P5uPGpTUrs/D/p6LTHsPNXVJWNpUlr23kIRDkq6jiZi8A9/83rwITz6/CKXZiqhw4MlXuSdvAnpEwZ8sSHvJQvdghaT90NARS+tlASIOie6KkD6hiZhqwX07F6EGJP3TvGheBdUnqL7XxLH2FHy1LlZ7xtIwW6VrsZnCu7tPV70WZK5vp2DNEmYmnaRzfgDnHxoBeDClPNqsNUqU/wN8K4JISmkALwLXCSHiAYQQY4DZwOunzR4ArhRC/Ovqzo0MBEO/+pXXeBBY+pVCi3cC24HNp226gSeB24QQ2mnX2Z3A6n911UWJAgOiqOGGEi5/5jZqfjeJp1+Yz9z0URy8azlNF4XYNTWR3jv7cLSEGTqiDhkXxvWOE/3lZB59/jzG/ekn/OqK1Rw/koU/SWH3zmKC+UE6Rkuy8ttZfM4OQg6F5B/W8KG3hLiTA6Fvzsxe9PwAoVgDV4VK7HX1ACze9GNirEHMPQLdCsN/cByTV5A1shmj0kl4WSojrz5MxduDECGByOmn81AShgVkYT9CB8MdJn9qLaGPE+mf6KO/y0bEKvCn6VQtfopjU14mdUojVU2JpH4mmPXTHdyS+wneLjvGO4moR52My61FNw/0bbth9Bc0TrPhrg4xb9FO4g71MNpZS1hXuWbKNvLfvol737yMN+pHs+nQEEbsvpRAooE+3YO5R+CrdCNG9CIL+nnPM5qWUgWtX1B/phnHWhdV3Qk4/uim7jyD7k1pKI4wzdNAjwtj2ecgpbCDm298F9OqeB6o+wHdgxTs4zoI7kggGCtI2tZKwo4WOkcIJl9wgL5swYd7R6B2a2R8GiJ7Q5CuYQLdqWPuk6RvhZj0PgLJkpjBXWRPbsCfbFCxaiy0W1BDUDtfEMwL0lUiGJdTh3Wnk5wNARxVJmoaErH8oA1zjZWY+gjqsYGaUAVn1BJudHBo4l+RAnZ35+Ie1MXCLUdIWOOhP1tn+/xlHL85HiUs+PyqscR8bqN/WvuXczG6OhTlmxJ1mX13iP9/fPJ/8oWEsAO/YSD93QfEMCCSHvnXIOjTgddLTx8PAEu+LmSEEL8ELmBgxanptE3bV46bGSjOOI2BBcEy4BYp5T+MmLZkZcn0O279r73RKP+nuPbMrXx8zzTq5kPmRoHZEyHh/hr65gYQORkcu8WN4lcx3GFi4nyEDsZhDOrHvt2BZ3QIJLjLzHhGhEnYrdE5IYKt1kTuyhP4JuRTu0AQn9lDV2MsMSc1HC0Gvos8HJywmtEPLMXRqtP4gwjWGgtZH/fjT7HSm6uS+sh2TFvT0IRBVXc84bI4sqfW0ftsJp4ChbBLolskF07bydoTIzCVO/Fn6MQeVgjHDKz+2K5oprYuEaVXw1mnEJnqISmmn+ZdaViG96DviiPikEgVLF2CQJIk8YCkZZpBwl4Vy4WtdPY6iNQ6Mef1Eeg382DpWi6J6QZg0Is3E47TcR3XmH/1Nl47PBbnPhu+dIluM8j+yKCnwMT4yw+y9dMRKEGQRf24PnHQWwBxR6FzhMTSpRAe4aXojg5Mf9Wp6EhE2+bGO9aPaLWQukPSvCiE0W8iaYfKObd/xhuvTwfAnxEh7VOF2CV1rC9eT966G3ls5qvc8skVpH8iaBunoNsNSsec5OC6IUScEqEPuL7y3g1TdYHK6JJqTnxYhC8nQuJulY7xOvlFLXhey2DodUfY05BDpNqJySNAgLvKoHVOGIsjhMkUIbw/jkBuiKQUD/0BM+YtbnrzDZz5HgJHYrG1CIQOw644Svvkni9Xg+amj6Lrukl0lUTvSt8Hmv78KMH6+v9S/5WY2Ew5Zuot39YlAfDZurv3SSmjNR6+Ad9m2r1PSnm3lHKMlPIMKeVIKeWyr2aESSmf/crxM//Wqo6U8vdSytFSylIp5XlfFUOnj4eklLdJKceeDpa+4ZuIoSjfT17YNIOmqSrVi56mP0Whc5gFRUhGbOund1mEjJxOTl20kuq5z2H6MJZBM6pIf8XMn299CtGnIXwqcQsbiU3uo6dYMuiZAMKAyZsb6Rpswn1MQ1UGeqRZZnXQOkmSebufwldvxjBBzMajZL2rkrnVTzDBQuNMsLcanFwxgcq2RMprMsiO7UHocLIinc4RAjUIhb85yIRxJ9n9i/EY9Q78qTop2wS9+QPBzv4kgf+VNGIPmDF5BXJGN5FjLlIdvdx2/nvEPRtDxCGZOPsI9mbB6z96GD05RMeCAO6jGia/RHk6icTX7ahBCNU4EYok3dT95e9uxBkVqF6Fj25/iLWVI8l7TtCXrxNJCjFqZBX9N3s448p9fLZpBGT6SSozMB9wIsVpd6BbIJUBgWHb5aRkXRMnthSQ+qR1oH1Hg5WsEc04a/px7bBhcgfpGiF59ch4dLtEmgCbTvcF/TS+m0vh1muILTNx6/ZLcB/RCF3dTdp2HUu7yq6qXNJ2BNDNkDapCWkxCMZppG8RHDiSR/qZ9Qirjv8HvSTsVwlENOxtOgdfL0HbFUMkMYw0geYHb4aC6DHhdvgJl8XxxNVPMfSBDoIbk/B12fGMDyDNEtO6WHSLJK4iTGK5jz1bhrChqYyCNUuYmz6KymWlUTEU5T+GBAz57W5RvjHR5q5R/ukxTJJr66aStHIHnhFhuu7O5rOHSmk7kMLdBR8x99gPyHv/RlI3NjEr8TiTH9jFDRuvRwkJsotbSbP3Egxr6A6D2vkxJO8L8eqJ8ZTfuZzeYp04qx+XFsDrt2BpV6m8Po3C1/rwjvNTtyqb2NvqKH74KC0TVW6YsZWeQQqOlH447iTrTY3gvSnoZkn6JwMCYuTio5x8upj6R4vw5GkomT7smV68WQMf17AD+kqCdJwVoGdckFCswTk5x3jmshWYlQirHlxAx3CNUEaIPRtKcLQYXP7HO3CWW4h4TfRN8NMy1aB5iqClVBBKOl2QcquVQ4Es8tfexFnHFrDveB5vnv8YZz15N+q+GILxJoQuwBCcaE/m8rzdbK0v5OoFm6mYsQrtplZiZ7bQmw9x+V1IBVwFPbScE8KbZfDOukkoIQjGa6glHlS/oHFfOo0zXDx112MAmPoUktZZyX2/n0BWCK3dTLDFTu/IIHqvmbSNzUi/hm6BiK6AgLMW7kHqCtULzQydUM2WYe8ycXglreMV2hYHianQqO+MxVJjQT/oJu5kgPnpR/AnqMTPa+Taaz4iPqkXZbgH3xg/IbfE3KXg/ziZvOk13Pz6DxHPB/nirmWgGagmg5x1BlKFystWUn+mirKt7MsA1sLbd34X0zzKPwvRoOrvjKggivK94IXsz+m8fhKDbtzD1c+/z44/r+Tk1Sv4Q+U5NPS4GbKsm7bHLeSb23gwpRy1X0EY4PFbMRCYPnehOMMYg/ppmmYiElYZ/eBSzp5wkLrPs6n3x2H7xMkfrnwJqULP/QES4r34Ou0cb0phy9qxTJx9hDeqR2Ma1U2gOoZgZpiuIRpCQtL4Vsx9BhGXzilPArLbzMJfbaZ7VARZZ6e/xUHCkTD25gG3jujXsJywQUBFDSh8WDuEm/ZdyRmxlXSWCJQJPcwacgI5tI+mWQa9ZwQweSWpWV3kpnVSvehprHl9GCZI3KUhFbBe2MqHbSXYGlVOHc4ga53guoduJTDCz+GfLqejROWh+X8lI6MLX5ed1+8/m9BJFy8em8j0H/4QfUUKbWUpGGaJ/CABZnajvRPHo5PXkLIbrMN7kAr4khTyErqI2CWSgVWZ39QsQjviwD62A3lFBw0znZjsYVxDOrFneMldI7Al+mibnkpcugepQsKjdnqu6ePA78egWcMcvOhRimNaATjVnYjuNEh5y4JvrJ9IrRN7kySQE6J6oYX3G0vomKDTtiWD5Rvm0HcoAWWnG7XGSiglwvhzDhOMhc4XcpA5ftYN+hCnYmXTrMcYntnIa888iuuCJvLfvun/tW65d8eX861yWel3MMujRInyXyEqiKJ8LyhYs4SE53bwTN02XmqYBMDsK68n7g6VnJ/5aH5Io6Mhlrtev5ri525Gd+tkjmki3uFjgrsG7zg/RkRhcHor2R8FkFLQUxLhcFca7gqDXeWFeLPhofuuoOiJOiJvJfHjgi2YYkKYygdyBLbtHUJPSwyZN3Zg6VAQfhXVD5U/VGisTaDpDI2il4K0dbqQDp2Vu6aTkdNJ0oGBf/Nqz5cDN3eHxFGj4p7SSkxaH0ZaAKspQqDDxi5PPrElnYTDKg2lXiZk1aL4FfSgSvdIgzsLPia0PI2zji1AP+hGCUP37ADuk4K2/SlUb8gjdVYDli6FlkkqPYMlSp11oC2GQ1IZTEGXAhFSaJ4TwX0CqHBQP0cQc7IHwzTQVkQJQ+LjdjqmhrnzzatpnmEgAWGArVNyYncuhlky+oyThNzgWZGN+5RBd2cMvTuSCQz1o55wEPo8EcMQNJ1hQtcV9v5uBQkOH0lnNlK9wIzYGkf7KJVIwMTM+25jy/JSBj97M949iaQVtsON7WS9rLHhoj8RcglSNmvoTp2S+BacpzRSdwZJ2gu2VoG3IEL8mDZiD5jY05DDlmsfYte/rODcwQfp0PspDwV4qWcint9mM+Wvd1JTnYw0Gzy96BlOPjWeMw/3fekqixLlP0s0qPq7IyqIonxvqFxWysw376Tt7Wzmpo/ik5efo+LqBB78eA1vjnqWzecsI5QcwdYuEH4Fb9BCTXk6t8dXkbx+oAhh08t5tEy0oQdV4spUgq+kkrv0JC/MeRYj10/bgiBHf5mB8+JmHjo6F6PBztkX7OTJ61dibVXJyu2g7roifFkRFpbuo6/AoCizjdVnrSSztJHaeXbuGrMRU6uJT+c8StPJJPrTFFAl9gozRoeFny1aiy/TwPRUApluD2nvmkm8WyFpp8owZxM/yDqMXuOk4omJlL1ZgmEzSP9II65coSkch6EJwrpK2GVgzvWyeEgZ3WMiRDKCTD73IHflbqB41imUoCBl54CISdxoBQHPHJpCz64UXMdV7KfMdEyKEErQITZM+C8+1IBAWg2S1pTTON0MEYGtRXBywQpcL7vQ+qF5qkRmBLC2Kew5XIASGSgI6bvYQ9Wc55g4/xC0W3CO6+CCy7ei6wq6FdTDTgo2X8uphiQWpR/E2qHQn2kQzArh2m8huKgHf7JA8wnCTkkoojExqYbeHI1Lf30X/Tk66hVtpG5VOdadgtkjaZhpJm5zFSavZPTQajoOJhOMg7EZ9dww7TKGPb6U9W9OYu7v7uTOy25i91UjKP6XI2j9guLCJlxHTfx5wfko/SojrXVRMRTlv060dcd3xreWZfZ/gWiWWRQYiPGoWDWWIfd38sG2d5jwi5uJe3EHBXusTHFVcHlMJ3e1jKY3YmXj4WFkv63QNkYjb2YNFbtziD8CfVkC0/huvDVuZHwIc7UVkxd+c+Mr3LnpEszdKlkfB7nhqbXc+94lmPoF4SI/eSskjdPtJE9voqnLxfS8U2wqGwoKJG7XMMzgmRbAfNRG/NQWdCno2puMsx56pgVQmqzEDu2koy4Wa5vG4kXbeGPDFKzFHvgiFmFAcKIXTdMJVsfgqhR0jwuj2nSMTjNpg9rpX5+Ks1Fn2+NPUfD6EqQiwR3md6Xv8pt9C8h+TsWbYcZ1dQNVhzJIHdzGzwo+YsWYsYj3Yji1PQfzUA+WD9wkX16LN2Sh8WgKSligJ4dwHLEgdHA2GXSc60M74iSQqjNhVAX7arMxDEH+Sgm/7aTHb8P8UjzWHzbR2hvD3UM3cP8HFzBoRSttM1IJJAr8yQbCAEehB6/XysS8GvZ8Ppj4I1B08zH2bB6C0AWWLrC3GbRMNSh6OUjVBTYGj62l9cVcuodJUnZJAvEK3VOCFGS0k+3opqHUS3DeeDSfjrivner9mVi6BEPmncQbtjApoZrOsIP3d4+m+tynmXDPzbirA/SnWrj7gVc41+FlbvpAAcaoGPp+861kmbkz5dhJP/m2LgmATzf8PJpl9g2JrhBF+d5RuayUcYU1fLDtHYq2XsPuP6ygYtVYPtw7gt8fOod3+p0cuTiPPS3ZaG0muoZq5KzvJd7iQ7cbeDMFgeIAfbVuDFeE7NUqSWU6/lSDfzl5NoWrw4iIoOoCE79/9lKKXunBXWEg2yx0ltiIOCUN7XHsmrKSo38ezvDB9cTt1/ClCmLqIpiP2bC3SszL4pFSEEoPowbB8GmoAQhGVDSviojA6j0TKZpYS+iIG98oP750g5DHgvqFGz0+gragA3SBHlBJ2yZobonDXR0hbFd4r9+O6hOMGlWF1mjhmdvOI+LTqLpcYL+yicqKNKytCmekVHHvymvwzB9GzSe5pG2PkPyEjWX3rMBlDlBfn4A0SyIuHUuVBW9BBF+6QX+KQsF1lYRiDXYvXEarL4ZIrxmtxkrrRDs1+zLp7HKS8KMaasvSkXvcPFA2j1/Nf5Pr139C9lWVJM1uRCaEiKlWUDfEMvi+LvY3ZhJJiOA7z0PDA0WoQYGzTsLMbhJuqsVZo9GXa2PQmDqOnMykc5REyfDTMkXQUywxV1upOpjBgVXDOfVwKd1FJmoWmOkLWvjVgjcwTFD/TBEnjmew/caxbHp3PIOW7mbYE0tZ8evHMDX3sv2Rlay89FyGbr+CS483RcVQlG+NqMvsuyMqiKJ8L9m3uwiAtyavpHD1Eoqu2UfKFwpZDytUBZM59YCTpEUniakB16wWTt2usfuzIaR/Cq/e+AjuXVbSvgCCCl1DTNz6x9VY8/pIujVMw0wbU+aWY072oQah4ko3Z971BZZOBSUCYbdOVnIX0x6/E+9lHrr89oE0+uIgBb8+hhjtwdIrmfXwNqxPxiFUiX5BJ1qXhogIIvvisBZ5SN0dAk0S1LWBPmT9GiavwtLJm7H0SIRfxbYyDlO3xsjCepw3NTB7yHF6CjVCbsHu/gKmzDpM/apC0nbo9GZrxO82IVSDutZ4ctcaIKDRH0tfUYS2BUEyt/gI/riLhhlm7rz/Ziq6EjE5wljaVdJzOxAGJGV1o9sNwk6ou2UUjnqFiW/cQVOXm5FDanHWQd+wENljGrEds1Lzbj7pw1txTWvFdNDJIV8W95Uv5ND2Qhr3pRO704JhBu80H+3T04lb6yB1s0rWPWF6czTOOXcnpgva+GjMMxytScfaIWk9Q3LsZAbCqpNW3EY4oIEBRozOk1c8ReHrPvpyQY+N0DsqiFSgZ28SDiWEEobOkZL18x7lnBc+R7dIhuzTcFUb3JM3gfVb32Ju+ihO3a7hsgf43frzvsupHCVKlG+JqCCK8r2lYM0SrvnjbVReuhKAjlFQe7tkY9sQjFoHJ1eMJxQr+GLE2yjVNiIxBi2TBD8rmIxnYoCkH1UTV67iaDHY2DMMucdN24xUnA2SrdtLcH7spD9dYlgkHzw7lS9uepi+PKg+92m2DHuXQ7cu5+CE1bR1xwCQvk6jO2RjXt5RPPkKbzw/i8ZpGjKg4jkZj8kr0G0S3SxJjvHiv60ba42Ztg2ZXLbwU1zHTQTjDHy6BTUoSdqpUDcfTEW9HNqfx6UZu9mybTiaTxKZ3cOaTVM41pVCxwSdt5Y/wuhrDtGfISgtrOac4qO0TjCjWyUnnhkCmkHq22YqL7bQcTCZUGqEmCsaGZXURM6TgtgTBl07Ukmc2oxtZRwA/qwIjiZJX56BpUtBCEn5wVwcrTrmZhNVNcmER3pZ8eMnaDiRTEtNApYeydYnSgeqhjsMwrE6fTnQOyRMzpOC3BtOEooRtI8RtJcm0j0mwrtbJ9C9O4UrTl5GzAEL4vwObpvxEVnrBVmva/RuSsXcZPJYvwUAACAASURBVOaq2Z+R8qnKr35+A9W3gqNpwLvhPGYh8YDgrgvXcr6zF19+GHujwrxNP8UqwqTs0dny0gRa54ZpeGsYc9NHcfnxBowWK+3HE7+byRvln5NvO+U+ukL0HyIqiKJ8r/EUyS9vcAV37uT4GS9zX+775KwPMvT3TRz+6XJK717C7quXsbh0D5mf6JQeCDAmr46Ox3PpGWLgquzn6B9GYJ3cgacQ+s72kjGsFc0vsXYKYo8oJO3rZ/znSxk+rYIjIf+/uQbNpCNV6B6kUt6QwfuVJQRG+FCDEltxD9YmDXuBByUM8Yclul1S3ZBEa5sb05huQm7Ji2WlWLolIiFIjT8BX6pC+3gDW6NGuMKFtV3h/p0LGPR0B/t+vQJvtx17k0LHsYEb+oT3bicsFY7fuJydp/JYv300/rwQwbQIPWf5ESaDniIVBNjaBMUFTTR1uxnmbMKwqATilYGMtU/SaButkb0eBhU14WwOE39IYO6F42e8TGZxG5l3VxBKjGA/ZUY74uSKLT9E8yqgSbonhOgeJmk7K4StScWe1I8SgeJnAjRNtbN/ZxGeYgmKxJcmEGad9M8NdKvE81oGWkDSURPPY+vn0ThToWO4hrcwwrCplax9dgbdgwWeAhXbDichFyhejeQ5DbSfGWT5I4uZcM/NZGR1En8sTO4bsLYkFe/1HjLWt5C4xUzm+UfY0FTGmqmjv4vpGuWfHAEIKb/VLco3JyqIonzvqVxWSoG5jcimbIqfu5n788dQdR10Tc1kyMql9KcpjH3tNjauLsWboVHnj6dsZxFSQPpnUHGFHW+aSs+peGbMLCfR1Y/vjVTaJg6kn3uKJRU/1LBYwhxfN4hhZtu/Ob+6y0UwPUzenGpmFlQQCpjQgyppHzXhXOPGsEj6WmIIxknaJktMHgVToxnHEStiSxxIsFRZ6c0TmC0RKh4eind4EGkzEKM8hF061k7J5EGnqL44mbwPb8BaZyZ2bjPukwIUScxJlX3vlTDs8aXIoIqW7Ef4Vc4ccRTrfjsAgSQDW5OKs8ngyoydhBodPP/K2VRdoOJo0fFlR2BSDz++5H3qzxbUfZ6NtbYHcW4n3lIfkw6ej+/1VHbtKSZxl0b6mfUEEwyemv4ikYwghAW2Sgs5I5ugx0woVpL8tI1QepjhTx3G2iUx9So4axXSvpCYvKA1W+go0dCdBp4C0C0CxTdQQ8pVqeDLD2Nr1Di8vRDNL4k7LnE2GMTNbyKYYGCYDXzPp5P6gRlPsaR9gk6Pz4Zxawf+W3rA0PEFzESSYgifN1DFe276KE78ouB/fJ5GiRLlv5eoIIoSBbjqnaVoZ9bhaIRfVpVRdebzCB2uu2gDP7r2XRIOClYufYKkHd1UPTCEjE91zL06bWMUlKCCN0diOHQ2nxxEy+FkukYbJOR1c+F1m7E3KSR+ZqYosQPdAm95XdzRPAaAgs3XEoyX5KyFk9tzsakhtAYL9goLDQszaJ0AoYSBfmLZH4eQ9ghCgqMe0ufWkbW4GhQID/ITKfQT6LSh9RtIv4rrqInnRr9E7nuSnqkBQobKvZe/hvOEmbBDovwlkdDZHuaPLqcv3yAyykvYJZk0tJL8q44xeGg9hx8djrcwguOQlYJRDcRMbaNzuOC+HefiPiFI2R3kzDFH6M1VWTR+P7Z33Kw4Pg2tT6FoZhXH7ogj9HEios7Gs0NewXJhK+dN24UvRVBZkUZ8ueA3916PUCSO1H6sXZKalgSUCEgF2keZyXpXYWPdYIJzeymYVY03x6BziErcgkbCyWH04V6sKf0kjGjnmpvWE3sCCsbXEYyDIXedJDTEh4iAOLcTQxVELAJVMTB7FAr/GqK1VNL6gyCOAg9KUGHfxFU0HknBszOZDU1lJLr62fjmiwT2xbOhqSwaQB3lvxfjW96ifGOigihKlNNULitlz31P8uulNwIQvKKLV5+ay6OHZxFyCqZYFaruM+NLUrF0BenNMbFk4QaS90rsTYK4lF4cB2zYCz1MH3uUjvpYtnUUoPmhu0RSXp5L9vQ6fvHW5Xz++ETmpo8iIc7Lp5f/iYaZGhnjmyjrzMTeLPAPDuDNNUjbLolL95C6oI66uWZUs0HccZ2U7d009brofDoHCvsxH7WRntiDCAnaxpqIy/DQl2twxc7rqTkPYnbZqOhM4oHyedimt6PoUH9JBF+tiw/2j0BNCuD6yEHSAUnZ+iF8VLubNFsvvbkKap/K9deup/JgJr5NyYyfeQz6NcS8TmrPMbO3JQtzj2TD+xNonxom1u7HWQftK3OZP7qc3pIw1g7BL2oXc076Ud7dWErELnlg5lsE4wTata0UprejKgZ33Po6QpEYJompT/CL616jYbEOW+MQO9y0rcrlktlfICRY1Ai2U2ZCPhORiIL6QgKrVs5DtwpOHM9AL/Fy7I+DsB60k/euF0NC3FX1eIqhfncGoQI/LZPtaMl+YrdZSXzcTtHIeuYeuZDEMkHELpk3/TwcZ1cx+4rryb5/e7RrfZT/dqIus++OqCCKEuUrDHptKR0lJgCuzN/NNUvWo1c7ueHH7/OcJ5VwQMMzCE5dZKMvH966fw7ada0k7/Oxa+xf+cGV24ixBtm5fjiu4xpydhMROxgWCRIqDmdiZAfozRNYPk2lvdXN7N1LUMJwReYuWrpjWH3Hw9iPWIkvF7SeF6S7yY3viQykIhn001pCV3fTPj4ObVMsvhSFcIsd5+R2Gprjyd5gDDR/3ZeAtOvkJHcxdkg1qU/vo6c1Br3SyZW5uwknhsl8zQQSTO4g1NnQLRB0CaydkoLXl7D9oxH4hwTIf8fPE+vPIX5QF+FJfeyqyUWJDdHVEYPu1OmtjsXcLxElvaj2CE2tsUTm9uApUPjg4HCSP9dwzm7lkdw36dOtJJRLwm6DP7xwMcn7AgyJa6HtrWx6W2JYXj2dB8etZd6kMmzjOnm/YySxu8wYZlCmdONPEqw+PA6zB6q/yCZ1TwjRrzE0rZWWyYJAIvSMCKOEFH476n0yPlawt0jaRzvpO5xA3efZ2FoF95z3Fo6YAP5RPkzlTtynQqxe9RdOVKdhmVOD0CX5P9+BrG8CQNu8L7oyFCXKPzlRQRQlytfwpRtMuXUJb/z2bG6Nq6HiyhX8KLaepx9cTPns5Vw67zPcJwSvXPwX/PGCQESjeYqd4reX8ta6KTQ1xFM8+xTCgLYfTcI33E/h6iDJuwWmVB+GLjhz3j6mJlTgPGbGssVFJD3EH8vmEu6zcN6qO9HH9dGXK9CO2zF1q7RcGES3G+iv23E+7sLWbeBLk5h6JUWrfTjMIc4edoSuYhMJRyIUzzqF1qVReyCDg9uLWHr4EJkfKoRTwmzpHERskhdfsopUJY5tTpJHtdI9Qufcn26ht1Bib1QIOw3S3zVRedPA14THayXU4CDSa0aptSJUieZROWfKAQKXdaPudKFW24jdYSGyLw5buyRur4m2STod5cnMWXsnb2ybyI6HV1J1wVPI8R7y/niCTz8cTd8UP+4jGjkx3Tz+80uo6E2iML6DgxsG05sPvhF+lI1xzLlkJwUrJL50iRIW1MzXSNivcGpdAYYrQu7UWlxHTSgBwWNVs2gfqaAFJF0Tw8SPbCeQHmbiRQcJGCa8tW70XjNxJ3Reff4xJr1xB2qPRsVjpfz+/mcA+LBqJyX7lKgYivI/QzTL7DslKoiiRPkbtJRC3q3HKdl5+Zf7dv1xBX/ty+fT1iKm3bCH+y6+jkCCoG9XEmGnZOHkfTx26fMMumEvfWELSefW0zM2iKoZVP0IrF06iiKRQZWtb47lpVVzOfvSHRgm+NPkN8h6VuOqCdsJJujYP3aSO7WWxMM6CHDssCPsOifrUmi8NkTcLbVYSnrwZguqz3VQU53Mh/uHEzu3GfmTdlr6Y4g/DDG1gkhimF89fg2+67qZP/wQ7Y/m038kjs6xBiI+RMYF1XRvSyUxu4f1v58BQPrnXhLLoPkMQdr7JnRXBOWkg6JX+hC2CBGXgXu3lUh8hAOdGV/+joacUUVfHli6YeINB+if2U/aVoUL53xB3GFB/tthSu9awuBtVxLwmdm2fiTBtDBGWKHsF8v5a94WGheFqSzPZO/+QhYu3o6S6aNghaQvT9IccNMxwgaF/SSVR5Bmg87RBqUXHMR1yEyjx42hgW43CIY11KCgc7jgxvGfEwxroEk2lQ9lSWwjVReupHhQI9Pu3cH1oxfy9KJn0LwKtsw+HioYzoamMgD+nLb/f3TuRYnyvxkhxD1CiP1CiJ1CiLeEEMnfYMxgIcRmIcTnQoh9Qogr/4bNJCHEDiHEZ0KI3UKIs/+GjVkI8QchREQIkft3znXD6XNsE0J8LIT4xhkQUUEUJcrfYcfOwRwufZX8tTdxxk9uAuCH7iYCL6YSMVTqzokheXoTeWta2X3tMir7knircxyey0tR74uj881MCrPb0FttpLxjxZNvwvxFDGZXkNBoL3fd8DqfPFOKVGBd10hS7q9i7YvTydwkmfLDvZyoSMeXpKD6Bd5SP2kfaLj3W0iP7+X49jz4PA7DLHGfHEhLz14naGyLxeO3snPUm1z7i/dw1UTIXKfizZJ0dznZ/+fR9BSqpOw1MHUr5D0nOLE7l1CcQU+vneYZBs46haapTiIWgWEzCMYqWOMDBFMjtJa6ycvoQDoi3PLjN4k5bqLbayesq/jSDY7syieupANDhaPdqWiazrxfbuW1w2NJKPfSf7eH1qkGZnMEl8tPbGkrwqpDn0bJX5ZS8pelqC0Wsjbp2BtVtj4y0Ii3/e4AeWMaOPnCYAKJAucnDurngKlHxdqm8snuEnpHhAged6MGwZbhxbEqlqSyCOZewcafTyPN1YvQDC4Yu/fLv3FNRzxlPZlc8kU5vzu1gGBKBMd7Lk69Opoxv7uZgjVLonFDUf4H+Zb7mH3LMURCiJ8CVwLTpJSlQDWw9h+McQIbgVellFOBRcBjQoi5X7HJAj4E7pNSTgNuAt4UQgz/ik0u8CmQDqh/51yLgAeB+VLKM4B3gY1CCOs3eX9RQRQlyr9DwZolaB4Fx1u7yFt3I3PTR+FoDvPh56PRh3h5ctBq9HgHE164nRM7c/lkTwmeQoVTSxW6h+tY1AjWVoXmaRJHs05oSh/js+qIdNh4oGwe/RnQn2WwvyWT3dsHo83opGG24NhtJWAIYhoi2Fsks4uO03yGIJAkafssnavmb8HSJTHMks5pIaoWO7He2kTe8wJ9ZxznVZ7FHz+bj/m2ZqQCi2btQqiSjkV+lMndNC0OET+2jVMXa1jbBLHHBUIxSPtUwTGvBW9eBG8OvDb3SUzntRHwmsl/zUDokqqaZGwxQVb8/nzOuHQ//g47UgoMdwRHvSD2QQeJh4MYTyaj7HSz6uMZGP0m4h5ppMdrR1h1nGvcJDr7sZvCFD0RRvUpOJokgRG+gXYDt7QzfvEhgrEC+w4H3mNxVNam0FsIUgV/iqD63KcJJ0YQEbA3qGSvVdC8gkdvXUn4pAvX/mZ6czW8uRHq5im0v5zDv5S+xebGIgrWLOGq2mmE/CaOl2fz6GMX0rIjneJnfQy66RiVM1/AUxT1N0T5n+d/a+sOIYQC3AMsl1J6T+/+EzBZCDH73xl6DWADXgCQUjYAa4B7v2JzC3BKSrnptM0B4DPg7q/YOBkQYy/8O+e6D3hZStly+uengETg8r8/5P8RFURRovwDdKtkQ1MZr521nLo3htMwy0z8IUFsjJ9F229G7fRy2cJPSSqT5K2NYJglOS+rOFL7iTX7SZ3dwOQxJ2g8SyKPOalcPhhri4ptlxM1IHDWKQTLYzFSggR3JJD9oYHqCxGf2QNA12idzZXFOOsUQvEG/vwQL3wyg748SN0uUdrNKGFB55osmkutxFbqtP0lH2uzRsuHWTRNhw/fLuXUrBcwHXLQ220HAaG1yVhaBwoXikWduD5x0LnYh29dKq4KDdUvuL7sauymMACWe5thXhcmZ4hggxOT36C8Mx3Vq+D3mVF7NJbc/C6nlghalgZQQpKQW3LlmZ+RvF1lb00OgR4rpgYz/SkKNXsz6ViXSX+2nYhbJ+36Kuz77MhsP/fkf8DBVSXoVjDN6SBueAcLRhxEDQikgEBBkCtqZmBu0Vh65fskHg6T9YuTMLqX+2+6FstgDzWXZpK8uI7xI05hTvbRMUHniD+TmJVutlzwMHX3FzPk1x3Ys/roGWZg9oDce5hXcrdy1sXXfpdTLkqU/42MAFKAL5dYpZStQB1w1r8z7kxgv5Tyq0UA9jAgpOxfsdn7tXF7vvq6UsrDUsrKv3cSIUQcMPZr1xcGyv7B9X1JVBBFifINKFizhPvyxpN94SGkBv2ZA7Vskt63cuqaFI70pbF92Urq5pg5ce0Kqi+ClCet7Nw+mLq9GezbOBRTt4qtTbD69w+jBsCfLEk4otM7LIRlRA/mGgsmL3Tc6EP7cxfivQR68k1gNkiM60Of7OHCKbtITO5FS/cRe0LSUqogVYkwoHPMQAHGsE2hbYyCbpMExvi4+8x1+AtCjNl7MYFkA1eZBdltpnukQSgniOpV6DmcgBIG52YHuhncZzfjTzXw1bpo2ZwJUlDdkcDdxRtxfO6EhCBNUwVdXjvW/D6GZTVj7lL483sLMQIaZ+WeIPCTbnQLrNpxBrbOCHmpHSj9Ksn7DXpLwggJB+9eTtP0garW7U/m4U+WxH5s47aDF+Gd7mPmRXtYlF0ObySy7uhwAtkhNB/MHnKcsveGctbZ+1m270xqz5eUt6WxfPRf6Syx4LQGUSb00PRhNoc3FJP/myCx5RqvvTuNoEtlZedktj73DEfvTSbjvCOIsEDoULzXRMGaJVQttnzXUy7K95X/vS6z/NOPzV/b3/KVY39v3N8aowC5/8Am5Sui6R+R95+8vi+JCqIoUf6DWNsFeomX9i4XO/68ElHkpefuTIYuX0radp2zcyaQuN1Ef5oJ0gO4K8E6uosFc3aRvLiOy+++k40/fYjUnTrNFwZJ+1ijvyKWYEaY2KowPq+FiKHgm9tH79ggQjNoaYklcszF+lcnY0iQpxyEYgTCgLShbRgmSfWip4nYBSU/OoQ2qA/VJ6DexrL3FjIopwXj4wQADAvkvxXGkuzjp+M/QXfp2Ab30D45QiBeELFD59Y0iAshdEjdGST7bYVpOZX85dcX0zM8gmuHDWkyUHe5MAzBya35pO4JkTKqlZy1cNyTwqC4NubMOIClTaW5VKNjbRZqqo/+a3qI36NR9GgVk+5YgsmjcLIiHTU4UDKg+yw/gfoYjEYb7+8ezerVs/AnCgoy2rHGBOnPD7O9IZeJ55azcdMYBj0cwOYKIIC/NM7GMEFrVSL+SjfBeEkw3qBzXAKe0gBxxySzf/YFP0nYTvFzN2OrMWOcMYpTF68kbeU+/pK+57udXFG+30gQxre7AYlCiL1f2X74n7w6x+nH4Nf2B4F/T7Q4/s4YvjLum9j8d13fl0QFUZQo35DKZaVsaCojfX4tJ6a+hB5UmTdsJkaVk54iO3nPnsJ5rJOKF4ax54EVdBcL3J/a0M0CT4ObTS+X0vVqFpPu3s05f7qb1vEqeSsE8365ldgToNkiNMxSEULS9WI2Rye/wuLhB8AQVM99DlngwzDBuJR6TMW9hOb0MnrKSRqb45CF/cw+upCIDfauGUEoaEIf6iXi1llw1i6SbF68uQZSk/Tn6Oi/7OSmoZ/z1z+dAyYDdWMcSRk9aBO7Obp0+UAbjm4zcUcE1VdL6i+M8Nm60fRlKZhiA0SckPe2jlTA/f+xd9/RUZX548ffz/TJTDLpPSGNhJ7QQ+9FQEBERAEFQQVcFQXdr669LGsBRVe6ggVFFhRUOkrvJaGZ3knvbfrc+/sDdo/rb7+77nfZjbr3dc49MDOf+8wtc+755KkmG45IJwXToeFQKEW3CvLORHP54y5crA1H0oIr2oHKKfNRnw3IewIwl3nw3uaiNlngDHETeEaN+jeVqB0CVbERn2wV2mYVobG1TJ1+mKjxhVR9E4XFZCPsoBqTwUnG210gxkrmQjP2UjOTYy+xIe4roreUEP+Zg5idDnQdmghIrKV2jJ2QoEYa41TsfWcgl51+ZM1dhT3Mw/4tGxkTnkLuho5KB2rFr1GNLMu9frCt/eGHQohXhBDyP9iGAq03dvlx9akesP6d72/9X/bhB/v9lJh/5P96fH+hJEQKxT8hfvN8sr+PZEx4Ci/2386uqwfRtAj87y3GFRdK1aBgQr7RM2rabMJTyzj//CqCV56g/UOnsaW2UN8RepsLaImWER6B5fclbNw/lMQ5mfjtM6KOtGK+YMRU6ab9Rwu40hAOwLA589BeNmEpkDizsTvOPB/GtMvgwun2dHrqGhEb9RReDsfWwY7DD/y+NeBq1LNj7DvYJB3HLyYS1Kka79zrgzOKygJYcWIUrtvq0ZXqaOjmJmi+DcdFPzq/u5D5Aw9irFDR7YHLqDQyok5H4GUPpjIJ3z0m2o0roL6DHkkD+7tu4pNhaykYux5tE/iGN+HdqQ5JC2X5gQwZfonEdxzUdfdw18EHkUbXUzZYzdmziYQfdaOr0mANEchvB4MMnkg79mHN2MPc1F4MZkt2DxweDQ5fqMoJxHTNTkOTFxWDZOKXOpjb7ygTB57jy4+G8GVLO9Yc/YxhK08S/GoB61M+wrMjECr0qNcH4uxsJXh3AcsSOtPj3J2EHhE8X92ZvWXpyBU/aSCKQvHv9Z9vMnsdiPoH20kg/0Z86I/2DwXy/k75+f/LPhJQ+A9iKmRZ/qkJUcH/8fj+QkmIFIp/knAKADY8OplRd83Bo5fhCV/EiYs4AgRVtzooH+hF8+fhJL+xkL1l6ZQ825+EZ1tI6FnMR3274QlwMWjsRcrfTSD4LJw92oHAe4vQnTXTHOvh2jAN2mZBzZYoZBlcZjWxowuQ763GYwSvUsHFJSmowm3kPBKH5okKBve7iuy5vrBpS5RAU6/hnmWPs/tKZ9StKiqzg1j5yB8xhLRiyDawdPBWmvJ8UTsEhjINzT3CERLYIjysPj6MqD2NHCuKI3C3Hku2oHSooHKQREMSjAzKQOWUMabWkHxwAbOOzQOgKVGi9aofcX61CAlMxRrWRR0na6ERnywNwqZGPnJ9ugBDlQq1TcJlvt75unisCqdFRm7Q4S4wo6tR4/L14Hapqfg2EpdFYlS/i6hbXcS97kZfqSbrAQtmtZ0pfue4tHgls32qiNSYeTowi09jDwIQuPYkUfs96Brd5A7diLu8gonf1xK41EDkwzmcStYqNUOK/1qyLDfJsnztH2wO4BJQCfT687435iCKBg78na/YD/S4MUrtz3oBJ36Q7Oz/Ybk/iPl75f74POq53qH6h8enBZJ/ajlKQqRQ/B/kLk/F60oZ+z/bQNbcVcjnr9K6J47Li1aStKQcY5VMwPqTBJ+zMXbiTCStzLAvL1JU64/vThVP9dvFkcJ4qnoIGtqrULmhbn00thAZ2dtN59R8nrvnM+o7yQQGNVM+QFDVasb+dQgt7Twsf2QNlY/Y6RRegSvQTdGZSE7s70L4Hg1Oy/URU2NGXAABQcFNJK24hqyXmHV0HrYGA/pGeOXqOGStjG+2hMcg47CosYe6uTR5BWgl7GFeiEwzVX1l6nu5kE1uTIUalt6+ifUfjyN2Vg5113yhWo9s1TDw0hSES6B2COqfb4e5zEPk6CLmFg9EpfNgLpPwT1fh8JdJ6HqNSXcc49pwHbLRg8YmCDkh8Pi68clSI6Ks6BoFqGTmdDmJ96Aq/C8KGl1Ganr4UPqczJUH/8jyMZtY5FfIQxfvZmreyL+6R4mH72XemofJXtMb3Z6zfPfR+/R/7Hris3bdrWjqWpkSdEGZhVrx8/Iznan6xiix3wMLhRB/7q+zBDgBfPfnOCHEYSHExh/s+iFgB+698XkEMB145Qcx7wAJQojhN2KSgcFcr736Z7wCzBJChNx4fT9QC2z6KTtr/skvUygUN2QujgZgTHgKv8tP5/XiMMaEp+A+oMP5tWBvWTqJH/Uj4fd5xG0Kpv/MHDbvH82nLxyk+ysLUQeAM9pF3HYHmqIqWj808Fz0MY42JrEu6jgAvy9W0dwQiDBJVBf54WUBr1I1C8/NwHjMzKUkbzquaiB/WgA+vapp6KSjX1gpx03t2XU6hfiLNuptgRStaIIqGNUxgwurUlC5ZeweFYHnVdgCBLomqOkuYQi0MeTlx0i4aqOukxFLrkzAfSUUfhuDd5FM622NXLJFQZ9Gvt+biLlnAwHrTAiPTGWfUPqPu0KT08jF6CgKbllDsbuFIfsX4XNJR+P0RlqrTGgbrjfbfXY6FeErIQweZLUGeVYN3geCCV17nipXD3ymlPFq7D7ea59Iwokajvf1p/ZCe7x9BNJZX5Ksc9Fo3WwMqSLAZCWtMIq4iw8ifJ2o1RLGUyZUHjCUaXksN4Ox42dQ91QrM56v5c1jLgKu+vK7nXe22e9Hofhbfs4Lssqy/I4Qwhs4JoRwAGXAbbL8VwftxfV5h/68T4sQYjSwSghxH9c7Pz8my/LeH8SUCCHGAcuEEK4b+98hy/LlP8cIIXRcn+DR98Zbm4UQZbIsT/lBOTuEEEHAbiGEleuJ2BhZlu0/5fyE/DO++DebPipKDl+8qK0PQ/ErE73PQ+FkQfznbjQtLvbs+BiALqdmcCV1ExOyb6Hywxg0DhnLlnOUPt4HQ7XMsmdX8UV9L94OO8fIGfcxbeUelh4ZT8Gt6/5Sdtz++zBdNuAxgr29nf4J+VTbzPA7f3JmGInZ4aYlUofhrgrk1cGUDxB4TBLBJ9RU9fOg9nXisWpQNWnQtArunfQd604MYVByJhlrO2NolPDOaiD3WT1Lkg+w9Mh44uIrsb4fTv1trQT4tFJzLgSVS+BKsuKxaQg6okU9vQqtSqL0aghxXzqoXGwn6B0vXN5qGmM16Dzi/wAAIABJREFUtERLeFWoaO1iR12mJ3qfk/yZAkORjs9mv8VjudMorbUgBDhrDehr1ET0L8Wis1H8UQK2IIHbJOMM8DAoJZO0ikiks77Ywj0Yy9TIPZrQaDw013vRLzGfs8c7kDNzFUPn3s+h99fR/ZWF2APBlWQjfkYaha/2w9ytlrVdPqHa4807fQeQ9bvENvrFKH6Nypa9jaOkRPwrZfiYI+TUrje3+Xb/qefOy7L84+Yoxd+gNJkpFP+i4tFqwr8VFNyqY8+OjxnXbQQpSxdiL/Cm35L5DArIxX/DSY69sRKSkxD967EHCuZuWcDh9/vQb8l8qrsZWPvmJEwF2r8qe1ynqwgZPHoZ3xN60sojKKj2p+opJ8bwFqofslE7wo685noyFHBR4B/ZQM1oO2qLizd7b8U/pAkEhPUt5/MNI9BVqyl+JQkxtYZrt7qp7uOPy6Zl3euT0PvZKbgajnZ2JY5aIzXnQtC2XK9BCv1Cj6pRQ8z92bT3rab6ZBg++Sr2f74Bz2k/6he1UDJOxm0ElVPwwrxPoFmLO8iF5rvzaCu02KOcTNu8CNXSQLQXzUxOvITW346QBIUVAVw9koC53E3IWQc+eSCcKuweLZZN3gycnIZs8OAyy6hUMj6f+IBdzYXSSLZNe4u+v13AoffXEfv1/URMKwABEZu15Lzbl5jfnaSuwoJd1vBWQkc8tXVt9GtRKP6Bn+88RL96SkKkUNwEFf0EU4aeBsDdPpKhs8/gsbj5ZOmbrD46jIrH+pP67ENk32vmk5QNWCMk4p46hRhbi/+pCoy1Eg8s2YE1ysN5hxOAMRkTOFSSgLWHjcjvnCBATrfgrjYSaG5FddqCUefCfN6IdU49iR/UUT3CSWO2P3kjNuBp1vLaczMR2wPwKlNRmhZGc4yEbw7o6pwk+NagrdDhua0OoZJpjRCE+jbjf1lg/o2gX7cc7pt0gNZEJy3REvWJavyvCEqafXk8dD/CA2G3FxL3p/m4U1qor/LGFGRFSGDs0MBTW2cQvUtCV6al+Pn+uALc4FbhVS4onHA98dvxdT88bjWWHAn/bw3Ef1xN0SSoSNXjnlyP7OMi88skqm63s+9CV/Q+DsJ7lfNkp72476tFX6XBd4eJxXl30BIpiN07F5+QFlreiCTu4zI0Vg/mAjVyv2TUXm5++8QC9palK/2GFD9PMtfHXt3MTfGTKQmRQnGTfHGoLwD7tn1IRk83/iFNxGvNPD5kL5E7ygi5pxBNkJ37Xn6MzDveo2FWKo25fpS/raeus2DZpZGYitQsWfAQHY/PIv9cFPq9PugyjHz7yftIGoG6RwMqfydlDT5cXrSS6jJfXGaoq7CQ+6wRdZUOJEHc/vsIOK+m8hYnthBBwBUXYSc8IGRqu0LxLV7cHnQe/+Rq2O2PplyPoVamvN6H+qF2iqeEcroghk/Xj+KenidBhqSxOQTNKsKwwp/bdzyKT4HEkMAcLFkqVBlmgo5q6RteRGs7N5FPezDUCCp7akkelo2hRx1qs5uAc2qcFpBMHlzdWmk3oIS4dXD/M9sJPFtPxiJ/jEVaIoaV4DzlT/f4YmwhMq5GPQduWY7brab6YDgrX57KqZStOKId2IJVLG63j5jPywgNbeBkrw9Jev4KjStVtETouLRkJepLuUgNOkzbTisjyhQKxd+kJEQKxU0Uv3k+c4oHAWA7E8gt7QfwsF8Rk3aewfOoL9lDPiTkuzJ6np3J6T+sQvKSsJ8NwFghyBr0EVHbK7k2VEPIBgPGSkHgmpP4ZXno8dICOk3PwNqqR6rR81nP9wFYMmAPHr1MSEQ9ln1ejBl2AVOpQF2uxzqmGW8fG6ZSmZZwDbX3tBLzjYfovU7cXjIvrpmJY2cwdn9Bwh++Z92TK1jc7QCqUgNCgvdSP6U53sOmA4MwF6tofCGauvXRlM1xoGkRaK0yq88MoSHFhb72+kKSR/ITMBVpmLztGM0JblwWifyNiTRUeuN13si5l1Zhi3The0lL+Ec6Cmv8KV7o4Y/vTaHoNn90dWq0Vng17gvsnWykfR+LiLJyS89LjPpqMaoiI5EHW/C72EDndxei0ki4DfDQV3PIuT8M/zktTMudRPFgmYpaC7IKOhybxeasAyQ9eUWpGVL8rAlkhHxzN8VPp3SqVij+DfyuCkyVHg6vXUvctgeJ/9xB8RgjLl+JwHMqRi46TrXTm/zmADQji3GN7In2wHla9sTh2hyCscYDApKfT+NYWRzOEwGYB1dRkxFIh55FfJO4m6tOGzOWL8YaKiNrQBbgkwctQ6x4H/Ki3YxcrpSG43aqUVXr8JgldDVqJo47xVc7Uwk74cb1aC3GP/hS28WALVjGEezBO7SZ5jJvDBUajNUy1lCBthksIysQa4NIfPIq5TYfsq6FoC414PJ3Y7mixT20EfURC1GTC3B51FR8HY3bCD5FEpaMZuzhXlR116JvgNaI688dczGonFDb24NwCfS1ajTJDbSUm1HbVHgsbm5JvsLpimjqC/zw/V5FY6LMnJGH+Oa1obRObeLWmCuceKYv+p1nqdjekdDJGUgDU9i/ZSNwfRRgzsaeiDrt37ljCsW/5mZ0qraYwuXUTg/erEMCYN+5F5RO1T+RUkOkUPwb1HeWsQWq6Xr6bj4ev4r9WzaSNXcVvldVnFm6ivPdVVxd3pXX47dStbA/DY+0UL64P9avQ1G5ZfR1Dq6NUHFgR29arvgjq6CyxA+1VZB1OoZnqroybc1iWiNlnr99C8INkpeEyg1SqZG63i5K1yfwTPedbBz0AZ/f/g5aXzs++XD2mV745kBNspaq9BCSl6Wja5IZMPoyfulqugRVoPVzsGPuG3j0gi7Ds3H2bsHzYTBH/7iGFO8SqlvNSK1aug3IAbdg2rxvebPbVlxmuFoQTmGNP57BjbgsMnY/FfbXWrk2XI0two1hfCXSjQTO4Ssw3FFJTHwlmhYV3r2rcWRZ0NWp8Xh7EBqZM2u605IewIi+V3D6CiSDxFfLhuE7t4TtPdYy2/8k9fe3kP9aPy72+YziF/pT9JvrnSeqPK3U3ddPSYYUvxxKp+o2oyRECsW/SW03Gd+PvXm1aAL9H5tP8pm7aBxgZ3zfCWSv7oNlRzr3XbwXtVPmkcSD2EJkXnnsAzR2mfInXIQel0mbv4LYviWEnbQTdEKDrAFjpWBLZnfun7mLUcPSeP6bOzB1rmdC3ws0xYMc4kBoJapSPZxriWXelgUszpnGjI7ncE5soHSwBrVD5vhvlhGSUsm+z1OZ+dtdnKuIwmtyJdkbOxC1XsO4rYuxBcukF0ehuWjm+Zc30P2VhWx+8Rb0GjcABR+3x5KpYX3aAJ575T5SbslAWDVIRSZSQkuJ3utA7ZQZGZpJ5LcevIo1NB8OwVQmMFYLokYV4dgSQmWjNx6jTGNaICFnJJzBboZ3y6BdRA1CgoT38jm2M5nWWBcrRn3CmaWrqPk0mscKpzLh0yWEvaJm4JArjAlPASB78EcABKtNMKW2zX4DCsU/TUmI2oySECkU/0ZlAwWeYWXc9dxuPkj+kJiNKqTqGrxzNDRMSSE1vJCZj+3mzQ1TSdjUwKu/m41PRgMmvRPLgWx6/PFRSg5Goz54gcZxrayathaVCzRXzGx/YhR5D7XH74qgJcuPK08nY+5ch+xRYTQ7SO5cRJJXBZ5IOyXfh7Lx5EAcl3xxhzgpHyJx2+zf8EDMEeyBMp8W96Kl0EJZThAJs7OoS9Ljd1Uw99YDIGRc3Vp4+46puHzAM7sG25chIEFjkozDDyJD6+nz0AXOlUQhayV0DYKM2mCao/TU9nGz6Yvh2P3UbHlwGWo7NHV10pzoIr8yELdRcGv8FXzj64h55iTXxkhE7lZxflM3Ss+H49GBNSWa4Atu1GY3b+aPIfHwvXhVe8g4Gcvs8d9ROsyH9I+7AnB47ht/dQ/qsv3b4tYrFIpfGKUPkULxH5Dw+Clsk/vgmV+D062hqdXAyl6b+F3WbazuuIkpOx/BO0dN2PEmKvv60Bopo2kRGKtlAteepHp+P5y+AmOljD1QEDCqjPKzYVhyoNdDaZyuiKbumi+z+x/jmxVDaI0QqFxgC5XQNgkcYW5MAVZaK0xomtUYqwTWUBnfbAg810T5UAvmUglLWhUlt4Xi6tVM3DNWipYa0H1nQUgyXtUS4YtyyaoJJnJ+HddW+/NA4jFcsprdcwaRe7eJpORi8o+2Q9ICN3pTGDo0YPrcQlOsCle3FkxHzagdMtLEOhpz/Yj70kH+FD3D+1/m26wkfE/qaUh1YDA5MRzwRj+pCqtDh+4r3+uLwGog9qMSmtdqKL8YyuSRp7iwpAf6qyXUjI3H78OTqH188DQ1KZ2oFf8xN6UPkVe4nJp0/806JAD2pb+k9CH6iZQaIoXiPyB3eSpHVq7leLcvONtjC9oLZjbXpGLbH8zvRtwJ3i6cFuiwJpOG7k7cYU4cQRIt0VA7rx/1KR6QQUhgDZcoKggi5ncnaYkQ7D7bjaaMANDIbMroxatPr8cW58DRzUr85zZcPjKmXC2tFSbQynglNdDSzQ7hduoGOCl8SsWnjy6jfJyLxnehJcZN2Pt68mcG43araejsJuRkI3VJas7ltUM+6kfmk7E01Xux5v1bWXFmBI5XmtG0CCq2tKPLsBy0TYLww27UdoF0yo+KwRKv3/cBXcLLcQ5vRJpYh/twABGdKykbaETWyJwsjcHnnIGYu3IJ/1qLz3YzslpQdyGY5hIfjHUerj68kqB0N7kPRGJepEEKt5PTHIzmu/PkPBZP9RAXFds7sivziJIMKX6RlFFmbUdJiBSK/5Afzn+jckE38zUuPrES44ct/E+fPXg6tXBoYx/ab3BDoxZ9tYqobx3YAwTmXA1qB1T39bBvypuYg1vJXZ6KX46EcAviexXTPakQd6UXLz49F9+zeuRyA/UdvfBq14SpXMZUrOGuXqfx0juhSUvEZi3aazrGxmfQWWdkcZ99jA+/CoDDV4PHAM4mPSExdeQ8qSPisBVDtgGPDrwLVAirBmsPGyM7ZVLZ6I2kg4ZOEkUftKfdjhpqu2qROzVfn2zO4KHZY+TaBwk4Cr2xnw5g1N2nqGk24fSTid4lIcvQlOghLSOG6jtstIapaI6TkFXQ8a1KKnuqmVM8CFkN7QcUIuUXE7VJg0HjovSLzugbBAVj19NcZGmbG6xQKH7RlIRIofgP+nNSFDa+mG1Pjmb0HbNpdhp4Y8ckRsdnobHK/PHT9xjW9woxf6okb7oaezcbhjoZ/ahqVE4Vo75ZzPykoyQ+fxWPDnShVnLPR9P0fBSmEhU9nzyPd5kbSSejcoH7oi+2QEHg8DI2H+9H9dUg1FYVUU9lE3hRwkt1fWbsh3xLWHdiCJEHBP4LirDkAG5BdVYgv0k+hOPZRpydrdjCPcTdkYM5T01saA3fnusCl71pv6EG76gmXN6C0X86Q/+pafh/YSJidDGjO3/P0pV3ETMvmx59c9DYYVnYBewVJhLfLaFpQRPOXB+MZWrQSYR8asBlhrBjMkHdK8mZF8bv7vgTBS91IHRJHs+1+4raGT0wVFg5nRHH8OhsIv5wgrh9cxEelMkXFb9cSqfqNqMkRArFf1j85vmkBhagfbwCe5CO1pURSAaZnWeTcXsJHsmbRtn8aMpHhqCv0KLLNGKb0MQnXTYScUhCFjJfLhxF5ookpLtq8fvShK5JUPygm0F3XuDs8p6UDlbhe1WFkGUkrYx6SB3FGaHkT1mDKsJGYLpM9ePR1HVWszWz+/XJC5v9KJi4lso+Kso/i8EWJPC9rAEB6zIHUNviRWJYFcLPSVpGDLZgmaLTkQSfFPjmSGQsttB8zYeY2/P4465b+H5pN2omW0myVBKqbwIg/UgiacVRAIyZPAtkePrwVzRn+mOoFiCD1uhi3mtf4Ah1UdNNzfMJX6N2CN5++w6ER6ZxYC0ztjzC2VdWUT7UAk4VKiETctKH9rPPt+WtVSgUv2BKQqRQtIFTyVryMsJpiNdg2nYaydvNrnFvE7E1n6zsCKp7Wmgd1MqkiScw9qlBq/bwcLsBGLef4dEh+8i/TYe2QkurXUffJef4fsFKzIdMfLe7OxVDPUR0rkTWCFa/vAJngAf2+hN27Ppfiz2iS6idZCVngQZNK0jXvJCzzTxzfhIDL03BJxekcfW4fGQCr9pQWwW6gxZsZWbizLWY0owIp4rkATmE9qygJgV0zRLIAn21motZ0Tw5YQdVvVS4GgwcvpbARxdSiZ2Sh8ol0F/ywmWCmu5mjGVqXisZR8QhNx4jxHxaQvbgj6h0Wwg6qcGSK7Fk1f14DDIXnltFbWcdVTs6kDNrFbF75mGokSmYvJavTvTkeG680m9I8Qt3k2uHlBqif4qSECkUbSB3eSr5U9YgDWik8NV+aKu1TH97CZ7IIAomruXIiysYl3gVm0eL9XwgPUKuYToSRNPueDJaw8i/YzWSXiZmSQtfnezJhOxbaOgsoWsUaGs0lBQE0djbzrTPF6Gv1OAZXc/TSz9k8IIHSDuUhC7djPmSAUeAzOwxB/EYZPx8rLTsCsUaIrBe8sPtJaF/sYJOA/JpiZSZ0O8C52sisQfIYHHR27eIa7nBqCJsmLJrSbz/LFEHrGhrNZQ7fVG5BOHfqXBc8iUwqJnsffE4Az0kjsvB6S/RbkYu+gaZgp1xVPXUYotyce/+I3Q+OYNVF4ZQPdBF1SgXppGVmDrVE//tHCJWXsB5yp9ezy+gYOx6YudnMWLWXIJPCxLnZ7f1bVUo/jUySkLUhpSESKFoI/Gb5xMx5Sp+3avJvmcVF59ciXz2MgBeKh3Zd0ax61BPzL1qyH2lE3nb2tN8KIRD+1OI3f4AklYm6xV/8HHxTeJujGVqhAdkDXR8Jo/4dRKPTvwGWSUT51fHE+lTaQ1RYyoBh6+MtacV2ll5/8Rg2qWUEfC8lqBJJbhNMuHHXLw77kMy0tvhllUEp1TyXMgh4i21WLrVkviOk0a3EZWvE1e9npKJIaRedFEyygtJA59+NQRNcgOVfcFYJQjzbsK7REYTYOPq4QQkowe7R4s9QBBw1YXLLOOdpeWZC5Nw5vkQH1ENAmRJcDJ5Gya9k4RZaWS+1Y3IpSeoS5bIcFrZHPsdDYuasXxyiuyXu7bxHVUoFL9kSkKkULSh3OWpnErZ+pfX8WcNZLta6f7qQpxRfvx+4mes7vwJ10aqsAfL3HH3ISzZUDB5LcZyNUnP1iNbNfR4aQHtx+YReqIZd5ATggPQXilgxaVhGGoE2dVBuN0q4u/JxhYs0DWJ62tnFHnhVaRhRHAWJaMtGDUuzF3qKJwu8+QH95H0Sg6VLd6oVwUyc+IDZGzsSLxfDUW/FRwoS0JyC9BJuHxkNu0fhDalHnOxQGMT2PJ8kIMdWENkvj8Xg8ot464y4vaWabcDMvLCSRiVT9EUGY1VYKiV0V40s33acowaF/P7HEbr5eRbm5qyrGAKft+PxwfvZW9ZOrJO4u43ljAmPIXAW7OVpjLFr4d0kzfFT6YkRApFG/vhiKjTFdFMTZtH8HsnKJik49Xvx/HE/Qvw/V4QcdDJiWQdKo9M3JcPYk1w8uDefXhna/DNdZJ9IJ6Kft6E7tOy+OttFDzSmezBH9EcJ2E2OvD/2otzee1wBHrwGGS6RJYhaWR8CiW+eGc4kg5yDsYRbWkg+KAOZMh4KZ66bH9KJnsoGWtBGl9P2qEkXE4NgU+qMF02oKnV4ohyYoxvQhz2o7mPDaevTNhxmfiVEmvuXsPowelUjHKRP3UNskamtouWDu+2cvVCDAgwlcm4vATxY/OZsWwxBd/EsW/RYFxNevKcIYQfkXEFuHn30lDiP59P1C5B+FfF1M7rpyRDCoXiplASIoXiZ+DPSdH5nlt4qfPXOMf2pv1v09Ds9qUmWc/5F1ZR2VtP+7N6kh+5SP5ta0ice45VUydhC5EpHqtF36uO+NtzqBjt4oIthmm3H+aZqq6YopuQtwXi9hJEb1YTGF+HI9CDQeNiyMArVPeEcy+twm2U0bTCldIw6juApAN1q5q86auZ1O0i1nZu3Ef9GToqHcMlIwVTA3B5g6ZVEHBcR0uNCZcXyHV6cmauonSyC0mv5snM29lztTNBh3WMmDkXr2tqfAolisf5MWfEIQrGrqe2u4eWgVZaXo3EZ0I51m42erxxgcQHzrKtYzBqp0zAOQ23Jl4m787VHFm5lk9P/on6TkofCcWvizIxY9tREiKF4mfiz0nRY0emY3mqGM+uYHzvKGXavd/R46UFLJv7PgcKEjm74foCpjkf9kDWqPAYZKL3uWks9KViZTzGXD2HxiTxzXuD2f7ZIMJf01AzyInLJKjtrKWxxUjiwjP466yc3NmNhWP3MfyeuWhbBU5fGU2OF6rra7fiMXvoeX4aGQ2hdO1YjMYOGfWh2IMkJK1Mj5EZeFXKGOskgo5riH7tDJHfSXR9eyGB3+nRlTZQfzUQ2aYmcPv3XBuuozXRSX17FdZ2bna8NYxO7y28/iQqMaLdd44u/uW0f8fN19+kokruyN6ydA6vWYtnQj0ZTaFMLxjOmPAUpkX2a6M7pVD8GymdqtuMkhApFD8j8ZvnU3DLekqbLWgeN6ObVMszgZkErT7J0kX34rPLTEM/B7Ff30/7ey/Q0NEbS7YKj17Fugnr2PL6m6g8kLugHQ0dZRwBMvmTTbT7k4rIr8po7eIga9BHmI4EcWh7D+yBEh98Mhb5tzV4lV9fP80vU8IR7CH4vBtDpYbGLH9uD79AxslYGns5KMkLIrBTDdoWwZPhe2gZ3krFVAfVqR72FJ+jZJJE6JgS3LfX0dAzGLfZA1qJ/kcqcUY4CQxuwtHRRsHEtUTNyUXfAF5FGmS1jCqlE4N9svB4aQjoU0nRs2pGzriP5NcXEja7kv4B+TTfaQRQmsoUCsVNpSRECsXPTPzm+fhPyKa6ty+7c44D4BrZE8M3Z3B6C4wmB+0/dFLwWTKuO+to6OrGeK2VZanDGbJtCT6FEqZrIFncqG2C+GfOUjrTRc3AMKI/VxH/+XxyvmqPI0BC5RQICep2RhBwVwmG/jXUJgtQy7SGqvns3rdQRdhYvmUySamF0KIhJKYO8XEglnyJB15YhPG4GY9TjbpZTezO+/Hxb6V6RxQNBX5UDJLxjWjiqQG7ODGlI/0S8/E3WukdW8SQBx4gb1t7Gjt6cFlknpnwBbt3fcrGmRPIvUuLt87Bkb5rKBto4OKTK6n+OIiDVYk444OVZEjx6yQDknxzN8VPpiRECsXPUO7yVDrcl0Fq+lS+takpHqMDwO0F+gM+uF5sIHvIh5zvuQVzrgbV8npqxiWQtLqGigEyrRGgrtXilylxbUkfVGqJpAVXsQVpMMU20m5CARqrwC8Dhk07i2ZEDXnpkdhOBeKJtBNyWM25l1Yx59K9uJ1qQs+4uJoZhcqmwiOpqJ7gwPdoIX73lOAygXeantwZq0AjMTo6E3ugDCqZsMOC5lYDK1dNxhXhS53Di7Ld0QCUTPWg8oCuToWhSvDBbyfT9fTdiMxCEuefobDGn63NiUS9fILpBcM522MLmpHF5E/St+WtUSj+jZSJGduSkhApFD9TJ0914FTKVkYYPeTOWMXesnSuPLoSp0XQw7/kL3FCBp3KQ6/fpEFVLfm3rwEBGqugdqINT89mpCITZ6+1w6vSRfjLKtxPBOKMcGL3Fxz5qDfNVwJAEnhVyVh8rAgPJHw6H5tDS0hwIzVdtCBDUJqMc38gfpZWch6JpeRgNJEji7GFySR+tABsao4u74sr1o5vuwZc99bSJaIc4YHCcQayS0JYPG8rm2O/w9vXymuL1qGvF1xaspJd771D+G3fIzU3k5IGjlYdHxf3RWh12N1a+i2er9QMKRSKfxslIVIofsZ+OCR/TPj1ztSBF138IfQs3X+/kIRNC7i0eCW2IZWcXtedFy7sY1TGrWyfuQxdEyQ8WIBR7yRxVRm6E97oa+2UD7awZ8fHhO7X4vSTaYqX0HdoxOPnojbVxfaU92mKUSHC7ague+P8MpjAEWUYyzRYQ1SohtdRU+VDyFkJv2wJ1cR6XMEuXH5uvHM12KY2Yko3MiIym7pGE7YhlTTHSST1LURVrePVC+P4qCkQa5YvK0aNuz7zNdBtx6O8X3wMgFKbL5oqHce7fYHscuKaY6Sq93/++isU/3FKDVGb0bT1ASgUir8vfvN8Eh4/xd6ydMaEp1C0QUIr1KQ9vZJxHQbT3rOAl7M2c7v5PPts3qie8mXyrYtRmSHz5U74fSPAVcyT8z9n1ZghXOz2CQDyzBpcuQEEpKswHjRj9lLhXWhjuPE3PDbra+rdJnZvH8rxt1eTuHEB7Y7YqOhjxOnUYs7U0RwBI2af4osRPdFWaghKk2mMBVuBBXWAzM4v+5G0Lo+Sx/sT+5UNc6oDXUwL4rwPVztGIqtlMl/yR30NxnUYTIfIem4peBLXi5BiPEHOrFWMCU+5cd4AoW16HxQKxa+bUkOkUPwC5C5PZfQdswE4O/IddloNDLvvfnJWx7H5zhW8cHECEyJ68sYjs2j/3vU1vYQb/C8L7IGCcfsu8XFpKmVZwX8ps8lqQNZLaGwy4U/k0tBehSavnF7tiln5ya18t2QgTpOg49qF6OsE+z/bgC1Yxp1vRtMKA+85T15LEMLood0eO+XjnZhLJYI7VGPJBY0Vvn8pmtvvPcS1YUby1icR+1gDMR8XcaEuivPT30JTaCD2qZO0Du7AAzt2cfnxleiS67lQF8Xwe+aS82EPRs64T2kqU/z3UGqI2oyQ/4sumD4qSg5fvKitD0Oh+JdoWgWxL52n+Le9iNlWjW51ExKCoi/j2LzoTZ4quo0NcV9yV84d5J2Nxm3xIDwCy1U1ac+spOtbC7GFSEhBTvYPfYd52TPY3nEzUzLvJD8/BLQSIQe02P1VNHXwYCpQM+hTnr0bAAAbtUlEQVTOC+w5nczEfufRCg+7Czshn7Pg6GzDmGaktaudLjFllDf74PgukJYUO5Eh9bg3hIAMcY9mcuZwR7wLIGh6MeVftyP0rRPclVnGbJ8qJueMYXv7vaQ+OZ9Tr6+m25sLsRR68PritJIMKX4Rypa9jaOkRPwrZVj0oXL/iJk365AA2FOw7Lwsy71uaqG/UkoNkULxC5M1dxWyy4lPoUTGYgvb2+8lb08c8tB6xu94nEuXYui/fgnlX7XDdE3w/NDt5N+2hvZ3Z9H7dwtYt+Bd7h/zLTEfC2759Aleif8Si8pIybkIDKVaorepETOraUqQCDinwjKygtPruyN7edid2wmHpCHC0og1xsX93Y7h8gbvCwbckgqxNYCw5SfwPaFHJWRUs6uoGOfkk5hDxD51koWPfQmLfdG2yORs7Mnq/MGMCU/B7tEwvt+tPPHcpwC0JNspneRib1l6G19thULx30JJiBSKX5j4zfOZllGBdFctifPOke5wYCqTsWX54lWmosNz2QRe9rDmkXcJPtfKmpemEPen+Vzen8SWF97grn0L2FrUHe2B84T3LOeyPYqeLyxAirbhVS5zbaSKwaF5CA8039JCeWYw8bOywa7i8z7rOP9aD7Lzwsgdv4aPPxmFpmsjzoHNlOyOoWVCM/Zb++Cf5aCmxUTz7lB+0/MQHVcvpHZuP6ySnqyHjTQkQvvZ53FtDwJAHl5K3WodH4weysCHH+S5vt8ganV/1alcofj1k0GWbu6m+MmUJjOF4hdK16jCHuXElKPjyqMrAYjbfx+hu3SIOVUIoKreG6nUSNQBD7o9Z8n+oBfRX6owfH2GqoX9cZug6+QMTp1LIuw4mLecIntlH347dCcxuhqWFY1mVsRJ3sgYzZbu65mycQn2UDcqmwrhgYjDEhV91HiVCyY+eJivVw/GGgaOIA/+aWr877yG490wWu5r5EKvzxnXfTSeyioA6ub0w+krCLu1iJz0KPKmr2ZC9i08EHGYx76+pw2vrELxz7s5TWYhcv+wu2/WIQGwp+htpcnsJ1JqiBSKXyinRWJAp1y0rdDh2Cziv51D1sh1uEyCsoJAyip9AXhq/HbqkrQ0zkzFkq7DdKYQlbc31iEtLLrvC1RCplNyEQ+//DnZG3rStXMx15z+LDw2k9zsMAwqF0LIjDvwCA4/iajYaizZAinIyeE1a9HYBLahzRx9oh9he8vxKoOg6Hrqu0nUb45EuMF1NICdVgO70vYx8ftaAOpH2Wjq6MIjq/jstncZ32c8Zq2Dl16/ty0vq0Kh+C+lJEQKxS/YyVMdCP7jCWJek0iKrEQr1Di9BdG74Mnee8ke8iEXW6MYds8ZLJ+cIijNRsZL7bAO7YizWccrhydydVMniur9uMNcS8GY96lqNXOb5TyaCh0je1zl6a/vwnPaDy8/G+ogOyV5QdT3dNH+PRcpSxciq+GF5G/Q/raCqqFhBK0+SXWRHyq7wBoiKL3LhbWbjfFedgAe8r0+qWT83emE71ehGlHCg28+SmvXcOoH1FHf8b+n1lqh+CvK0h1tSkmIFIpfuNzlqfi+U07R/hieqerKoSVvcnjNWlZsmcSY8BT0KjdXG8Jonp6K6mgahlIthq/PgABzrobg905g2ubDo2X9aH9oNq2Hgjnc2oHP7lzBuqjjaBtVqB1ga9UjSQJLpobk9iVkP6AnYNI1Ir9tZcULd+LwaHBObKD88f5EfCfw+LqRVTChw2USZqWR52oBrk8wadsbizohluMr1pCSBoZ6CZdZpYwoUygUbUZJiBSKX4GzpxPpOj6TV4Iv02PnIoY8+ACfz17O3rJ0vtqZytvxW2hqpyJvWSrhxx0EHPdDtKqJ3nqN1j1xVA53k9PHyXt9NtFxYhYh2kbm/HER3V9diMtbxmtkFRHbtRiMTkzjKpgfcYigo1oqvo2kcKIXq36/guKcEC72+YygcdcwbT2N4ZoOIUMHYzlqPz8WthtI0oYFqDu2x/yYlpx5oYzrOhxvtZ3q7oLjK9a09WVUKNqeMg9Rm1FmqlYofiXqB9TR594FJH548sbszqmI3l35eMs7aIWExgqJS3Oo+CCQyq86ElYgUTY+Ep/33BSsWUvCm/N57oW+VA6QOKuNRdu7hV5RJZxKS+RUylYS+s8nYLsP5X3NPNJwJ1IX0DaDK8TJkvsWEPk/VbQ/NJucoRuhDAY+3BfTttN8M6IbnvpynPvb4cx3kz89iMz7VzJi1lyK7++Ap1sduWWrlBFlCgUoSUwbUmqIFIpfidzlqdQmy6h9LXQ/O529ZelkzTVy146HSdSa8JtYymfp3/BQ+8O0+/waAMHvnWDC0m9Jen8B5+98CzGjmn7dcgiNqOf0gNUU/DGJHsl59L4wjRGDL+IxgCVTjafci9wZq4j61oquTMeL76/HsSkUT42euAP3AXDs3TXUzuuHe0wdzdNTqf86gqHJmQgJEj9cgMrhoffky2Sv660kQwqFos0pCZFC8SuT+XIHgidlkpo+lcT5Zwg7JpN09B4Odt7BtMh+bJs0gJ0nvsLmL2h/Vs+neb2JefYk3b9YRFVGEPkrk2iyGpiaNZ3uj6cTbGihqdVAo8tIczTE35mNJb6efkvm0xRjIChN4ov6Xjz17Md0Sy5E9gjGhKfQ938WcO6lVewpOI3vuUqs/Vu4WBVO9N4Wwo+6kfRqjh/sgqpZ3daXTKH4mbjJzWVKbdM/RWkyUyh+ZYRLsLLoGPHadPrdPR/hga7hZfR5egHu+eDRC4bPnoe3zkNabQSfJn/AxawItlQG4ZbVXA0Kw3zEguq8lpwXZYpPRGIugcef2suScwnklSdiCwEvh4fK3ira9y3mVt905uyfR8BZNaKbzLSMCoZ6vUHShieIf+0qXjubiR1U8JdjNMW24/tng1A1KA9sheIvZEBSJlNsK0oNkULxKzR62xISDs2mNVSFV6UT2x1qmtsJdM0yaofMdxvXc3jtWnxnt7K0fCxvv3QnaXnRfNV+D1q9m9cefp+aFC/CvBqJef4M/nde4/fFE+j3whk6z/weY0odDh8VOfesovC7GF4fNwWvoFaaRlp5YcxW5loqmP7CE8R83cquzCNYp6rYW5ZO9treCL2e6iHhRH+hPH4UCsXPx09+IgkhdEKIpUIItxAi5m98Pk8IcV4IcUwIsV8IEf83Yp4WQlwQQpwSQmwTQgT/6HOtEGL5jXLOCiHWCSFMP4rxEUJsvPH5BSHEa0IIpaZLofgRucLApSUraYzV4+gYQcaDK9FaZYJWn6TrWwtJODSb/PnxnD7QGZdZoK3Q8fuaJJzlJmo9ZpAh0tCAOBBK0ZlIZocfZ+vV7pzMi8V52h/vUhe9L0zjlsmnqF6mQk6zkD3kQ7ZW9qLzuws5++oq9n7xES9WdyL30euPA1WLmqwVydgDBNeGKwmRQvH/UZrM2sxPeiLdSIAOA+HA/9fgL4SYBPweGC/L8kBgB7BPCGH4QcwjwCxgsCzLqUAB8OWPinoN6A70BfoAvsC6H8VsBNSyLPcGUoFBwEs/5TwUiv828ZvnM+qR4xzY9AHjOgzm6Htr0ISFcvmxlcTfnU70CyeI+s5ByMEq/DLAorESvcfDxnkTefThrWzbNYD5UYfRNgkeOzydhJlp9IsvwD/Dg76yldpCP3bu7EvQ4x6iXjlBwqYFlHwexxP3bmVMeAqLy3twqp8vT0/ZBsDu25cxKDmT1gilWUChUPy8/NQ/0cxcT2Y2/C+fPwt8LMtyxY3Xa4BAYAaAEEIFPA2slGW55UbMG0B/IcSIGzF+wG+A5bIsu+Xri6y9AdwlhEi4EdMFuA14HUCWZSfwNrBICGH+ieeiUPxX+fxgfwAyXu9A17cXsvP8Hqo8rQBEnzZROE/Gk53H4mc+5atOAViDNZheKeP1TVMJ6llJviMYt0nGL7iZ1IsuzhzpiM/JQoTNiSbQRua8VXiy83CN7EnujFXommVm+1RR9kR/loVdoHR+CrN9qhiwaD7T/7CE4yc7teHVUCh+5pQaojbzkxIiWZavyLKc+7c+u5HI9ATO/SDeBaQDo2681Q0I+VFMJVD8g5ghgPaHMUAa4AFG3ng9ErADV34QcxYwAgN/yrkoFP+N4jfPp9PScrQDr68jFqw2kfdmKsV9W8kdtoH81/rx3uI7cYzrjXxHLZcuxtB3/GUqrgZjl7TEv/k9YQ80sPf1Qayftoq8BfF4svNo/1gVHY7Nwjm2N9oD5xkT0Z3tr77BzMKhJN/2PWPCU/AfW8a4YVMxbzlF0KqTbXwlFIqfs5u8bIeydMc/5Wb0vYm98W/5j96vAOJu/D/uJ8bIN94DridWQojaH8VU3qg9+mEZP/yOvyKEeAB4AEDt5/ePzkWh+NXKfDSChImnSPzsXlQ5Xug7N5H7VipjwkH/RRMt+T5YClwA5E9dQ+JHC1DFtHJscCj5a6ORJIEqS8Vvs24n8/6V9ClYgEcPktRKTTctgYvb0fhFOGGaNKr7N7C3LB3KANKJf2YOeSPSlfmGFArFz9bN6NX4507Pjh+97wC8/skY14+Snb8V87fK4Acxf0WW5bWyLPeSZbmX2mT6WyEKxX+N3OWpZA/5kP+Zto2r/Tbx/9q79ygpyjOP49/fIBAFvKBcRCMwXLxEVHRQ8BZWMbNyFI+XbFhXj8TVSBBFs2ZjzEU9iiZsYJHESMSIl2PETfCagAISQFFwuIkkWbmDLDCIICogMPDsH+87WDQ9080wdDPTz+ecOt311lM1VQ8zzdNV9b516yUTWDqkBwu6P0ezlRVMfuoJ7KWjAahovZ32LTawo0sxTSc0pWL9oezqvJnDL13C+fOvovn7n9Fy+kYWXvgMX3/8b3z68nFYAwGgklMBuLTTefT+sDf3dnvNiyHnMjEw21Wrk8tebRREm+Nr45T2xsCWfYxpKEkZYtJtg0SMc64aHcb05+XyrhSPvYU3bjqfRdc9xjbbQfMfL6fL8AEUVUDvLhfRclJDuGQNq88/lAbb4aT7l9JwfhP+9X9X83FZK8p7HMHCG48E4B8Pn8jcn/yWTSfv5NXNh7Gy9HC6lvVl64WnMO7Ecdw/7po8H7VzdYRfMsub2iiIKkdba53S3hpYEt8vzTJGhHuNAIjd6Y9OiWmZUjRVbnMJzrmsfDC7PUuv/h1rz2lKyb3fp89x3fiyz3bevX0YzVZth2OaM2PISJYNPpvHb/4Nh35SwfBZL3P8w+/w/Elt6HTBchpfto6OP5jB7au7cfJvNnHqIwOYe8Vw+jTZQtFZm5jbbQwbbvmC0jZn5PtwnXMuo/0uiMxsI+FG6JLKNkkNgdOBSbFpPlCeEtMSOCERMxXYnowhdMFvALwZ5ycSbqD+RiKmBNgKTN/fY3GukHQY058GF39C0XZYOPJsFv34FC6Y3Y+NnRpxx19eodOz36fjk+VM+LwLU0aN4ra257F4WHcA/tx5PBvntmDVPefyYckOtrVqysAbXuGIokP51tU30KBoF6VtzuDYhw7ZvY5zLgveyyxvamtktAeB6yVVnt25GfgEeA7AwoXMh4ABiYEW7wLeASbHmI3Ao8Cdkg6JZ4HuAp6v7OFmZn8jjF30Q9hdeA0Chie68zvnsrRp8VG89/BjtH3F6DxiBbtM3PcfTzPwTzdx7Ds7Wfi9Vry24lSKX7pld2GzeFh3OozpT/Hg92lz8Ue8sXoeh0yezZCJlwPw5X2fcV3H90JsX79vzzlXN2TVy0xSI2ACYaBEgDGSVpvZVQBm9oqkFsB4SVsIXeNLzezLym2Y2QhJzYC3JW0j9D+5MuUm6rsJgzPOJPQ4m0coeJL6Ab+WVEY4ezQJ+Pk+HLNzLqHDmP7sumIX1wxew19ajwPg0bvfDb3EKmMW731D9MIHT+Oqo2ZS2uYMKi46i06DZsC3YVqXl+jx/tWsG9YiZ8fgXL1g5s8yy6OsCqI4AGLPDDFPAE9kiBkMDM7wc+7MsI3PgBuqi3HO7ZuiLUXMPwtK7Yw9CiGg2t5hL045h47MYPWFjdnRuzsnPtmDncVbsbVfq3Id51w1/DJX3vgzwJxzACweeg5HdtpA+3EllH30CAP/7yLGv9U183rDutNoE3S4awa3LlrIo506+31Dzrk6xwsi59xuny5qzrK+Iyltcx6Lh2UuhiptPyKc5vdiyLn9Y37JLG+8IHLO7aHDmP4wbN/X80LIuf3lPcPyqbZ6mTnnnHOunpN0j6Q5kmZIGhuH0Mm0zkmSJkt6S9JsSdeniekh6V1J0yS9J+mf08Q0kvSwpApJ7dIsv0/SPElTEtP4bI/NzxA555xzBwPjoB5dWtLtwPVANzP7QtKvCEPhnFfNOk0JvdTvN7PfSzoemC9pnZm9EWO+DowHrjGzSZK6Am9J6mFmH8SYdsDzwEJCD/Oq3GFmU2pyfH6GyDnnnHPVklQE3AP8NjHu338B50q6uJpV+xEGVB4NYGargDHATxMxg4AlZjYpxswFpgH/mYhpSijGRu/3wVTBCyLnnHPuYGG7aneqPacRHq01a/eumpUDK4FLqlmvFzDH9nzSbBmhkDosETMrZb2y5HbNbEHlIM0HihdEzjnn3EHAANtltToBx0ialZi+V8PdK46va1La1yaWVbVeunWKgHYZYloliqZs3RjvHZou6VlJnbNd0e8hcs455+qv9WZWkjkso8rn8GxLad8GVFe0NKliHRLrZYrZkuU+rgQ+B24EdhKefjFb0mlmtqzaNfEzRM4559zBwSznl8wkPSjJMkw9gc1xlcYpm2hM9QXL5irWIbFeNjEZmdmTZjbUzCriY8F+AWxg70eApeVniJxzzrmDhOW+l9kQYGSGmI+Bk+P71sDyxLLWxIe0V2FpjElqDexKbKeqmLVmlnVBlMrMTNIyoGM28X6GyDnnnCtQZvaZma3KMG0D5gPlwO7Lb3EMohMID1mvykTgzNhLrVIJ8E6i2JmY3G4iprrt7kXSI2majyNcSsuooM4QbV+16ovld971Yb73o445Blif752ogzxvNeN523ees5qp7by1rZWt1G7PsFpjZrskPQQMkDTazDYDdwHvkDhDJGkqsMzM+sWmp4EfER7KPlrScUBf4NrE5kcAN0u6yMwmSzoduBA4dx93s4+kN83s1bgv1xFu3B6VzcqyAhomXNKsWrq5rGB4zmrG81Yznrd95zmrmYMxb5JeJxRqtWm9me016nNNSfoJcA3hpufVQH8zW5dYXgYsNbPvJNpOAh4jnIRpAgw3s2dStnsuMBTYQRi36Gdm9npieSPCAI9HAqcDM4HVZnZVIuZa4CbC1a9GQAVwr5n9Natj84LIVcdzVjOet5rxvO07z1nNeN5cKr+HyDnnnHMFr9AKosfzvQN1kOesZjxvNeN523ees5rxvLk9FNQlM+ecc865dArtDJFzzjnn3F68IHLOOedcwSuIgkhSH0llkqbFB74VTM8CSZdJGifpTUkzJI2XdFqauJskzZb0tqSJkjqkiblH0py4nbFxUK7k8oaShsXtlEkaJalJ6nbqGkm3JYavT7Z7ztKQ1FbSC5ImS/ogHts/JZZ73lJIaizpvyXNkzRV0kxJV6bEFHzeJDWS9LCkCknt0izPSY4kHS7pqbh8jqRfSiqocf3qJTOr1xNwFvAFcEqcvwz4BGid733L0fGvB65NzP+CMAx7q0TbFcC6ypwAA4ElwNcSMbcD/wCaxvlfAdNTftYw4K+EsSYE/BH4Q75zsJ/5awOsIDyIuqfnLGO+jol56BnnK49poOet2rw9QHh8QbM435Uwzsvpnrfd+94OeJcw0J8B7VKW5yxHwIvAs/F9I8LghA/lO0c+7efvWL534IAfIPwJGJvS9nfggXzvW46O/8WU+Rbxw+T6RNssYGhiviGwCfj3OF8ErAVuS8S0itu5OM4fBWwHLk/EnB1jOuY7D/uRv7FAf/YuiDxn6fM1JM1/HidU/ufleasyb68BL6S0rQPu9Lzt3tdTCc+k6kn6gignOYr7YUCXRMy/EB5C2jTfefKp5lMhXDLrRfhDSSoDLsnDvuScJUbxjLbG18YAko4inEWblVhnBzCPr3J0GuGDIxlTTng+TGXMNwkfQMlczwV2Ev4N6hxJlxNGTX09pd1zVrWrgWnJBjNbaWbLPW/VGgtcIOl4AEmlhC8v5Z63wMwWmNnidMtynKNewJfAgkRMGWF05fNrcGjuIFGvr3lKag4cAaxJWbQWuDT3e3RQ6EH4Y341zrePr+lyVBzfF2cZY7ENCB9Ikj5JxNQZ8Z6BwUApsXhM8JylEXNWDDSQ9BzhEscWYJSZ/Q+etyqZ2VOSDgMWSFoDdCZcqvkj0CWGed6qlsvfrWKg3MwsZRvJn+HqoHpdEBGemQLhWnzSNuCwHO9L3kkS8DPgp/bVs2eyyVG2MTtSPiRSY+qSB4CRZrYmzc2bnrP0joyvDxIuQcyRdDYwNd5w+lFc7nlLIekm4B6gxMwWK3R86EU4M+G/b5nlMkdNqtgG1P08FrT6fslsc3xN/YbfmPDNtdA8BKwws6GJtmxylG1Mw1h0VRVTJ0jqCpwDjKwixHOW3s74+mczmwNgZu8BLwE/wPOWVjyOIYQzaYsBzGw+0IdQJHneMstljjZXsQ2o+3ksaPW6IDKzDcCnQOuURa0JvQ8KhqQ7gJOB76YsWhZfq8vR0ixjRLhGX/kzDwGOpu7l+jLC/QCTJU0BxsT24XG+YZz3nO3pY8I35VUp7SsIlzT8dy29FoSbeZentC8j3JPlecsslzlaCrRMKZoqt1nX81jQ6nVBFE0CUscdKontBSGeju8NfMfMKiQVS+oFYGYbCTcQliTiGwKn81WO5gPlKTEtCb2HKmOmEnpnJHPdFWgAvHkADuuAMbMHzOxMM+tpZj2BvnHRHbFtJp6zvZjZTmA6cGzKolbASv9dq9J6QiGZmrdjgS2et8xynKOJhC9M30jElBA6rEyvlQNy+ZHvbm4HeiL0PPgcODnO9wY2UDjjEPUlfHv6JuGPtgS4BbgvEXMF4YOiVZwfQPrxO/4ONInzQwh//ErEDCN8aFSO3/ECdWSMkww5bEf6cYg8Z3vn6lvARqB9nG8b57/reas2b78DFgLN4/yZhB6Ogzxve+WqJ1WPQ5STHBHGIXo6vm8IvI2PQ1Tnp/p+UzVmNlvSvwHPSNpKqPRLzWxthlXri2cJf9hTUtrvr3xjZq9IagGMl7SF0Aut1My+TMSMkNQMeFvSNmA1cKXFT4TobuCXwEzCB9Y8YFDtH1LuSBoOdI+zwyUtMrNve87SM7MJkm4Fxsa8HAL80MxGx+Wet/TuBO4DJsW8NAN+BIwAzxuEUaqBCXx18/4YSastDi2S4xz1A34tqYzwf8ok4Oe1esAu5/xp984555wreIVwD5FzzjnnXLW8IHLOOedcwfOCyDnnnHMFzwsi55xzzhU8L4icc845V/C8IHLOOedcwfOCyDnnnHMFzwsi55xzzhW8/wfQgLsNfN9vcgAAAABJRU5ErkJggg==\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "clim = (0, 1e-2)\n", + "\n", + "v_data_plot = np.copy(v_data)\n", + "v_data_plot[v_data_plot <= 0] = 1e-10\n", + "\n", + "norm_image = ImageNormalize(v_data, interval=ZScaleInterval())\n", + "fitsplot = ax1.imshow(v_data, norm=norm_image)#, clim=clim)\n", + "\n", + "plt.colorbar(fitsplot, fraction=0.046, pad=0.04)\n", + "\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Circular Apertures" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Creating Apertures" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Performing Aperture Photometry" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Calculating Aperture Corrections" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Elliptical Apertures" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Creating Apertures" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Performing Aperture Photometry" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Calculating Aperture Corrections" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## Exercises" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "May 2018\n", + "\n", + "Author: Lauren Chambers (lchambers@stsci.edu)\n", + "\n", + "For more examples and details, please visit the [photutils](http://photutils.readthedocs.io/en/stable/index.html) documentation." + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.5.5" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notebooks/photutils/04_psf_photometry/requirements.txt b/notebooks/photutils/04_psf_photometry/requirements.txt new file mode 100644 index 00000000..a0d3505e --- /dev/null +++ b/notebooks/photutils/04_psf_photometry/requirements.txt @@ -0,0 +1,4 @@ +astropy>=3.1.2 +matplotlib>2.2.2 +numpy>=1.13.3 +photutils>=0.4 diff --git a/notebooks/photutils/photutils_notebook_style.mplstyle b/notebooks/photutils/photutils_notebook_style.mplstyle new file mode 100644 index 00000000..2f1f6450 --- /dev/null +++ b/notebooks/photutils/photutils_notebook_style.mplstyle @@ -0,0 +1,20 @@ +# ---- Matplotlib formatting ---- + +font.family : serif +font.weight : light +mathtext.bf : serif:normal + +font.size : 12 +axes.titlesize : 20 +axes.titlepad : 12 +axes.labelsize : 18 +xtick.labelsize : 16 +ytick.labelsize : 16 + +text.usetex : True + +figure.subplot.bottom : 0.15 +figure.dpi : 200 + +savefig.dpi : 300 +savefig.transparent : True From 907b6370c5ac435cb255b780fc795ad5c14932e7 Mon Sep 17 00:00:00 2001 From: Lauren Chambers Date: Mon, 20 May 2019 16:38:36 -0400 Subject: [PATCH 02/14] Rename notebooks, and implement previous edits across all notebooks --- ...n.ipynb => 01_background_estimation.ipynb} | 122 ++++--- ...ection.ipynb => 02_source_detection.ipynb} | 52 +-- .../03_aperture_photometry.ipynb} | 84 +++-- .../requirements.txt | 0 .../04_photutils_psf_photometry.ipynb | 253 ------------- .../04_psf_photometry/04_psf_photometry.ipynb | 333 ++++++++++++++++++ 6 files changed, 491 insertions(+), 353 deletions(-) rename notebooks/photutils/01_background_estimation/{01_photutils_background_estimation.ipynb => 01_background_estimation.ipynb} (82%) rename notebooks/photutils/02_source_detection/{02_photutils_source_detection.ipynb => 02_source_detection.ipynb} (93%) rename notebooks/photutils/{03_photutils_aperture_photometry/03_photutils_aperture_photometry.ipynb => 03_aperture_photometry/03_aperture_photometry.ipynb} (88%) rename notebooks/photutils/{03_photutils_aperture_photometry => 03_aperture_photometry}/requirements.txt (100%) delete mode 100644 notebooks/photutils/04_psf_photometry/04_photutils_psf_photometry.ipynb create mode 100644 notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb diff --git a/notebooks/photutils/01_background_estimation/01_photutils_background_estimation.ipynb b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb similarity index 82% rename from notebooks/photutils/01_background_estimation/01_photutils_background_estimation.ipynb rename to notebooks/photutils/01_background_estimation/01_background_estimation.ipynb index 46991a0e..c02e4078 100644 --- a/notebooks/photutils/01_background_estimation/01_photutils_background_estimation.ipynb +++ b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb @@ -4,7 +4,11 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "
![photutils logo](photutils_banner.svg)
\n", + "\n", + "\n", + "[\n", + "\n", + "](http://photutils.readthedocs.io/en/stable/index.html)\n", "\n", "# Background Estimation with `photutils`\n", "---" @@ -29,16 +33,24 @@ "\n", "Background subtraction is essential for accurate photometric analysis of astronomical data like the XDF.\n", "\n", - "##### The methods demonstrated here are available in narrative form within the `photutils.background` [documentation](http://photutils.readthedocs.io/en/stable/background.html)." + "*The methods demonstrated here are available in narrative form within the `photutils.background` [documentation](http://photutils.readthedocs.io/en/stable/background.html).*" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "
**Note:** This notebook focuses on global background estimation. Local background subtraction with annulus apertures is demonstrated in the [aperture photometry notebook](03_photutils_aperture_photometry.ipynb).
\n", + "
\n", + "\n", + "**Note:** This notebook focuses on global background estimation. Local background subtraction with **annulus apertures** is demonstrated in the [aperture photometry notebook](../03_aperture_photometry/03_aperture_photometry.ipynb).\n", + "\n", + "
\n", "\n", - "
**Important:** Before proceeding, please be sure to update your versions of `astropy`, `matplotlib`, and `photutils`, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the [AstroConda](https://astroconda.readthedocs.io) distribution.
\n", + "
\n", + " \n", + " **Important:** Before proceeding, please be sure to update your versions of `astropy`, `matplotlib`, and `photutils`, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the [AstroConda](https://astroconda.readthedocs.io) distribution.\n", + " \n", + "
\n", "\n", "---" ] @@ -73,14 +85,14 @@ "from matplotlib.ticker import LogLocator\n", "\n", "# Show plots in the notebook\n", - "% matplotlib inline" + "%matplotlib inline" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [shared style file](photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" + "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [style file shared with the other photutils tutorials](../photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" ] }, { @@ -89,7 +101,7 @@ "metadata": {}, "outputs": [], "source": [ - "plt.style.use('photutils_notebook_style.mplstyle')" + "plt.style.use('../photutils_notebook_style.mplstyle')" ] }, { @@ -105,7 +117,7 @@ "source": [ "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", "\n", - "(Generally, the best package for web queries of astronomical data is `astroquery`; however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with `astroquery`.)" + "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)" ] }, { @@ -183,10 +195,17 @@ "# Set up the figure with subplots\n", "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", "\n", - "# Plot the data\n", - "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", + "# Set up the normalization and colormap\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch(), clip=False)\n", + "cmap = plt.get_cmap('viridis')\n", + "cmap.set_over(cmap.colors[-1])\n", + "cmap.set_under(cmap.colors[0])\n", + "cmap.set_bad('white') # Show masked data as white\n", "xdf_image_clipped = np.clip(xdf_image, 1e-4, None) # clip to plot with logarithmic stretch\n", - "fitsplot = ax1.imshow(xdf_image_clipped, norm=norm_image)\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", "\n", "# Define the colorbar and fix the labels\n", "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", @@ -194,7 +213,8 @@ "cbar.ax.set_yticklabels(labels)\n", "\n", "# Define labels\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')" ] @@ -203,7 +223,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "*Note: Double-click on any inline plot to zoom in.*" + "*Tip: Double-click on any inline plot to zoom in.*" ] }, { @@ -219,7 +239,7 @@ "source": [ "You probably noticed that a large portion of the data is equal to zero. The data we are using is a reduced mosaic that combines many different exposures, and that has been rotated such that not all of the array holds data. \n", "\n", - "We want to **mask** out the non-data, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data." + "We want to **mask** out the non-data portions of the image array,, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data." ] }, { @@ -251,15 +271,16 @@ "ax1.set_title('Mask')\n", "\n", "# Plot the masked data\n", - "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", - "fitsplot = ax2.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", + "fitsplot = ax2.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", "\n", "# Define the colorbar and fix the labels\n", "cbar_ax = fig.add_axes([1, 0.09, 0.03, 0.87])\n", "cbar = fig.colorbar(fitsplot, cbar_ax, ticks=LogLocator(subs=range(10)))\n", "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", "cbar.ax.set_yticklabels(labels)\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", "ax2.set_xlabel('X (pixels)')\n", "ax2.set_title('Masked Data')" ] @@ -288,7 +309,7 @@ "metadata": {}, "outputs": [], "source": [ - "mean, median, std = sigma_clipped_stats(xdf_image.data, sigma=3.0, iters=5, mask=xdf_image.mask)" + "mean, median, std = sigma_clipped_stats(xdf_image.data, sigma=3.0, maxiters=5, mask=xdf_image.mask)" ] }, { @@ -305,7 +326,7 @@ "outputs": [], "source": [ "# Calculate the data without masking\n", - "stats_nomask = sigma_clipped_stats(xdf_image.data, sigma=3.0, iters=5)" + "stats_nomask = sigma_clipped_stats(xdf_image.data, sigma=3.0, maxiters=5)" ] }, { @@ -329,7 +350,7 @@ "ax1.axvline(np.average(xdf_image), label='Neither', c='C6', ls=':', lw=3)\n", "\n", "ax1.set_xlim(flux_range)\n", - "ax1.set_xlabel(r'Flux Count Rate ($e^{-1}/s$)', fontsize=14)\n", + "ax1.set_xlabel(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), fontsize=14)\n", "ax1.set_ylabel('Frequency', fontsize=14)\n", "ax1.set_title('Effect of Sigma-Clipping and Masking on Mean', fontsize=16)\n", "ax1.legend(fontsize=11)\n", @@ -343,7 +364,7 @@ "ax2.axvline(np.ma.median(xdf_image), label='Neither', c='C6', ls=':', lw=3)\n", "\n", "ax2.set_xlim(flux_range)\n", - "ax2.set_xlabel(r'Flux Count Rate ($e^{-1}/s$)', fontsize=14)\n", + "ax2.set_xlabel(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), fontsize=14)\n", "ax2.set_title('Effect of Sigma-Clipping and Masking on Median', fontsize=16)\n", "ax2.legend(fontsize=11)" ] @@ -377,10 +398,6 @@ "fig, [ax1, ax2] = plt.subplots(1, 2, figsize=(12, 6), sharey=True)\n", "plt.tight_layout()\n", "\n", - "# Define the normalization\n", - "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", - "xdf_scalar_bkgdsub_clipped = np.clip(xdf_scalar_bkgdsub, 1e-4, None) # clip to plot with logarithmic stretch\n", - "\n", "# Plot the original data\n", "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", "ax1.set_xlabel('X (pixels)')\n", @@ -388,6 +405,7 @@ "ax1.set_title('Original Data')\n", "\n", "# Plot the subtracted data\n", + "xdf_scalar_bkgdsub_clipped = np.clip(xdf_scalar_bkgdsub, 1e-4, None) # clip to plot with logarithmic stretch\n", "fitsplot = ax2.imshow(np.ma.masked_where(xdf_scalar_bkgdsub.mask, xdf_scalar_bkgdsub_clipped), norm=norm_image)\n", "ax2.set_xlabel('X (pixels)')\n", "ax2.set_title('Scalar Background-Subtracted Data')\n", @@ -397,7 +415,8 @@ "cbar = fig.colorbar(fitsplot, cbar_ax, ticks=LogLocator(subs=range(10)))\n", "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", "cbar.ax.set_yticklabels(labels)\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)" + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)" ] }, { @@ -414,10 +433,13 @@ "metadata": {}, "source": [ "
\n", - "

