data-open-david.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import scipy.signal\n",
    "from scipy.signal import butter, filtfilt, deconvolve\n",
    "import scipy.io as sio\n",
    "from scipy.stats.stats import zscore\n",
    "from scipy import ndimage\n",
    "from scipy import stats\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.pyplot import plot, xlabel, ylabel, title, imshow\n",
    "from matplotlib import gridspec\n",
    "from matplotlib.patches import Ellipse\n",
    "%matplotlib inline\n",
    "\n",
    "import pyret.spiketools as spk\n",
    "import pyret.stimulustools as st\n",
    "import pyret.filtertools as ft\n",
    "import pyret.visualizations as vis\n",
    "import pyret.nonlinearities as nl\n",
    "# ft.linear_response has a different dimension before version '0.6.0'\n",
    "pyret.__version__\n",
    "#pyret.visualizations.__all__\n",
    "\n",
    "import h5py\n",
    "import warnings\n",
    "\n",
    "import pylab\n",
    "from pylab import rcParams\n",
    "rcParams['figure.figsize'] = (10.0, 8.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from os import listdir\n",
    "from os.path import isfile, join"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rcParams['animation.html'] = 'jshtml'\n",
    "from IPython.display import HTML\n",
    "import my"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 0419 David intracellular h5 files:\n",
    "data_dir = '/Users/peterfish/Modules/1d-whitenoise-analysis/raw-data-david/'\n",
    "files = [f for f in listdir(data_dir) if isfile(join(data_dir, f))]\n",
    "files = [f for f in files if '.h5' in f]\n",
    "files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "data = [dict() for i in range(len(files))]\n",
    "# data[file id][cell id][flash or lines] = dic() ..\n",
    "# data[file id][flash or lines] = dic() ..\n",
    "\n",
    "i = 0 # file id\n",
    "filename = data_dir + files[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load dataset by sequentially visiting all possible dataset groups.\n",
    "all_obj = []\n",
    "with h5py.File(filename, 'r') as f:\n",