**Exercises:**


\n", + " \n", + "

Exercises:


\n", + "\n", "Perform a median scalar background subtraction on our sigma-clipped data. Plot it and visually inspect it. How does it compare to the original data?\n", "

\n", "Compare the median background subtraction to the mean background subtraction. Which is better?\n", + "\n", "
" ] }, @@ -480,7 +502,6 @@ "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", "\n", "# Plot the data\n", - "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", "background_clipped = np.clip(bkg.background, 1e-4, None) # clip to plot with logarithmic stretch\n", "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, background_clipped), norm=norm_image)\n", "\n", @@ -493,7 +514,8 @@ "cbar.ax.set_yticklabels(labels)\n", "\n", "# Define labels\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')\n", "ax1.set_title('2D Estimated Background')" @@ -529,12 +551,12 @@ "plt.tight_layout()\n", "\n", "# Define the normalization\n", - "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", "xdf_2d_bkgdsub_clipped = np.clip(xdf_2d_bkgdsub, 1e-4, None) # clip to plot with logarithmic stretch\n", "\n", "# Plot the scalar-subtracted data\n", "fitsplot = ax1.imshow(np.ma.masked_where(xdf_scalar_bkgdsub.mask, xdf_scalar_bkgdsub_clipped), norm=norm_image)\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", "ax1.set_ylabel('Y (pixels)')\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_title('Scalar Background-Subtracted Data')\n", @@ -549,7 +571,8 @@ "cbar = fig.colorbar(fitsplot, cbar_ax, ticks=LogLocator(subs=range(10)))\n", "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", "cbar.ax.set_yticklabels(labels)\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)" + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)" ] }, { @@ -564,10 +587,13 @@ "metadata": {}, "source": [ "
\n", - "

**Exercises:**


\n", + " \n", + "

Exercises:


\n", + "\n", "Calculate the standard deviation (with sigma-clipping and masking!) for the original data, the scalar background-subtracted data, and the 2D background-subtracted data. How do the values compare? Which has the smallest standard deviation?

\n", "\n", "Notice that the difference between each dataset's standard deviation is small - why might this be?\n", + "\n", "
" ] }, @@ -575,11 +601,12 @@ "cell_type": "markdown", "metadata": {}, "source": [ + "---\n", "## Conclusions\n", "\n", "The `photutils` package provides a powerful tool in the `Background2D` class, allowing users to easily estimate and subtract spatially variant background signals from their data.\n", "\n", - "**To continue with this `photutils` tutorial, go on to the [source detection notebook](02_photutils_source_detection.ipynb).**" + "**To continue with this `photutils` tutorial, go on to the [source detection notebook](../02_source_detection/02_source_detection.ipynb).**" ] }, { @@ -587,11 +614,26 @@ "metadata": {}, "source": [ "---\n", - "September 2018\n", - "\n", - "Author: Lauren Chambers (lchambers@stsci.edu), Erik Tollerud (etollerud@stsci.edu)\n", - "\n", - "For more examples and details, please visit the [photutils](http://photutils.readthedocs.io/en/stable/index.html) documentation." + "## Additional Resources\n", + "For more examples and details, please visit the [photutils](http://photutils.readthedocs.io/en/stable/index.html) documentation." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## About this Notebook\n", + "**Authors:** Lauren Chambers (lchambers@stsci.edu), Erik Tollerud (etollerud@stsci.edu)\n", + "
**Updated:** May 2019" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "[Top of Page](#title_ID)\n", + "\"STScI" ] } ], @@ -611,7 +653,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.5.5" + "version": "3.6.8" } }, "nbformat": 4, diff --git a/notebooks/photutils/02_source_detection/02_photutils_source_detection.ipynb b/notebooks/photutils/02_source_detection/02_source_detection.ipynb similarity index 93% rename from notebooks/photutils/02_source_detection/02_photutils_source_detection.ipynb rename to notebooks/photutils/02_source_detection/02_source_detection.ipynb index ead20219..b40fad3d 100644 --- a/notebooks/photutils/02_source_detection/02_photutils_source_detection.ipynb +++ b/notebooks/photutils/02_source_detection/02_source_detection.ipynb @@ -34,7 +34,7 @@ "##### Which data are used in this tutorial?\n", "We will be manipulating Hubble eXtreme Deep Field (XDF) data, which was collected using the Advanced Camera for Surveys (ACS) on Hubble between 2002 and 2012. The image we use here is the result of 1.8 million seconds (500 hours!) of exposure time, and includes some of the faintest and most distant galaxies that have ever been observed. \n", "\n", - "##### The methods demonstrated here are available in narrative form within the `photutils.detection` [documentation](http://photutils.readthedocs.io/en/stable/detection.html) and `photutils.segmentation` [documentation](http://photutils.readthedocs.io/en/stable/segmentation.html)." + "*The methods demonstrated here are available in narrative form within the `photutils.detection` [documentation](http://photutils.readthedocs.io/en/stable/detection.html) and `photutils.segmentation` [documentation](http://photutils.readthedocs.io/en/stable/segmentation.html).*" ] }, { @@ -88,7 +88,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [style file shared with the other photutils tutorials](photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" + "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [style file shared with the other photutils tutorials](../photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" ] }, { @@ -97,7 +97,7 @@ "metadata": {}, "outputs": [], "source": [ - "plt.style.use('photutils_notebook_style.mplstyle')" + "plt.style.use('../photutils_notebook_style.mplstyle')" ] }, { @@ -133,7 +133,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "As explained in the [previous notebook](01_photutils_background_estimation.ipynb) on background estimation, it is important to **mask** these data, as a large portion of the values are equal to zero. We will mask out the non-data portions of the image array, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data. " + "As explained in a [previous notebook](../01_background_estimation/01_background_estimation.ipynb) on background estimation, it is important to **mask** these data, as a large portion of the values are equal to zero. We will mask out the non-data portions of the image array, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data. " ] }, { @@ -465,15 +465,15 @@ "\n", "The answer comes down to the default settings for the two algorithms: (1) there are differences in the upper and lower bounds on the requirements for source roundness and sharpness, and (2) `IRAFStarFinder` includes a minimum separation between sources that `DAOStarFinder` does not have:\n", "\n", - "| | `IRAFStarFinder` | `DAOStarFinder` |\n", - "|------|------|------|\n", - "| sharplo | 0.5 | 0.2 |\n", - "| sharphi | 2.0 | 1.0 |\n", - "| roundlo | 0.0 | -1.0 |\n", - "| roundhi | 0.2 | 1.0 |\n", + "|   | `IRAFStarFinder` | `DAOStarFinder` |\n", + "|----------------|-------|------|\n", + "| sharplo | 0.5 | 0.2 |\n", + "| sharphi | 2.0 | 1.0 |\n", + "| roundlo | 0.0 | -1.0 |\n", + "| roundhi | 0.2 | 1.0 |\n", "| minsep_fwhm | 1.5 * FWHM | N/A |\n", "\n", - "Thinking about this, *it then makes sense* that `IRAFStarFinder` would find fewer sources. It has stricter restrictions on source roundness, meaning that it eliminates more elliptical galactic sources (this is the eXtreme Deep Field, after all!), and the minimum separation requirement further rules out sources that are too close to one another.\n", + "Thinking about this, *it then makes sense* that `IRAFStarFinder` would find fewer sources. It has tighter restrictions on source roundness and ``sharplo``, meaning that it eliminates more elliptical galactic sources (this is the eXtreme Deep Field, after all!), and the minimum separation requirement further rules out sources that are too close to one another.\n", "\n", "If we set all these parameters to be equivalent, though, we should find much better agreement between the two methods:" ] @@ -555,7 +555,7 @@ "source": [ "For more general source detection cases that do not require comparison with models, `photutils` offers the `find_peaks` function. \n", "\n", - "This function simply finds sources by identifying local maxima above a given threshold and separated by a given distance, rather than trying to fit data to a given model. Unlike the previous detection algorithms, `find_peaks` does not necessarily calculate objects' centroids. Unless the `subpixel` argument is set to `True`, `find_peaks` will return just the integer value of the peak pixel for each source.\n", + "This function simply finds sources by identifying local maxima above a given threshold and separated by a given distance, rather than trying to fit data to a given model. Unlike the previous detection algorithms, `find_peaks` does not necessarily calculate objects' centroids. Unless the `centroid_func` argument is passed a function like `photutils.centroids.centroid_2dg` that can handle source position centroiding, `find_peaks` will return just the integer value of the peak pixel for each source.\n", "\n", "This algorithm is particularly useful for identifying non-stellar sources or heavily distorted sources in image data.\n", "\n", @@ -568,7 +568,8 @@ "metadata": {}, "outputs": [], "source": [ - "from photutils import find_peaks" + "from photutils import find_peaks\n", + "from photutils.centroids import centroid_2dg" ] }, { @@ -578,7 +579,8 @@ "outputs": [], "source": [ "sources_findpeaks = find_peaks(xdf_image.data, mask=xdf_image.mask, \n", - " threshold=20.*std, box_size=30, subpixel=True) \n", + " threshold=20.*std, box_size=30, \n", + " centroid_func=centroid_2dg) \n", "print(sources_findpeaks)" ] }, @@ -659,7 +661,7 @@ "source": [ "print('''DAOStarFinder: {} sources\n", "IRAFStarFinder: {} sources\n", - "find_peaks: {} sources'''.format(len(sources_dao), len(sources_iraf), len(sources_findpeaks)))" + "find_peaks: {} sources'''.format(len(sources_dao), len(sources_iraf_match), len(sources_findpeaks)))" ] }, { @@ -679,13 +681,13 @@ "source": [ "# Make lists of centroid coordinates\n", "centroids_dao = [(x, y) for x, y in sources_dao['xcentroid', 'ycentroid']]\n", - "centroids_iraf = [(x, y) for x, y in sources_iraf['xcentroid', 'ycentroid']]\n", + "centroids_iraf = [(x, y) for x, y in sources_iraf_match['xcentroid', 'ycentroid']]\n", "centroids_findpeaks = [(x, y) for x, y in sources_findpeaks['x_centroid', 'y_centroid']]\n", "\n", - "# Round those coordinates to the first decimal place and convert them to be sets\n", - "rounded_centroids_dao = set([(round(x, 1), round(y, 1)) for x, y in centroids_dao])\n", - "rounded_centroids_iraf = set([(round(x, 1), round(y, 1)) for x, y in centroids_iraf])\n", - "rounded_centroids_findpeaks = set([(round(x, 1), round(y, 1)) for x, y in centroids_findpeaks])\n", + "# Round those coordinates to the ones place and convert them to be sets\n", + "rounded_centroids_dao = set([(round(x, 0), round(y, 0)) for x, y in centroids_dao])\n", + "rounded_centroids_iraf = set([(round(x, 0), round(y, 0)) for x, y in centroids_iraf])\n", + "rounded_centroids_findpeaks = set([(round(x, 0), round(y, 0)) for x, y in centroids_findpeaks])\n", "\n", "# Examine the intersections of different sets to determine which sources are shared\n", "all_match = rounded_centroids_dao.intersection(rounded_centroids_iraf).intersection(rounded_centroids_findpeaks)\n", @@ -721,18 +723,18 @@ "# Plot the data\n", "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", "ax1.scatter([x for x, y in list(dao_findpeaks_match)], [y for x, y in list(dao_findpeaks_match)],\n", - " s=30, marker='s', lw=1.5, facecolor='None', edgecolor='#EE7733',\n", + " s=30, marker='s', lw=1, facecolor='None', edgecolor='#EE7733',\n", " label='Found by DAO \\& find\\_peaks')\n", "ax1.scatter([x for x, y in list(dao_iraf_match)], [y for x, y in list(dao_iraf_match)],\n", - " s=30, marker='D', lw=1.5, facecolor='None', edgecolor='#EE3377',\n", + " s=30, marker='D', lw=1, facecolor='None', edgecolor='#EE3377',\n", " label='Found by DAO \\& IRAF')\n", "ax1.scatter([x for x, y in list(iraf_findpeaks_match)], [y for x, y in list(iraf_findpeaks_match)],\n", - " s=30, marker='o', lw=1.5, facecolor='None', edgecolor='#0077BB',\n", + " s=30, marker='o', lw=1, facecolor='None', edgecolor='#0077BB',\n", " label='Found by IRAF \\& find\\_peaks')\n", "ax1.scatter([x for x, y in list(all_match)], [y for x, y in list(all_match)],\n", " s=30, marker='o', lw=1.2, linestyle=':',facecolor='None', edgecolor='#BBBBBB',\n", " label='Found by all methods')\n", - "ax1.legend()\n", + "ax1.legend(ncol=2)\n", "\n", "# Define the colorbar\n", "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", @@ -1045,7 +1047,7 @@ "\n", "The `photutils` package provides users with a variety of methods for detecting sources in their data, from familar algorithms such as `DAOFind` and `starfind`, to more complex and customizable image segmentation algorithms. These methods allow for easy creation of a diverse array of apertures that can be used for photometric analysis.\n", "\n", - "**To continue with this `photutils` tutorial, go on to the [aperture photometry](03_photutils_aperture_photometry.ipynb) or [PSF photometry notebook](04_photutils_psf_photometry.ipynb).**" + "**To continue with this `photutils` tutorial, go on to the [aperture photometry](../03_aperture_photometry/03_aperture_photometry.ipynb) or [PSF photometry notebook](../04_psf_photometry/04_psf_photometry.ipynb).**" ] }, { diff --git a/notebooks/photutils/03_photutils_aperture_photometry/03_photutils_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb similarity index 88% rename from notebooks/photutils/03_photutils_aperture_photometry/03_photutils_aperture_photometry.ipynb rename to notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index 0368dbe4..2363a079 100644 --- a/notebooks/photutils/03_photutils_aperture_photometry/03_photutils_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -35,7 +35,7 @@ "##### Which data are used in this tutorial?\n", "We will be manipulating Hubble eXtreme Deep Field (XDF) data, which was collected using the Advanced Camera for Surveys (ACS) on Hubble between 2002 and 2012. The image we use here is the result of 1.8 million seconds (500 hours!) of exposure time, and includes some of the faintest and most distant galaxies that have ever been observed. \n", "\n", - "##### The methods demonstrated here are available in narrative form within the `photutils.aperture` [documentation](http://photutils.readthedocs.io/en/stable/aperture.html)." + "*The methods demonstrated here are available in narrative form within the `photutils.aperture` [documentation]( http://photutils.readthedocs.io/en/stable/aperture.html).*" ] }, { @@ -89,7 +89,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [shared style file](photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" + "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [style file shared with the other photutils tutorials](../photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" ] }, { @@ -98,7 +98,7 @@ "metadata": {}, "outputs": [], "source": [ - "plt.style.use('photutils_notebook_style.mplstyle')" + "plt.style.use('../photutils_notebook_style.mplstyle')" ] }, { @@ -114,7 +114,7 @@ "source": [ "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", "\n", - "(Generally, the best package for web queries of astronomical data is `astroquery`; however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with `astroquery`.)" + "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)" ] }, { @@ -134,7 +134,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "As explained in the [previous notebook](01_photutils_background_estimation.ipynb) on background estimation, it is important to **mask** these data, as a large portion of the values are equal to zero. We will mask out the non-data, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data. " + "As explained in a [previous notebook](../01_background_estimation/01_background_estimation.ipynb) on background estimation, it is important to **mask** these data, as a large portion of the values are equal to zero. We will mask out the non-data portions of the image array, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data. " ] }, { @@ -160,7 +160,7 @@ "metadata": {}, "outputs": [], "source": [ - "unit = u.ct / u.s\n", + "unit = u.electron / u.s\n", "xdf_image = CCDData(data, unit=unit, meta=header, mask=mask)" ] }, @@ -182,10 +182,17 @@ "# Set up the figure with subplots\n", "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", "\n", - "# Plot the data\n", - "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch())\n", + "# Set up the normalization and colormap\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch(), clip=False)\n", + "cmap = plt.get_cmap('viridis')\n", + "cmap.set_over(cmap.colors[-1])\n", + "cmap.set_under(cmap.colors[0])\n", + "cmap.set_bad('white') # Show masked data as white\n", "xdf_image_clipped = np.clip(xdf_image, 1e-4, None) # clip to plot with logarithmic stretch\n", - "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", "\n", "# Define the colorbar and fix the labels\n", "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", @@ -193,7 +200,8 @@ "cbar.ax.set_yticklabels(labels)\n", "\n", "# Define labels\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')" ] @@ -224,7 +232,7 @@ "\n", "It is also possible for users to create custom aperture shapes.\n", "\n", - "Any aperture object is created by defining its position and size (and, if applicable, orientation). Let's use the `find_peaks` method that we learned about in a [previous notebook](02_photutils_source_detection.ipynb) to get the positions of sources in our data:" + "Any aperture object is created by defining its position and size (and, if applicable, orientation). Let's use the `find_peaks` method that we learned about in a [previous notebook](../02_source_detection/02_source_detection.ipynb) to get the positions of sources in our data:" ] }, { @@ -233,7 +241,8 @@ "metadata": {}, "outputs": [], "source": [ - "from photutils import find_peaks" + "from photutils import find_peaks\n", + "from photutils.centroids import centroid_2dg" ] }, { @@ -243,7 +252,7 @@ "outputs": [], "source": [ "# Calculate statistics\n", - "mean, median, std = sigma_clipped_stats(xdf_image.data, sigma=3.0, iters=5, mask=xdf_image.mask)" + "mean, median, std = sigma_clipped_stats(xdf_image.data, sigma=3.0, maxiters=5, mask=xdf_image.mask)" ] }, { @@ -253,7 +262,8 @@ "outputs": [], "source": [ "sources_findpeaks = find_peaks(xdf_image.data, mask=xdf_image.mask, \n", - " threshold=20.*std, box_size=30, subpixel=True) \n", + " threshold=20.*std, box_size=30, \n", + " centroid_func=centroid_2dg) \n", "# Display the table\n", "sources_findpeaks" ] @@ -285,7 +295,8 @@ "cbar.ax.set_yticklabels(labels)\n", "\n", "# Define labels\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')\n", "ax1.set_title('find\\_peaks Sources')" @@ -342,7 +353,8 @@ "cbar.ax.set_yticklabels(labels)\n", "\n", "# Define labels\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')\n", "ax1.set_title('Circular Apertures')\n", @@ -382,7 +394,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "In a [previous notebook](02_photutils_source_detection.ipynb), we showed how you can use the `photutils.detect_sources` feature to generate segmentation maps, which identify and label contiguous (connected) objects within an image. Then, with `source_properties`, you can access descriptive properties for each unique object - not just their centroid positions, but also their pixel areas, eccentricities, orientations with respect to the coordinate frame of the image, and more.\n", + "In a [previous notebook](../02_source_detection/02_source_detection.ipynb), we showed how you can use the `photutils.detect_sources` [feature](https://photutils.readthedocs.io/en/stable/api/photutils.detect_sources.html) to generate segmentation maps, which identify and label contiguous (connected) objects within an image. Then, with `source_properties` [feature](https://photutils.readthedocs.io/en/stable/api/photutils.segmentation.source_properties.html?highlight=source_properties), you can access descriptive properties for each unique object - not just their centroid positions, but also their pixel areas, eccentricities, orientations with respect to the coordinate frame of the image, and more.\n", "\n", "Here we'll use the centroid, semimajor axis, semiminor axis, and orientation values from `source_properties` to generate elliptical apertures for each of the sources in our image." ] @@ -450,7 +462,8 @@ "cbar.ax.set_yticklabels(labels)\n", "\n", "# Define labels\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')\n", "ax1.set_title('Elliptical Apertures')\n", @@ -607,7 +620,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "The `aperture_sum` value is what reports the number of counts within the aperture: 3.47 ct/s.\n", + "The `aperture_sum` value is what reports the number of electron counts within the aperture: 3.47 e–/s.\n", "\n", "And, just as a check, to make sure our sky apertures give basically the same answer..." ] @@ -693,7 +706,7 @@ "plt.yscale('log')\n", "plt.xscale('log')\n", "plt.title('Histogram of Source Photometry')\n", - "plt.xlabel('Count Rate [ct/s]')\n", + "plt.xlabel(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')))\n", "plt.ylabel('Number of Sources')" ] }, @@ -713,7 +726,7 @@ "plt.xscale('log')\n", "plt.title('Count Rate v. Aperture Area')\n", "plt.xlabel('Aperture Area [pixels$^2$]')\n", - "plt.ylabel('Count Rate [ct/s]')" + "plt.xlabel(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')))" ] }, { @@ -745,7 +758,7 @@ "\n", "## Local Background Subtraction\n", "\n", - "In the [background estimation notebook](01_photutils_background_estimation.ipynb), we explored how to perform global background subtraction of image data with `photutils`. However, you can also use `photutils` to perform local background estimations for aperture corrections.\n", + "In the [background estimation notebook](../01_background_estimation/01_background_estimation.ipynb), we explored how to perform global background subtraction of image data with `photutils`. However, you can also use `photutils` to perform local background estimations for aperture corrections.\n", "\n", "To estimate the local background for each aperture, measure the counts within annulus apertures around (but not including!) each source. In our example, we defined elliptical apertures with `r = 3` to measure the counts within each source. To calculate the background for each source, let's measure the counts elliptical annuli between `r = 3.5` and `r = 5`." ] @@ -795,7 +808,8 @@ "cbar.ax.set_yticklabels(labels)\n", "\n", "# Define labels\n", - "cbar.set_label(r'Flux Count Rate ($e^{-1}/s$)', rotation=270, labelpad=30)\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')\n", "ax1.set_title('Elliptical Annuli')\n", @@ -847,13 +861,6 @@ "bkg_phot_table" ] }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "You might have noticed that these background count rates are *really* small. In this case, this is to be expected - since our example XDF data is a high-level science product (HLSP) that already has already been background-subtracted." - ] - }, { "cell_type": "code", "execution_count": null, @@ -863,7 +870,7 @@ "# Calculate the mean background level (per pixel) in the annuli \n", "bkg_area = [annulus.area() for annulus in elliptical_annuli]\n", "bkg_mean_per_aperture = bkg_phot_table['aperture_sum'].value / bkg_area\n", - "bkg_mean = np.average(bkg_mean_per_aperture) * (u.ct / u.s)\n", + "bkg_mean = np.average(bkg_mean_per_aperture) * (u.electron / u.s)\n", "print('Background mean:', bkg_mean)\n", "\n", "# Calculate the total background within each elliptical aperture\n", @@ -876,6 +883,13 @@ "phot_table['aperture_sum_bkgsub'] = flux_bkgsub" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "You might have noticed that these background count rates are *really* small. In this case, this is to be expected - since our example XDF data is a high-level science product (HLSP) that already has already been background-subtracted." + ] + }, { "cell_type": "markdown", "metadata": {}, @@ -910,7 +924,7 @@ "plt.yscale('log')\n", "plt.xscale('log')\n", "plt.title('Histogram of Source Photometry')\n", - "plt.xlabel('Count Rate [ct/s]')\n", + "plt.xlabel(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')))\n", "plt.legend()" ] }, @@ -923,7 +937,7 @@ "\n", "The `photutils` package provides a comprehensive toolkit for astronomers to perform aperture photometry, including customizable aperture shapes that allow for more precise photometry and easy photometric correction.\n", "\n", - "**To continue with this `photutils` tutorial, go on to the [PSF photometry notebook](04_photutils_psf_photometry.ipynb).**" + "**To continue with this `photutils` tutorial, go on to the [PSF photometry notebook](../04_psf_photometry/04_psf_photometry.ipynb).**" ] }, { @@ -941,7 +955,7 @@ "source": [ "---\n", "## About this Notebook\n", - "**Authors:** Lauren Chambers (lchambers@stsci.edu), Erik Tollerud (etollerud@stsci.edu), Clare Shanahan (cshanahan@stsci.edu)\n", + "**Authors:** Lauren Chambers (lchambers@stsci.edu), Erik Tollerud (etollerud@stsci.edu), Tom Wilson (towilson@stsci.edu) Clare Shanahan (cshanahan@stsci.edu)\n", "
**Updated:** May 2019" ] }, @@ -970,7 +984,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.5.5" + "version": "3.6.8" } }, "nbformat": 4, diff --git a/notebooks/photutils/03_photutils_aperture_photometry/requirements.txt b/notebooks/photutils/03_aperture_photometry/requirements.txt similarity index 100% rename from notebooks/photutils/03_photutils_aperture_photometry/requirements.txt rename to notebooks/photutils/03_aperture_photometry/requirements.txt diff --git a/notebooks/photutils/04_psf_photometry/04_photutils_psf_photometry.ipynb b/notebooks/photutils/04_psf_photometry/04_photutils_psf_photometry.ipynb deleted file mode 100644 index 0a6aa615..00000000 --- a/notebooks/photutils/04_psf_photometry/04_photutils_psf_photometry.ipynb +++ /dev/null @@ -1,253 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "[](http://photutils.readthedocs.io/en/stable/index.html)\n", - "\n", - "# PSF Photometry with `photutils`\n", - "---" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "##### What is PSF photometry?\n", - "A more specific form of photometry than aperture photometry, PSF photometry takes into account the shape of a source's point spread function (PSF). The PSF is a model that represents the distribution of light from a point source as it falls onto a detector. An example of a basic PSF is simply a 2-D Gaussian, while more complex PSFs can include distortion, diffraction, or interference effects associated with a particular telescope. For instance, the PSFs from the Hubble Space Telescope and the James Webb Space Telescope have been meticulously modeled, and can be simulated with the [Tiny Tim](http://www.stsci.edu/hst/observatory/focus/TinyTim) and [WebbPSF](https://github.com/mperrin/webbpsf) software packages, respectively. However, for datasets that do not have readily available PSF models, such models can be statistically generated by analyzing the image itself.\n", - "\n", - "The `photutils` package provides tools that combine background estimation, source detection, and model-fitting to perform PSF photometry on image data.\n", - "\n", - "##### What does this tutorial include?\n", - "This tutorial covers how to perform PSF photometry with `photutils`, including the following methods:\n", - "* Gaussian PSF Photometry\n", - "* Iterative Subtraction\n", - "* Point Response Function (PRF) Photometry\n", - "\n", - "The methods demonstrated here are available in narrative form within the `photutils.psf` [documentation](http://photutils.readthedocs.io/en/stable/psf.html)." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "
**Warning:** The PSF photometry API is currently considered experimental and may change in the future. The photutils development team will aim to keep compatibility where practical, but will not finalize the API until sufficient user feedback has been accumulated.
**Important:** Before proceeding, please be sure to install or update your [AstroConda](https://astroconda.readthedocs.io) distribution. This notebook may not work properly with older versions of AstroConda.
\n", - "\n", - "---" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Import necessary packages" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "First, let's import packages that we will use to perform arithmetic functions and visualize data:" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [], - "source": [ - "import numpy as np\n", - "from astropy.io import fits\n", - "from astropy.stats import sigma_clipped_stats, SigmaClip\n", - "from astropy.visualization import ZScaleInterval, ImageNormalize\n", - "from photutils import make_source_mask\n", - "from photutils.background import Background2D, MedianBackground\n", - "import matplotlib\n", - "import matplotlib.pyplot as plt\n", - "from matplotlib.colors import LogNorm\n", - "% matplotlib inline" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's also define some `matplotlib` parameters, to make sure our plots look nice. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [], - "source": [ - "matplotlib.rc('font', family='serif', weight='light', size=12)\n", - "matplotlib.rc('mathtext', bf='serif:normal')\n", - "matplotlib.rc('axes', titlesize=18, titlepad=12, labelsize=16)\n", - "matplotlib.rc('xtick', labelsize=14)\n", - "matplotlib.rc('ytick', labelsize=14)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Retrieve data" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We have place the data for this tutorial in the github repository, for easy access. The data were originally retrieved from the STScI archive: https://archive.stsci.edu/prepds/udf/udf_hlsp.html." - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [], - "source": [ - "with fits.open('data/h_udf_wfc_v_drz_img.fits') as hdulist:\n", - " v_data = hdulist[0].data" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's look at the data:" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", - "\n", - "clim = (0, 1e-2)\n", - "\n", - "v_data_plot = np.copy(v_data)\n", - "v_data_plot[v_data_plot <= 0] = 1e-10\n", - "\n", - "norm_image = ImageNormalize(v_data, interval=ZScaleInterval())\n", - "fitsplot = ax1.imshow(v_data, norm=norm_image)#, clim=clim)\n", - "\n", - "plt.colorbar(fitsplot, fraction=0.046, pad=0.04)\n", - "\n", - "plt.show()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Circular Apertures" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Creating Apertures" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Performing Aperture Photometry" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Calculating Aperture Corrections" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Elliptical Apertures" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Creating Apertures" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Performing Aperture Photometry" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Calculating Aperture Corrections" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "---\n", - "## Exercises" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "---\n", - "May 2018\n", - "\n", - "Author: Lauren Chambers (lchambers@stsci.edu)\n", - "\n", - "For more examples and details, please visit the [photutils](http://photutils.readthedocs.io/en/stable/index.html) documentation." - ] - } - ], - "metadata": { - "kernelspec": { - "display_name": "Python 3", - "language": "python", - "name": "python3" - }, - "language_info": { - "codemirror_mode": { - "name": "ipython", - "version": 3 - }, - "file_extension": ".py", - "mimetype": "text/x-python", - "name": "python", - "nbconvert_exporter": "python", - "pygments_lexer": "ipython3", - "version": "3.5.5" - } - }, - "nbformat": 4, - "nbformat_minor": 2 -} diff --git a/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb b/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb new file mode 100644 index 00000000..27e39968 --- /dev/null +++ b/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb @@ -0,0 +1,333 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "\n", + "\n", + "[\n", + "\n", + "](http://photutils.readthedocs.io/en/stable/index.html)\n", + "\n", + "# PSF Photometry with `photutils`\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "##### What is PSF photometry?\n", + "A more specific form of photometry than aperture photometry, PSF photometry takes into account the shape of a source's point spread function (PSF). The PSF is a model that represents the distribution of light from a point source as it falls onto a detector. An example of a basic PSF is simply a 2-D Gaussian, while more complex PSFs can include distortion, diffraction, or interference effects associated with a particular telescope. For instance, the PSFs from the Hubble Space Telescope and the James Webb Space Telescope have been meticulously modeled, and can be simulated with the [Tiny Tim](http://www.stsci.edu/hst/observatory/focus/TinyTim) and [WebbPSF](https://github.com/mperrin/webbpsf) software packages, respectively. However, for datasets that do not have readily available PSF models, such models can be statistically generated by analyzing the image itself.\n", + "\n", + "The `photutils` package provides tools that combine background estimation, source detection, and model-fitting to perform PSF photometry on image data.\n", + "\n", + "##### What does this tutorial include?\n", + "This tutorial covers how to perform PSF photometry with `photutils`, including the following methods:\n", + "* Gaussian PSF Photometry\n", + "* Iterative Subtraction\n", + "* Point Response Function (PRF) Photometry\n", + "\n", + "##### Which data are used in this tutorial?\n", + "We will be manipulating Hubble eXtreme Deep Field (XDF) data, which was collected using the Advanced Camera for Surveys (ACS) on Hubble between 2002 and 2012. The image we use here is the result of 1.8 million seconds (500 hours!) of exposure time, and includes some of the faintest and most distant galaxies that have ever been observed. \n", + "\n", + "*The methods demonstrated here are available in narrative form within the `photutils.psf` [documentation](http://photutils.readthedocs.io/en/stable/psf.html).*" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "
\n", + " \n", + "**Warning:** The PSF photometry API is currently considered experimental and may change in the future. The photutils development team will aim to keep compatibility where practical, but will not finalize the API until sufficient user feedback has been accumulated.\n", + "\n", + "
\n", + "\n", + "
\n", + "\n", + "**Important:** Before proceeding, please be sure to install or update your [AstroConda](https://astroconda.readthedocs.io) distribution. This notebook may not work properly with older versions of AstroConda.\n", + "\n", + "
\n", + "\n", + "---" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Import necessary packages" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "First, let's import packages that we will use to perform arithmetic functions and visualize data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from astropy.io import fits\n", + "import astropy.units as u\n", + "from astropy.nddata import CCDData\n", + "# from astropy.stats import sigma_clipped_stats\n", + "from astropy.visualization import ImageNormalize, LogStretch\n", + "import matplotlib\n", + "from matplotlib.colors import LogNorm\n", + "import matplotlib.pyplot as plt\n", + "from matplotlib.ticker import LogLocator\n", + "import numpy as np\n", + "\n", + "# Show plots in the notebook\n", + "%matplotlib inline" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [style file shared with the other photutils tutorials](../photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "plt.style.use('../photutils_notebook_style.mplstyle')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Retrieve data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", + "\n", + "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "url = 'https://archive.stsci.edu/pub/hlsp/xdf/hlsp_xdf_hst_acswfc-60mas_hudf_f435w_v1_sci.fits'\n", + "with fits.open(url) as hdulist:\n", + " hdulist.info()\n", + " data = hdulist[0].data\n", + " header = hdulist[0].header" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "As explained in a [previous notebook](../01_background_estimation/01_background_estimation.ipynb) on background estimation, it is important to **mask** these data, as a large portion of the values are equal to zero. We will mask out the non-data portions of the image array, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Define the mask\n", + "mask = data == 0" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons (counts) per second." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "unit = u.electron / u.s\n", + "xdf_image = CCDData(data, unit=unit, meta=header, mask=mask)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's look at the data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "# Set up the figure with subplots\n", + "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", + "\n", + "# Set up the normalization and colormap\n", + "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch(), clip=False)\n", + "cmap = plt.get_cmap('viridis')\n", + "cmap.set_over(cmap.colors[-1])\n", + "cmap.set_under(cmap.colors[0])\n", + "cmap.set_bad('white') # Show masked data as white\n", + "xdf_image_clipped = np.clip(xdf_image, 1e-4, None) # clip to plot with logarithmic stretch\n", + "\n", + "# Plot the data\n", + "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", + " norm=norm_image, cmap=cmap)\n", + "\n", + "# Define the colorbar and fix the labels\n", + "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", + "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", + "cbar.ax.set_yticklabels(labels)\n", + "\n", + "# Define labels\n", + "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", + " rotation=270, labelpad=30)\n", + "ax1.set_xlabel('X (pixels)')\n", + "ax1.set_ylabel('Y (pixels)')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "*Tip: Double-click on any inline plot to zoom in.*" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Circular Apertures" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Creating Apertures" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Performing Aperture Photometry" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Calculating Aperture Corrections" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Elliptical Apertures" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Creating Apertures" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Performing Aperture Photometry" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Calculating Aperture Corrections" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## Exercises" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## Additional Resources\n", + "For more examples and details, please visit the [photutils](http://photutils.readthedocs.io/en/stable/index.html) documentation." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "---\n", + "## About this Notebook\n", + "**Authors:** Lauren Chambers (lchambers@stsci.edu)\n", + "
**Updated:** May 2019" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "[Top of Page](#title_ID)\n", + "\"STScI" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.6.8" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} From ef462b5daea49dc53fcae39759d93f43a5e23d62 Mon Sep 17 00:00:00 2001 From: Lauren Chambers Date: Mon, 20 May 2019 16:58:30 -0400 Subject: [PATCH 03/14] Fixes so HTML renders correctly --- .../01_background_estimation.ipynb | 8 +++----- .../02_source_detection.ipynb | 18 ++++++++---------- .../03_aperture_photometry.ipynb | 13 ++++++------- .../04_psf_photometry/04_psf_photometry.ipynb | 10 ++++------ 4 files changed, 21 insertions(+), 28 deletions(-) diff --git a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb index c02e4078..9136a730 100644 --- a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb +++ b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb @@ -6,9 +6,7 @@ "source": [ "\n", "\n", - "[\n", - "\n", - "](http://photutils.readthedocs.io/en/stable/index.html)\n", + "\n", "\n", "# Background Estimation with `photutils`\n", "---" @@ -42,13 +40,13 @@ "source": [ "
\n", "\n", - "**Note:** This notebook focuses on global background estimation. Local background subtraction with **annulus apertures** is demonstrated in the [aperture photometry notebook](../03_aperture_photometry/03_aperture_photometry.ipynb).\n", + "Note: This notebook focuses on global background estimation. Local background subtraction with annulus apertures is demonstrated in the aperture photometry notebook.\n", "\n", "
\n", "\n", "
\n", " \n", - " **Important:** Before proceeding, please be sure to update your versions of `astropy`, `matplotlib`, and `photutils`, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the [AstroConda](https://astroconda.readthedocs.io) distribution.\n", + "Important: Before proceeding, please be sure to update your versions of astropy, matplotlib, and photutils, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the AstroConda distribution.\n", " \n", "
\n", "\n", diff --git a/notebooks/photutils/02_source_detection/02_source_detection.ipynb b/notebooks/photutils/02_source_detection/02_source_detection.ipynb index b40fad3d..2de6f6da 100644 --- a/notebooks/photutils/02_source_detection/02_source_detection.ipynb +++ b/notebooks/photutils/02_source_detection/02_source_detection.ipynb @@ -6,9 +6,7 @@ "source": [ "\n", "\n", - "[\n", - "\n", - "](http://photutils.readthedocs.io/en/stable/index.html)\n", + "\n", "\n", "\n", "# Source Detection with `photutils`\n", @@ -42,9 +40,9 @@ "metadata": {}, "source": [ "
\n", - "\n", - "**Important:** Before proceeding, please be sure to update your versions of `astropy`, `matplotlib`, and `photutils`, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the [AstroConda](https://astroconda.readthedocs.io) distribution.\n", - "\n", + " \n", + "Important: Before proceeding, please be sure to update your versions of astropy, matplotlib, and photutils, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the AstroConda distribution.\n", + " \n", "
\n", "\n", "---" @@ -336,7 +334,7 @@ "\n", "

Exercises:


\n", "\n", - "Re-run the `DAOStarFinder` algorithm with a smaller threshold (like 5σ), and plot the sources that it finds. Do the same, but with a larger threshold (like 100σ). How did changing the threshold affect the results?\n", + "Re-run the DAOStarFinder algorithm with a smaller threshold (like 5σ), and plot the sources that it finds. Do the same, but with a larger threshold (like 100σ). How did changing the threshold affect the results?\n", "\n", "" ] @@ -445,7 +443,7 @@ "\n", "

Exercises:


\n", "\n", - "Re-run the `IRAFStarFinder` algorithm with a smaller full-width-half-max (FWHM) – try 3 pixels – and plot the sources that it finds. Do the same, but with a larger FWHM (like 10 pixels). How did changing the FWHM affect the results? What astronomical objects might be better captures by smaller FWHM? Larger?\n", + "Re-run the IRAFStarFinder algorithm with a smaller full-width-half-max (FWHM) – try 3 pixels – and plot the sources that it finds. Do the same, but with a larger FWHM (like 10 pixels). How did changing the FWHM affect the results? What astronomical objects might be better captures by smaller FWHM? Larger?\n", "\n", "" ] @@ -910,7 +908,7 @@ "\n", "

Exercises:


\n", "\n", - "Recompute the `SegmentationImage`, but alter the threshold and the minimum number of pixels in a source. How does changing the threshold affect the results? What about changing the number of pixels?\n", + "Recompute the SegmentationImage, but alter the threshold and the minimum number of pixels in a source. How does changing the threshold affect the results? What about changing the number of pixels?\n", "\n", "" ] @@ -1026,7 +1024,7 @@ "\n", "

Exercises:


\n", "\n", - "Play with the isophotal extent of the elliptical apertures (defined above as `r`). Observe how changing this value affects the apertures that are created.\n", + "Play with the isophotal extent of the elliptical apertures (defined above as r). Observe how changing this value affects the apertures that are created.\n", "\n", "" ] diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index 2363a079..17ec279c 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -6,9 +6,7 @@ "source": [ "\n", "\n", - "[\n", - "\n", - "](http://photutils.readthedocs.io/en/stable/index.html)\n", + "\n", "\n", "# Aperture Photometry with `photutils`\n", "---" @@ -43,9 +41,9 @@ "metadata": {}, "source": [ "
\n", - "\n", - "**Important:** Before proceeding, please be sure to install or update your [AstroConda](https://astroconda.readthedocs.io) distribution. This notebook may not work properly with older versions of AstroConda.\n", - "\n", + " \n", + "Important: Before proceeding, please be sure to update your versions of astropy, matplotlib, and photutils, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the AstroConda distribution.\n", + " \n", "
\n", "\n", "---" @@ -226,6 +224,7 @@ "metadata": {}, "source": [ "With `photutils`, users can create apertures with the following shapes:\n", + "\n", "\"Examples\n", "\n", "Each of these can be defined either in pixel coordinates or in celestial coordinates (using a WCS transformation).\n", @@ -737,7 +736,7 @@ "\n", "

Exercise:


\n", "\n", - "Re-calculate the photometry for these elliptical apertures - or just a subset of them - using the `subpixel` aperture placement method instead of the default `exact` method. How does this affect the count sum calculated for those apertures?\n", + "Re-calculate the photometry for these elliptical apertures - or just a subset of them - using the subpixel aperture placement method instead of the default exact method. How does this affect the count sum calculated for those apertures?\n", "\n", "" ] diff --git a/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb b/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb index 27e39968..7ebdfd1d 100644 --- a/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb +++ b/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb @@ -6,9 +6,7 @@ "source": [ "\n", "\n", - "[\n", - "\n", - "](http://photutils.readthedocs.io/en/stable/index.html)\n", + "\n", "\n", "# PSF Photometry with `photutils`\n", "---" @@ -46,9 +44,9 @@ "\n", "\n", "
\n", - "\n", - "**Important:** Before proceeding, please be sure to install or update your [AstroConda](https://astroconda.readthedocs.io) distribution. This notebook may not work properly with older versions of AstroConda.\n", - "\n", + " \n", + "Important: Before proceeding, please be sure to update your versions of astropy, matplotlib, and photutils, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the AstroConda distribution.\n", + " \n", "
\n", "\n", "---" From d0a8698573075b3cf909cb917a584f1359c6387d Mon Sep 17 00:00:00 2001 From: Lauren Chambers Date: Mon, 20 May 2019 17:03:01 -0400 Subject: [PATCH 04/14] Add .gitignore from #99 and add apertures PNG --- notebooks/.gitignore | 39 ++++++++++++++++++ .../03_aperture_photometry.ipynb | 2 +- .../03_aperture_photometry/apertures.png | Bin 0 -> 76112 bytes 3 files changed, 40 insertions(+), 1 deletion(-) create mode 100644 notebooks/.gitignore create mode 100644 notebooks/photutils/03_aperture_photometry/apertures.png diff --git a/notebooks/.gitignore b/notebooks/.gitignore new file mode 100644 index 00000000..27a68f88 --- /dev/null +++ b/notebooks/.gitignore @@ -0,0 +1,39 @@ +# Files generated by convert.py +# ----------------------------- +exec_*.ipynb +*.html +*.fits +*.coo +*.match +*.log +*.png +*.list +*.txt +*.cat +*.jpg +*.gif +*.zip +DrizzlePac/Initialization/reference_files/ +DrizzlePac/drizzle_wfpc2/reference_files/ +MAST/Kepler/Kepler_Lightcurve/mastDownload/ +MAST/Kepler/Kepler_TPF/mastDownload/ + +# Exceptions committed to repo +# ---------------------------- +!DrizzlePac/align_sparse_fields/input_flc.list +!DrizzlePac/mask_satellite/sat.jpeg +!DrizzlePac/sky_matching/drz.list +!DrizzlePac/sky_matching/labeled_local_globalmatch_match.gif +!DrizzlePac/sky_matching/MDRIZSKY_Values.png +!MAST/HSC/HSC_TAP/smc_colormag.png +!MAST/Kepler/Kepler_FFI/ffi_tic_plot.png +!MAST/Kepler/Kepler_TPF/tpf_fluxplot0.png +!MAST/Kepler/Kepler_TPF/tpf_fluxplot_28-29.png +!MAST/Kepler/Kepler_Lightcurve/light_curve_tres2.png +!MAST/K2/K2_FFI/kepler_focal_plane_layout_channels_color.png +!MAST/PanSTARRS/PS1_DR2_TAP/stsci_pri_combo_mark_horizonal_white_bkgd.png +!photutils/03_aperture_photometry/apertures.png +!**/requirements.txt +!**/skyfile.txt +!**/exclusions.txt +!**/inclusions*.txt diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index 17ec279c..ef35021a 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -225,7 +225,7 @@ "source": [ "With `photutils`, users can create apertures with the following shapes:\n", "\n", - "\"Examples\n", + "\"Examples\n", "\n", "Each of these can be defined either in pixel coordinates or in celestial coordinates (using a WCS transformation).\n", "\n", diff --git a/notebooks/photutils/03_aperture_photometry/apertures.png b/notebooks/photutils/03_aperture_photometry/apertures.png new file mode 100644 index 0000000000000000000000000000000000000000..e79a4af6a5bbae8ec43e5b9bb2a90c85736acee2 GIT binary patch literal 76112 zcmeFYWmH^S);044Sl*ma*c(ksFJj(D4CLjovDSj2@H%>XhIT#8rCju|B;+~-YYR-nP2kJ6zIb4XxU`M z)}KFP%AopXe~`r0uxBa?G0>V&M+Xy%TZe!bbLZ~-89~3lX;$T!BDj8TJj%FQaeLzB zSv^|vble%`Imv%5Z`rTFMc|UekB|p6EMklHqXd z2ec2K2Fa=t%_hlMk~fzF=>ztp5qB^HPJ;TWAYqTFZQf>o?6T(>fz@9$X)%rV!>e1! z5?FegU^b;wQqF>I(ECh7pVVl!?YOXTxNh#T!BHziAp`BX*3$4k>rc%@PueN?HPu!x z4c;myzy}%S!8e$y@1FVNFe;>zyy>G@AGpU*NB~QGsU};b@s0@eTwP;J%$Icg!lwVm zyMi_tkGF5qZ9VLEBVw04>{gl|Ok;8<9`UheAd9EZtvH6wD4iiJQ?PG-7o&62(G%OP zm{DiKv1VHTV^qrgW*gG==Io&K7gVAn@^=$atlwu9WAG&De#$zu;&8pfH{5(MgQGp2 zZ^j67bgaU&j+nZBYq;s$g}wKT936%z#2Yx-Ak_Yvpya2#WNn2L(y zYrVM&>5bza(4I6AAE7ogdnRgU z*z{S^IY{RXU9_~Lq~rBEDX5__!4I68jS^%4|IPp*at%3i4dE?1atu0xqcBWybf~i+ zDQ-71?dK1gFVb>|K6Tj^`3Km;5v)_C2Zr0@*g}Y1;WEabu^7S#LeTxms6P9#V#tfV zV3T0P;QJijMCmPsh#A>KjwFRlAz?i5!T6O*L_FE{0P|M@-UuSeQHrFP=UcFof#;NV zF)Xyt^Ya+h5V+pb+_aC~|-A&6>!^lP2SI>_%*j&kVil8 zKoK)Qor;+e;@9)~1Hng(k7ytHsw3w__Q^hBAxKljQh$%#l0bfSBF+0%Wb)l4>?HoA zyvmDRtQ3hE%Eg$6?4_F?16(27fGNoh&Wi&_-u6rL*u z6_u!|mJo8qrb_6(Wy+gTVNoJgU7axJe3r^#11~S8t-z2+oU5DbJRvys;}@&eJehHpOR}` zwx7>0$*b6`c2jifysbfxje~_liqk%*l7O7BNq?mVubxuUP?A%UQevfcRjT*yK$WsM zrRb&@=?C)kr_v17J=2ZkAUVxed?e@b{hfDfe6US+Ja_UFD@xtK3BnKBr*`TFOlyPfJbGP8Lly<(-Jf zRXz*F)f*DA3YzgOCw`AtrzdYL@03BCtjT#Y&=b5+)DZZo>5}4Ld11B= z!P&_fG(zUOXJH% zd>>9Wu242DlS%VQvl7EWjuDP>HZSWw)-%>2_EOFk8wJjt)WS47o(@C5_pdd-zE{)K zX%u53W)e?aWa6X7pypM?R4hq%8Vj@CHc7BjH~L{R((t@`w9&TCTVUSY;eHD(Zjsuj z@nhBZs(GWcfp;U^1Kg_(^OxlrffmLQMsDQo)&t;088&&=>7lZcWinSH*!Z!Ty# zbU8?~e8Xk&JF|G%7T(sgQP(!(kmw5TDr=cf;9TIEB!Og%A59=I(?h_cZMbc+^|@=? zWy49L`>co0`NtEjjj4_BGqhu-Q=2Q*v+z@fE05oHYgrrmn?mSo=vB1YucFak5;c(@ zlPCt$tf`;D`=|RS!ezWDK#+t>LJ&f!MvMvKR1_@)%P+n@IC0wC-Prw-gFc9|{PLD+ zn3IuVzXx?Aq6!~}BQtn0h~>FmxPip*mpRfC@<~1i-K{*@T3RhNt2&4CvR}V{MX|NB zy<*F#jx$Q?(fcwLhSkf}8@G|SdDZt~^gONE&2)@0SSzshaFjcv&ERsI_q`+g&-Bz% z{v@F*a}YT>wV_X!hBs|TtWffR>{8-uvSSQuj8<%SWFDluxTRQ`SogT)SlQS+l_eGK zg7ZT6uZacim^UGV;&zjusZRx7axdgplbFXsz84KqnNb^);FX3>_3rHKj%u@3ihiJ@ zi+3_@Hntfhh=!NFGNqiatxa>#OrCG1=7@VB+Z%KqD<62^6mox@0GFWjpj0r`X*RhV z+GYBzu!J9n!-c=c+hftIZn6rUjn$Qn$q$v)PQqtq(448|H*oub{5$L%r{m}1R?=37 zK7MUYvv8q-+uN%butL+=^f)poN97K>^=FdzQMsgx8jF=H-)=o(s(#5-%YU>undS25 z((tZ%DBh>JS=g#S*dv-f&{Wi6(erT$=npEv{1Mw4d-!FFo7;NBnttKj`)p|>NwzTl za7kpDVsTOnS7oN|hxy)Zx&Dyq(2<&}#C666x#H>3tU4i(9?q@U zFYmffq3sZk5UcVjte!WyT_zrthc~)4R<5LY7jD{4FW+66F7LInw60!}ZFZ&wH@^rb zVe*l4)j1d2mflr8C`{-`d3rayJLgt1)VQ}$8)6pq4ZO_aGxqpkrP%4%cQ&clw(?uB zr9JN9@qtHg*mL4c<<#Mq`Z{TekF|%*4$h|k;=YAJp4aaEc}?h~D2gP8QkaPwRhu#k*H=_CnGr$h67FS#R7|Gt z!Xm>4>7YVGBVoCCe>(MgoX(>Hg6}lyu{?d!3!9{&j)TrrFPQOIuHPfU`QctP!)oBq zn=W@!c=x4`pbyB%$<%$YZUBfyvX|0wgn_|*1^tDUR-rlu@NK4rnx>PcyxbcjI~!(0 zW4jL~%&s=}z-ky6e%Cj^OB)j>Lo!zzYg@-Rt^yQ)R=fdTL*HhhAp5h#$x47iQ(lQo z)Xu?#jEk9*nUz8inT(8#-@(}Qjf$AWUzY>_2~e0jIoZEqVR3PBVRqqQwsSCJVdLfH zWnpD!VP|IoRxmlb**Y1zGTAy({^KV9ypNcPqmhG!y_1EVEgAH_h9B&lodhT-pbz?= zpMUJr#MR6L-YQSF*R5!p+AMi{2*I(eZCoIL~*fb1`FpRX=TQyhM{bhuB?A`_VH5XGe zM?D8jDh5n!W#zY2bRbYHSX|g4o|acSo-8&15ivQ$9}`FHEm&DuNlgj8_j48R(eP^9 zy`J)o}B zu0Lc%T3{h8VIaBke-{YFZU5PS7X@9(7sU<4jVzk`-vt5;65+p#LiT?R^1t}a|24?} zvHSni)Bk^LkQ~j*v8q!s$?;*s`$WbjCX}*I85v*piD*F2Kzr!+#T=ZgQAq2!r9y1 zOgsz>T>qL^D0%lI;>s0j81XF5yDr~z!3K?s{4B{D!0jgyf4PhiKOHi*-(N;>b3er2 z-R`LhQF1YLPZOO7ww~TLeT`Cl6|O^&S2uf`gyH%haw+5$aF}j$Sm%K)&mtq?1Fwx=$g)4y9FH5QRVZPq+tCGDE-3y8%U3GVgq^e2S>kBt(xe1?GGzNyua(;Q zboq@je^JaLdJuH+kXxPE=$JeQ3wp`jdRR~P(>*8%{?Vrzx_v%%=gn;8z=U`D&RD10 z+q(?x7aBE>aWG~wAf4h+;ZL|a3u8~-Rz@u&IX1fYPeQj=+ETBrF(#@cdmy;pyNv=B ztot`%kj##v%e;rkt?Ocu4#VfztRj&7FG9S4*iZNQQHmYiDSX_$Rvmzu0h)XisuUF5 z*5MYk=zI_moCh&~QEjLSdHN!o)K{!Iw~74lG^d6QyYMl>8l$yhff7AW6l;* zJwKR(37Th(OX37sgtN4)rKKZ+numG%AC-dDf z($>8LCa;iB)*av7PFyqAx`e3}%5!nyiAhA-cet>z zy&?v0Hp|RR@SW6!J28{`XuPcH3m?qGJ*p61s_~Wn9i93n1JJESrp4IQ#3&9p_)mKX zukW|Nm!=!GH6r(j=Zr2l=GcxbD(zIrgoGkDl7Ki0d|^P}i+5ce?-L{m(6hNXhrE`xh>g@$R)UR)=PemQczx zpZPWZnD#y#Y2)ZSN~i$=Vz7XCJF;%fJa2zSH|W~M>i=ddQF<6{gJ>TCZ%*Rg{x6S& zSVFg6)_Z|;s!n)ixVzWFIC5cU4J|?X$Z! zI%qN>rakQ;Mwl0d6V@`QE*}h2g1b}7#Qw~ zo&99A^4leQm1(%!1NNag>-PHGyp3RqWX?k&Q5e<)d~U8n39sKo?dXT&u(^WJFAbF0%~-dD9BFPnV0-+i7!?I;IGsJA9?;*xYk1L*GZRh@t=5izUkKf zOTrN4$LF7?eTX`_2tcr~Jg?J8ReeDuC{5`D_xE03kF=8RAR%9`QoQ@(KW`uOv0BRl&C#0b!44VXCROKQqnM;fmj zVJ$eq<}G=1+;)z9W{`kBhR`&Gg3bU9)@%d(wWAiad(;N2;U2~nd+C_iP$l(w+9qIJ zBEv8|;=<8}G4p@fCtg*LNT&DB@;;6>+ue4#{TnyD%%hf=yH}51Nh?z}zeRm2NM0ao zs*^D=5E}h#4+27oR)TO-YaV6S_%~&FCkx#dJa{a$I-aN@AnaouNJ8por-Q?mx{T7C zj-TrYelQ3kmW{&ZQ?^w*I~k+i=%z7Y$giKa6aok`*_#{GSh-WW>WJhVi9`e*x66qDwMhXN(E0Mp|+xXsB3n7eQqgrHdv86s?bfwyeHrmNveTwarVK1{GX0tZh0zoB zSKDL+GyW%!@rZ_j3|61793q4+f!;i*fQzj?fop1`8$6gM8yJ5QcTqbdYrArGrCv?r zUXUQvv5+4Bot^M40u3^6d0iaf^9%k0r*{lUJ1Fz5)^AK^Wz1@6kM9-Ow4|_5pS~|V z@j{kjIDX98{Kfu(TxIII+-jbn+MwP1;Q0ppuX9ae?{(pZKBLw|tbsS5(CI1#qta_v zwU}Knup0WYg*At~5oa2fKaVhf2#E*<1tA%G*E}n>8-6>QB}Dj+hT&U-uy9Tq`phR* z|HCq$gScy?fB7{GSPomsf&8eOl4jiP+LNmAbRI-~$y8Z7la)97wuP^9cJoD!Ik{jr z#7YU5>@(!2K=8QFhU~!LCBx)ryJn@5ZW~}zw}ifk-@>xfj5saSymrTG7;h4HIrgw^g;Bpn`n~G z2>xEHUCRqM2ev9kdPjYFwnHXBIb~*B-#3@YVi+Il1zFaAyr^W$rc^d!dy_gis*grV*dawGK7v zU(N!d2gxvP?Jin0?DJts(^A9e<2cOOrh8<;Y0Oc3t;O9L)GW$WnC_xA}|H znuNW!W?L}`UxyqWljw6{-2wNV;6#WvNI$2Whjz`!FJ0e9?6gLCK+tXaDWMuZ3u*2Q zZ|ZQAFbcAGc%hZT7y`;g0&p0+i{~KkRe*U8Tjm~6MxaSLc|4C)NvLqd75Y21;>!ja zWZZhNzjWa|V8g-Ni2D*_#8iL;2Yz$h)Uf^2w~!#u{_*A1hCwg(Kq;iKUm-{SNlQw0 zC!@=aEAFBds2ljp1y)s8F<-x7mz&{y$phZPuy2lT-% z)e}L0m9aB+w4}jNIPo^pyG|)yLAtpjx$kD6ng~T1#;x74E~{y0c7@EFA7*F+{ylhr zyNb~qJWuvpNKC7q@RUfnW-=W?^tn5uAo+bt^mxtCnF@nD@&>3s1s1r(#xBt?3YZR0 z0JrpHS!=yUh)s%;n9)G+)PF6@sY8}+ZXtG_Ua|b(P`d71D}sFeo@}RaN=lEPXkLiu z%z1)w3=URu-g4oZ6AczDo2xBZw1i;u-5sel&z-Hr}*|ZEQ70y7tzu16-oU7>e|;xRGhO{p20qL1)ke zsdy2UJ5}WtG&_Xci`&%UE*?QTj+-#(U`1W69X1h4yC(O|cPmgd#`)d|quWSM5EqeF z2RY_JUs(9WL6AA6R~d2z8Wr-9m~1tR_PVuN#T&&8#IlcXM~ivT^9Az!`UfN7pUC#P zOFBL?%niPbSlb^M+)0T0X~qs*^KYRhI?24Xt3AWzX-U!kaFTx87K-mqHJ;n4`?7Qy zR<%jEOmg;&0@;-wJuBHC;Z4N0;|V`l?f#8d_y~>mbm#Hm2nI2Jfop?Y6sR3zK+uaw zn&mv;tm~+qE;?U|+yDi6;&3=EBD;k&kwOpNLf)^;2qC7x$N18fu*Ci{YDbB}_xDp5$ah*Hi~0D_enttUe>3^y_ej zlr+g0IbS}@vh>e9ARvn#8&hq{@0BUPYRaFhyNlL*%SNycxAB3Z?!cF{dsE)3({|?* zq>Bsv8Kwl;S#6t?9w&B11gGJL4`lE_OwA9jC-5CFSVk?2S63#p1MPaOPB%=Y!`TNX zxB3--O1#t=3t`4W*C}nbOkpE^0e-D~NKV93`x$zoCEW1+H0s)kZZ{%O(;N(JRTX*& zdPCgzQC436TrE6a`I{KI9Kqak=NnJ@@VX{93zGYN?YjNDRyWux(_Ogp1K&c_L*L&C zGO~Dl&kB4yz>_W&5`%ASF-_ub=H%LitT*#<^DJt04hK+ANwWeW%PV<#rylY z(o5ePP)^h(G>MLrOxt?1MnKeH72?wd#|FW8(I&Y%_7YLuo7Y=x#4L5bzKol3@hEPJ z&ohFT5hCAb6S9fxqi~jc&`=9Zkp!NK-+ePrl!OHcDE?_TibyY?L;`GkSxABWJ^#c# zY{c&$m;eT~)g6kX%l_)yV4rk4i2e1*P5t|Ma=aPHuW_YZfX9y6+At;)2U1GA_6%0z z9ega_3(24B%Yw9D#EU4rL;MW1V{I%3vEJewYNM*3{U*r>y0Wa6|RGwqz;?d0Fr5njJK&}j4ud8_*F zHs2c85)Jl27JdxPok*Uf%=XLYlvm&i>TDb3q}DAhW6%51AA;K+xOn#Q-*A*gle%!f zn~`a^G9<^gGYH^FJKmD*5!&Y1$~;5VvoP4cL+4!+|21~CWhf6ErFQLdit`FI42@}X zpJw!)NVjtv_4EN=FDTs@eWjEneSO-4zxh6vi;e+G#0(QXQVp)^sLd0lB!2{HY0Wp4 z{;0h^7<1jEH<2Gi=4`5@FL=J|`yp^6ek3MTdCd}3u`fq8Mhf@P?8;AT&*bO(!33S_ z{?(ijvd$hhP)25!A*nG>O$mJUnV(gIG>tzaV* z7aOYa9|h~zUoD5&S@eN}6_l=zzUnMo%?8q9yE$p%TR$Br{D60|jCfp#dQ3$C^W&^7 zyJilae6vFO91L$1|LN*jHR{Lua`&vr#+?jO6Z>TJxrmT4@sm_>YjT7S90)m4c)FIs zU~G)V_(n20Uym6Yq`u-WM|=J3*|t050$0$lH8o@A1~CONrAIk7(MNaLIppfB_nBDg zp#Z8rbk6n1g`7x_2N`OL{I$hGyW(QIasa%v<2IGw%O~@FL$PJjUYZrJZLNE|h{FG3aYcJ2&5L7YjL?y6FthGwQR;KWds>2sx=yD}1 zXK>HLzGtCwpuZE=?~zSkLgL-=M(37*r_O{zx)74)*l@QEt}ptO8)D7+Zdy%Spc=V& z*-_7tV_resxFAv`7-)!O+#k>J z({HqUr!?s$2d0E@2cMnw>WpiRI59NBU>_&=wr7}5S|6mS_CLQZkxDco_gmsSlduSP zs|GUxXp>8Z%!O~HSn))GKM#UaN^>>Je=urjTU)Hztw%$w5U-CAOsb{#Cngex{XHg* zh^!p%%oH=(ycMYtQw~g0q}$ z!l|uqIlJEH7lN(T+ypx~M046@U&^&!MO3D7(Nks%aj>{=(F06f!uQb7WR^7G;7CzX??;97xjvyb448_`A3|F#*vEOo zdlFY42G{yAF#7CV?_yHK@07_*4miIadGKNCZa!hBK_;9p*=O= zF@n<7(X$qF{QI}j+NsG)g-qIQ?IBlPznF^(nSP&SLwvx$T~d8FSF;|D7VpzQL5T5; z?_f~+TtqKq^3jT;5|khspxu^P^6iB0?evk`N^3r#&MO{Ku`9Z#yEGQswF=?iR<+TW zqsT5Sn>W}^Il|_(nTS^NM?pI$aBhHdne7(Y9G?ETt`Cx<<=VAzB%f7+C^Xq!2ew8G zGpc!-!%T9pZ=1VG6qot7kNli#FUP?CrNk{s{WBBNyvma89&UK3+fz%oW&|rK3@55$=G>n}0%OEc`I1kv?7Z?mJ=-aa|h z>pP>1HuDRyjl<9>b;}&v33WiS&(*28{7~123>=$nS=h(=kDlzKsa)S_pm^b4WCbYx z1D^-%49-4#Sse>fr%&vu*>PVl$h}Dzj*@PssWKh%M$KqnXbX4c2H#U1Lv$N|u4E*! z1Zl(Gg(mmj3X|OxVo7-(A~j9kEE$p7*MtDIiuN3M*4i_J1K^h)nWCNd$uP|&$Jnpm^D&0vHluphDJBX-1@y+buN{_be+~`x zQ(Ciw)sO*g^w;^H-8XlvZrCAT%R_KqMihPATWc!LUHxs)`5W=|(>?x)k#=`NY>|S> zu$U&SwixOM>hfR!8X6iI8)O0pxWvJuXM*+**G2og>mH-HT{RMZ0l>`9aBf37i&!0|^U!Sg+ENb%E48#o%DW+K;93L!7O@ zLcgUiWH7fyUu*~GN)BFQV44&5(D{zFe_;UL}$+g zE4e!uLh^vThS)ko=cz9Oq+!`PbJTfO=r+-wu{BXdD#8cUzA1)ch0l5evT_O|UV@oJ z+p#4l_+`#;hppwPiSoPP2j-U~Gnm&ur^kx!dR(nujacVnn%nEE50hST=2D2{hoyCv z&_y>`kgM=8DRJ){i;n-{DDXQz|tStmtD=wXuk4~W@R{}s{S^5gzw#>&p>Q@A6lTRRyCYZW@d>?^CjTSD>ItC&zD2Gk$6@Bn|fw;4da z|Kkr&aKlU~n{myLWxCdX?d@BnPKe8;7xgDujyK3U)Y&Xs@)Nk{p!z@o5=#{RrEOj! zfsgMGs8TZ6LASN1nY&AHqTd7ucYe}Ml{iqvo{Q)kdpBrA8sO(of?xw}2QwHO!HzA4 z`GjvOzi)4m8Gi6Eesbln)HvpcI*W1zL9S#ud>sq%akMr7p&S}X^R6nPkp5AfwbCh4 z@#lQ4=?VBBCRTd29w5qKdKBoc>p{$gw1V~phjl2>P}VplE&g3Ra@LI8bO<~^~@DV)kwh>8Q|)I>zKlq zTf+)fAwD3(@>|OD36jTB-&cebKn*OKtQGq84=5S*=&s@wT#gQ*(bR+nk6(eU>=(CW zK(n=LIiz@qUMW|~yDVj%blZ-Z-U>P5?hObJEB2#Rp5H8Q0I9ODj^E;70_6A3U7l2@ z=EneI<4KgQ`;WLtd87dh8Tzi#?QA%*WB{gwQ}+Q>&-MUW|4eJCpyU2 z8e|Z^UqEo7RsY^`QRUPDEh{Tq8Fd_>hW5&=-SnC$t3tZ_1uYC1mpUxd?R)Pm91Yeq@UHLoNKXc);wm4>SMz=0K!Iod=HYG+O3fnDNVz5< z?8sCK2N7j<&%eKXWAsi3LZS-8Hr3dPOOWGQAa_&(n%4@MZ8wf~JPa=N9_FDlDENct zQkqB**KpHJ8v=No9dnlm1@ub4{VCBJp8dTqSu-%KkvD?ejRuy+jkcvu)x#P)Lw2_J1-X%2VYQNM%JlT zNi?(-__k{R*WTo00!@S)Jzx;03LHpgx}aUP!y+x#=8o9YY3+Ors}r?@C+hQi#4}rHJU+ z)rFEAlqDaCv5#fQs$`hN(6si|@muGJpyll5==*JSXmRsg{wiY#ZT49ri954!|Rb3 zKg|PZ&5-a=VQVM2F3|vtuT1U;=8IFNW@eNe3<$-$_#ujNoxPup?<)i(FOpW7*OFGB zx^BXygZj!WO`^D330Wd$+O4zekzOFXQba#I;7KJPTf7Trl8`T*kOs%h2cP_w0H;cfOVu__2yPo)b++quTaI2RKR> zi7K@EB~pj}89WIz)P?IUQ#;4YQ9JhgEm`*GJ@`aVkBsc>wXJkBv9p5qNslH0g3Tz! ziEsv!dPr1qo<|6;{mTJwxT`e<# z8JSiGlTCGX@{r)CEi-#PAI+eNB~zbe9n`NST3)J?_S58j)ryMv>lEa*>R`|G=9!Ha zIZTn4l(N;TL9Q2pZ^wPkq7}2pu1)5e_ONb1JailW3mdwn!B)8+Gx+T9%6SNDyo^lW zaYuSP5|DbG5I^_)1S0i$%=o0C-yjWgP}k1S1w{87ARP*|X?{PtWbW1&4kWH+N563i zMe?!-r7@JXgf5h?ZMC4{$`MRKg<4RZWH0$~yKD`}OZz?GNC2ob$M;Qy1jw0oT9XC?3ecD*7E51eo1tV$#?0K*}_$rauTBFl9GmO^w8Q#t3&qgTnRr526Vg3jMSY*u``wh z4hbSxg2=<#SV5947SWZen$qx2o$%pw7G@&xg1P+}5*F?CSM+X)KfQGkidef#rKl2d z7H+{Bi|av**p69TDd^LtdW{eMjEnfK1%u*l82gO`{7#DSU}~|{YJFdAI8Lj~Dm{l z73wN*=*6(Dh6DlobK4Oc7B4;pk|a10(p~?mPXXV_LjnE0vY-vJ2!racVC|(6T|dpU zXsy7ye_H4JDiV}g2Xe&)A6QiWG0+5tp(dXhS&0UH#*US1ZZk-rpScrg?~b5}Ko&$r z{aI@?WSuOu@IJN3Ll7%b>G)_6?gV)wXB&tC>bOJr-KWt68m;RL3b0XqZ)@lH(fHS6an@j}DG1!Xudh8-!jaQbqUWS>@sNZ38EGP5`bn^fGN6cO5g z^uxjPl#h)e`VSxhvZ{qFx8Y$)KA&Hfmqp*LQ3TC>Si{^mrRSSR>dcu)f ztVS;0GrhmEd6b9Uh4VjhgTrWZGb?sNbII^koly|I9o49C47w%dm3b$ox!8%=dZ0%a zH*$OW>}#taA+K3S9J{wZn0TS7bS5&uQ=Mq&BoENlLW{vVpu{g)LLYw!3p=<4g2)mgtr&tD|N1QSS| zZ~fYt7vfLPTh3SJukaJw+7YrWwe3$~KIC@QoBc$43k-1#K>Gz-D_;ti4>sgpKXqJP z`-3Kt&BqYKQ!a5qVM~lQ$f)+b_|ID!&jYj8-j57>JZ4m`)$jbu zKU)1f&GQq;_(!EG`Skrj>ts%=#UOmEHeB59wL*OOib%wm92Av}aGshi6J^~O57n3e zVnWlnUHvD?p5vVx{%7+~H=JsCrFwFG9e(7n=zi4SuO>HEp8g?^(gE+xmT8FZY{_5sdh1-blFVd;GmyjB z@*Ni9OeP`jgyXj#lkeE|9~2B9A`KZxI_tH`S|60=y7CfaRuh~s=-#)O61mD2x1OfT z0UdutD*@KZ!ZZFjXr{yDFHHdW^iT&nRslpH(;=9*_OSouMK%jwh;98$^>V?6x!Bj* z!ak#O-{zlvRIESs>F8bcM;g|zC62?hMQ5bn>~{Mm?*)+Wqt{{bMJGTBo>_oj#>EYt zYxXcid47S%G<4wy*qfK{07X2|9s)>OcqK#652TotE+ylYPU>2>i2#y0@;b+PgPO^p z_rgRBn^joY$8@FCG4t&gdK0vwFX~MU$Adqw*sO!AD41 zz_{Q$*711qzL55*{c*Sz?JPJ~rW4nvq^h`!K6% z%B5`PaH(b{BjJ|gn8IVEYIr96cBc4yco4$g^wM9=`a8Kv#i88QAZ~ziJURZit zq*q)AH}RgKN0*dDywLEhEaBW_{7IMq#(Q+$@=Ng13<5|NolVz9e6Q>zG65$-p^z1TCNs} z{9&&qv>(mZn-@PUWULnx{v@~AV|S^Hzw~^-&t#iwu(T=vnRq$LJ#6Av%AVt;OZ51VXjE`VXHYTdBpLLIakx7&l%<+To)i_3UO9(VgP0NmsL^z}?lr1#9m`bu0`QVV6rZME> zJ|XgP?%TE2uMRg&S+cwHS*e$c3-6^lv}KV#yO0D&BP?pto2kc+@Sifr{(QgrB)R|! zz?kqt_CNI2O!#?58x;|-7+vFusxzNLqRX-`x6Zi^{@{Xp^LY*C6TF`UV|})wH7DQM>Y&owXrwy zm{NLz1o(G6DZnsfzV8zl6qY1996$0{8Nzj~L-x^K!twp%cve)=Erwb>Hw6NVih#VunW~A1>H@sMnh3N?V&=@n>nvs!tbXA(0ekOM zjQk)|7z2pV7Ohd(t?7BU$t&c{g)0l|1}B5d;LEg-!4g;}E~U zwxN>00{e9~(7~Qhv0`joHU;K~tGpl4FT)vHKF1z4O{#vZetHHV#BAR|(8Q*GaNfBM zePNM~?daK#{YQII$hHb|;}BMjT2xcM#$2OK)LY!NcW?SGZNBJ52^r_dFOPC@Fc6~59zuZO ztC)n0&C81#J$wC&Fijp0GJg5|00SEe6dwmw2U&gw*1N$217hu%zDLQ-*?_n2MxzQ# zLla3t&upy9|0`gPqqwQ!+CW-08p3B~@iD-1m_^FXs4M9_a6l%c7Kd0q;r`@r3FCcYkfA0`pk;ZdHN9eM@E~)a9Aw z!)R+yaYn+5sjB|Kmz4Gcc+C4bR9JOt&XFpkjNkt#x@C)?#uPfsr~XUhTM46=EI^~k zx4g&!aP!863gFvVV1n7=NRv2^F4Ilp=Fkz}8>s#8LUASmFTd9&_AvE0GDL%fFBio6}st-L4PFQVFs5UivqlYMV znG%}aLg=OpVgVXHdp+w*eAN*pW53+3{W?PkPHKP_Oo@aBlZn?Y$JzPLD+)f@!Rs?j zC!mV=l=z^RjVdUafFvI;?YIg` z<+apb3E$^HdF%4g!_kG{_2wEk6${DX=qr^B!pSPiF22ZUdQ%3tQox{a7Q%Q5SfFlk z!rnJs1a%8zIsK1?xPV(^R{l{3|HF-l$?waCX~&?YoBl4xawb+2dju!jo9ucFx#7q3 z23pm&UREqh){5VAzYvK$Jf1hoo%{2X^Wr?vw13tmEqPkf@Uj|7D!ALqrSHyb4`tVTDcuQM#|>hTFMKYu|9%2{B_SF;q6-Pz#x zd5(v`RaFNY*H|EI+`)@jGBKX|T{hPFkUK2y$I>HT6@YFDTBWqhs%0k$`~!D~-)=eJjW<*)34m6BHqV5k zUazd3x2&D8W(UPstn>t~-G-^`tK;yqRlY!OhSnuj0S@cJE< zn$EfWYP!6iHEBxDzNE1CYoguPVS^V(?z;_(+)k#AZa3mmnUUvV1HF@nb^Hs~XYmv| z%%sxz#n(^8v^zRyjwb9e9mvUUyfCe~eav)3=n`qJ^r^HyUZo3m`ST^t&ozpUM|^a?xz zzP*qt97=Ihy_r7&FhbeEouKfUgPxunI|7~3lOHxSnNco0Sy4Nxt`BZ*x{{1N6Ig{_ z?<>p$)qzIq?HP`!e}EXumy;&C=;I}}Y)mE@os?AOehJ_WgI-R6p;= zD(~4UOUdczUc@pL+4?%U?=)w%SJ2+N>hDXleAqx)3e!Hm$1;DD7CLC`O|v+2Lf(ST z!JP8HqFk5$lgKxm`2n3&&;tp?oBXIzh-Kr_RJ3=mI3fi23dWL$B)C=3g_Q_)=~MrJ z>CO2Z9p{1Ij{$o5;=NxpTC~Y53-RC7D>x}RP`vtB`WNpai_)n{tXz3;BaRB3B1>p? zWMumSSFS+3yd{N$_q6uvenl}17@F06>Oz}oKascYZz(r|dDwUmg{Db#3UPqJ$Yas( zGfeDSi@j7psXqwYGrEA}0ed(~A=6#Lo~&Huzq&<jf8g;F=( zt~)1)ZdLX1)uk`%=&@4i8_*ot|Bw`C`S)0Ue=N%9Hi#!5fbQq_06tV<+(JvO!K}2s z8&C7@wW2dX@=F=}BoZvy433KVviu2Ry)SfMAEQq2USl1^cI`PNC%RYn!DOi3tTQt0 zj^I2{0Ov)^a*#&mQ54X1E=|rn<;^eie!lCWr!J^SE#+s0^jbzk;{1BH;(~3JN14n* zfz;T$eXIx3Q{Nfr;v%+pq#U7`hDxnAlF%{q(Y*q`NkL9=H=;~WlM<&4z^R|y2KyV( z7AbQ@woA-E;M6R_{l|Jh5iS6n0{Y91Hb)pvguQ$s_a`qRO-f!tXB?oA5P(Y4GIDFy ziLKvIHh0>{?LV&szN>a(+t=+b&uCR0Ui`7R`h?on&m!s`l!SKuDOkuwmS zaF-Cx%7Mr%H1qo$&+qjAboCH_aPJ0-qA3gNoI7i%^OG)av^9tg>qjP+8X*=Mg{C>p zEBn(CtD-5-x{%_{EyZF@_&;j-(`<<%mSsX>9b$o9cF}1m47N#x{RLiwI;EEe-}n!t zV7WI?t+no7;=poKiB*NgfIKmS2-r~soF*^ykm=^X`22|#i)Q*$|dw7q;EUB_M7C6`G(v#^8YfNqY+f~CMB0Lc*yW29nvNRA9dRs2?d z2efpauK+Cs3>2JzwKt<*EYqvB1DGiwEb>F@ety(U@m+H@~;#_;1EK1FX;QWu@$;gc=?B5h(NR6VyRw z^2>(^d#uPl47V(`2RxdB0!f{o&QH8gc4!<81^o45>*+HNxn=ZM@B6;OYMZd_W2ciBk%d8ZXWJXjRf5yfi*zL%Q zNXMa8GPwudbMv}dCAPaFe*QY!!+Keam}jUO#5Sj~1Gkq!vjQJD5#A8%0GoA-F|se* zyrJr=L%?vF1d}fS;&EzQa=guMS>nC3&5A%&C<*nQZvhiQ#7y7^cW(0S_k*F(hhpoX6xZy)k3p$DQ@`rW-znyl2<{D z)0PsjhH1r!AIuL++4o~d!=*&QjrGKz(5WlFc#9eTOusXh$~u`ptl<}W7Jbt(lXhpa zH6hL&kB%oXZwT@4-ECCOR{uyzz4tC?;$BA%d1Yxyi3<7bl$K)xfGm_Bq_jz_l>AGv zIXo%V1rkfClp8B4N)VWPzaYj<{l(uif16%vCB~HGsBh{>XHVBtVJy~575u#gM31-V z?~l=9KK|vMrbz4e=>I@(9V$O9TuU)P312yUm`3CUs^maNQjm=Le!3?>*?>FKLVDU?qH+(Se6sW5C*f{}omufC_19njH*g84= zWj?u!2N`bFu2mAB&ycI(;1zwuGWn*P!@;9|X9&L=ZzyaHfvRA7?We0e_gz8s`;QV<5R8Bxeaw2I@Z)j)Ps#(EX+_t zkqvx{vd1tDz@|bm)TeyT&H4N$qYEqADiHk^AE>f%!0w*MBR%1M?)RP_4;LuGoARsb zpsVH2Q2X1+V~)pb*t|0zPPZN|xOfFSf*D;D{Gw*sKh*N=t!Ri0%$B{GWbDx9@A-*g zL&B?8e6GQxiE;MbQHh39E%viXj^;G0&Y~`Y@N4y;EK`&q+%%D#p~n+;^zZf2d<)($ zdg;00G>rZ3o5j<7XgLURZjBLp7z%sUG!fFZS(rNDfs)Fi9fMXx1!3MHm##8@1L;EM zKY;{Hq+r`SmZ;Khcby*I0a*-8{T^J5ZtX^$ueE^1C5)+**8+ky_X|8JEIxnnF4(P^}TN7cP<i2kW>Jznx&Kfs6q3lm19kA7C*|A``L0SjQ6sU=Swlw-^>PHkyM zv%T}7f!#R~)dhisA+EWTC2v_E95NWD7F9XZFk_jbvd{&u93q%ULm+J4_P z=^TkNXXH=IKk&vZq_c#HDo@0qEf4oqZ*MLCM#@W{ z`gv?ehi3nDSX;LCy!`=xXQLG(Hr~%U%&rYqK{gMV&J^LoTQ2Yye=4^>9}U8bOG*q6 z@5-lP=rrnW54|r^Z1BT8DS4y6cZ(B!IgJkSMLl~!PCbI~S|&?BI9EG?2Bk!a0gAbd zMcO$9F%JT40VLdw3RvQO{DM!dNQmSMvASD7>3mC}f)vhRzcTf&e`Oiae}g$dmXXz( zHs?X-khvlHM8C`@BB~GLBj1Ox`rr@?@SZBcmVhni>rw^hx_(8W{c|S+f5hI^32Ejm z#+qYPsns6`uw!212-Uu)A5!p?hIEH>)~e>rwGw)VdC_y8#4Csdw_!Jst%OUVqhk}bC`>xx9^s@ z$LIoV2@=8~d9>iZaxmHJiU)6qg2G#+yNvq1n#N4zJ=#r0`B=amC5rDg-px`5fP;8o zI}~Y_a6}zM7rtLn$H&qQlGDHO1#BzB=nU(5fVCb9CWOpzTBc`DPecV_$4O?_Np8@6 zRu7TX@lY#SL5|@PvF+#^Er-?~7(_eGKRFODzfJ6R{Qp(_aZb-AY&~ky0P4S*%x^76 zd}-#ixunR_VFYtrV)ncdxg<*8vbFueGmG)}Zm0`)FUIvnWT!T?$tp#mAV3#-%Y zdH+=ig4Nidd;BDyv-i^uJ6n-F05}V&q*mm0No4zsJ}<4V!1EP&kHR!Cy!GL-qFq0; zE!SFwtj7Ybiu^Y%btR#L-w0J`D3xNyGI8(Zwk6mfAJmlOhdctdVJmt2TLX|S$Na`3 zN$v`*kSpS+KJ&l!ac3lg0bRR)tySL_tW2@aCr?polifouo6ABTWYRTfO7%d%+zx#M-09DLLdHn7{F#4UO8j4!f;>9x_7|5H z5a7CEkju5j1=s8QZNnT7l>%|OIkn4)Hq~xTL9%USkd>Ru9mlHvUnV^%dHb5&XTwXy1wfY=#9hzYt8d z7g0?Wkl)TyKc+S-B=gudEuN_H7b646O0iuOkMsWz`TBt9LL;MjdrO2foZA@={SRlY z?67zq7I=p^ZC<$Lpow0Z?iN7HvPov$?exvwE*FX_ZC|;X&nr7ZnAU<-+uV^uwXYQD zM2JxIC(+*G+@NE=HzUk%Y!7G+=QzeteKkwC`>AG{sMv$=SijTp=5yG()z8rwnGnU| zZod=cS!giSd()F6*+`*CrGNX=1q?l@-ME-M-$QCVj>70&(Q}-Dk6_yw1b{e69T;zn zQyZ8Vb5Y5t===uCK#XJf40kszBb-Inc_!$*i3Orrj>oB3igyZf+Way$oddSHP2~IB z`e3YYupNXRGh?%;P|nRtPq5{{K?khU;lNG7%-8p|OYB zvdZX`Z&1kK%|5=!js%vEOtrR9d?v%sq8nVI`9fGLK8X>ZmS1-_-KtJtK_I18^qZ9| zZ`B2XfToehzz$+c$Wft)M>%3(hCX+#Elxab$ko=w}}nOa~nMUIxlTSt~*{#h)> zcoW{P|8aEeW)9{QIxSGV6C#M)h$c0hP5@0d2^UK(yxDrKtIL~ZW@1LbR!imi7u7{=3PoUh+L zJh~Q&UnDP!@j5%8OJdU#D=jL+UoIZuSj(nH4-Cb)-Bk}SNUf-!rElgkc!gFzkaw1U ze1IO-jYnObPT685Yn*18(Vm)bt}iH|H#64)jtsB#`+ONGq=l3m{5!5A#ec`OiGBL> z(UtJZY$4ddu!Px$dYMl_G_R#I?5pf*@D@r@M#0*RhvPjg%OPvuvFn4{*NGM<-CaAL z2Q`@COF^9`N0nvMzQZAn(xrHnM5j%cG1^l{llvgC=(FKsaQa9smCk#gCWFjwYy#8j zvOb9`!7DuL{K627DA;H}*4>aQ)^FTRF^Shw!1j!N-jU`(foJhf=$CR0 zQ}b3!#rc&3;N&LS!;}}q;df5O}B1XF2L*Gt4k|+OX>;{eN z$RpO&sJ_J;08!@B2-m=lOt7uj27_So^AnSv=Xv#K0q~DGCf|y>IVv$JEu>W=oZnq{ zsKX8C3D|o-rR!`xJ^i#f9c=XTA$b0vpI9OL;Pr38x!!hluonAZ{`A+NpS~B84rS)@}V4<<(``z8~&2_1b{O*V}T*ipg(sT zl-6<;4w+H;)PG`Fi=Bwq)sBV$%JFXuJ6o=~(y7+212KFx_p31Vt6XBic%Oy8HwpoV zkMe^dP5fSZV|>?iK3Uw5M4djmk-gCLd&|*x>a9$KXE4C>XpqKz^^K1zyYKaybkx~? zq&y<~kg~}6tiy8|hS$>|0~`~;N07_m)}^qD;e?(02B0K`OGll1-DiorAmi%uUx|AK z#r=yfAaSz+5;rwqdaxq9w4T{Mo@z95UlLOzgrOs=a@Iv*Db?=zaDx|bY^pALSj+A8 zu%7c->9#F{|G*Obg)vf~Z@lkb3y0QPwsX^Vt=%#AWPNUBsa5W{_QIdrLm)EH3mTw{ z0$Y>|4+`v3`QE*M@e+39F28AR4i1pB=ehr-IrcV&q~XN2p@Lb1f0)*^{S(h{6y>54d^9m5hMn z@nwo*CV~fgECfHb^Ve%ge!Y{*UFwKZS(MXRSSH(HH6249jopQA>(~jUjqrYv+}$8t zcKbH#Ha&~xZTP-o|0#8wqNGHTro6JW;B9LqJQl{cY;QQgoJ~vO+Ko@@r#tsXaTbI& zk6msv%nL01_0}9v$^WjHB-h+=i4-Ss!Ib)yKA%wj&}>jD=FHn74k+40HdSvB?jd8J zQ!kkEUqpX%)#3j(aaR&qtDr6gm9IgoF#a z>xK}u(1zxD--5zTS0aFgkfO{Yc-+S<3x{i+ew%@jA=YfiC&gmM3iEMnFEhohguq9% zOYo#&TfzFTu_GQv{!ZFV%P%Og877Q6`lo?$?RLO8R3?!R&5$382oZY5JS<3y*_A6W6#C}Mh%_E6t^oU_^q9|kPW?j%_?vF#|n`b&25tR8>WZj^{q>I7&GF$zbPn~TRqC;uj zN_v^Bb!43NMRrr8*9F@k;KB@{?(2pOoK9NmtRjAAO5m9ZAP}$bJa)ZHJz&Gh!ywd)n0c@adYNHLr)!)LzPsFzdz{;lUn<#ev1!$P7_X-!DG{Uz zIDf6Ht>{T~zoKHILN5H+Y81VU?2l%u2*2@h#uwm!@f(}akDjG_A;!542Xb|h=1%Uz zQM%OQ)9c7&Fg(D6+_JlX+<|-u4OlH%z*G#34?XK=m7Hb&$`~0%|6~l{H74*kH<8`m z+fJ+h(~f+w6x?#hUOF|oV9Ab_cK&m6yYPGoIXkydQ;|6e1@8QcC)J{b=CHpt1s3bH zn-!fZ-h89MlA=jWoP+K^fGubc&m-Rs`&PrQJ%Dx8|Q4CAPJhlMltykXo%%Vcw)J zpyWK|VIdgUQ^!-L@Ok6*WXLfv)?$QsHBj6Be9!^z9LWR;!uQhiAZdsv$-P&5pMMHcF3INjSd28uwi=+tPa)HIBkir(S zC&Y95nRR-8y$EA{Q>1bc>G?V zoi^&-XMS72(7;JnwuCxZD@S%;cR(9`yLTo@j&Z7E1~^A60j2s~G}s*T1Me)I#;i&> z+kJ2}oXM}XlFSq7b!bRJ$b@}V`**X7@P8-7mx0!#sIV5?=h?|^ue?zKAJ*mU?!Z|w09T|?9YXvi8JVp0ui*mLzGk@qY+!B)Msyr z`2R?&315JkWb%|yr* zJXkyx_=G!Rg_JtQkFnL}(^q-;q13fwn#;)^r_NrXnYvtl21|5IwsE><2K zTsMZC0NS|^)$A^nsyDGx2na-y4M&>J4jC7;u`N=|nBKFa8Jl_zg0Jka*uh*iFgaVR|9;<2SdD z=sa|FFoQSQJ!I159;!okn-}~g%0mE*l@|83z?z*{$M%eW6|CnS%M3e)N7j2C)u_{E zCj?i`Fb-Xv$@2M=2p0%Gvp!%m;El!{S7IsXDM=_lHX&nZBQE~21%uU33wx6jq-HktZ4bp9w8$0F}ZE!RY(vVljQ2aJQKH8HB#sTcnDwClSCD~v)A(Yb`{VECCSzxp zCX+)A#Xyr%vgme?>Y`I|9JH!bZ?Hi+*x@afBOLiCURaqP*DU-%m$-W%z3=`6GC`eS z<$@6L{y7x6*kyGg4UWSGz1L>BSMiQ4e;kzbgM>UZDf+5GmZtD!4F%dQWzrMwe;Y#re_v6JB*FgBy~II_MkkZ#0}_$Myr1cRT9 zs@=7;D5$GoJRSa&7t(HnV+}0w2 zZB6d*UYoc6G23@@x=p-F+EmhI!GpYZs1gc%U+L)rf1luY5HiGP+7G;HD?@_%{E8($ zTR=I8-xCBzirbD7v(LVf)BVY(v!b3(ivwLU8aJISu$re-arnypLF=$9{AK zMPLLp_-$@sq2)(Ykm?XY#bSVrhyn(OXWDNSrSL)6h=&(0XboYC7WXErefdKQ$M~(- zGOx_#F3CebWDOFkg@{8#UL>^h_A8$7 zP9{|#;dIUIG9({-!5i4p8=HepE+4gKvYD&5gV-0$H=@Oeg1vK~6QtjIv*GjR7D<0l zW`c4t`{8u!w>}c7f>_ae6Wb!IA~F*9NVR#?(~REGli6sB^L-@l6+8@JPYv&iouWY~ zlhZIMevc7VZ4}=iGl-9-6XY#5HGhm>o01(w6?X?6lY2|DztMh|D0m+?${9P#otwk$ z6XvlW-Qrw#8sJc8?3oSkek->72CDg|6p$1@z$rV-XmTMh7+-Em?i&1vfP@sPncBPW zbD@yJ6iRXWc*zx&mXhH{IMr9$@o)hD1s#esDzUR)K$@w*s}pY(2zd;M7zlx}R>b!`dL}SM-Q}jz?SX#Temz!Q8*(k#NUm>4`mM#_7j`UnMPG}@=BGV@*N~Pw_ z-?#bH*1>ykd*H-!^d3rrde>_yPgd9)Zi9@tG9%;*6X^<1i(F=B@DbzMgKQx&Eti;w z5vL2>`PKLFphxF+<&Wv4s0$XEpUq*2I zT+kwXp_nm_uV9MSgjf!eu_;1$8g2mO!{o3AJ%1LrIchYBff~rrtdM$lh#F5_M3RSkw!j9(^|9cGiZ{5Lu$-$7u z?>?YWwcPyK`Q7Wx+t(S0Ge$c{0+KB0MYO}|!%6{OezUD~RAZ0o^QZ>=3Q~p}!|czk zu3qFhIcA4t&`u=_hNLXVgtbUIUIK0I>wVdU|74b!aOW&E>j<7NPG;#>b2!f8C--?t zI;{>B%A@(_hvNfLIK&%NLuTYG8i6lbgEzgYnFwt;XS{`a`;_a|Vp1y|d6}=#2ZBNc z_J*us%YW?)d-H)Tk<~j}k1!uWY;wKDwB6Cw%%bdlGM@QsL=Es1nJ?DrOsEG^oNim5A zq`^Xz^nF9+u%SgFk-^^Y;gO!2=wI1en0z|nENxIN-}#_VaY^`~wJkpB^z`Y+kSgXh z4%TqI{7UjFVLR6yqWUHJK3%qN%Aeseo)fK!S1>#UcYWlDz3GR_plNet=3qzZCx$AXCK^!wqHoUB=p~T}hg< zE>_og!$S&(27Cs;24dttXh6qHu^Z6su4>Pj*1{sEWo-aS0rTQQ8Za}>X}|Gc~OB2#%R z^tzysemR61ku8)_QGT}dLE{ZqRU1j2>AMd$id-9-*&Qe*eW~|5vYq;lVx46|OwBGk zL`WiXOi-yte-cRc{vK;!E0g;@Q&>QQ$gO>&}nI( zn$9oinp%_laE7b8&k`%7v{~WAY)v_PKGWc+hxZftgDrH+_ado{CSN2lXgwf@`?fQA zkMofyygu%_s{R}>R_ZS|T)r>;+1qLf*vHcw)~L0ZgM09feml3vNPBhfm&hydMT#wRbNvQ%n}ZZ1=ggEe zx_5o=WKi&tgi`Qq}`hcAIw_%kA~~KNU1ex2gFzr zKc}W(JbY7#n>1*hnW1J}~It;li`zw`cy{&htf=p@oV9qu<>RL`y$S_6Fc|aiFLNN73H5K zWQU2n!G~(m_wlBBnL<~U_okN&LO`51p5&g3+5ZqI0H;V^b5&XPe@05HQst!h&Q7Z> z85E+0k;dPOjf}=~O4g-x`pznVrYB(R0o!`2t@;@3P9 zl^B%0WdtF&`oYj|ytMqp2HA14uwgU(RNfJvtOYxpe`6e@Ec8m$VH#?IB-VJMJT4c& zjehrZR<*nMoEn^DwntnZY-6-;d$b?Ve&hKZrtd{(F@?yDusv z=TR3?2L9$smYPEc2~fpQpIf^f@g`2$T=?fPGlDAqN_414crdwzMx)kzdjx_8wCJen zx0_O@TC6zHF%}C&*!6E*{28;kpgNR{S2su^WZ1VkKAhzlcAEs~Vjfu$HZ_vaZ-`f? zJXjXB&HDMtF$Hbl^*tJK1}E-tu0AsCr)n;(E^#4k-Tr6_(=N_pjx-nA>t@s$9+V#< zpKx@C9rfR5u3LCf8cq)6d8s=!LpwZ>=ResteRFXjc}~;hap6h|Jo}&L#%RYh@vkLt z8T)Hdg6`3adf#RK9|zTX^OS_QyMwo zGico?ESf_~Z<fn{IV z8ob0AdkK#+cA&ECJ7u*K4J%hCe@q} zxpnnY68Br-&J7GXVd?jT8W=~U20u~tGr1@Nh`E%U9p1uUvHAWflb$*!CI-X>!aoCr z*MAY$yZ?o_G6Xwb#%@`5h9w;!sF5)XdozNbmKZHgNXxx*x(T65rxM&Uwr`tq|G*S} z70!R@j=ETm#o2L&Fk*OOckyR*=Gd}>8t{otM163pAcQ~fOSYJpMf~I!loyYB@-4N? zbvppgRQx5{^*8TYW)P`U7j+c@+FFt~Jw5J8S7lGa*D}7B>GzUdrTxxH``~VyeXO@F zU=P#1oUtm3721{>{*RwBc*jlGtN?`qRXpBV9#ViS4aCt9p9`>0Jl^}~0$d!88!?9o z6kyPE0p>`kyjdpXGvhI=3>tQdWC-#`gS)TdLCC-N2qTV{3db(Rdgy29{l3)S9Hu>( zHJotTLphGn_R~TO%@J6v!l@uN{$;-?Ev}6IHnG9Dud3xoLbVx(eB#9P(4IWSzB;1TF-g(%^qz+f7 z46gHNxFFeol1bRLAcr!Q?A8T-?Tgf_v{hAq+w8kLSBcUyyxcpt%3OC`n}+P$tp%cA zZX24wz=8@k)K02SSGDa`oq+){NgZ`ru+yZSU_!9BGIZWq-#w+3kNYbutVj3(4Mqw- zR-Z*#+-mk4=G`zt=;l;3{_6nB%;9x^DcTgsFH1(H_dRB5cN7>NS-Keu~`lNFznci&L6 z^Jk3Uv)I5-AFvCuKgxFY&B4=amj*IeXo&!G#HLNsMvLe%vTCRgkBraUqt`iMuysiZ6o@-2M|WS zyBi_Ns|fbnoR{Ses8k&xq|dMA$`4kf`Gb+=V|%#}L^>G>VNFJArH7HwAp z@TB{-BN=$14r(ZVYa4$QhhZg(!nmV*&@M52(H+K=ul^(YX+nwOBuP{&4dVvAYHXN; zax^E)V2LI4CyMJpivT-+(-wdJh8A8n)7gadVzA9mx)$ye0`7%#AG3qRrXX%W z$jY-y`IaVS)D)!hgIL00a{06Y|fa0hLimG;1z19$P<5Iy}SFFrlQwzv;+uKWwF zKJM7Y*|^4efEfx?R(Y&#J0DWdhlCV55;EJ-ZfV#LDHcKqLJ_gR-OCiVP#e#O{ufy}^@!Eg?q^U);i|7( zu#f;CLF4l_R5a&jtdej%Wm!cKOe*Td6zkUuYLZXOuC-s=Qj962ciJ9xEfhV^(cX=4 z>aQo7b?A+?R94&W3+U?`a_JAlF=o!OB1b~s;t7;k;#qL0i4#oYB`nN1g!&L*_@HB0 z+rj49tg zL{H%(JsMFCx;W;zCY--?X7b%Bv^po*lej%*kz+`Q#k^k)ovmddl14p;;f*i$j_)m{ z4_<5@m5g`dV=%=jrqwMhvgM)&K^F!#873b>V=P@cP2i(pGPxCkQ5gA||qdG-tienFROi9>Z zFOufxq8!X&V~miM7m(*DK_M8&#>>bV0z#OLQm^JnqS`&)k1LDSV4&(nv){bxu0Pyo zn(@S~kevz~lv-l!n3czf_PEs0m)Kr0$h?20Y*G_H)C2GAfcM(T1m+!r!~i zvZ^r?(^Qki&l}9v*$u1zLl9aMfUHw4e=|N#?r6YGTP|L%Ru3vsalr&Qv@e*akZf}+ z9s2%w!pr8*z&Y6W;^2%|pKJqB`3Nl-U42CTY#dNqG zvCfUnK<|O{x_OZfieJ68U{m(5#O^R;V3`%AK%OMK9T=H67 zDlxiJvV%($6SfaV8=8+$bK{w43Hvxu_!jV5n6lfp@oRi0h^v3(6_s3nq5`38RP>$- za|=R-eiE?zgK{1vNTT#b#X;b6S(fO+ueTmvYc!kzp8~-(z6fZj!JS^LqN`&y_Aj_b zEn1$g=N1|9^9qbk>AnTeE{#+)$GTt*!Q==g^Fv>|NIkiA%?!UhWmcKtd!=8KuQU(r zO^J)k2)XMbST4(S)~_6{o>n$mGm`kj(}Q72WQ(-8m{!(;m6kb-vAlkc7Dll@3Y9NL z6OpbmL8%}7W)ieS`c?9^KV@i;5>+ODxQ99`tWI~7(P=7UaOy3EiJ5xpIgxG6*WsUj z^dI3H;-FR6t$JLqUX?f4y=VovFVl=$4;12g2_aC7G*R|8VsRkK+gW#Qla2h7qNc9m ziuk``v|a{2-M4^-k^Ocb3%s;{XMUwo*F=vrCP%=O&22yJH4{0W6UNT)s{mJc)K1OUSLD3Jb znP~#O(q+e?th4ybBx|wBLUt&W0$>X+W0Kcp0uLCO2cnfszknsDLp>wE2XPsDV=+>_z@6=2MIhz3;w2N`BWlp(Lv z7%qQ{SJCZ7F#~R~cIkCf2zlF4=aIFt6F)xfQ{Oi&v3H7Ab?=D=xQsykJAo;2!MyCl&eWj5g_BhkbdoU}Ap%}zQlZ*2IPi}*Pv3Hi3RDDp>p^Q3eL>dPLUtN~ zYx_X}b#veFYw*2l<2syO^=E$JPA!nd@;xP>*jrdRloS&C`X$pv`3 ze@}ecC}(nUrqzooI~L*R9H#%e(OR{5?O)kMtuBOWU7`5)aeaCsp@X226*U7kluJ7B~#k)w!13GVBlvp6DqPlKXi!L>VqMp?iSVU zwDC#Ni2@hU?z^|GeDC{S0P!5Z#<*+!0AWAT5pttOR zIRINpBZ%eY%|RSEvSc=bMxByRmo%*0t^z!upA25+HyWEJ7w%C8Q}Dm|oA-hSq52`u zw*AH?J6qvNk&4jO*cws*o?gCq@$}^4K8N+6AA18+%%;iS+wUC;Me@P|RWE30XmrG? zwcILad^aC2=BxAs{cr<<@K1{jN%{R*j<{}3{LA` zrEkyGy4A!cTY1KUxst6UlMQ^CFw;8wGrcG&wK@5>7*#c`HH=@t`;q*5%KNqf#&`{9 zhp_%VtTV{z`!L50AI)%rO5ccWipC!FVavJmZrjRbenhs;ruO?=7d7tdA>Yl>FbU#l z{)sQI6XPQ~Xik`(7zIF5+D$46ZQNPVQa?$>;`z^dZY679$_AIX)X@_Ae7WUCW1`57 z=HEqgKFWBC)5<1O)ZJ6?qxn`K9Ko-7UK|zy-qGaZd$=u3vp#t^{c*)7)=#A_2%?7L zww@Za9mrV73+k&}mwC}*e=l-1JX-q{=Jwgx3m(n+rVG(vUwVcy8(pljDD;q1p@i3q z(KTQ!Cx%RJtUb1_-Lfg}qlvPufss!vYEgMS%nZT1kId~7r|(dr`90wkD>{(yd2gNJ$*6x7(yzhz!92i$40^7Ez~e_0F_yp!_gcX3Jnnxs#|`9fx*c7nl_zFy9VPZPgyp?+^*0RIxBD4jeH(m`Ib78G1nF8IZ-WCL@!-uzr{R(>HoNe=}0ura9+ z;824jz(HCGS9~V{i%hF9#ii>*g=poZDF$$LremMl@f{-LfPdwQ1p|jt8f$jwA@hp& zPB}O(sV|Jyr((a1BsUXGQL%*+zci)oMTcm`rPNzOTVh+ybVBm${^c%4&qRvW>Tx0_hjheOYJUv6+Y)JtkX)*J+c+;CuryIQ`94nnM*Gc2<=Yi*zq_q? z{e%A8Ut%e3CxIM5mYwH+aCSb_A+`S+hyI09eqXtdq{0Og8+%i4GO=<8$V~fvj>kU3g z`2s=B`NwuMx4G1z!`+v zQ;1e14~@nE*^E(gp4!Sk%t_a^6ew4?t-9b|BPG@hGpH3LQ2XE#VE1gIKTHs>PYpKh@X!n|<7TOC=E=^CGCShj zXTk^^y+mxECi^%XvvRr2ka|P43Ukz;P~NfV#PIo_3V2rT$Np({OSv^VC*$0k1LoqL zsrUk@Jt##$4OK+r6;K%q4jM>&_(kLElaWyQZHIl~;rznf3XZSkxf!oY1G?$-Fgauwv zQL+^Al|*iMLqvf`9Ss?9u%mE9vJAE$n_Vi#Y11rTEVCaqnsdacI377w6&x@0nnE#22!~84<^VRtI;vdjj zS;VthQX3Hz@V+hw%BLjwXgH_!Cwj4-k;_YV?kFtK4-_7$l(+f>AAs)=(|LN85&JGu z3+t3ll+tGvT0PCZjb-j+6$n|wG9bITm8V4X&VYz~yXo>hrtwdatc%c(FkSJ#A&SIU zkuuNOd=QFhE$ z4%f`UZd2G4sEBDILinNgaqi>&x zmOM72DYUY7dAK|4JvKEXV`z#=9i>la45d}fW^Mol!Da_}6fYE=(oD${fAo;(E@d|} z@YhrQF^_G@Dd$7kpP?jf&?)EhDpebman&9%JSAgN!EZOmN-@+i-$Fij7`>Vh6a z46+5vn$W*H9c(PD7>bg`rTi;sGLqmXq{bly>a0@Cv8S1kyUdhI==T? zEWgZUSbSbA_qRxAi55wyC6`hOr57G!kBJ32Jj-5nTOH?>=j#ta`Um4P%Gc>PLt5q4Be3Ig?T^C-RYaNYEceNy- zj5RUyfm1b9yzA*%PQZT!rW#gJ#2P>_)R2Vc^}KXdYt*}wr?CYOt837P9`c9*-I8s2 zGwTB8om{mcj)}+My8LvFw7`MRR-Z_xdb=X_vTe-d{jq@!R#^TZt&ixJ#<&E-0&6wB z3jG+BW0;|^x0^7aQFA>3NQhQ@CkI4N7P4fW#Kn+2CsJh)!{EvmL)4=ob!#KO^JNVFH;M}e1GFm4XOFnljOtEh>+tQ6`9M8qPU<&sOQJek^Gxcij`SN zfAB7E-$3$3-$%_c)(5&qdUfmZDOwL@j$>i)L)-sh@2$h4Y`e8_X=w!&kyJ{hMQH{Q zX=#;CN$F;QAw@*Go1sBkx{>bg?(P_np}vdf*?YgwyT5NA$NvBKJC6BdhGUrfy05j? zxz2N~b>BA;wY;)X=WnC=uy{&mg-;(oMhp3=wzv{IEuEGaRG7Q0@TspwW1xN-wNbS* zrTUrl>M8C|c0a|Z^7T?!Ca_rRr83Qu4?lWr9(f4Orb3GkzTHUevXTdUxO4fdw>Whf z(=hYkvtqg8_C%EDH{V-2a?s->wt4TnG+6uCUBpin)2P)hOwY;(3xus!;&6ZL0fn=>^cT4$D(cS)72PN2E>BO>`IHTvKV7I>2 zNn|JUMvXCT_Q?xl?GW6L|Y z-#OEUhF!W8YKbZrC^FuQn$&QehZ^?Ro=N3zFj?M3oggk8^$C_Z@7jHv(3(vfLW<%N z+dD1$#$`>Whv{jO<_P+Zz+s<@N4Lo%duA7sAd1yFvp@K^3$0Z#{ek?D_omwh+(PIt zxH21YXc<;1GXAMyJ8ye*l1t@hpnAU|YE3j!$hVv^zV5IfdaXC^-RQfiha#vk02nLCukXg(M!}-_UuE{1S9V#Bc7~;JbsSZ|Sqf}^ zy(YbHTzbekRJ=c{I96*Qwt1jW)bZ8hW8K?Fgq89+Z4HrQc4$^OPzgS7I`5n>?H0!mrIVX|sYmXDmm!nfDSH$3s9pPoqJruawF=Vs9$Rh#6YP{g^%S{a zJR!n}Bb7YqnI&;7BBP$3mzV6ocHBe^;9zmvTRZ<@n`DT5mBNaP$!IXT06(3m(5SdL|hp>JML#Ey%u{kJ1>_< zhf8BbhTlbTO9Z96OZFnf$jlRK-pAPN+%DDOqnP(G;3UHT*AVl`x_9V61hZC0yi1Bv%pUTE7RHXe&gKNI$# zSPjj_d~#pmFQV|gKA<-hVh|95{Qz7pd5gHBuJ^w&Pc^l6uV|Gx&L!(wcK^uPD$3$; z3%zeVR2f>!d!W~N?Sw}wZ;aZ|uSs12xh&?t>JnE}Z?7xH(WLU}YmnJ~bV}|0X1Y+( zSSiN69y_M)D5~D7kjcd$WRk1D%WqOqeObUAbfYF~ zwLaC}yj}BS_vs)qt~Rw)-4rD8|G@bpFYtAq75apN7WZin4-$Q8NP!X-wy7!slL{v5 ze~JG$d5~Im#{1Z%?H13#MevxVd7j(|*)bvc2uJI_X#M&T@#+DroH>=~*Yj2(oh_hj zs_1x!z`y(5v*rM`)BJ&EW92bdRC~?vD%!Cak+L zSDA5CZb7f7&zz2^Jtv4%4jZkeQ;a?`2b+@`JBa}K8rLaxQ};H`%w4CM75WwZS{!V6 z?Soqt0906=^u2w{Y`Mb%K{dGgf9M+*iWXi(Vb67m@6jHA#!cLA-Fbf9szr^0#Z-6~uO^9c@O>clUa0g|Wls z{D5hgT+kins^jtMuFQVsJ@3~_eeZ{E^KN06UAwavO*>3d$-PT~;60}ldeDR;U)4o1 zajHI%`HR;mB{shSCP!uT&#@$2DQ<488xF6)J>|Ju=oc z6ipRjK!f>XWK?WGC@vyQNYv-G;2%2^hu@vgRv*4d<3PnON=q2OgY85Nl1Y^#_R8?7eA`y~@1El<6m8EMBDLq>UbH)_oCJlVkxjg z?)>ZZnpQ|$ON5?3%D5b#QSZo%^WwvjJBi9SYqo zNBi#B-YpQ^z!IL-JJjsTii$cxXt`ky%dk%CdiEUOKZFAnFm`vx7$(>r%daSwE>+I= z+_>mHOwooe6b!xBux}A>EaH67W6`nSB?!rsF}dvCcX3GiSodtHEvg(kkZytxuSbnS^skra^LxoOtT<)&l*l0GdOv13^wR?<+v}+G)O>;_1*#pQviAAKKWSEF z*alEZP#2)HiR|YGH-x$P*7lLMSm(cso>kd93ni0Vf3HhI>~_K+U$G#kdejoA_+}6P z=KEziHTugMvuDgf4N@<8?U^O6dm_+rB$H1o;QF1P$0Y!TdoYBzwkw=GNY5wXzswSS zrj01_qHr8}_hWGAJzf&kY0u{jYvf9N${c6= z&b9VS&=}@Zk2-;U(xbTAo6QR0R1VpeyX3{~V>V~23{q)QLEnQ_Mq8Q^wj@96cMexb z?e;5u*0uaardn^qXX3V@I;5W^@~*4$oyD(luI_a?pdX_Mz|C`R7I09n(~pkLMK5Fx z;eBpHPPwzk;ZJR!{@Y%Lpl@{Gf|^zxcF&)4F4EUK6wp&du)*dl7Sd&zm%Q`jmFjX; z`#8greXCokN-s)j8PMs$gg>ZemDgULW+{Hu3=ijzF?PUCaBbPMK}^h^rD%yc)6HW zPO2rF-xxVT#gS^#$KG<0nKe}AQG+W-vluy;hSI3CLjU=?=W&MuibBmU#f^$f!WzjS z?)uyt?yhXu8%wsQ-`0og`Qyl@>L5a7rRQd+SN5p(o)Nvm(}Rl{jH5sPX$Ie+S3GbR zW6=z)8Hd*j=r!L8 z-V4hTweQ5Fk?2zX9%X-6RlyzdikD`Md@j~kv-WmTFl(sjdI>gY3>eG0vA`@$-Rue|Wp!QIOf)v6K zTz^pDikJ^dUPd-95YGu(ssfMADa z)qMQKucHtgdj`FzVZC^qlAWMVSMMXUqkC(=>N5}3Mj1!Gp&lZLaF=?1SjLA~yni79 zK!uegu)9@8J>+EI>NOX%K@ zAEYcYY-pA@;pwiUqqfoamFPX6fnd}rV4^zDX7UxWJP&)uYVP(!d8{_$^(^A9Np-(_ zMlBtM(|Jmt#y|J9Skfb?)@(>$?FF{ZqaM>VK`z<4v(<3c%pV4-FZy-A_F$KCQY*I(@>-bGhA7@mwbq+*BLRcH(RZCX`)EcJ>r_?N}4pb&>ePz-NpX|Os zK~-rh*a{!-!K+JAyf_{qdTxE{uQHceH>us#V>P*Y_l4;%n^O**AGGqR=xhr*t-h>r z!dP-ryZeIC&wYA+Xbz^7yq;Ya1e~Nell*xEsZ5g0e>ySnW_bxzyOoDwZhq@m$2)?n zJ_?^#&$-&BI8LR=#ruA^FIWhyepznW9*9abL2ubgtTD_i5`ett$>Ou*FS%7 z*TtR^wh(knL2VvsJ$U(*Qv$Ifw@YxGK#5;OEikj(vgr{#R1bZ&c{n#yVk!KoCjuwTuougfqBW;yQ&f`$kw5dF^6?VR)dw@$2Lb4dnWUOOe*-U)&ZsDpKh^ zGCTzz-{jYx&Y_bc2BTlOqnusfsV_d(vAXud>SC#@_7#XcluV{Z|6=H7%37z2n4DI; zXwUPSkl%R!OT%!P*&&j??le?v);V^PdDVy0?>zht_G8phC3=a`iSdjtO?--H<5YB= z=+YOQb6>d9rX3n1t_$f#VO3SKM5 z5oQZBT6XVkcg5aR6z&+D#-^gNK7Z3j<9Tv-WJGVE@qs#A48|7fO`xbV1KqZ;S^IU- zajI?>^NhZM-gAQ}``o(tL!f)NT^w1~9P4nZsqrJG62A7p;{xk`f%L&|>V6Tnn>|cA z21K+WWT(itlAl}2 z>~83pSFgqAu2xK;jyqfqSjZnaUm7;(4mHEu0`(g-%OgG$UPqgiY$E9t3?GG|r%Eiu zQ@L>>Uw4jW_tz)UzbPwIc+FABS*Gzbmb|uotQ=xLpRA;HY-mUuVv1^iye5H2XJ143 zeSjVNsraw0^L8g>YLBJIesGwTdg3_WH>aMX*Nd=;qatQNxfLi=^KN!0*e>pB(#v(` z!m7tnL&j<5X69e?U*iPvWBz;oeggtCQZ0}~5*l5Y9TqGIe@5ks%nx8M(W`%xxM2^* zj3QX<*`4innZ)-h>d4C-SxrV$pkA!aMl{bB@T9W@kt-%F?skcfvvwC6DU z+tEi}2|$#k45DSy2c6?3LD}uyOES?>jVV!7+7K4cYPFxSJgQ$ZukvNy4A=j-oO8As zxyqLlOG{Qt>@{=i!;wA1S%Cg4cM743z6E&s>}Q7}Yul~V8ZNA_I2#vfKOa}>kr;_v zioLI-rA4ofP?0;GHL;ys*H)LvdNUQ}99eQndyfp&_C}3%nf@(2b0Vm&bJ*?v4wf2I zih_ZUS%s+bSNxo0-bM~t!mqnt7_ZQqU{BWbB1MhVaq&V#iByV7K{8pudtE&cAM_Vud819 z2IsREBpkah$2RorIyZ*nGy}x)StAoUBZ{NUe|Ws||1XO0*PBT?fB+8AS7qPtW#<&I zmz)@OM$h*~_|5mR?F&wo=q-&v=Q5PAYHZ0VR`|@SLQBnuUp=!k!301Nbc&i=4 z{KAbo-lfTC%O<@TcE9d@Kj*i%L*WQz-?msrjD_uT?r#}CFv6_CKI8!8-^07mV$ba7 zo5E^A_sDAK8f?