    "    print('Keys :', list(f.keys()))\n",
    "    \n",
    "    # visit all dataset\n",
    "    f.visit(all_obj.append)\n",
    "    all_datasets = [obj for obj in all_obj if isinstance(f[obj], h5py.Dataset)]   \n",
    "    all_groups = [obj for obj in all_obj if isinstance(f[obj], h5py.Group)]\n",
    "    print('Groups :', all_groups)\n",
    "    \n",
    "    for key in all_datasets:\n",
    "        data[i][key] = np.array(f[key])    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def visit_h5_keys(d):\n",
    "    # d is a reference to h5 file.\n",
    "    dataset = dict()\n",
    "    \n",
    "    for key in d.keys():\n",
    "        if isinstance(d[key], h5py.Dataset):\n",
    "            dataset[key] = np.array(d[key])\n",
    "        elif isinstance(d[key], h5py.Group):\n",
    "            dataset[key] = visit_keys(d[key])\n",
    "        #else:\n",
    "            # do somthing            \n",
    "    return dataset\n",
    "\n",
    "with h5py.File(filename, 'r') as f:\n",
    "    d = visit_h5_keys(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(list(d['boxes'].keys()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "rf = d['boxes']['rfs']\n",
    "#rf = d['lines']['rfs']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "t = d['boxes']['time']\n",
    "t = t[0:38]\n",
    "time = - t[::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "011914_cell_1\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x576 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAl8AAAHVCAYAAADPfJ4+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3XmU1dWd7/3P93fGmpipYhQEFEEGUcQ5zkaTGO0YE013YmcyccV7u9ddd90na93n9s3qdf+4t3vd5/bq3HTUpGPmGBOTaIyJA06ogIAgAqIiiDIXU0ENZ/zt549TVZYIMpxhn+H9WstFnVO/Or9PoZQf9t5nb3POCQAAAJUR+A4AAADQSChfAAAAFUT5AgAAqCDKFwAAQAVRvgAAACqI8gUAAFBBlC8AAIAKonwBAABUEOULAACggqK+AxzLmDFj3NSpU33HAAAAOK5Vq1btdc6NPZFrq7Z8TZ06VStXrvQdAwAA4LjMbOuJXsu0IwAAQAVRvgAAACqI8gUAAFBBlC8AAIAKonwBAABUEOULAACggihfAAAAFUT5AgAAqCDKFwAAQAVRvgAAACqI8gUAAFBBlC8AAIAKonwBAABUEOULAACggihfAAAAFUT5AgAAqCDKFwAUITx4UOHBg75jAKghUd8BAKAWuTBUeskSpZ9/XgoCtX7lK4qMH+87