*Zy3vSPi`u?z(1=;{xuPcXr+n47wQaTWB!*{^x@|KzT7NXg?C4&Y z4ytu2tDWzkJz~T)oQf>0u8yeCY_9Eb>Hb(7xX5*Dz=dZz%-7pg=5e?%q|8=NapPkT zd7~5TdG+q7%Y5m9tgV~|L}w}P2YA=Z>#?83Ys~3pbZK&8(PoK?FfUHhz#O*yKS`ed zz!-q-)Q_sJi7jzs{p0q@v|L-X$oB&v98!+QQ^hVM zs$k0g$3o8M``}+aTcx9?z46Tgk%Jq)kFj3!>*}npn*Q0LBbBgRJk`T;J-R_`w;9cI8zc?Akv!!P=Fo;mEbLFpR35Qt#YHt2 ziXLZ{bC#LZg_WvjMB42fRIuotYZ;I5iox&s{#(db)S zOw+NxvF(fjoWjw`3QofCC-JbIC#1|?lpmkue;F(NC6G&SG0px@0r%hDzl#PGZZMxr zP3#|Rd`i*)~s%={1M z2fiQ`19p#2{WmB6zd!!}ZvF4u|Nj}c?qR&?pb3uH+WInSQL~$_ z-QXrEecVha*x^H>W$f;_aV_35N>1=bAsAJ^sd={tWna75~Kd|?(fB~LWJtE<=wZ4v#v+h8*e12f1t1RB{IU0C%x;w4B3hMD%Z+N&IWRlYh8BxoUE!SCf@chwLDiOOiN$3u;1hcPkOtQ|N(u2OwILxZ+h@htnmAf9B+b zfwR&0hEEnXvUa%K-aV#nY3}ZLHtQ%oUk?j!o2Hmm-&|i{c>DY|?ulU^%2h6W8{6b6nVf$2@~E4&cudW7(%>6pku)d; zf-=fLezg)V2L`qd4+L-hP(n&)s?@{_<=I_jEog>@{piP2qwCx9lIL4up$1N5#o-P% ztjWJI9-!%*ncktyFUNFN?Rt%cL~-Nk^88Zc!KMnWiO@ZNU+plVPq#)`PX8|>jD%uM zJD=`Iife^QRqxkdtmKvSfzc6cGRkM38nKm>^O8pXv9L9TX^;iM^Hsv#D1AhMd&XUW5tcF+?gkN7@>{GyA1=VU8kzAbc(6IWXacpTD*@^|+7i^S} z{~E|S)scjvym8cUu~$<(ZBa96s6*JW94t6Dp57VG&<%ILHff(UE9p%)=#HvCT)w1> z9t*JDedxP80!Jbw{>yK4c#f=>+a@zp0zJx6K`11ID#NjEXgpo-TWu46dOGQP+UVBc9v0tXifTYEf@39$f8RBW>RRYWkLTb7H0%dMNrIGD^o&Yq1Z;dbFn#P-M1SE`6@zgXjZlMXVX)C& zVK3{QM&L`iN>Sg@BY>5qy!JufRiFX6Z{*9@(+C`pe3mqa!O%&~wQ-#w_;h*x?n<5a zsA=>b1$5C#l!cRTh(?(X%JH3I^LE^a|E|0{|L=(3x?6JCA71&Z6yK7e$sD6bN$R89 z^y~U?@b|M@Vp7W(6;$gqPL+Sg076#~p6xCzu2uO2I3;;uN2{T@^GXh_CFU<00n6;z zP%MR>RM>wvnML)w?FULIuuH>k4!^cgHOln z#0bmH4qRmsyA_}1$jFI_n+4zt5!&uoA8%19H?R0d-mkVN#jTIjB|so&^QTUP)>|#U zZPZdQW}+^&oYxuy+r2rIA)Q_P)TTaqEP=~1$(-*L=bb>KMP6Agaw}I_RaMpaf>=(a zSl{pRWLWOh^mJ$P6!MjV&mp2Ok*~Xpu`uEB?3uem`o+>>G$nerAf-+idFC~(=TU>m z2$d!0B|R7VmoHydatrH=r)4K_D@1$=PWL(7vdFkB65`(rDCXzphFjL3x6LV6nha+a zANTROv|O&U^~Jt63A;rhx0@?RRr9jZojSv!dAy1I3PT!?^@@2n({bT6u|WNCO~+ob;`iR@L(lGggf(52ceB)_GbtsUyigL? zsz+e=pH|Pibp6ozszR(%ZTE9H*kk^bQouQ#i4cidYVHk*aM+!iY>19GE0+9(q|~{C z)F0>!tY_G%!=4^!7_7G6jySEN(eDoRb?_{|KS23bJM`37XMf2IV};#*>qE_qF{4HG zW@gHTdaYUG@#`(~z6745b}IJ?#JBo_0?isJD70$a1VbR{CynLdQmaD^frk6VE>}&D z@%ZAyQ?(;B&lBsq)WV(Yw4fvZuKl<7IzGUy=F82eIkS*oFR0GG@FF9a#kGEI5>N4G zS7dK1Wd_x{ZWEW@COWo6VvVdigN8t+ZK+w7g-5!9myqioy4DZ3L~M`utJ(IxNwl3gOWr*ARX*#Wj=TGndJ*Js>nR?P z;_+;L5Mvo|5mKC<-<-MMtdq-KTJ6`JP3_`N_+5ITAKx*Uwe1&(g8q?JlTA0E9M~<7 zIC>wsM=Zy$JN<(<=9R$@3RvrphhvCwZOd3J7y}$^Jl99X%OrH0w|9_sZKOu}GH@OJ z$FQd3kU6=5EYrfe@+1g_-GFfY>NiND`}O$?#Ea~)O2j7kg|lA1$!Ot|PLLmx@Y9mz zve7Cai>l-Ka>m;14+v2%lhXQn7w%`t8{R3(M{A2+ykwaLP~3Oc33<#AGFszkX_^uF z$)?Y`0>6h^FV}LuKFZpTj!$6q*VL5Sb`alQQGf`W39w1{C^vs->#inUwSOu*Ok7Ooh=_GRRG`?p{SvV&)J zw_?7$=gznE^MyHn@41Hy#g*n>wZk|WGPj^+kGI<{iLcY`@~}N9(al_7 z6*~G+nP(mI%AU=_m46f*-Oc^p{ozKwQrEhib>D;rhaXZxZYfy zktM))?(E2``{I!oi-#`ChOU-lhAmseg-bWG~yRDd6luhtt{R;@moM^BLx zC%1mrh}cZ=rIeC}_t28AKk-k6#G)ii?q`{85TEpFj_aOa$Hg%%Obxkk5(CpDL_=WU zP(cukUYc>_Yr743EOGE3!s&gEK7XivI!YfLlM6VP8iPHfYKXwc7 z{%Xd1v{${&(jEmkoQq+@auDCt;XtU5N83CGiH6S2cT>{Zu5U+gD|?<(_JeyfABwdT zI6C{mpDOyH2x>j)XC*v?8XKoAX0LC$6jz&!<;6bypb~=hX5Ton)vaMvWD+0Z=fsI%n&}V zXSmvzj-c?)v8q?aXgZE43aVXf_EpaPd9SKPxzRS7z z@Dp6#VKl3kYo7du9$%LNy16k7$b9(PE+iZIwfHT(r{^Vzm(wrFV4$?>XY0)|;6Iky z^Bd1aJ(N)U$m;eGv;CpIjl)(8CK@ogT6b%5w7Nmh$Mnf*55_h1oVCmu>?oVhBboRC zmB!!s;4-^k=ZuhYsD|y-?AMp{NKxYNG@fg5A~uKHIRvgy$w$8d=F$sE$kS#BMe7rG zC&M)J!KM)lDC)tHl1++pT;e-wUSKXM_D|%;d}Exz0CyeYkPckt8tB?yvF+n&?`+Pc zARi0$6@E5bZIAoW(#j?i$BM{A6#;3aj%aaydqXZ_g=)N<3t8k9^{}N?fd5vlVvrUH zfW$_w=7B9_QGJ!v&NitB0LVgqUQ6u`tSQ}Je-TI*>jJvE%&@w&tEGoMQZ_t;Sy@TW zUFG<-!-H8M7tM69q_;P35GUo&yX*__QA}VO77a_ot;|glJ_ieMnp|sPSr=>(9~}&3 z6$SCXckA*2M~OY1%qN-@s@YiJprH?Cc7{4;-ZUE!npxf7K#dtfkYo!ogN<&j~T4kqDF zJDxjU@$cuV{CLLvrDO|5M>q<$E4t}bHAzL)jexP46|#5+nU7#lmJb-Bp)oT|ef?=T zJmb8b@8CMZ$lEo$tDo-edm%!=Lu+T0K%!2`;F*l>wMER!5BNZ@-Y0K3Rsr8_EZYTNrPyfU7?otvQ?7X{SX*+ zL0$i>^Z3)$Z^Z9=33m#VMj2-MEB8Qbe3B8-&KdLBB2M7*Lw-fEb2eY#b?nbODdD!h zbN>W;J#&b`_~T2M7l(?u%{Fm?IxFmLJc9z961ozaj{G```x_nVp^$T}b9-V`jx*LT z5S_Dy1TdW9Vw5ck4l`l|w)q2ot{Mb(xFuUQs)0NVXSt&ps$jLE+h&^z(FGj|*xQC` zNLK2c;%4}CD3-6|2ZUb`zhFjZlcOPpyEq=Z9tszZMY%poySD40s8Kqhg_xbjA0KK$ zmnyrJ6a{axCk{F%soIkZMemj$9Ui_#n=t*7A4R&9cIX`tgty#bdvhnHaht2_0lp>o z^U+6lbH_XGmCerH5jO|DE~iKy-}!2gba3H!EX%k1x?hi)!kiBS2L@#3-8g1Cr!4uyt-hSFdJZTMvN*8bo(78 z=K&n3Z6-1}#kCNp{h_j|I_ud%8#&k32!oj4_?OEx6sGgaartyOeK*Yz<5I}arAzhLBw zcur;gaFET(XPqM2J!}o_KZ(9H1mmIO_G^1t9%&?u)X>&e>zVi|>j0F7pP$IQKDP>d z#-nI=Q;$pL_|WZb6LU4|dSPhT6SH{f)(j1Q`FnD>#7J(P+{}OJqgJ}lKHFUlEMZ}7 zYbqCTy#5_NvvsumxA9$!ROI6AkS>3ca%;(lGlzo&JllCCJJqrdltbD@I-VI<@AK8l ziYv+oFKZ2JshI=4g;8|9T@Gr-1ESeDj@UNd?ed(>*a|?KWFw&8syK_#A$`I%Bn@O* zjk_Hp+Rr|nH$knAQm;dk>T8D<4!@!Qe!p+D;ju-$N2Af~hWu))(66@s7%##c_EuX& z9_A%5z~Jcqa_?hd{pS|Ag<(j}mGBAnRgEqJd)!_grYKHn<^>VvTX^U11!)|_u9-oi zU(}vpyGEL)P(X|5@*s;-cXQ+{#FNzF$|1HvPoGwg0BL7-{f398Qgov7AD=_Y|P(~OgQ?uvq=cm4+v7fcT6=X(pM%O zqQUa+5$N-{l6Pn%1{;0mo7Q^cv7Q9pbvgX)_lKL1kVY#}(CN-Vxzasx%AsazvQkGN z-{|N2@FqNi7!w17eT@ozpS9Ry*=gzgj-J|o6T=`C#J@{>Xa57udRg8=A;-2=lJDmUIZ+5L5k4SsFP3C-;v< z9o$!@df%+=HtGUn(PQZD(^g(EvC*`!CP3F>PwN6ruXjrN%cp7F4i6~|CsMtM*3+w% zAknSd z>9&YZGYN{@nUOy~!r7)@e}_D6NJf6GjDh8iiv#VtrP{pP(VXo0p>bB6@&jC+0GzQm zOt8*tI@H(3ac_nv1oD({=g`kK#1{_HjVjwBUec4l-zWDlS?=+NZ@0;>J{s~=Ccow_ z-x1rkthuB(Twb<;bZ@IegmS|4{=UKpxF}CibCuM!uw+#Jd?K6UedR`~e**5vwl4PO z&jqBwPL*iazA*jHP@FRw#(gQrNp=oQ8ha}U|FWBO=}K3qZ{SL(An#fcrR9Lfw}8Hv z??`LK_yxY4x+y@6^)d19eYU$?)Ck}{!?jyEA6#@Cfv=3;K``);kY`*wl`L|d(u(gxag)$YhGVo0mn+Ew7u%k$uqP?HJ-E2Fs*$v4^Pq{{cYY%oW!lXcy0CDB(;>2q zVZEYj<=H0g>ptM|Y76RiPRn~VU`_eDvJ6KSqtJsuyR6mno@4PbI~%h#+^8_fCZX{B z!BlA01#L5Bk3PW@w(pP4-z4~wdi{k;8a;|Qyc@HOGL$zLJWcs>*|>SeVK0RkXV7*O z^Lf`~3D>uZ&nM0RDV+&w$}gXKX%1@&%hh_CM@WXynFG z#AeBz7Cg07T__z{3-9=97F(SRicdd7F`|lEz#%Tm!!T?9VA%jFfK-V?*Fnf-?Hf=O zgZy0>J3}wZiem}&n`F`QHmGH4ELN($yK>!1=6rTsYg31ZivoT34decwEw%#RmUP&JUh+r zH4E$gNuN?cdAISU2Yq($j;?MKa4+1ch($nZD4;4hB<2apjdjjMNb@j^^sG4S>&=PzABU9`cRYs$M}plUHesU)U9nK*0nX&8@j{A?NX%d^T956~d^wdq z_U{ng5L*@mw!=?I$}H6<03Bbj>G_>Dl^%;H@yRfq0d;8wM)$O}ivjpBSn0+6=e+>j z)DuWv>x^R72?p`3#iVC7HYE2a44996o>C?gnrqPH+2>2BHKcr2FdZ8G#VS0r*O92>p{MHx zv6(XLX#2pizXNFaUDSu9{Y=Qqa7NgVA2h6>8uIE>SEuRPQx9Of7~S9UPtZ%I$Vyo08a?L zI8nhJJJIcOazBIHm5uIp@GX8NZqnB!ELJwm9@a-8FP&@ZYa2cp7|e6Af5r@8re(Yq z2Y^78g5|_8li7d)GE6foE(#Fl_V+n;R0#h13*z6$GAghpo*X$Vwe&PO2@%AL$tb}} zI6|vgr}zDd$-zQnR5ZlLdc{+}8i~H1y%j!>CtN|png^+RGi%C*wkJ0lP*iewBK5YR zd`pR(99qPu32U;?K}PKGexwJav|u`W3y>VO&#G>YtbwHZ5&FeDvt&VfInKq&sT*fS z(bD2`-|y3e#YP@IM!!dcMGetP;|-NyA%|8w;)_L!G3%t6J=6yc@lSk*zkC5)D1)5h znk@*Lg~r!+bw04Tp&hc_*O&9PWBS{wUyemPP zbQ;##fI*glBXs+f3=S(wrqEzzBYwln9174Im8{_Id$l6LNDx$Ohr4whh_}^&%X^Pb zGhS+wpMM_2%`EvlxK+>=@f79e2JsLeo@p%XI%G6;!8SkTKyk;<0rQ;7K(t1vdpUzb z0uZche}GV~AK(u2u(R~cDlDUnT)7VM=%>Dk0PCI9*53AYNaH+P5Uzj#!GS^1K{oI; zPb`gwHJAHFatCxnf3Z3^Ra!4*5zk0-JM*A<^Y5$sn;rQ7{5gnJ_lpWcn>+j&?F zoJljAVa`(Z1*=Q3JvreI(5otCqq%p%&}8CYcSfE3GR$tpz~5mrL*zaWw*bYvbYG@YbDX%nb@!3sk*2wf-kRh zG%LCqH|PzFgUYDIeq&0BTUkEwNGORz;bxHrDI8t#zC5g{ES?eOg`Xn_0vV9-%^XlD z)7Vr>Acq1?s>4D?0fUmy-?M3j9l6LM6J9LT$k;m}gz! zOkMfSMVD=n)F?rBUp##mtriRH+axa2uY@Uuy;0KO z*P~Huno*Yt5IWiJF%8nuEX26@7*IG1!f18)YP;=_tNOe7PSA|ZaS4-jUP04PD zZK3b=zWWWINzx?bJHr20Chuv&=4?$xT?*?K{`j~0507W}fr)GAr5FT&ajP@0<=qvv z=>kIKP@N6I?SQ1Vq*6f)ngjB^VBfk5oAcWd!z_ zR>8@OoLeo>lYpCcYx@xf4c(0UrD^16gP_kBSv;_T^;_%ncf#G@SBZVn(KaHyF^fwC zd$6DZXhJR<4H^*2t%D3c!UB)2szI@NDwo64uU@{qMOj{4i#PY3iFg%=)6EItp5*L0`nEnohj>Pi9Ti;cLqPErH@QBr_zk8@o5*6lm64PSan< z+C6CmPxL$Y>}!?q106`dz(oCVlqVuF^MuW zKSqc%<#i^#+%SI4CR?7wiC}mr?55p?eh`uzFaP<|y8_fBgk-MrzKj=W z%J#~`j$F!ZS@TKX!F7s&xE=<9IzkQ|U|`}34(g%?H6g2+zvH!QlllJ7ma?>3$#8nl zohGf%NSEk*$CKbnG(&RZDJr)=DHpkUZb_lvXb9CftQLfFVHcQ=1v^O2Q(+ zar<#iqd+pQzOA7Qf6bueApR3~c75MFTknC(DTR1EY&xm#ZsRK;u7EVsk21JMOT(nv zLrFftZ~K|fx3<;U|0_hTlweY89nEKd8&?3-9!}qQ*nt|r`6l8%2GHxB@8DkCN55&h zZZ9>x=OaxM#L_$X>$CVYkM|oh4k*-|lk9KV-}AU-!^aM7tmTvN!4vi{bFitXqx{1r zv>42chL81DSkQ%zx2MW-?E!&NSAmnHy*gM3?1dkqp|TGp6HeK!ytrTQODA+CDHDJM zXbg*F6N6wcILc!cC!2@o1{wYxpZtZZy|b<%ER^L3@xk-gRkrKZ4SiGV?vO3q#00MxKj4edyDRwqQ z_YR2B{}#1hK-7X8Vs8;A`P4Rnw%$6wI>=DG$peH76KP%O=50&H8BE07me^(}OXDVj%RqH@+5x07hsaHJG!W+{(Gbs)Eu|E4XtoIVLj}i$IDFV!UXYDn zZQF+h-AXm6(`L8}t7rqv*BZ!nU5AutNDAIx=}ch^==2$>sMh{UZmVCPU!$|cY5Y~Z zv{^}U^;61fjv!V#4sFZHE!cvpl>=FzCMtS2QKkw<<*m|jWAk&jyF_HXTF~LpbGYykl6~Ki;R(3GYe{#eq_8OLLawI$zSWQ4$$ZZ&44HBO8D~E|T-Ii%|6{ua z=$l7=QtK>dR}!B~;IU1>QZ047zL>w6y0G@J%yF!C3;UQ+^U12r4hS_4m!0;mwE-d|lqWM}#3yrYLJ1l^q)*AA8)qKGvb#-p)(6wX`OusdC$rBip8`pwgkj);#yD$reZI~| z4nR4yych)Gy4Ge_ZW_PDH0~}}U7Ic;^1&Qvmbi*7uAG;NNGR~ew9IQ{gCPvL! zL^t@W>3%)+s`+GbMsK!Je^SyQ$g(10yy_;2_G5=VT;Z&@?cuCF_TmR=aZO+MW^1O` zv%c))(lp6^Z(8ZDY!&f9hF7!v6zn8ba1XOx9GoE*LL=j-8nE_gE17@q|32?8>t)gq za^B9b1zlA=v5)fuDR#Lm?K!bH<>P91qoQ`~APoF)204eFg@D@Rk;jDt;F%fgS%sRmm?R;dlUpg8Z+brTDeO(LeSU2c%n&CcRAX5zb(|a zoh->|8p{(DlE%}-3ERislS3`C4l?5{zUYdX#1x%QnNB-^Y2a^a4oj&N1;__& zph7WH3q4NQ2zN0J#bv3a=JSyWlL8VI6i{a!rsa|7eSeut*Pd%M%dRSwDMt=e_Rpbv+($D8l!9)q&%@o2n+#d}AULX`Dc*_O@uKXem5r%&zX*nd(x;XNLrfhJ z=Wpz^aTWTY+S}m=$@RLkIa~wGC|A+x2e)>`(fssm`CJU~B7jD2|Ko@z3#gsvJDXcH zT%DRkg0jR!=5y7%TODW&jU%i&;6j9csyCg(**MfLDGnGoz*C{0+_Wx;|(obV*9W6;HCL{Be0GM8TP?kX8E7Yz`l8cwg>(Ti^Uuie0~M)f1i zzOUFSHX?jhU_`GsokseQI@z3ozXOVqSm!0?J|K{tDVz9ZpV;?RGVFl^I2((xR$NepXR;M03!vEMG4VsGeYc>_NLv=b`H4Jo`CW z2ZU-&CawDZ~TRbFw|%M=85c@9JuyGI|nvgbj#&1AQ&L?6V@l=~9_ zIt6@mB0DZU`CuU6JoRNKYgrn901Y8Uzq3r{@R>)x@Bn-8sYH##ZXV$Cqn`2)>*+zaB@Y{~;%?GdkkGrAqUsVWF`?_PEDvdGWUD(#Q@sQ8_V zY$07MfO%C`i$6Dv`-;6W@JY+*Z0k8;aVdQ%iOE8R-e9VzZcovfpb{Urn%akjF*F73 zuRPfCILIZx0$g-9?zct!2_ikhz#sn;BHY}xMckEv{CW)90>G0H*wZ=&skfl*>Ayu> z`PJSmWe+qKUIWqv{RufQ>kl#1vpP$pA!LQ&EOh{{`I_+l_7-3*YD+f+D0d%`v)>cd zDAI-@B9Apg2xD5ayg;Mc*Exv@@UZA0{)z8>1#Bp3Z7&%acaQ9MCI*%HXDPbx#eF~t zdb9cst4be*M#gSTotLl}%n{^-V;P>(Vg`Km#y7r`vPeOn;IdaG3&@zy)AmXW)DbGK zKqUO9AOc!>lXhKca$x)PE_enf$Lt~5kMd!E*80a?cxOknQK>b7>JNKu*~N(Tk$MFHtWK?yz5 zL3#^KkS>U!h>9r1NUs6u9qC1ifb`ybM>;5k76{xG&w0;xzf;C{|J)yUjQb9MyvAhj zwbz*Xu;98tbOp?XS zMOG?_H8B*eCfw#tfWfedty-C{MJX6Ev^d_MjM3u953s=(_iy@U88y2hpC zdkmi8<ZnT-q%;NV`GR@e=-3e{PyJf3xea|3Yla6#Ol z{ElkE)zzke9EMK{+}PmM_Y{x_ysD>q!11SzyeoM(1n-7t6}Gf&ghP^yKdlmrCA$Ow z|KaECvaUufP@zY(R~+L2VNytf#tV0JwY+Q_2ykn}Ar!M#DPPE$LWtV-r@XYgbIS3a zm3+1O`La--lu)t1WgHJ<&m`FD>8sU8{r;GT&HbtJjyNMPvseW<#amSr{Lw?7t{i0D zqLP-wXi&Zq5XL#{-PmQ(Wq*+M?2~_33E)ajt}ecqbZB9x+y4{0k$Zx0c&zuicb0k* zuX`ds@tWdLM#JwEA&Zvx0HgJtak#|D_b~eVck#6|twOK97DN!rJ%_Ul=eOUdOB`&8 z;ZE4SCl}`OMeb6c)k|y@C4^ygpu4J}4#e`>0C$MKrCt*zB~^6GFY;E1OltF^1LD<} z>r~IchEh!gDV$p|TZ!urAK+fn0UNWo0~EHxxbnX&RRo4f^;g*-5tF#rnpy>lWYMnC zW3Pn@Dn`B|VR(&v-1i}*?35TWcXC!(g{9$-31DgXm~eC&+c6V0u9tqJYv{3K>(xM5 zf7Q{|HMSR$@B3*NsXx_MB)I}?9=+3RRz_ccym(c>ZEK0faf8V$GIx7*c8lKnmqSBi zGGOvduYZF;u2;b}+O=lAVrzcJJG|-&68Jh6vg4St%`m>8r#H;iPU=-h15`k?bvUf-nOVja?AH|Q?+`r=rNuS9Ak?pgK+x$c8x^QFFo z@d0zMVG-H$p3@OM({g9V+m0jZuMQ;yfLFj<9ntMZ1%fV;VV>TBv+~(9f^^Fa!JcLZ zS%n6v6Ju9g+ub}Y0)OlNQNGj}6X5`dkxedCj3Y`0>?C8rPm*ZEByd+9vB)iwdoBp*R9$k^9j!>$Iagl2;D6}7})H+Ml_@*X--Aw zQq0(tgM%E%Zj!GoQ7|e}({*8@ldAc~zG$JP#~ZalruYI&!Y}uy#izK^@}vDd{BWY} zJmeikG|nT%3WB?g-cDc65<@_gUP5^Fi65vQ?2t2k7({+9!-qrzl0pr+ z8a=$lG__Cq>PO@ITeYM zIoO&Q(nCH`_*2pbTr9{$KZpAn{y-M!%Sw4DPPKWt4;V&q9~4O@vBmM9DR1o-cWgg~%xCkeksN+xg6E6@7XcJ&tkh?oOE{HzyyC_~|L$ek?E2zC#z}9^U z+pCtE#J9sDLPQ8IGx}_ORYyC2H~_q~@12)ySwk5mN1L9BZU;U<673{tp$5$BZcl@R02G~6WJWd3J?)FGm-HLM4Il`5n_&P^J;8W{o zIM4Qoc`_r$WJ9d4AM%rl-+J;9Oa%~{tf5AI z<&2^6zpk~r4Rz1j41e6v(uXkR1CG=N75a^XXq=3>zp6f5j-DB`8cZr@K&uYT z2p2`4&q}Raapr6QFo%jPTyCxw@IL{M$2k(z)DuAvY8oaLNq>-<|7#};x|6}53Y8GM zzIw7wwfDU=d6XGN{4z!rd6!i)>9n6CY5k+rHxw7x_%Z||9O={`JSzZK8HKw+kv~b65Qz8JOLqGh0h+ zA{P&{o9M5mcF*A@dUtA0p^`p-i-uLOxZ1{o2t;Vr&hBi5q#*>X-%!QO9?KPs>M9{3 z0mi;K>Jl&7ZVx`FQKTYMdruf>WZ~3`rS3|6teM!8U zo=a@C`Li=LtN85?*pYBN1OW_#RQ5)-A|al?Gl?Uc)B|we5OvN-V`d{EfPE`$^`v+= zl1XuzXAqPPJyEYDJ!W!bai4D_JDTSYk!oxHm93GsK@DhkFU5|bZ-^cs-R;-t!}`6h{G%Sk8`LnLHL@HYZcmMs;=ncAnK7=Dm4vls zkf{nsMnZp|1`ngIHS*?rFH%_h)L4CxF5nQ!SZWWvqcM`D&C~GRGja})nx#430f5Jc zV*AQ+S{Rqn1xTX>Aj#3vpoZ(a!%4Wa%7R{o<;UODMpA&U1Kj$_DEL15`}4~iJp~aa zSV(U%NptbSl8t$ksg*qG{1>|qgm46O@I5!Refa4z0Km!t5ug_-J9myan`?~)@)N)vMmC7XWSyVQ+-T=^1>Sw36?Hj{WhQO+zVwl# z0>l!QJQ<+|Vd^{rF{U^p;m)fW_4y{%xDpuce5XnRd&qu7`N&?5YS=o zdM~`aiMvkKnS;jOVUIW$x^@Zp%eKM6EN}|<>Yb)QcTPl?T3y3d~_L#>ea9LUvcTTs7zl6lQq8DVj^idx)Ieq{m0xm2(Y+eQ+jq^|-FaO)uW*b6XFNG67} zPInvTMmd&K(UGKJ_OF154umX1!80JHx$7{)xVYHP6F8t?M!WZ@(cPjQv=ei}?9}e{ z(%(zEYa!v)!kb~$(5+~HkM%a5;ZmWD+pU+$S-;N~=a*T2drsPM)RO%Ct{noI?x0%z z@+z}gv0GEw0(SjFwaX9T-j&J_3q3=LB~nq zd!d^5LIv+7eYO~lydku&*_~ZQupX}vM?;uTgP?R4VXfQpQq&3@r9LOu(v?Q;Wb*u)x>boeeWt)X<;YQnquN#!OBPDdfbd-*L!ajS7Q}=;mVjNp zbkAq=!$gB#H>IbAMD|w2=RhK{91&qH_FAKQ6RD;;dFhkHu z6l}u-dMSSVVo-oWpL;lE5WNB40JG!>OP_3+lE0)Bt}hSA|1!AAiTvbg6^~P zlfEAdcUPCB$~^j4-@1O65sEeLR=!%1nX)szPMcK_tL`ZIVhL5U$?`oi{o;$`8xh7y zJuh1D3Oji0%I}SGpTOwHOV!&d@zvt*tEyoAo{zm<_!eFUc*@ld+zdiPRj*=5vmFvY zI?x=v^V*?kB|Gvy+k!r*zO8FfcZ@W+Fiek7MgDpguqinOx)+JyOZ^-cs2_f0-I+;oXpr%IOoeU_k%bgoOiU@641`Am#2dy%`P-e{ESD>WcE(%^xBEQM4+|?f$0=IW zTWVJ;9`A_3`fy{UwbRR!M!wOn8!a z8)v$iz9zY>N6POjMMEUkGu+g#KT~XTN{$c@cr^^K*jE&-bMNa)bUoDSFJW1+nhXfA zjn&47zmAyg?p5z1>ftb8vbz)8s-#hTi?nn|=55%wvVo|b;D?ffDlrr)K(gZ#4>$*~ zZO;EHBf{cO{rPTlH?KD@z$JBTIrWvAwmv>XR&Ey7PL>|aaWql=QH~)F0L9)_%Zl4g zfXM~xn^pfzO8)p0=r+EVADNxd5{Akiqu6E*SOhgyuGt#vj(?UyVYz}|{D`6A$HT_G z1^st?s&{=tJMhS*%#eGaEARb8GjI2L$R8C(C1NuII@EevxJ-%76#Vy(o2j5@!g9iN zjS4mxN^=j$Aw;?hlLMH!=Vfr2&X!`h`$4}LC8s@jz%PX+H^Tl|7vzfiJXWaTBY#bM z#nM_@cr<9H5-e5|opF!ARt7v@ZJoS@?_8-y8JRWHB7jED4g6`kh+r9WQ@O2Db@+$J zjaZpfMsAsOIqsyh7?l$_$=ZlCj9Z^Zud-S(8t0A35U4g{u%gF;RZw|PK+ajmwl zE_%`hWTYL`dPcWuo)4uw3|t2F#64-CPJmTzpaD{o{E?28Go!R@&Z7o4mj|)^o4SSI z?b?RO`ET@M&MXI9B*)-gF{}c(Q52$7;|Qioy0kjBCsKg2{zEH1CfqQbq}FcGcK{m=Rf~N z4Jbp|tTU49P%uuQ{`gB^qXVr((%Z3b-fq=E{$b_yRsK(Vd#_GsiYK5qG@&jB)&?Zt!Liov2vhwdZ({@=oS)`Q@n7_4K55}P z_lMF7DLB;~=>BO7PZZa(cW|Izo~ZMAG2wTCSpo2VgnE+1{o`c9KgK zo_SBqqd^HqYy72eQW%7Aq5@FXXh{Y%?L(b*T%f=4-~K$!7T=O_9X?X_fN`o6nIH96@Th@A&x2T+;D0upf8H;DG<){2 za!6-;FsFarw*Tf6iuYhU2Hd?z`QHcmPh3@6YfBuk#0*D>pnp?2_ zca50;O_}F^m+om70N>v0N~r!n(OUZSfCb0}U2bYo{5Q|^uX;|OOu@HTVnp5kPb8vP z0v-V6gR5_X5&!%^|6UWR0erjb-`VoN9a;cZi;#$l_3uyl0ESOMO2-Czz00V)_c}N- z3`${fciKz;OZY0KowyjIy-eM`_>YOPO7gEjc6N3y2XiEwU1*IG-(-J?RrgJ*#yW;e z#E~=sLcL=0OkVRiwZm6$@P5)Qcp626b0g-z45;av*(n!;Kr8!StHJ4r8*SYInmO%` zWWOVh0Lkp@MUEmrvBcho^xG=sG=68lgLn-}TKLZ=9g2)|Y4n}I4x|CE*w7~74rc%S z06}4PnytwF`3G22iVCFHH876lkC&#PC%1cEfr)g*YN8qq@;w7_Id!zGTi>Tf1cZsJ zPv$fqcsgYd507^_zyBy_wKL*)K3WOm(*k2V4!XoN9VW5CSqmS|)?fv)H0V_o-`m*T zAB%kg7`1|%67VMkAneg z`IIH$FUo_l91m0T-2s|0YW~|JStDb;0dK zDl>VW)%S5Y!1G2E6|-`yuwYec{jpa2>h~{||3^n8rL*4j_UcL&()Tm7-;1q7`-uS?3Nhkg^8v}H8_x;mxNeZx6g)<3)B1bq&l0ak0U z`&$3Qm#8(wh!kjf;%FyaSs%%Z$*=wyPr4PZCGP;lCj`CstWJQAK!(FuVF^%k5aEO_ z20Z?FqXXEQBQmfrub?6x-s8SIYYa@AGHu_FlISsuX^pUDE6E99i#yio+a!5d4XKGn zjTRUNm5d1NeZHfzb&5}CXkx+-`qAUu*9Y&o9w}-|egQfWR-^d_f^f9^si=LFjxGW# zBmxYV4^+YSjf0w`iVnZ>Lp>?~TiJ9*bDudb58Ln6Wrr-~C{H%GGyMtn}mM zfLH4jW|7LL#Magql$?N%OGYU~?tLlq?a2Z7h!g2)MdSZ=9IR7p#^-9+1FMmjMbSuJy8O zDWHKkvw0%SGVy8ey%izzCj3=~R}bjWGZ1FD_HHf91-1fscmfu=#s(z=M~}Na;V1nG zX^{ZEjrlkDo{FttkMD!c^j$0&)1&=XkUlt+JbGPB*V^Cm z*MSb}-n#ov(_$Q(qO)GjlXh82pG}%um8=-uy6=sa*$pKg`!WOSiGU>eP~6g^1j!eL z9)KOA-^w-4LO%T*5Wxx#KveZKGj%?3>0luzE7`TvjA6%QKq@@1!pQezCD|o~@g=i> zeld&Px91I*(XUIuF)7it?ap^_AfF-qK%X}HacM)KZz7PW&@F}aFw_pqz~=y2mT&ku z#k-5SS#Lt9X5@umy(!+q>zp0$Go1Y&-QoE@NXc9Nh~kaF?brT9L2P&$9ZJ_bQa+b0 zUpm2ih;t1g`TBF<6aVaaB?_@d%_m}w%=a67`%Z+VYlTXOY9%*qY)p+kjwX`UlXtw; zpu4rmEcrf?;7b7)6xmZ`jv(YHKR|5KpkLT%jc5J@6$CISgM9IabUMIS%A~-bx&DOpt$-)rp9ViSECIQ%B;YpqT11+(W}91&j%m zeCGrZ*>;AW1}?%huhVrMl*Hc?Ir9M)m1VDL`8wfoZ09u^69}?v?-P!)Lb=HTwlpb>Li-GoJOu2-lT@>+!+u)9AZ%2uloU%*Tn_tpfe~-b+7}rC9>0m*hy?1S+%%|vV+ua1mL@MemxMu=$%7=H2~>AxCl-J@ zf|N2dqig%E0rVO$@!&y-tJT3EWaE0A%xAnO=D@!iqZJ*K*I6@i(!wJ%wV3T#0>{^l z@?|vZGs8t@+=ue>lP$j*9(zs}i1cTKNBTv{NbG*wx8FNZ0ysUwOJwby$Q))_d=cCGP)*`K%$F3~`QBb&v}Us~q34QAmoZEr zoV(UAnWgUJXF{og&O&!-X|F_{;I4Y={i#q`^jA6W4xG$5N5R}$>#ae+sLd3-0zB13 z4$Ma18>N?aRX9FHy?^U2tqE_P)B)YLZ!@WAx|4Xi_B-%;#9H1coNiban0FJ-(8TsN7kIp=xsqI z@p~4PDVpJ=^xZux$y519nb0?of#xJuw|alSd{Y%@n#?`ZezS7{iYLlI#vzl&mL$QOgVZWiQdyy;z973h&G*gib!ZF-X{94ckvQ*{C(i&al#jWC zsz%IzoQ@RvQs1cDEG6a|f%l+F4ZIu*t6`(;?O%_ujRbJ?dRFN*L>uD^BXo4-w;!vi z-!JRn(Xe+=WXus(TESWszNnY_-G49Vm`Srfs=X6EHi-U-r+o7ze5Er2sPA zTuaOv>&TD^niJ6bW(Ia)vS`J$w9O4xyL?|1f?-Jqttc@7{>R(N;IRGxq3cUjj*xk= zk)|QJH_XK|2r)fwx_RYb09%ockr{mcQMZ-lff)tv8V4$3gc%iqnBUDN7M>t8H&y;B zU~Woxz$1O_gx(rxb|9?Si_}CRu8PkD-3J}CEV%so36f*Qz7Fwtk8OQ zFVv~;Z(W56=aaS)ch|v=R4alPe0pq_-Ugq(DiLsh^y_~}81=X(EFl&2KZ{pI zIw2n4Q1biz(KOv+g2$FvJp&<>r$(?^tT6VPz&Fgs46HMR9**(upYHE zX6HpiLT)`xso&WWbJTBf@oM~drx-cN?BYj(HKjIIhCHE-En$S=bA=emgv$_JnyG(6pn_ z?b|$wLy)J8sl0E(zCB^KyWz+PPe)~jr&GOr3)W7R*{)Gn0G{qce1PyZv_K(ZQ^Q}5 zToKOu;60;CAn}D(YxJ(IuJTMu=d*R;l``Hpa$Se_2lLCs-F`;i>0pfEkV?$> zg3--w;;-FW?83$n+K&%bP2dMQvQrAD+IH;f26sVq#agpD8?pUVqJ^-MyhU6**O1dF`Qh$7~0U36o~OSj{NprEW{5*%pCx+fQw*=p}y$sJfL5PNE- z^*gH2zUT@4HS2@Aoie}B{t{s>EK4Hp6I%-5O?vYxO6CfqcV>9mZ8>9jx(<~pGglEeJe@3fo%KgSUbt{d z8;n(jBXXV9xB+{=V_6 z?H~x4*i&E2@}0dsNTLj$fap~7oH))m-abrf2s$3K*Z1+cR>XrLOnpCsqpktO?qgt| zgpWSoNkufHouCY0RPw}g{z!*gPt`=pPK zF}a|()_*V^mR;gAuyz`ISLbc?Ot?bt7_p&Kwe_JY3h&leCaN-$*u5dmKtU@?mVPDh z5IIAL7wZ`@o;Olawb2u0M~{V`cFJ+43vY_D=htlH8ZZgk#`J_FU58+m;MTnL%6pPt z0S@UZZ_*?upQ*C{16~}5AnhPv%(?_m=u_Qw@)+KPNqv-kFEa^%N9E@lr_!-*Nwl#2K$co#VVsjlcm{aA zPT9v>Dmi1=HSq|1en6~WJEGkn0D4HmX->7o`fIzHzW6?}^&9kc^avaT`+}!4astWz z1Chm7t%{5~fJ;t%NCiO~Vas_W$D6;c2J_iGVX1O1LeL}n9f?pGVt%qHo7epc-;`G= zukV1%eKwxLDhplpy6f>7deKj&V1o$gCnLkuO; zyD!@b%#WwkjjJ+uXy~PfYcmZ&=ol#S!Xsx<2-wa|uq}uWG`7sRGeqISBe3NhzR~<` z3G`ZJ1E%14?R`8!k#;+KN52u4>hVoVKUbnZb4i8~mH|ygL(AWew6@g~oPvr4R*+Te zSvexC(3b>YTNwOyw5NO`+cH^@iH4`LJ_l)FFz#=_NIpG(7>Ch0GCSlDzhb!sRdzV! z$lxpeeY;KF%^a$r`Y*L zqt-Q8%BX=EcShH1GdBTuLvQJ)`4l7;!)UD*Dl)$%RwU~*G<6~BiCSAzwPod&!i$Z7G8%qp?(JM ztGvYzl2kl@+>;;CzWx*nl(Y1_Ka$x7b4edPRQL!)k#{EUgThI@wyXp`Q40l4nei1; z4Oz@O)Uu>^G5%97kRYC6hSJWrcgcPRYONAjg-JMl(IF&(2&n&_U<)|$;oYJ{A@3g~ zd1mLz-|TIVDBKXW6_BW-LRSKQy@&Bn4%rhWPj#S!bfBR>dfO(&w@ruu& zmm>oZtT87z)!6aA=B+O@$L3G!pnx^zjKCgX9uO z4>NMM+w;};@I?*cVwpp7@NIO_f_B)A=Ru3X;%no<^o@l8SI&QkW?S4#>06PV=0`8L%H_bI%EbD7E>-+%&Nm6tnrQ(_NfjvFztn4B=y zBIY2mux~pr_ZsoYfv;^H9J|Kg4L8qIXT$pewql(-DCP+We~$mIUD`j)ndsa^oD*&T z^x4-RCC569^-pZKL54SSIY%t7Rb}y!wSiqmRE>pwbaIp%jbC}0hQ?j_5d!*WPuy!( zzvgh4p|&>h@57qs=}qK&EphQBKz_pi_Fc9wEISr!o11M9!dZ(-MY>P@^0X5fs!BawjA|paB$u3XZmv>yu5avV|R`0_J8>+DA6~pzYE+fMyUX`XWS66xm1RyR`e209@Aq5oPZlBbe|Hoh5~ToNC)ll? z9pIovSGCq;oWA{TeROmHq<^VfiOFjQ77UF0?-?J(T80rrj}E-S>8^AGH2&2Oe$Gz9 zrej!F&*3!cNYF=|)6&!V9`^(sn$j)`$NAgtLpxW8YG|1|Kqn^UYP-txV9Q&VPO_Ts zlpPNV86%T)r?9%zFCI$+%=L58s}-J!-=G1B$G~U1ymm2?(b?jzOIYz1YE|Z_+Nw2# z_N_+4*@^%M01k9LcR4F1|{u2J_z3g(WdoFIUj^zy?Ba_eBQ0b1`O?Co%rM zn+LV@dEsnY2d6ynWk5Qh_W)hG8^xQE)~ncvuVwqP5lzw9cdT5eOY3Le_yj}@kaX$d z&@cdCbAu}LdeSifx5nSgXv*xqT9p)n4;v7e#(p4I7N}!|ryq@Y#B*D4bO7;)#Al#~ zEO=)fRN6xZA)PAQr648oEoV)7=;F>i2&oaMeDmBUvTzqFb7GYl;~_VuGbLpEfSFX; zd2cWc*$=I=mqzE@ywywok1Pe)@P5B2N=uX~Oq0?v9gA1aL@GPeL$C7|Mx7 zYd_67?TUro6fqD*xx@T&OZr}9mhL*Scw=>wZB{_9$mx}BD5%D}?g5bG{zq1N>`y)E zCK{cj0T8uzxM&5Jr39QmpgeZ*+ z+Z#%BitpLagee4$KmFc2VaP*@V4;vVyS_b`9?m&d?>d?+0or+&CV-$bTxzwQCuH?a zWcU(;_gj}-vfOPAgGAOM?x%fj7iX)oLm<$O zWvs6q1@LLNW$M;-DWDgh8ICqCPu-7y74(1vfQS1&Bxy(sdo~viL?h?qtu2+4cx!ad zI;}VGQPU@*Ot@)lM{4L0DN(bjDO)hq;=Sw(YjA3gfUFbXT|SWX$&Z&EsCB4vv*#^SHQvg3h9M}M)XDkQn8^H$!=CD@(&Z{s%A?*?d;sYL9wgB`p1 z%I9Pbbc!)H0ac3y9ygHiW;)$SYiN@vmll@lu^~dS%_|(8pqF5w3LrA3P-PzRSnj|X zE|Q5CcZ2EbdcdZ6PYQk9gfM3(nSZkK0qqydZ&l35h7cSCAipo;#=%Wv_xH7*1yM4c zM*(|-3+vqvo&Zq;uovb*b|xd3pVn>1^lu)-Az$W1*TbYLvTx z=@Y=B5W~z}fY&6R0G!-)^%ysR8`rg6Uy^NAKVRl0NE6R$4aN{zuh0V!?mU2Oq$x7b z`0UozOM$wG;Yh&|+-VprmpE-+IB$V_cQF*s^{Yfto+vQYr8a<=FVCFtY0q$hr#(55A*W+Oiy@ z=T>$jF!GJ|(uZ-}ZJ-czQ^%CApi|>#`{vj=DoYmAF@@AMa`}<%S}qM}&vo)N~ zOnj^YnNj9bf?Mow7^`U@vo z(S#5yX8(L^P&4VVd1RQBN7G2-8q%<|DqAe0?QqIx-{^7Lh04_fB-Kli zT(_EcUiRFQIJgR6$7_!J{%{h>cYW%!e0a0m;swUT$I9AcdK+ke7LMW6?H-0zW2&#D z)xs-ZF%UP$7K4HZx-5|H3pfe=P8@!aeG&WJLIs}`ge3DtRX5^ZO@CU?!d@k9Qo8Bz z#(wL$k*Z9K<4qiowR}d`(-nm=#?9b zlFN&u+lTQkh+ETe#J(a9uqBzVjc_SukhYwb&^gISSyZCg;C!;>#%5}rV6wF16Ot*! z5fHS@#0Gtya7;>Mfw55wv`uaaJ|LK5y)Gt0i?1-iU;drq0AThfFQhW-B5;hhi_6IY zsaKn)JS4I3W@&nI;#Cmqy%B)K0SQi+bsbQ3Ni%s(t*d`aCC^j23B=?Q_}Y5;j310| zYuL%duP?O^kqqj?-9hMedbFfsiq>4;ev9Vx%$dM;VBH7K{QNe+r$pe_hsg8;qGk`h za!Mg4*?f8Ksm62X_I`O_p@tAyQrIr_qVV(r%L9Z5|A@ko(38%ch=vHTy$aY*JpeL4 zWL0lG?P2_9Uy-KB;2vh6u`A}`Ao!d8ZDTeX?&!RXIX`MR!t}>;CFn2tH3ce*n$_LsY&A(hD7E4Pas?(Mq_-uC1keB!q|6`Vbyx8G2y zw@0OQBX0Ie4Vx5t*kr%ydgJ)@$Bm?keMa02e72eb#=++vAu=TcAupLx0zft4cq3$Z z1hSzT$Bx*~Q{Hc1vqmfad5G5lNQaeDU%ppb0@Qu9BnNpoWrrX&NC^OcNgNCSPuIyh z3<2*A81tF=AX}t1uf}ZWc?aE|wji&DJL;6f7i?E)fz6=29cyXQkoLN&U(~MueKXaC zkH`EJ7v9ts!GgOR{kPSph!txMYe%daihv&nm9Uqf?gsbRf-X+k`O&?@0`Skhn(6JQ zkP-(@2++wr17q5#v<5x;xsHCC7fjoieEJUW9a(uf29f zw022nPK*N5+KBIJYX4Kf^QhzS1xXD64iO?DksMQdQ^2!!mHcjN)>w^yt6?*_2jsjK z_MZe|D_*CY9=>VAmL{Rj(*wf4BiX*0CY+Y zr8~_1HLL~Ac*--rV=A;yM6t&9=XiR@>2j5{b?&%TR6^rt{U#=IKxr3QK7*DbksprI zGg+T=Al>>D@s)dRrt+|(Q}bs#h|0$Xv9Aa*^#gMJxi#*dJK%dtKr8*|RM;LuO!5BK zukGi4Q)A2?GB5c;Z^1$FlH){Jx9HA!*X!uGin-<|%mtv{D19&h3h#GtaemG5R3$5l3DKBTj2*x# zcTPTY`T>wlAPB=1P69DWZhJ5%9OwuTu0w&wHkre(D3Ek@8EYwM^Nnz(y{5IZ@*jak zIplv;h}I=5y=mfgOJz*_v()vcf{pe$ESqNQF9j*FM|aNNSC+tcFb7*G+sC|;oNsln z_sB&JqVH|{D};&n#13ucI15}$GKIxi4sqL0Yqu`)U{Re_BKzxtxE6yE@kjgZvzE35w)l5*amMktZflcWrf!G9}MlE%((Wqqhp^jrS zHWb~>B;WzBiet?)IxHmXLS$K=OHARGS5qp|p#~;?+ILhtR~SdJLNC{Ejp1n(|belg=dos?8yR%%dW&aoWG2~P*{%w9XgR6oLo z6=QKQ0IX-MN=9?SXFbhaY_G84{N=^@JGA2Ic6*M&bu2zH3F4!0W#*jx%^d@tUtYtG`u|7T1yw3A7APi)K0UdaHEA2xE8B_qZ zRWfyoYKQ@nIgd$`?<^AotaYEJv+_?oF9UxWo7psS{|BjrXQU-BRV zs9Sjlf+N}wvuWxkim{>J1xxAb+CAT64cNdngs=y@iw;fXP^AOphtP`c=oBm}MBhq3_w-2P6SC+n-=bK@xIl$8zv=32d9#&k}yuWvaiFFxPh)Lh;_9#PA%550&=;64b@XRun^~q6|O<(pt}x$=`}?m zVX7j3`)a+2_@1^(ve#m8G63oQ&_lE5!*5ujEB$*(VB^xE96>DHyafeaBQ^EtO>lhI zupe?<;~~39YuJK%jrwu_mc*`3%t7h#=un2#F?DQ#5OJ4g65 z;;`@U2o_3bb2@}9Dlo4uRKQ6Io$8e($myQytbu> zPF44*T}DW`kQCX9_qpgU;u{1CfMGk-i+k}5Bgy)wDM$T1k-d|IqMM?plJS~s!9pgd zuGvyf?5fch;Eae9yLI3BUSi7Ln4Y8FhqrN=2giGOn37_Rs0D@v|rJY{sRy+RR zKfxgixrVc)Vu>o|s=tj9he%lc62xTbeaAmW9eu5_ez=6NbNg=>$pgNCzO?OQ zKV2!Ig3GPHU)(nFQrNfFYY{kq4VjPCAD&}P9yt~8=h1b+6NtjXZ>HXdk%uhG{r%$o zV&HK@S=sUa8uC+xOQ)J$Oq`W?f5u$E?$7$Xn@K>q>MCH`=ND2&m{=XSb(MWhqLg3#A^4Apo;|G6+2R^0I!!F1{hlbN% z5BAr;{I!CA{WJhOF=HETT|E!~`siP-_V?d}@PNr$LN9ZQ{x!#cJ^$a=`ycmsfL$Yp zgQb<9|M=)Xjkq-mteXqmHIn`xCi%}-`)k&*=2DLS?dtZQZu^gWEGdHpsFD%yB>hj% z^4C57SngAdSGD;6I8iQlfGrpy8CKr?<39gkAh!3xL>2gcn`Zc*CaSTD5}0KnG|W~0 zf13TWcra1AvL;Ss|IKug&q#i~f&U!CIA7rc=AaDX%=+^s@SoCS^+$zrX0QJjXfC1R literal 0 HcmV?d00001 From fe62145c40864dfee1baedb9d64ca17d7ddc8387 Mon Sep 17 00:00:00 2001 From: Lauren Chambers Date: Mon, 20 May 2019 17:05:41 -0400 Subject: [PATCH 05/14] Maybe adding a tags will help...? --- .../03_aperture_photometry/03_aperture_photometry.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index ef35021a..6dc76b08 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -225,7 +225,7 @@ "source": [ "With `photutils`, users can create apertures with the following shapes:\n", "\n", - "\"Examples\n", + "\"Examples\n", "\n", "Each of these can be defined either in pixel coordinates or in celestial coordinates (using a WCS transformation).\n", "\n", From 4f0b732e76f90eaf3186cd5af5a84c0aea40dee1 Mon Sep 17 00:00:00 2001 From: obviousrebel Date: Tue, 21 May 2019 11:04:42 -0400 Subject: [PATCH 06/14] attempting to fix ci errors with latex --- notebooks/photutils/photutils_notebook_style.mplstyle | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/notebooks/photutils/photutils_notebook_style.mplstyle b/notebooks/photutils/photutils_notebook_style.mplstyle index 2f1f6450..7b9b10d6 100644 --- a/notebooks/photutils/photutils_notebook_style.mplstyle +++ b/notebooks/photutils/photutils_notebook_style.mplstyle @@ -11,7 +11,7 @@ axes.labelsize : 18 xtick.labelsize : 16 ytick.labelsize : 16 -text.usetex : True + figure.subplot.bottom : 0.15 figure.dpi : 200 From dc17e7c78b9fcec1a4824a0733a8ee6185376f50 Mon Sep 17 00:00:00 2001 From: Lauren Chambers Date: Wed, 31 Jul 2019 10:17:48 -0400 Subject: [PATCH 07/14] First round of edits --- .../01_background_estimation.ipynb | 8 +++++--- .../02_source_detection/02_source_detection.ipynb | 8 +++++--- .../03_aperture_photometry/03_aperture_photometry.ipynb | 3 +++ .../photutils/04_psf_photometry/04_psf_photometry.ipynb | 3 +++ 4 files changed, 16 insertions(+), 6 deletions(-) diff --git a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb index 9136a730..5037a5c1 100644 --- a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb +++ b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb @@ -4,6 +4,8 @@ "cell_type": "markdown", "metadata": {}, "source": [ + "
Made possible by the Astropy Project and ScienceBetter Consulting through financial support from the Community Software Initiative at the Space Telescope Science Institute.
\n", + "\n", "\n", "\n", "\n", @@ -136,7 +138,7 @@ "metadata": {}, "source": [ "#### Modifying data\n", - "For the purposes of this notebook example, we're going to add a background effect to this data, but don't worry about this. (Pay no attention to that man behind the curtain!)" + "For the purposes of this notebook example, we're going to add a linear background effect from the top to the bottom of these data. But don't worry about this (pay no attention to that man behind the curtain!)." ] }, { @@ -171,7 +173,7 @@ "outputs": [], "source": [ "from astropy.nddata import CCDData\n", - "unit = u.ct / u.s\n", + "unit = u.electron / u.s\n", "xdf_image = CCDData(modified_data, unit=unit, meta=header)" ] }, @@ -452,7 +454,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "The `Background2D` class allows users to model 2-dimensional backgrounds, by evaluating the background signal in small boxes, and smoothing these boxes to reconstruct a continuous 2D background. The class includes the following arguments/attributes:\n", + "The `Background2D` class allows users to model 2-dimensional backgrounds, by calculating the mean or median in small boxes, and smoothing these boxes to reconstruct a continuous 2D background. The class includes the following arguments/attributes:\n", "* **`box_size`** - the size of the boxes used to calculate the background. This should be larger than individual sources, yet still small enough to encompass changes in the background.\n", "* **`filter_size`** - the size of the median filter used to smooth the final 2D background.\n", "* **`filter_threshold`** - threshold below which the smoothing median filter will not be applied.\n", diff --git a/notebooks/photutils/02_source_detection/02_source_detection.ipynb b/notebooks/photutils/02_source_detection/02_source_detection.ipynb index 2de6f6da..20e658c1 100644 --- a/notebooks/photutils/02_source_detection/02_source_detection.ipynb +++ b/notebooks/photutils/02_source_detection/02_source_detection.ipynb @@ -4,6 +4,8 @@ "cell_type": "markdown", "metadata": {}, "source": [ + "
Made possible by the Astropy Project and ScienceBetter Consulting through financial support from the Community Software Initiative at the Space Telescope Science Institute.
\n", + "\n", "\n", "\n", "\n", @@ -148,7 +150,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons (counts) per second." + "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons per second." ] }, { @@ -157,7 +159,7 @@ "metadata": {}, "outputs": [], "source": [ - "unit = u.electron/ u.s\n", + "unit = u.electron / u.s\n", "xdf_image = CCDData(data, unit=unit, meta=header, mask=mask)" ] }, @@ -452,7 +454,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "### Note: Comparing `DAOStarFinder` and `IRAFStarFinder`" + "## Note: Comparing `DAOStarFinder` and `IRAFStarFinder`" ] }, { diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index 6dc76b08..e430a215 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -4,10 +4,13 @@ "cell_type": "markdown", "metadata": {}, "source": [ + "
Made possible by the Astropy Project and ScienceBetter Consulting through financial support from the Community Software Initiative at the Space Telescope Science Institute.
\n", + "\n", "\n", "\n", "\n", "\n", + "\n", "# Aperture Photometry with `photutils`\n", "---" ] diff --git a/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb b/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb index 7ebdfd1d..3d4c77fa 100644 --- a/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb +++ b/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb @@ -4,10 +4,13 @@ "cell_type": "markdown", "metadata": {}, "source": [ + "
Made possible by the Astropy Project and ScienceBetter Consulting through financial support from the Community Software Initiative at the Space Telescope Science Institute.
\n", + "\n", "\n", "\n", "\n", "\n", + "\n", "# PSF Photometry with `photutils`\n", "---" ] From 852a3f27eac99c6059d9984d6dace3af25010c88 Mon Sep 17 00:00:00 2001 From: Lauren Chambers Date: Wed, 31 Jul 2019 10:22:15 -0400 Subject: [PATCH 08/14] More fixes --- .../01_background_estimation.ipynb | 2 +- .../03_aperture_photometry.ipynb | 16 ++++++++-------- 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb index 5037a5c1..f39076cc 100644 --- a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb +++ b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb @@ -239,7 +239,7 @@ "source": [ "You probably noticed that a large portion of the data is equal to zero. The data we are using is a reduced mosaic that combines many different exposures, and that has been rotated such that not all of the array holds data. \n", "\n", - "We want to **mask** out the non-data portions of the image array,, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data." + "We want to **mask** out the non-data portions of the image array, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data." ] }, { diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index e430a215..3a27a501 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -301,7 +301,7 @@ " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')\n", - "ax1.set_title('find\\_peaks Sources')" + "ax1.set_title('find_peaks Sources')" ] }, { @@ -661,11 +661,11 @@ "source": [ "# The CCDData mask will be automatically applied\n", "phot_table = aperture_photometry(xdf_image, elliptical_apertures[0])\n", - "id = 1\n", + "idx = 1\n", "for aperture in elliptical_apertures[1:]:\n", - " id += 1\n", + " idx += 1\n", " phot_row = aperture_photometry(xdf_image, aperture)[0]\n", - " phot_row[0] = id\n", + " phot_row[0] = idx\n", " phot_table.add_row(phot_row)" ] }, @@ -728,7 +728,7 @@ "plt.xscale('log')\n", "plt.title('Count Rate v. Aperture Area')\n", "plt.xlabel('Aperture Area [pixels$^2$]')\n", - "plt.xlabel(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')))" + "plt.ylabel(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')))" ] }, { @@ -845,11 +845,11 @@ "source": [ "# The CCDData mask will be automatically applied\n", "bkg_phot_table = aperture_photometry(xdf_image, elliptical_annuli[0])\n", - "id = 1\n", + "idx = 1\n", "for aperture in elliptical_annuli[1:]:\n", - " id += 1\n", + " idx += 1\n", " phot_row = aperture_photometry(xdf_image, aperture)[0]\n", - " phot_row[0] = id\n", + " phot_row[0] = idx\n", " bkg_phot_table.add_row(phot_row)" ] }, From b19ad49f10fd2df7f38582e850e6cc9610f0a49c Mon Sep 17 00:00:00 2001 From: Lauren Chambers Date: Wed, 31 Jul 2019 16:11:40 -0400 Subject: [PATCH 09/14] More edits from @mwcraig --- .../01_background_estimation.ipynb | 69 +++++++++++-------- .../02_source_detection.ipynb | 62 ++++++----------- .../03_aperture_photometry.ipynb | 60 +++++----------- .../04_psf_photometry/04_psf_photometry.ipynb | 29 ++------ 4 files changed, 84 insertions(+), 136 deletions(-) diff --git a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb index f39076cc..60d47c41 100644 --- a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb +++ b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb @@ -285,6 +285,15 @@ "ax2.set_title('Masked Data')" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "On the left we have plotted this mask, which has a value of 1 (or True) shown in black where the data is bad, and 0 (or False) shown in white where the data is good. \n", + "\n", + "After the mask is applied to the data - on the right above - the data values \"behind\" the masked values are shown in white." + ] + }, { "cell_type": "markdown", "metadata": {}, @@ -300,6 +309,8 @@ "\n", "By \"scalar estimation\", we mean the calculation of a single value (such as the mean or median) to represent the value of the background for our entire two-dimensional dataset. This is in contrast to a two-dimensional background, where the estimated background is represented as an array of values that can vary spatially with the dataset. We will calculate a 2D background in the upcoming section.\n", "\n", + "### Calculate scalar background value\n", + "\n", "Here we will calculate the mean, median, and mode using sigma clipping. With sigma clipping, the data is iteratively clipped to exclude data points outside of a certain sigma (standard deviation), thus removing some of the noise from the data before determining statistical values." ] }, @@ -309,7 +320,11 @@ "metadata": {}, "outputs": [], "source": [ - "mean, median, std = sigma_clipped_stats(xdf_image.data, sigma=3.0, maxiters=5, mask=xdf_image.mask)" + "# Calculate statistics with masking\n", + "mean, median, std = sigma_clipped_stats(xdf_image.data, sigma=3.0, maxiters=5, mask=xdf_image.mask)\n", + "\n", + "# Calculate statistics without masking\n", + "stats_nomask = sigma_clipped_stats(xdf_image.data, sigma=3.0, maxiters=5)" ] }, { @@ -319,16 +334,6 @@ "But what difference does this sigma clipping make? And how important is masking, anyway? Let's visualize these statistics to get an idea:" ] }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Calculate the data without masking\n", - "stats_nomask = sigma_clipped_stats(xdf_image.data, sigma=3.0, maxiters=5)" - ] - }, { "cell_type": "code", "execution_count": null, @@ -344,36 +349,38 @@ "ax2.hist(xdf_image[~xdf_image.mask], bins=100, range=flux_range)\n", "\n", "# Plot lines for each kind of mean\n", - "ax1.axvline(mean, label='Masked \\& Clipped', c='C1', lw=3)\n", - "ax1.axvline(np.average(xdf_image[~xdf_image.mask]), label='Masked', c='C2', ls='--', lw=3)\n", - "ax1.axvline(stats_nomask[0], label='Clipped', c='C3', ls='-.', lw=3)\n", - "ax1.axvline(np.average(xdf_image), label='Neither', c='C6', ls=':', lw=3)\n", + "ax1.axvline(mean, label='Masked & Clipped', c='C1', ls='-.', lw=3)\n", + "ax1.axvline(np.average(xdf_image[~xdf_image.mask]), label='Masked', c='C2', lw=3)\n", + "ax1.axvline(stats_nomask[0], label='Clipped', c='C3', ls=':', lw=3)\n", + "ax1.axvline(np.average(xdf_image), label='Neither', c='C5', ls='--', lw=3)\n", "\n", "ax1.set_xlim(flux_range)\n", "ax1.set_xlabel(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), fontsize=14)\n", "ax1.set_ylabel('Frequency', fontsize=14)\n", - "ax1.set_title('Effect of Sigma-Clipping and Masking on Mean', fontsize=16)\n", - "ax1.legend(fontsize=11)\n", - "\n", + "ax1.set_title('Effect of Sigma-Clipping \\n and Masking on Mean', fontsize=16)\n", "\n", "# Plot lines for each kind of median\n", "# Note: use np.ma.median rather than np.median for masked arrays\n", - "ax2.axvline(median, label='Masked \\& Clipped', c='C1', lw=3)\n", - "ax2.axvline(np.ma.median(xdf_image[~xdf_image.mask]), label='Masked', c='C2', ls='--', lw=3)\n", - "ax2.axvline(stats_nomask[1], label='Clipped', c='C3', ls='-.', lw=3)\n", - "ax2.axvline(np.ma.median(xdf_image), label='Neither', c='C6', ls=':', lw=3)\n", + "ax2.axvline(median, label='Masked & Clipped', c='C1', ls='-.', lw=3)\n", + "ax2.axvline(np.ma.median(xdf_image[~xdf_image.mask]), label='Masked', c='C2', lw=3)\n", + "ax2.axvline(stats_nomask[1], label='Clipped', c='C3', ls=':', lw=3)\n", + "ax2.axvline(np.ma.median(xdf_image), label='Neither', c='C5', ls='--', lw=3)\n", "\n", "ax2.set_xlim(flux_range)\n", "ax2.set_xlabel(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), fontsize=14)\n", - "ax2.set_title('Effect of Sigma-Clipping and Masking on Median', fontsize=16)\n", - "ax2.legend(fontsize=11)" + "ax2.set_title('Effect of Sigma-Clipping \\n and Masking on Median', fontsize=16)\n", + "\n", + "# Add legend\n", + "ax1.legend(fontsize=11, loc='lower center', bbox_to_anchor=(1.1, -0.45), ncol=2, handlelength=6)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "Just from simply looking at the distribution of the data, it is pretty easy to see how sigma-clipping and masking improve the calculation of the mean and median.\n", + "Just from simply looking at the distribution of the data, it is pretty easy to see how sigma-clipping and masking improve the calculation of the mean and median: the masked & sigma-clipped values are closest to the center of the distribution in both cases. It's also worthwhile to note that the median does a better job even without masking or clipping!\n", + "\n", + "### Subtract scalar background value\n", "\n", "But enough looking at numbers, let's actually remove the background from the data. By using the `subtract()` method of the `CCDData` class, we can subtract the mean background while maintaining the metadata and mask of our original CCDData object:" ] @@ -385,7 +392,7 @@ "outputs": [], "source": [ "# Calculate the scalar background subtraction, maintaining metadata, unit, and mask\n", - "xdf_scalar_bkgdsub = xdf_image.subtract(mean * u.ct / u.s)" + "xdf_scalar_bkgdsub = xdf_image.subtract(mean * u.electron / u.s)" ] }, { @@ -495,13 +502,15 @@ { "cell_type": "code", "execution_count": null, - "metadata": {}, + "metadata": { + "scrolled": false + }, "outputs": [], "source": [ "# Set up the figure with subplots\n", "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", "\n", - "# Plot the data\n", + "# Plot the background\n", "background_clipped = np.clip(bkg.background, 1e-4, None) # clip to plot with logarithmic stretch\n", "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, background_clipped), norm=norm_image)\n", "\n", @@ -525,6 +534,8 @@ "cell_type": "markdown", "metadata": {}, "source": [ + "You might notice that not all areas of the background array have mesh boxes over them (look for those boxes that do not have a `+`). If you compare this background array with the original data, you'll see that these un-boxed areas contain particularly bright sources, and thus are not being included in the background estimate .\n", + "\n", "And how does the data look if we use this background subtraction method (again maintaining the attributes of the CCDData object)?" ] }, @@ -535,7 +546,7 @@ "outputs": [], "source": [ "# Calculate the 2D background subtraction, maintaining metadata, unit, and mask\n", - "xdf_2d_bkgdsub = xdf_image.subtract(bkg.background * u.ct / u.s)" + "xdf_2d_bkgdsub = xdf_image.subtract(bkg.background * u.electron / u.s)" ] }, { diff --git a/notebooks/photutils/02_source_detection/02_source_detection.ipynb b/notebooks/photutils/02_source_detection/02_source_detection.ipynb index 20e658c1..ee395477 100644 --- a/notebooks/photutils/02_source_detection/02_source_detection.ipynb +++ b/notebooks/photutils/02_source_detection/02_source_detection.ipynb @@ -113,7 +113,9 @@ "source": [ "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", "\n", - "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)" + "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)\n", + "\n", + "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons per second." ] }, { @@ -123,10 +125,7 @@ "outputs": [], "source": [ "url = 'https://archive.stsci.edu/pub/hlsp/xdf/hlsp_xdf_hst_acswfc-60mas_hudf_f435w_v1_sci.fits'\n", - "with fits.open(url) as hdulist:\n", - " hdulist.info()\n", - " data = hdulist[0].data\n", - " header = hdulist[0].header" + "xdf_image = CCDData.read(url, unit=u.electron / u.s)" ] }, { @@ -143,24 +142,8 @@ "outputs": [], "source": [ "# Define the mask\n", - "mask = data == 0" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons per second." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "unit = u.electron / u.s\n", - "xdf_image = CCDData(data, unit=unit, meta=header, mask=mask)" + "mask = xdf_image.data == 0\n", + "xdf_image.mask = mask" ] }, { @@ -267,8 +250,8 @@ "metadata": {}, "outputs": [], "source": [ - "daofind = DAOStarFinder(fwhm=5.0, threshold=20.*std)\n", - "sources_dao = daofind(xdf_image * ~xdf_image.mask) \n", + "daofind = DAOStarFinder(fwhm=5.0, threshold=20. * std)\n", + "sources_dao = daofind(np.ma.masked_where(xdf_image.mask, xdf_image)) \n", "print(sources_dao)" ] }, @@ -376,8 +359,8 @@ "metadata": {}, "outputs": [], "source": [ - "iraffind = IRAFStarFinder(fwhm=5.0, threshold=20.*std)\n", - "sources_iraf = iraffind(xdf_image * ~xdf_image.mask) \n", + "iraffind = IRAFStarFinder(fwhm=5.0, threshold=20. * std)\n", + "sources_iraf = iraffind(np.ma.masked_where(xdf_image.mask, xdf_image)) \n", "print(sources_iraf)" ] }, @@ -484,11 +467,11 @@ "metadata": {}, "outputs": [], "source": [ - "iraffind_match = IRAFStarFinder(fwhm=5.0, threshold=20.*std, \n", + "iraffind_match = IRAFStarFinder(fwhm=5.0, threshold=20. * std, \n", " sharplo=0.2, sharphi=1.0, \n", " roundlo=-1.0, roundhi=1.0,\n", " minsep_fwhm=0.0)\n", - "sources_iraf_match = iraffind_match(xdf_image * ~xdf_image.mask) \n", + "sources_iraf_match = iraffind_match(np.ma.masked_where(xdf_image.mask, xdf_image)) \n", "print(sources_iraf_match)" ] }, @@ -579,7 +562,7 @@ "outputs": [], "source": [ "sources_findpeaks = find_peaks(xdf_image.data, mask=xdf_image.mask, \n", - " threshold=20.*std, box_size=30, \n", + " threshold=20. * std, box_size=30, \n", " centroid_func=centroid_2dg) \n", "print(sources_findpeaks)" ] @@ -609,7 +592,7 @@ " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')\n", - "ax1.set_title('find\\_peaks Sources')" + "ax1.set_title('find_peaks Sources')" ] }, { @@ -724,26 +707,21 @@ "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", "ax1.scatter([x for x, y in list(dao_findpeaks_match)], [y for x, y in list(dao_findpeaks_match)],\n", " s=30, marker='s', lw=1, facecolor='None', edgecolor='#EE7733',\n", - " label='Found by DAO \\& find\\_peaks')\n", + " label='Found by DAO & find_peaks')\n", "ax1.scatter([x for x, y in list(dao_iraf_match)], [y for x, y in list(dao_iraf_match)],\n", " s=30, marker='D', lw=1, facecolor='None', edgecolor='#EE3377',\n", - " label='Found by DAO \\& IRAF')\n", + " label='Found by DAO & IRAF')\n", "ax1.scatter([x for x, y in list(iraf_findpeaks_match)], [y for x, y in list(iraf_findpeaks_match)],\n", " s=30, marker='o', lw=1, facecolor='None', edgecolor='#0077BB',\n", - " label='Found by IRAF \\& find\\_peaks')\n", + " label='Found by IRAF & find_peaks')\n", "ax1.scatter([x for x, y in list(all_match)], [y for x, y in list(all_match)],\n", " s=30, marker='o', lw=1.2, linestyle=':',facecolor='None', edgecolor='#BBBBBB',\n", " label='Found by all methods')\n", - "ax1.legend(ncol=2)\n", "\n", - "# Define the colorbar\n", - "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", - "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", - "cbar.ax.set_yticklabels(labels)\n", + "# Add legend\n", + "ax1.legend(ncol=2, loc='lower center', bbox_to_anchor=(0.5, -0.25))\n", "\n", "# Define labels\n", - "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", - " rotation=270, labelpad=30)\n", "ax1.set_xlabel('X (pixels)')\n", "ax1.set_ylabel('Y (pixels)')\n", "ax1.set_title('Sources Found by Different Methods')" @@ -976,7 +954,7 @@ "outputs": [], "source": [ "# Define the approximate isophotal extent\n", - "r = 4.\n", + "r = 4. # pixels\n", "\n", "# Create the apertures\n", "apertures = []\n", diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index 3a27a501..3f585cce 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -115,7 +115,9 @@ "source": [ "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", "\n", - "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)" + "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)\n", + "\n", + "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons per second." ] }, { @@ -125,10 +127,7 @@ "outputs": [], "source": [ "url = 'https://archive.stsci.edu/pub/hlsp/xdf/hlsp_xdf_hst_acswfc-60mas_hudf_f435w_v1_sci.fits'\n", - "with fits.open(url) as hdulist:\n", - " hdulist.info()\n", - " data = hdulist[0].data\n", - " header = hdulist[0].header" + "xdf_image = CCDData.read(url, unit=u.electron / u.s)" ] }, { @@ -145,24 +144,8 @@ "outputs": [], "source": [ "# Define the mask\n", - "mask = data == 0" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons (counts) per second." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "unit = u.electron / u.s\n", - "xdf_image = CCDData(data, unit=unit, meta=header, mask=mask)" + "mask = xdf_image.data == 0\n", + "xdf_image.mask = mask" ] }, { @@ -308,7 +291,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "So thanks to `find_peaks`, we now we know the positions of all our sources. Next, we need to define apertures for each source. First, as the simplest example, let's try using circular apertures of a fixed size.\n", + "So thanks to `find_peaks`, we now we know the positions of all our sources. Next, we need to define apertures for each source. First, as the simplest example, let's try using circular apertures of a fixed radius: 10 pixels.\n", "\n", "### Circular Apertures" ] @@ -328,9 +311,9 @@ "metadata": {}, "outputs": [], "source": [ - "# define the aperture\n", + "# Define the aperture\n", "position = (sources_findpeaks['x_centroid'], sources_findpeaks['y_centroid'])\n", - "radius = 10.\n", + "radius = 10. # pixels\n", "circular_aperture = CircularAperture(position, r=radius)" ] }, @@ -430,7 +413,7 @@ "metadata": {}, "outputs": [], "source": [ - "r = 3. # approximate isophotal extent of semimajor axis\n", + "r = 3. # pixels; approximate isophotal extent of semimajor axis\n", "\n", "# Create the apertures\n", "elliptical_apertures = []\n", @@ -490,16 +473,9 @@ "\n", "At the moment, the positions of our apertures are in pixels, relative to our data array. However, if you need aperture positions in terms of celestial coordinates, `photutils` also includes aperture objects that can be integrated with Astropy's `SkyCoords`.\n", "\n", - "Fortunately this is extremely easy when we use the [World Coordinate System (WCS)](http://docs.astropy.org/en/stable/wcs/) to produce a WCS object from the header of the FITS file containing our image, and then the `to_sky()` method to transform our `EllipticalAperture` objects into `SkyEllipticalAperture` objects." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "from astropy.wcs import WCS" + "Fortunately this is extremely easy when we use the [World Coordinate System (WCS)](http://docs.astropy.org/en/stable/wcs/) to decipher a WCS object from the header of the FITS file containing our image, and then the `to_sky()` method to transform our `EllipticalAperture` objects into `SkyEllipticalAperture` objects. \n", + "\n", + "And, better yet, our `CCDData` object has held onto its WCS information ever since we created it!" ] }, { @@ -508,7 +484,7 @@ "metadata": {}, "outputs": [], "source": [ - "wcs = WCS(header)\n", + "wcs = xdf_image.wcs\n", "sky_elliptical_apertures = [ap.to_sky(wcs) for ap in elliptical_apertures]" ] }, @@ -528,7 +504,7 @@ "from photutils import SkyEllipticalAperture\n", "from astropy.coordinates import SkyCoord\n", "\n", - "r = 3. # approximate isophotal extent of semimajor axis\n", + "r = 3. # pixels; approximate isophotal extent of semimajor axis\n", "\n", "# Create the apertures\n", "sky_elliptical_apertures = []\n", @@ -645,8 +621,8 @@ "Woohoo!\n", "\n", "Unfortunately for our purposes, the `aperture_photometry` function can be only used alone for one of the two cases:\n", - "* Identical apertures at distinct positions (e.g. circular apertures with `r = 3` for many sources)\n", - "* Distinct apertures at identical positions (e.g. two circular apertures with `r = 3` and `r = 5` for one source)\n", + "* Identical apertures at distinct positions (e.g. circular apertures with `r = 3` pixels for many sources)\n", + "* Distinct apertures at identical positions (e.g. two circular apertures with `r = 3` pixels and `r = 5` pixels for one source)\n", "\n", "Since our elliptical apertures are distinct apertures at distinct positions, we need to do a little more work to get a single table of photometric values.\n", "\n", @@ -762,7 +738,7 @@ "\n", "In the [background estimation notebook](../01_background_estimation/01_background_estimation.ipynb), we explored how to perform global background subtraction of image data with `photutils`. However, you can also use `photutils` to perform local background estimations for aperture corrections.\n", "\n", - "To estimate the local background for each aperture, measure the counts within annulus apertures around (but not including!) each source. In our example, we defined elliptical apertures with `r = 3` to measure the counts within each source. To calculate the background for each source, let's measure the counts elliptical annuli between `r = 3.5` and `r = 5`." + "To estimate the local background for each aperture, measure the counts within annulus apertures around (but not including!) each source. In our example, we defined elliptical apertures with `r = 3` pixels to measure the counts within each source. To calculate the background for each source, let's measure the counts elliptical annuli between `r = 3.5` pixels and `r = 5` pixels." ] }, { diff --git a/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb b/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb index 3d4c77fa..1a6687f3 100644 --- a/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb +++ b/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb @@ -119,7 +119,9 @@ "source": [ "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", "\n", - "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)" + "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)\n", + "\n", + "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons per second." ] }, { @@ -129,10 +131,7 @@ "outputs": [], "source": [ "url = 'https://archive.stsci.edu/pub/hlsp/xdf/hlsp_xdf_hst_acswfc-60mas_hudf_f435w_v1_sci.fits'\n", - "with fits.open(url) as hdulist:\n", - " hdulist.info()\n", - " data = hdulist[0].data\n", - " header = hdulist[0].header" + "xdf_image = CCDData.read(url, unit=u.electron / u.s)" ] }, { @@ -149,24 +148,8 @@ "outputs": [], "source": [ "# Define the mask\n", - "mask = data == 0" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons (counts) per second." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "unit = u.electron / u.s\n", - "xdf_image = CCDData(data, unit=unit, meta=header, mask=mask)" + "mask = xdf_image.data == 0\n", + "xdf_image.mask = mask" ] }, { From 01b488feed6f6fae3309a5d9af165eb4a871af3e Mon Sep 17 00:00:00 2001 From: Matt Craig Date: Sun, 5 Jan 2020 15:11:34 -0600 Subject: [PATCH 10/14] Address review comments from @onoddil --- .../01_background_estimation.ipynb | 28 +++++------ .../02_source_detection.ipynb | 24 ++++----- .../03_aperture_photometry.ipynb | 50 +++++++++---------- 3 files changed, 51 insertions(+), 51 deletions(-) diff --git a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb index 60d47c41..eb8ca774 100644 --- a/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb +++ b/notebooks/photutils/01_background_estimation/01_background_estimation.ipynb @@ -21,7 +21,7 @@ "##### What is background estimation?\n", "In order to most accurately do photometric analysis of celestial sources in image data, it is important to estimate and subtract the image background. Any astronomical image will have background noise, due to both detector effects and background emission from the night sky. This noise can be modeled as uniform, or as varying with position on the detector. \n", "\n", - "The `photutils` package provides tools for estimating 2-dimensional background noise, which can then be subtracted from an image to ensure the most accurate photometry possible.\n", + "The `photutils` package provides tools for estimating 2-dimensional background flux, which can then be subtracted from an image to ensure the most accurate photometry possible.\n", "\n", "##### What does this tutorial include?\n", "This tutorial covers the basics of background estimation and subtraction, including the following methods:\n", @@ -163,7 +163,7 @@ "source": [ "#### Data representation\n", "\n", - "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons (counts) per second." + "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, each image has units electrons (counts) per second." ] }, { @@ -291,7 +291,7 @@ "source": [ "On the left we have plotted this mask, which has a value of 1 (or True) shown in black where the data is bad, and 0 (or False) shown in white where the data is good. \n", "\n", - "After the mask is applied to the data - on the right above - the data values \"behind\" the masked values are shown in white." + "After the mask is applied to the data (on the right above) the data values \"behind\" the masked values are shown in white." ] }, { @@ -311,7 +311,7 @@ "\n", "### Calculate scalar background value\n", "\n", - "Here we will calculate the mean, median, and mode using sigma clipping. With sigma clipping, the data is iteratively clipped to exclude data points outside of a certain sigma (standard deviation), thus removing some of the noise from the data before determining statistical values." + "Here we will calculate the mean, median, and mode of the dataset using sigma clipping. With sigma clipping, the data is iteratively clipped to exclude data points outside of a certain sigma (standard deviation), thus removing some of the noise from the data before determining statistical values." ] }, { @@ -349,7 +349,7 @@ "ax2.hist(xdf_image[~xdf_image.mask], bins=100, range=flux_range)\n", "\n", "# Plot lines for each kind of mean\n", - "ax1.axvline(mean, label='Masked & Clipped', c='C1', ls='-.', lw=3)\n", + "ax1.axvline(mean, label='Masked and Clipped', c='C1', ls='-.', lw=3)\n", "ax1.axvline(np.average(xdf_image[~xdf_image.mask]), label='Masked', c='C2', lw=3)\n", "ax1.axvline(stats_nomask[0], label='Clipped', c='C3', ls=':', lw=3)\n", "ax1.axvline(np.average(xdf_image), label='Neither', c='C5', ls='--', lw=3)\n", @@ -361,7 +361,7 @@ "\n", "# Plot lines for each kind of median\n", "# Note: use np.ma.median rather than np.median for masked arrays\n", - "ax2.axvline(median, label='Masked & Clipped', c='C1', ls='-.', lw=3)\n", + "ax2.axvline(median, label='Masked and Clipped', c='C1', ls='-.', lw=3)\n", "ax2.axvline(np.ma.median(xdf_image[~xdf_image.mask]), label='Masked', c='C2', lw=3)\n", "ax2.axvline(stats_nomask[1], label='Clipped', c='C3', ls=':', lw=3)\n", "ax2.axvline(np.ma.median(xdf_image), label='Neither', c='C5', ls='--', lw=3)\n", @@ -430,7 +430,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Note that both plots above use the same normalization scheme, represented by the colorbar at right. That is to say, if two pixels have the same color in both arrays, they have the same value.\n", + "Note that both plots above use the same normalization scheme, represented by the colorbar on the right. That is to say, if two pixels have the same color in both arrays, they have the same value.\n", "\n", "That looks better! You can tell that the background is darker, especially in the top corner. However, the background still does not seem to be completely removed. In this case, the background varies spatially; it is two-dimensional. Thankfully, `photutils` includes functions to remove background like this." ] @@ -462,11 +462,11 @@ "metadata": {}, "source": [ "The `Background2D` class allows users to model 2-dimensional backgrounds, by calculating the mean or median in small boxes, and smoothing these boxes to reconstruct a continuous 2D background. The class includes the following arguments/attributes:\n", - "* **`box_size`** - the size of the boxes used to calculate the background. This should be larger than individual sources, yet still small enough to encompass changes in the background.\n", - "* **`filter_size`** - the size of the median filter used to smooth the final 2D background.\n", - "* **`filter_threshold`** - threshold below which the smoothing median filter will not be applied.\n", - "* **`sigma_clip`** - an ` astropy.stats.SigmaClip` object that is used to specify the sigma and number of iterations used to sigma-clip the data before background calculations are performed.\n", - "* **`bkg_estimator`** - the method used to perform the background calculation in each box (mean, median, SExtractor algorithm, etc.).\n", + "* **`box_size`** — the size of the boxes used to calculate the background. This should be larger than individual sources, yet still small enough to encompass changes in the background.\n", + "* **`filter_size`** — the size of the median filter used to smooth the final 2D background.\n", + "* **`filter_threshold`** — threshold below which the smoothing median filter will not be applied.\n", + "* **`sigma_clip`** — an ` astropy.stats.SigmaClip` object that is used to specify the sigma and number of iterations used to sigma-clip the data before background calculations are performed.\n", + "* **`bkg_estimator`** — the method used to perform the background calculation in each box (mean, median, SExtractor algorithm, etc.).\n", "\n", "For this example, we will use the `MeanBackground` estimator." ] @@ -486,7 +486,7 @@ "metadata": {}, "outputs": [], "source": [ - "sigma_clip = SigmaClip(sigma=3., iters=5)\n", + "sigma_clip = SigmaClip(sigma=3., maxiters=5)\n", "bkg_estimator = MeanBackground()\n", "bkg = Background2D(xdf_image, box_size=200, filter_size=(10, 10), mask=xdf_image.mask,\n", " sigma_clip=sigma_clip, bkg_estimator=bkg_estimator)" @@ -664,7 +664,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.6.8" + "version": "3.7.5" } }, "nbformat": 4, diff --git a/notebooks/photutils/02_source_detection/02_source_detection.ipynb b/notebooks/photutils/02_source_detection/02_source_detection.ipynb index ee395477..636ecbc7 100644 --- a/notebooks/photutils/02_source_detection/02_source_detection.ipynb +++ b/notebooks/photutils/02_source_detection/02_source_detection.ipynb @@ -444,7 +444,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "You might have noticed that the `IRAFStarFinder` algorithm only found 211 sources in our data - 14% of what `DAOStarFinder` found. Why is this?\n", + "You might have noticed that the `IRAFStarFinder` algorithm only found 211 sources in our data – 14% of what `DAOStarFinder` found. Why is this?\n", "\n", "The answer comes down to the default settings for the two algorithms: (1) there are differences in the upper and lower bounds on the requirements for source roundness and sharpness, and (2) `IRAFStarFinder` includes a minimum separation between sources that `DAOStarFinder` does not have:\n", "\n", @@ -479,7 +479,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "The number of detected sources are in much better agreement now - 1415 versus 1470 - but the improved agreement can also be seen by plotting the location of these sources:" + "The number of detected sources are in much better agreement now – 1415 versus 1470 – but the improved agreement can also be seen by plotting the location of these sources:" ] }, { @@ -651,7 +651,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Next, how many of these sources match? We can answer this question by using [sets](https://docs.python.org/3/library/stdtypes.html#set-types-set-frozenset) to compare the centroids of the different sources (rounding to the first decimal place)." + "Next, how many of these sources match? We can answer this question by using [sets](https://docs.python.org/3/library/stdtypes.html#set-types-set-frozenset) to compare the centroids of the different sources (rounding to the nearest integer)." ] }, { @@ -707,13 +707,13 @@ "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), norm=norm_image)\n", "ax1.scatter([x for x, y in list(dao_findpeaks_match)], [y for x, y in list(dao_findpeaks_match)],\n", " s=30, marker='s', lw=1, facecolor='None', edgecolor='#EE7733',\n", - " label='Found by DAO & find_peaks')\n", + " label='Found by DAO and find_peaks')\n", "ax1.scatter([x for x, y in list(dao_iraf_match)], [y for x, y in list(dao_iraf_match)],\n", " s=30, marker='D', lw=1, facecolor='None', edgecolor='#EE3377',\n", - " label='Found by DAO & IRAF')\n", + " label='Found by DAO and IRAF')\n", "ax1.scatter([x for x, y in list(iraf_findpeaks_match)], [y for x, y in list(iraf_findpeaks_match)],\n", " s=30, marker='o', lw=1, facecolor='None', edgecolor='#0077BB',\n", - " label='Found by IRAF & find_peaks')\n", + " label='Found by IRAF and find_peaks')\n", "ax1.scatter([x for x, y in list(all_match)], [y for x, y in list(all_match)],\n", " s=30, marker='o', lw=1.2, linestyle=':',facecolor='None', edgecolor='#BBBBBB',\n", " label='Found by all methods')\n", @@ -742,7 +742,7 @@ "\n", "If none of the algorithms we've reviewed above do exactly what you need, `photutils` also provides infrastructure for you to generate and use your own source detection algorithm: the `StarFinderBase` object can be inherited and used to develop new star-finding classes. Take a look at the [documentation](https://photutils.readthedocs.io/en/latest/api/photutils.detection.StarFinderBase.html#photutils.detection.StarFinderBase) for more information.\n", "\n", - "If you do go that route, remember that `photutils` is open-developed; you would be very welcome to [open a pull request](https://github.com/astropy/photutils/blob/master/CONTRIBUTING.rst) and incorporate your new star finder into the `photutils` source code - for everyone to use!" + "If you do go that route, remember that `photutils` is open-developed; you would be very welcome to [open a pull request](https://github.com/astropy/photutils/blob/master/CONTRIBUTING.rst) and incorporate your new star finder into the `photutils` source code – for everyone to use!" ] }, { @@ -791,7 +791,7 @@ "outputs": [], "source": [ "from photutils import detect_sources\n", - "from photutils.utils import random_cmap" + "from photutils.utils import make_random_cmap" ] }, { @@ -827,7 +827,7 @@ "ax1.set_title('Original Data')\n", "\n", "# Plot the segmentation image\n", - "rand_cmap = random_cmap(random_state=12345)\n", + "rand_cmap = make_random_cmap(random_state=12345)\n", "rand_cmap.set_under('black')\n", "segplot = ax2.imshow(np.ma.masked_where(xdf_image.mask, segm), vmin=1, cmap=rand_cmap)\n", "ax2.set_xlabel('X (pixels)')\n", @@ -981,7 +981,7 @@ "\n", "# Plot the apertures\n", "for aperture in apertures:\n", - " aperture.plot(color='red', lw=1, alpha=0.7, ax=ax1)\n", + " aperture.plot(color='red', lw=1, alpha=0.7, axes=ax1)\n", "\n", "# Define the colorbar\n", "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", @@ -1013,7 +1013,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "It is clear that using `photutils` for image segmentation can allow users to generate highly customized apertures - great for complex data that contain many different kinds of celestial sources." + "It is clear that using `photutils` for image segmentation can allow users to generate highly customized apertures – great for complex data that contain many different kinds of celestial sources." ] }, { @@ -1072,7 +1072,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.6.8" + "version": "3.7.5" } }, "nbformat": 4, diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index 3f585cce..6318ab4f 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -421,7 +421,7 @@ " position = (obj.xcentroid.value, obj.ycentroid.value)\n", " a = obj.semimajor_axis_sigma.value * r\n", " b = obj.semiminor_axis_sigma.value * r\n", - " theta = obj.orientation.value\n", + " theta = obj.orientation.to(u.rad).value\n", " elliptical_apertures.append(EllipticalAperture(position, a, b, theta=theta))" ] }, @@ -439,7 +439,7 @@ "\n", "# Plot the apertures\n", "for aperture in elliptical_apertures:\n", - " aperture.plot(color='red', alpha=0.7, ax=ax1)\n", + " aperture.plot(color='red', alpha=0.7, axes=ax1)\n", "\n", "# Define the colorbar\n", "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", @@ -473,7 +473,7 @@ "\n", "At the moment, the positions of our apertures are in pixels, relative to our data array. However, if you need aperture positions in terms of celestial coordinates, `photutils` also includes aperture objects that can be integrated with Astropy's `SkyCoords`.\n", "\n", - "Fortunately this is extremely easy when we use the [World Coordinate System (WCS)](http://docs.astropy.org/en/stable/wcs/) to decipher a WCS object from the header of the FITS file containing our image, and then the `to_sky()` method to transform our `EllipticalAperture` objects into `SkyEllipticalAperture` objects. \n", + "Fortunately this is extremely easy when we use the [World Coordinate System (WCS)](http://docs.astropy.org/en/stable/wcs/) to decipher a WCS object from the header of the FITS file containing our image, and then use the `to_sky()` method to transform our `EllipticalAperture` objects into `SkyEllipticalAperture` objects. \n", "\n", "And, better yet, our `CCDData` object has held onto its WCS information ever since we created it!" ] @@ -520,7 +520,7 @@ " theta = obj.orientation.value * u.rad\n", " \n", " # Convert the theta from radians from X axis to the radians from North \n", - " x_to_north_angle = (90. + header['ORIENTAT']) * u.deg\n", + " x_to_north_angle = (90. + xdf_image.header['ORIENTAT']) * u.deg\n", " x_to_north_angle_rad = x_to_north_angle.to_value(u.rad) * u.rad\n", " theta -= x_to_north_angle_rad\n", " \n", @@ -550,20 +550,20 @@ "source": [ "Now that we have aperture objects that fit our data reasonably well, we can finally perform photometry with the `aperture_photometry` function. This function takes the following arguments:\n", "\n", - "* **`data`** - the background-subtracted data array on which to perform photometry.\n", - "* **`apertures`** - an aperture object containing the aperture(s) to use for the photometry.\n", - "* **`error`** (optional) - an array of values that represent the pixel-wise Gaussian 1-sigma errors of the input data.\n", - "* **`mask`** (optional) - a mask for the `data` to exclude certain pixels from calculations.\n", - "* **`method`** (optional) - how to place the aperture(s) onto the pixel grid (see below).\n", - "* **`unit`** (optional) - unit of `data` and `error`.\n", - "* **`wcs`** (optional) - the WCS transformation to use if `apertures` is a `SkyAperture` object. \n", - "* **`subpixels`** (optional) - the factor by which pixels are resampled (see below).\n", + "* **`data`** – the background-subtracted data array on which to perform photometry.\n", + "* **`apertures`** – an aperture object containing the aperture(s) to use for the photometry.\n", + "* **`error`** (optional) – an array of values that represent the pixel-wise Gaussian 1-sigma errors of the input data.\n", + "* **`mask`** (optional) – a mask for the `data` to exclude certain pixels from calculations.\n", + "* **`method`** (optional) – how to place the aperture(s) onto the pixel grid (see below).\n", + "* **`unit`** (optional) – unit of `data` and `error`.\n", + "* **`wcs`** (optional) – the WCS transformation to use if `apertures` is a `SkyAperture` object. \n", + "* **`subpixels`** (optional) – the factor by which pixels are resampled (see below).\n", "\n", "The following methods are the options for how to place apertures onto the data pixel grid:\n", "\n", - "* **exact** (default) - calculate the exact fractional overlap of each aperture for each overlapping pixel. This method is the most accurate, but will also take the longest. \n", - "* **center** - a pixel is either entirely in or entirely out of the aperture, depending on whether the pixel center is inside or outside of the aperture.\n", - "* **subpixel** - a pixel is divided into `subpixels` x `subpixels` subpixels, each of which are considered to be entirely in or out of the aperture depending on whether its center is in or out of the aperture. " + "* **exact** (default) – calculate the exact fractional overlap of each aperture for each overlapping pixel. This method is the most accurate, but will also take the longest. \n", + "* **center** – a pixel is either entirely in or entirely out of the aperture, depending on whether the pixel center is inside or outside of the aperture.\n", + "* **subpixel** – a pixel is divided into `subpixels` x `subpixels` subpixels, each of which are considered to be entirely in or out of the aperture depending on whether its center is in or out of the aperture. " ] }, { @@ -715,7 +715,7 @@ "\n", "

Exercise:


\n", "\n", - "Re-calculate the photometry for these elliptical apertures - or just a subset of them - using the subpixel aperture placement method instead of the default exact method. How does this affect the count sum calculated for those apertures?\n", + "Re-calculate the photometry for these elliptical apertures – or just a subset of them – using the subpixel aperture placement method instead of the default exact method. How does this affect the count sum calculated for those apertures?\n", "\n", "" ] @@ -738,7 +738,7 @@ "\n", "In the [background estimation notebook](../01_background_estimation/01_background_estimation.ipynb), we explored how to perform global background subtraction of image data with `photutils`. However, you can also use `photutils` to perform local background estimations for aperture corrections.\n", "\n", - "To estimate the local background for each aperture, measure the counts within annulus apertures around (but not including!) each source. In our example, we defined elliptical apertures with `r = 3` pixels to measure the counts within each source. To calculate the background for each source, let's measure the counts elliptical annuli between `r = 3.5` pixels and `r = 5` pixels." + "To estimate the local background for each aperture, measure the counts within annulus apertures around (but not including!) each source. Since our sources vary quite a bit in size, we should define apertures that are scaled by the size of the source. In our example, we defined elliptical apertures with `r = 3` pixels to measure the counts within each source. To calculate the background for each source, let's measure the counts elliptical annuli between `r = 3.5` pixels and `r = 5` pixels." ] }, { @@ -776,9 +776,9 @@ "\n", "# Plot the apertures\n", "for aperture in elliptical_annuli:\n", - " aperture.plot(color='red', alpha=0.4, ax=ax1, fill=True)\n", + " aperture.plot(color='red', alpha=0.4, axes=ax1, fill=True)\n", "for aperture in elliptical_apertures:\n", - " aperture.plot(color='white', alpha=0.7, ax=ax1)\n", + " aperture.plot(color='white', alpha=0.7, axes=ax1)\n", "\n", "# Define the colorbar\n", "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", @@ -805,9 +805,9 @@ "\n", "Now that our apertures have been defined, we can do photometry with them to estimate and account for the background. The aperture correction is calculated by:\n", "- Calculating the count rate within each annulus using `aperture_photometry`\n", - "- Dividing each annulus' count rate by each annulus' area to get the mean background value for each annulus\n", + "- Dividing the count rate of each annulus by the area of each annulus to get the mean background value for each annulus\n", "- Taking the mean of those annulus means to get a mean background value for the entire image\n", - "- Multiplying the global background mean value times the area of each elliptical photometric aperture, to get the estimated background count rate within each aperture\n", + "- Multiplying the global background mean value by the area of each elliptical photometric aperture, to get the estimated background count rate within each aperture\n", "- Subtracting the estimated background count rate from the photometric count rate for each aperture\n", "\n", "(Just like when we did photometry with the elliptical apertures above, the below step will take almost 5 minutes.)" @@ -846,7 +846,7 @@ "outputs": [], "source": [ "# Calculate the mean background level (per pixel) in the annuli \n", - "bkg_area = [annulus.area() for annulus in elliptical_annuli]\n", + "bkg_area = [annulus.area for annulus in elliptical_annuli]\n", "bkg_mean_per_aperture = bkg_phot_table['aperture_sum'].value / bkg_area\n", "bkg_mean = np.average(bkg_mean_per_aperture) * (u.electron / u.s)\n", "print('Background mean:', bkg_mean)\n", @@ -865,7 +865,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "You might have noticed that these background count rates are *really* small. In this case, this is to be expected - since our example XDF data is a high-level science product (HLSP) that already has already been background-subtracted." + "You might have noticed that these background count rates are *really* small. In this case, this is to be expected – since our example XDF data is a high-level science product (HLSP) that already has already been background-subtracted." ] }, { @@ -933,7 +933,7 @@ "source": [ "---\n", "## About this Notebook\n", - "**Authors:** Lauren Chambers (lchambers@stsci.edu), Erik Tollerud (etollerud@stsci.edu), Tom Wilson (towilson@stsci.edu) Clare Shanahan (cshanahan@stsci.edu)\n", + "**Authors:** Lauren Chambers (lchambers@stsci.edu), Erik Tollerud (etollerud@stsci.edu), Tom Wilson (towilson@stsci.edu), Clare Shanahan (cshanahan@stsci.edu)\n", "
**Updated:** May 2019" ] }, @@ -962,7 +962,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.6.8" + "version": "3.7.5" } }, "nbformat": 4, From 916b99803c4259dd25fc9a58b93cfe4b3b73901c Mon Sep 17 00:00:00 2001 From: Matt Craig Date: Wed, 15 Jan 2020 08:05:53 -0600 Subject: [PATCH 11/14] Eliminate warning from CCDData by removing unnecessary wcs argument --- .../03_aperture_photometry/03_aperture_photometry.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index 6318ab4f..fa531093 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -610,7 +610,7 @@ "outputs": [], "source": [ "# The CCDData mask will be automatically applied\n", - "sky_phot_datum = aperture_photometry(xdf_image, sky_elliptical_apertures[0], wcs=wcs)\n", + "sky_phot_datum = aperture_photometry(xdf_image, sky_elliptical_apertures[0])\n", "sky_phot_datum" ] }, From b686565ba6bc8e7d387b31202e29fc8208ef7a4a Mon Sep 17 00:00:00 2001 From: Matt Craig Date: Wed, 15 Jan 2020 08:41:47 -0600 Subject: [PATCH 12/14] Fix another angle unit --- .../03_aperture_photometry/03_aperture_photometry.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index fa531093..3892653a 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -757,7 +757,7 @@ " a_in = obj.semimajor_axis_sigma.value * r_in\n", " a_out = obj.semimajor_axis_sigma.value * r_out\n", " b_out = obj.semiminor_axis_sigma.value * r_out\n", - " theta = obj.orientation.value\n", + " theta = obj.orientation.to(u.rad).value\n", " elliptical_annuli.append(EllipticalAnnulus(position, a_in, a_out, b_out, theta=theta))" ] }, From c0d60bc9958f77faef5bdef4581b65d08c2e0dbd Mon Sep 17 00:00:00 2001 From: Matt Craig Date: Wed, 15 Jan 2020 08:42:41 -0600 Subject: [PATCH 13/14] Rewrite description of the radii in defining elliptical apertures --- .../03_aperture_photometry.ipynb | 57 ++++++++++++++++--- 1 file changed, 48 insertions(+), 9 deletions(-) diff --git a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb index 3892653a..498914a5 100644 --- a/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb +++ b/notebooks/photutils/03_aperture_photometry/03_aperture_photometry.ipynb @@ -407,20 +407,39 @@ "table" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Defining apertures for sources of different sizes\n", + "\n", + "The galaxies in this HST image vary widely in size, from a few hundred pixels across to a few dozen pixels. Two of the \n", + "properties calculated by `source_properties` are the 1-$\\sigma$ standard deviations of a 2D Gaussian fit to each \n", + "source. Those deviations are called the `semimajor_axis_sigma` and the `semiminor_axis_sigma`. \n", + "\n", + "In what follows the elliptical aperture for each source will be a fixed multiple of that source's \n", + "`semimajor_axis_sigma` and `semiminor_axis_sigma`. The multiplier will be denoted by the variable `r` and we use a \n", + "value of `3` in most cases below. The multiplier value of 3 was chosen here because it is the approximate isphotal \n", + "extent of the galaxy.\n", + "\n", + "There are several accepted ways of perfforming preciion photometry on galaxies. Though this notebook illustrates one \n", + "way to do it you should make sure that the method you use is appropriate for the science you are doing." + ] + }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ - "r = 3. # pixels; approximate isophotal extent of semimajor axis\n", + "r = 3. # Multiplier of semimajor and semiminor axes of source\n", "\n", "# Create the apertures\n", "elliptical_apertures = []\n", "for obj in catalog:\n", " position = (obj.xcentroid.value, obj.ycentroid.value)\n", - " a = obj.semimajor_axis_sigma.value * r\n", - " b = obj.semiminor_axis_sigma.value * r\n", + " a = obj.semimajor_axis_sigma.value * r # pixel\n", + " b = obj.semiminor_axis_sigma.value * r # pixel\n", " theta = obj.orientation.to(u.rad).value\n", " elliptical_apertures.append(EllipticalAperture(position, a, b, theta=theta))" ] @@ -462,7 +481,8 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Clearly, these custom-made elliptical apertures fit our XDF galaxies much better than the one-size-fits-all circular apertures from before." + "Clearly, these custom-made elliptical apertures fit our XDF galaxies much better than the one-size-fits-all circular \n", + "apertures from before. It is also clear that our apertures do not quite enclose all of the light from each galaxy." ] }, { @@ -621,8 +641,8 @@ "Woohoo!\n", "\n", "Unfortunately for our purposes, the `aperture_photometry` function can be only used alone for one of the two cases:\n", - "* Identical apertures at distinct positions (e.g. circular apertures with `r = 3` pixels for many sources)\n", - "* Distinct apertures at identical positions (e.g. two circular apertures with `r = 3` pixels and `r = 5` pixels for one source)\n", + "* Identical apertures at distinct positions (e.g. circular apertures with radius of 3 pixels for many sources)\n", + "* Distinct apertures at identical positions (e.g. two circular apertures with radius of 3 pixels and radius of 5 pixels for one source)\n", "\n", "Since our elliptical apertures are distinct apertures at distinct positions, we need to do a little more work to get a single table of photometric values.\n", "\n", @@ -738,7 +758,17 @@ "\n", "In the [background estimation notebook](../01_background_estimation/01_background_estimation.ipynb), we explored how to perform global background subtraction of image data with `photutils`. However, you can also use `photutils` to perform local background estimations for aperture corrections.\n", "\n", - "To estimate the local background for each aperture, measure the counts within annulus apertures around (but not including!) each source. Since our sources vary quite a bit in size, we should define apertures that are scaled by the size of the source. In our example, we defined elliptical apertures with `r = 3` pixels to measure the counts within each source. To calculate the background for each source, let's measure the counts elliptical annuli between `r = 3.5` pixels and `r = 5` pixels." + "To estimate the local background for each aperture, measure the counts within annulus apertures around (but not including!) each source. Since our sources vary quite a bit in size, we should define apertures that are scaled by the size of the source. \n", + "\n", + "As we did above, the radii defining the annulus are calculated from the 1D standard deviations of a 2D Gaussian fit to \n", + "source. Above, we used a multiplier of 3 to calculate the axes of the elliptical aperture. \n", + "\n", + "Here we need to construct an annulus around each source to meach the local background. The inner diameter should be \n", + "large enough that it does not include much light from the galaxy on which photometry is being performed. The outer \n", + "diameter should be sufficiently large that there are many pixels in the annulus. \n", + "\n", + "Below, we set the multiplier for the inner annulus radius to be one more than the aperture multiplier, and the outer \n", + "radius to be 1.5 larer than the inner radius." ] }, { @@ -747,8 +777,8 @@ "metadata": {}, "outputs": [], "source": [ - "r_in = 3.5 # approximate isophotal extent of inner semimajor axis\n", - "r_out = 5. # approximate isophotal extent of inner semimajor axis\n", + "r_in = r + 1 # approximate isophotal extent of inner semimajor axis\n", + "r_out = r_in + 1.5 # approximate isophotal extent of inner semimajor axis\n", "\n", "# Create the apertures\n", "elliptical_annuli = []\n", @@ -797,6 +827,15 @@ "ax1.set_ylim(2000, 1000)" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Note that in this approach the annulus of each galaxy is \"scaled\" to the size of the galaxy. That in turn means that \n", + "the number of pixels in the background annulus differs from galaxy to galaxy. The area of each annulus by photutils \n", + "and the annulus radii increased if needed." + ] + }, { "cell_type": "markdown", "metadata": {}, From a55786040e01c264faee23cbdb0e6de0591ddb0f Mon Sep 17 00:00:00 2001 From: Matt Craig Date: Wed, 15 Jan 2020 08:43:14 -0600 Subject: [PATCH 14/14] Remove PSF photometry notebook --- .../04_psf_photometry/04_psf_photometry.ipynb | 317 ------------------ .../04_psf_photometry/requirements.txt | 4 - 2 files changed, 321 deletions(-) delete mode 100644 notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb delete mode 100644 notebooks/photutils/04_psf_photometry/requirements.txt diff --git a/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb b/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb deleted file mode 100644 index 1a6687f3..00000000 --- a/notebooks/photutils/04_psf_photometry/04_psf_photometry.ipynb +++ /dev/null @@ -1,317 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "
Made possible by the Astropy Project and ScienceBetter Consulting through financial support from the Community Software Initiative at the Space Telescope Science Institute.
\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "# PSF Photometry with `photutils`\n", - "---" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "##### What is PSF photometry?\n", - "A more specific form of photometry than aperture photometry, PSF photometry takes into account the shape of a source's point spread function (PSF). The PSF is a model that represents the distribution of light from a point source as it falls onto a detector. An example of a basic PSF is simply a 2-D Gaussian, while more complex PSFs can include distortion, diffraction, or interference effects associated with a particular telescope. For instance, the PSFs from the Hubble Space Telescope and the James Webb Space Telescope have been meticulously modeled, and can be simulated with the [Tiny Tim](http://www.stsci.edu/hst/observatory/focus/TinyTim) and [WebbPSF](https://github.com/mperrin/webbpsf) software packages, respectively. However, for datasets that do not have readily available PSF models, such models can be statistically generated by analyzing the image itself.\n", - "\n", - "The `photutils` package provides tools that combine background estimation, source detection, and model-fitting to perform PSF photometry on image data.\n", - "\n", - "##### What does this tutorial include?\n", - "This tutorial covers how to perform PSF photometry with `photutils`, including the following methods:\n", - "* Gaussian PSF Photometry\n", - "* Iterative Subtraction\n", - "* Point Response Function (PRF) Photometry\n", - "\n", - "##### Which data are used in this tutorial?\n", - "We will be manipulating Hubble eXtreme Deep Field (XDF) data, which was collected using the Advanced Camera for Surveys (ACS) on Hubble between 2002 and 2012. The image we use here is the result of 1.8 million seconds (500 hours!) of exposure time, and includes some of the faintest and most distant galaxies that have ever been observed. \n", - "\n", - "*The methods demonstrated here are available in narrative form within the `photutils.psf` [documentation](http://photutils.readthedocs.io/en/stable/psf.html).*" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "
\n", - " \n", - "**Warning:** The PSF photometry API is currently considered experimental and may change in the future. The photutils development team will aim to keep compatibility where practical, but will not finalize the API until sufficient user feedback has been accumulated.\n", - "\n", - "
\n", - "\n", - "
\n", - " \n", - "Important: Before proceeding, please be sure to update your versions of astropy, matplotlib, and photutils, or this notebook may not work properly. Or, if you don't want to handle packages individually, you can always use (and keep updated!) the AstroConda distribution.\n", - " \n", - "
\n", - "\n", - "---" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Import necessary packages" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "First, let's import packages that we will use to perform arithmetic functions and visualize data:" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "from astropy.io import fits\n", - "import astropy.units as u\n", - "from astropy.nddata import CCDData\n", - "# from astropy.stats import sigma_clipped_stats\n", - "from astropy.visualization import ImageNormalize, LogStretch\n", - "import matplotlib\n", - "from matplotlib.colors import LogNorm\n", - "import matplotlib.pyplot as plt\n", - "from matplotlib.ticker import LogLocator\n", - "import numpy as np\n", - "\n", - "# Show plots in the notebook\n", - "%matplotlib inline" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's also define some `matplotlib` parameters, such as title font size and the dpi, to make sure our plots look nice. To make it quick, we'll do this by loading a [style file shared with the other photutils tutorials](../photutils_notebook_style.mplstyle) into `pyplot`. We will use this style file for all the notebook tutorials. (See [here](https://matplotlib.org/users/customizing.html) to learn more about customizing `matplotlib`.)" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "plt.style.use('../photutils_notebook_style.mplstyle')" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Retrieve data" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "As described in the introduction, we will be using Hubble eXtreme Deep Field (XDF) data. Since this file is too large to store on GitHub, we will just use `astropy` to directly download the file from the STScI archive: https://archive.stsci.edu/prepds/xdf/ \n", - "\n", - "(Generally, the best package for web queries of astronomical data is [Astroquery](https://astroquery.readthedocs.io/en/latest/); however, the dataset we are using is a High Level Science Product (HLSP) and thus is not located within a catalog that could be queried with Astroquery.)\n", - "\n", - "Throughout this notebook, we are going to store our images in Python using a `CCDData` object (see [Astropy documentation](http://docs.astropy.org/en/stable/nddata/index.html#ccddata-class-for-images)), which contains a `numpy` array in addition to metadata such as uncertainty, masks, or units. In this case, our data is in electrons per second." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "url = 'https://archive.stsci.edu/pub/hlsp/xdf/hlsp_xdf_hst_acswfc-60mas_hudf_f435w_v1_sci.fits'\n", - "xdf_image = CCDData.read(url, unit=u.electron / u.s)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "As explained in a [previous notebook](../01_background_estimation/01_background_estimation.ipynb) on background estimation, it is important to **mask** these data, as a large portion of the values are equal to zero. We will mask out the non-data portions of the image array, so all of those pixels that have a value of zero don't interfere with our statistics and analyses of the data. " - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Define the mask\n", - "mask = xdf_image.data == 0\n", - "xdf_image.mask = mask" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's look at the data:" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "scrolled": false - }, - "outputs": [], - "source": [ - "# Set up the figure with subplots\n", - "fig, ax1 = plt.subplots(1, 1, figsize=(8, 8))\n", - "\n", - "# Set up the normalization and colormap\n", - "norm_image = ImageNormalize(vmin=1e-4, vmax=5e-2, stretch=LogStretch(), clip=False)\n", - "cmap = plt.get_cmap('viridis')\n", - "cmap.set_over(cmap.colors[-1])\n", - "cmap.set_under(cmap.colors[0])\n", - "cmap.set_bad('white') # Show masked data as white\n", - "xdf_image_clipped = np.clip(xdf_image, 1e-4, None) # clip to plot with logarithmic stretch\n", - "\n", - "# Plot the data\n", - "fitsplot = ax1.imshow(np.ma.masked_where(xdf_image.mask, xdf_image_clipped), \n", - " norm=norm_image, cmap=cmap)\n", - "\n", - "# Define the colorbar and fix the labels\n", - "cbar = plt.colorbar(fitsplot, fraction=0.046, pad=0.04, ticks=LogLocator(subs=range(10)))\n", - "labels = ['$10^{-4}$'] + [''] * 8 + ['$10^{-3}$'] + [''] * 8 + ['$10^{-2}$']\n", - "cbar.ax.set_yticklabels(labels)\n", - "\n", - "# Define labels\n", - "cbar.set_label(r'Flux Count Rate ({})'.format(xdf_image.unit.to_string('latex')), \n", - " rotation=270, labelpad=30)\n", - "ax1.set_xlabel('X (pixels)')\n", - "ax1.set_ylabel('Y (pixels)')" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "*Tip: Double-click on any inline plot to zoom in.*" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Circular Apertures" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Creating Apertures" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Performing Aperture Photometry" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Calculating Aperture Corrections" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Elliptical Apertures" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Creating Apertures" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Performing Aperture Photometry" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Calculating Aperture Corrections" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "---\n", - "## Exercises" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "---\n", - "## Additional Resources\n", - "For more examples and details, please visit the [photutils](http://photutils.readthedocs.io/en/stable/index.html) documentation." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "---\n", - "## About this Notebook\n", - "**Authors:** Lauren Chambers (lchambers@stsci.edu)\n", - "
**Updated:** May 2019" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "[Top of Page](#title_ID)\n", - "\"STScI" - ] - } - ], - "metadata": { - "kernelspec": { - "display_name": "Python 3", - "language": "python", - "name": "python3" - }, - "language_info": { - "codemirror_mode": { - "name": "ipython", - "version": 3 - }, - "file_extension": ".py", - "mimetype": "text/x-python", - "name": "python", - "nbconvert_exporter": "python", - "pygments_lexer": "ipython3", - "version": "3.6.8" - } - }, - "nbformat": 4, - "nbformat_minor": 2 -} diff --git a/notebooks/photutils/04_psf_photometry/requirements.txt b/notebooks/photutils/04_psf_photometry/requirements.txt deleted file mode 100644 index a0d3505e..00000000 --- a/notebooks/photutils/04_psf_photometry/requirements.txt +++ /dev/null @@ -1,4 +0,0 @@ -astropy>=3.1.2 -matplotlib>2.2.2 -numpy>=1.13.3 -photutils>=0.4