FoAawMgXAJyk/N696vnRj5R+9lkpDKVcTj0PPKCwp8d3NAA1gPIFACfIOaf08uXqvvde5bdvlw0bpuYvfEGRSZPkDh1S74MPyuXzvmMCqHJMOwLACQi7utT78MPKb9kiSYrNn6+m66+XJZOKjBun7h/8QPl331Xqz39W06c+5TktgGpWkpEvM7vezN4ws01m9u2jfP6bZvaama0xsxfMbHYp7gsA5eacU+bVV3X4+99XfssWWXOzmj/3OTXffLMsmZQkBW1tav7856VIRJlVq5ReudJzagDVrOjyZWYRSd+TdIOk2ZJuP0q5+qVzbq5z7hxJ/yTp/yv2vgBQbmFPj3offFB9f/iDlE4rOnOmWu+6S7FZsz50bXTiRDXdeKMkKfXnPyu3dWul4wKoEaUY+VokaZNzbrNzLiPpAUk3Db3AOXdoyMMWSa4E9wWAsslu3Kju739fuY0bpXhcTTfdpObPf15Ba+sxvyY+f77iF14ohaF6H3yQLSgAHFUp1nxNlPTekMfbJF1w5EVm9i1J/0lSXNJVR3shM7tT0p2SdNppp5UgGgCcHJdKqe/xx5Vds0aSFJk6Vc033aRgxIgT+vrktdcq7OxU7u231fPrX6v1K1+RxWLljAygxlTs3Y7Oue8556ZL+n8k/b/HuOY+59xC59zCsWPHVioaAEiSclu26PA99xSKVzSq5Mc/rpYvfemEi5ckWRCo+ZZbFIwapXDXLvU9/LCcY7AfwPtKUb62S5o85PGk/ueO5QFJN5fgvgBQMully9Tz05/KdXUpMn68Wu+8U4kLL5SZnfRrWVOTmm+7TYrHlV2/XukXXihDYgC1qhTla4WkM8zsdDOLS7pN0iNDLzCzM4Y8/KSkt0pwXwAoifyuXUo9+aQkKXH55Wr56lcVKXL0PTJ2rJo/8xlJUvrpp5V9882icwKoD0WXL+dcTtLdkh6X9LqkB51z683sH83s0/2X3W1m681sjQrrvu4o9r4AUAoun1fvH/4ghaHi55+v5BVXyCKRkrx2bOZMJa68UpLU+9BDynd2luR1AdS2kmyy6px7TNJjRzz3D0M+/rtS3AcASi39/PMKd+9WMHKkktdcU/LXT1x2mcLdu5XdsEG9Dzyg1q99TdbUVPL7AKgdHC8EoGHld+5UeskSSVLTTTfJ4vGS38PM1HTTTQo6OhTu36/ehx6SC8OS3wdA7aB8AWhILpcrTDc6p/gFFyg6ZUrZ7mXxuFpuu03W3Kzc228rtXhx2e4FoPpRvgA0pPTzzyvcs0fBqFFKXn112e8XjBih5ltvlYJAmZdeUmbt2rLfE0B1onwBaDi57dsHt39ouummim2CGp06Vcnrr5ck9T3yiHLbP2pXHgD1ivIFoKG4XE59Dz9cmG688EJFK3yaRnzhQsXPPVfK59X3u9/J5XIVvT8A/yhfABpK+tlnFXZ2Khg9WsmrjnrSWVmZmZI33KBgzBiF+/cr/eKLFc8AwC/KF4CGkdu2TemXXpL634Ho68xFi0bV9MlPSpLSS5Yov2+flxwA/KB8AWgIH5huvOgiRSdPPv4XlVF06lTF5s+X8nmlHnuM8x+BBkL5AtAQUs88o3DvXgVjxijZv+u8b8lrr5Ulk8pt3qzs+vW+4wCoEMoXgLqXe+89ZZYufX+6MVqSwz2KFrS0KHnttZKk1OOPy6VSnhMBqATKF4C65rLZwenGxMUXKzppku9IHxBbsECRyZPluruVevpp33EAVADlC0BdSz3zjMJ9+xSMHavEFVf4jvMhZlZYfG+mzIoV7P0FNADKF4C6lXv33aqcbjxSpKND8YsukiT1PfooZz8CdY7yBaAuDU43SkpccomiEyd6TvTRkpdfLhs+XOGuXcqsWOE7DoAyonwBqEupxYsV7t+voL1dicsv9x3nuCweV9MNN0iSUk8/rfDQIc+JAJQL5QtA3clt3arM8uWSmZqreLrxSLGZMxWdOVPKZJR6/HHfcQCUCeULQF1xmcz7042XXabIhAmeE52cphtukGIxZTdsUHbTJt9xAJQB5QtAXcmsXq3wwAEFHR1KfOxjvuOctGD4cCX735WZeuwxuWzWbyAAJUf5AlBXcm+9JUlKXHSRLBLxnObUxC+4QEF7u8IDB5RessR3HAAlRvkCUDdcJqPcO+9IkqIzZvgNUwSLRN4/ePvFF5Xfu9dzIgClRPkCUDdyW7ZI+bwiEycqaGnxHaco0dNOU2zBAikM1fenP3HwNlBHKF8A6sbAlGP0jDM8JymN5DXXyJqblX/nHWXXrvUdB0CJUL4A1AXn3OC7A+ulfAXNze8fvP3EEwr7+jwnAlAKlC8AdSHs7JTr6pK1tCgyfrzvOCUTmz9fkSlT5Hp7lX7qKd9xAJQA5QtAXRiccpwxQ2bmOU3pDB68HQTKvPKKcu+95zsSgCJRvgDUhXqbchwqMnasEhdfLKn/4O183nMiAMWgfAGoeS6VUv7ddyUzxaZN8x2nLBIf+5hsxAiFe/Yos3q17zgAikD5AlDzcps3S2GoyOTJsqYm33HKwmIxJa++WpKUWb6crSeAGkb5AlDzsnW2xcSxxGbNkrW1Kdy7V/ktW3zHAXCKKF8AappzTrn+9V6xOi9fFokoft55kqT0yy97TgPgVFG+ANS0cNcuue5uWVubgvZ233HKLn7eeVIkotwbbyg8cMB3HACngPIFoKYNnXKspy0mjiVobVXs7LMlSekVKzynAXAqKF8AalqjTDkOFV+0SJKUXb1aLpv1nAbAyaJ8AahZYW+v8tu2SUGg6Omn+45TMdGJExWZOFEuleLMR6AGUb4A1Kzc229LzikyZYoskfAdp6IGRr/SL7/MthNAjaF8AahZA0cKNdKU44DY7NmylhaFe/Yov3Wr7zgATgLlC0BNcmE4uN6r3vf3OhqLRtl2AqhRlC8ANSm/Y4dcX59sxAgFo0f7juNFfOFCKQiU27hRYVeX7zgAThDlC0BNGjrl2AhbTBxN0Nam2KxZknPKrFzpOw6AE0T5AlCTcg1ypNDxDCy8z7zyilwu5zkNgBNB+QJQc8LubuV37pSiUUWnTvUdx6vI5MkKxo2T6+1Vdt0633EAnADKF4CaM7jQfupUWSzmOY1fZqYE204ANYXyBaDmMOX4QbE5c2RNTQp37ixsOgugqlG+ANQUl88r+/bbkhpzf6+jsVhM8XPPlSRl2HYCqHqULwA1Jb9tm5ROKxgzRsHIkb7jVI34+edLZspu2KDw8GHfcQB8BMoXgJoyOOU4Y4bnJNUlGD5c0bPOksKQbSeAKkf5AlBTsg18pNDxDCy8z6xaJZfPe04D4FgoXwBqRtjVpXDPHikeV+S003zHqTqRKVMUtLfL9fQou3697zgAjoHyBaBmDIx6RadNk0WjntNUn6HbTrDwHqhelC8ANWNgf68Y672OKTZ3rpRMKr99u3Lbt/uOA+AoKF8AaoLL5ZTbvFkS+3t9FIvHFV+wQBKjX0C1onwBqAm5rVulbFZBR4eCYcN8x6lqifPPlyRl169X2NPjOQ2AI1G+ANSEgS0mmHI8vmDkSEXPPFPK55VZtcp3HABHoHwBqAmD5zky5XhC4gML71euZNsJoMpQvgBUvfz+/Qr37ZMSCUUmT/YdpyZEp01TMGaM3OHDym7c6DsOgCFKUr7M7Hoze8PMNpnZt4/y+f9kZhvMbK2ZLTazKaW4L4DGMHTK0QL+zngizKxw5JBYeA9Um6J/iplZRNL3JN0gabak281s9hGXrZa00Dk3T9JvJf1TsfcF0DgGpxxZ73VS4vPnS/G48u++q/yuXb7jAOhXir9CLpK0yTm32TmXkfSApJuGXuCce8Y519v/cJmkSSW4L4AG4LJZ5bZskUT5OlmWSCh+zjmSpPTy5Z7TABhQivI1UdJ7Qx5v63/uWL4q6c9H+4SZ3WlmK81sZWdnZwmiAah1uS1bpHxekQkTFLS2+o5TcwYW3mfXrVPY23ucqwFUQkUXT5jZ30haKOmfj/Z559x9zrmFzrmFY8eOrWQ0AFVqYL0Xo16nJjJ6dOH3Lpdj2wmgSpSifG2XNPTtR5P6n/sAM7tG0n+V9GnnXLoE9wVQ55xzyrLFRNHiF14oqbDwnm0nAP9KUb5WSDrDzE43s7ik2yQ9MvQCM1sg6V4ViteeEtwTQAMI9+6VO3hQ1tysyIQJvuPUrOi0aQra2+W6u5Vdt853HKDhFV2+nHM5SXdLelzS65IedM6tN7N/NLNP91/2z5JaJf3GzNaY2SPHeDkAGDR0ypEtJk6dmSlxwQWSpPSyZXLOeU4ENLZoKV7EOfeYpMeOeO4fhnx8TSnuA6CxZAfKF1OORYvNm6fU4sUKd+1SfutWRadO9R0JaFj8VRJAVXLptPLvviuZKTp9uu84Nc+i0cFNV9PLlnlOAzQ2yheAqpTfuVMKQwXjxiloavIdpy7EFy6UIhHl3nhD+f37fccBGhblC0BVyu/cKUmKjB/vOUn9CFpbFZs7V5KUYfQL8IbyBaAqDRyHQ/kqrcTAthNr1silUp7TAI2J8gWgKjHyVR6Rjg5Fp02Tslk2XQU8oXwBqDouk1G4d69kpkhHh+84dWdg09U0m64CXlC+AFSd/O7dknMK2ttl0ZLsiIMhojNmKBg9Wu7QIWVff913HKDhUL4AVB2mHMvLzN4/cohNV4GKo3wBqDqD5WvcOM9J6ld8/nxZU5Py27crv22b7zhAQ6F8Aag6jHyVn8Viip93niQpvXSp5zRAY6F8AagqLpdT2NkpiZGvcosvWiQFgXIbNyo8cMB3HKBhUL4AVJX8nj2Fne3HjJHF477j1LWgrU2xOXMk55R++WXfcYCGQfkCUFVCphwranDT1VdekUunPacBGgPlC0BVYb1XZUXGj1dkyhQpk1Fm9WrfcYCGQPkCUFUoX5U3MPqVXr5cLgw9pwHqH+ULQNVw+Xxhg1Wx2L6SomeeqWDkSLmDB5XbuNF3HKDuUb4AVI2ws1PK5xWMHClLJn3HaRgWBIpfcIEkKb1smec0QP2jfAGoGkw5+hNfsEBKJJR/7z3ltm/3HQeoa5QvAFVjoHwFlK+Ks3h8cNPVDKNfQFlRvgBUjfyuXZIY+fIlsWiRZKbs+vUKu7p8xwHqFuULQFVwYUj58iwYPlyx2bPZdBUoM8oXgKoQ7tsnZbOy4cMVNDf7jtOw4kM3Xc1kPKcB6hPlC0BVYLF9dYhOmqTI5MlSKqXMmjW+4wB1ifIFoCoMli/29/Ju8Mih5cvlnPOcBqg/lC8AVYH1XtUjetZZsuHDFe7fr9ybb/qOA9QdyhcA75xzTDtWEQsCJfo3XU09+ywHbgMlRvkC4F144ICUTstaWxW0tfmOA0nxc8+VDRumcNcu9fzylyy+B0qI8gXAO0a9qo8lEmq54w5ZW5vy776rnl/8ggIGlAjlC4B3IeWrKkVGjfpgAWMEDCgJyhcA7xj5ql6R0aMLBay1VfmtW9Xzq1/JZbO+YwE1jfIFwCsW21e/DxSwd96hgAFFonwB8ModOiTX1ydrapING+Y7Do4hMmbM+wVsyxYKGFAEyhcAr4aOepmZ5zT4KIMFrKVF+S1b1PvAAxQw4BRQvgB4xZRjbRlawHKbN1PAgFNA+QLg1UD5CihfNSMydmyhgDU3FwrYr38tl8v5jgXUDMoXAK8Y+apNHyhgb79NAQNOAuULgDfh4cNy3d1SIqFg5EjfcXCSIu3t7xewTZvU++CDFDDgBFC+AHjDYvvaF2lvV8uXviRralLurbcoYMAJoHwB8GawfI0b5zkJihHp6CiMgA0UsN/8hgIGfATKFwBv8rt2SWK9Vz2IdHS8PwL25ptKPfOM70hA1aJ8AfCGxfb1JTJunJpvv12SlHn5ZYXd3Z4TAdWJ8gXAi7C3V66rS4rFFIwe7TsOSiQ6ebKiM2dKuZzSL73kOw5QlShfALwYut7LAn4U1ZPkxz4mScqsWKGwp8dzGqD68BMPgBdMOdavyIQJip55JqNfwDFQvgB4EVK+6lry8sslMfoFHA3lC4AXjHzVt8iECYqecYaUzTL6BRyB8gWg4lwqpfDAASkSUTBmjO84KJPE0NGv3l7PaYDqQfkCUHGD+3t1dMgiEc9pUC7RiRMHR78yjH4BgyhfACqOKcfGMTD6lX75ZUa/gH6ULwAVR/lqHNGJExWdMaMw+rV0qe84QFWgfAGoOMpXY2H0C/ggyheAinKZjMK9e6UgUNDe7jsOKiA6aZKi06dLmYwyy5b5jgN4R/kCUFH53bslSUF7uywa9ZwGlTI4+rV8ucK+Ps9pAL8oXwAqKr9jh6TCsUJoHNHJkxWdNq0w+sXaLzQ4yheAihrcZoL1Xg1n6Novx+gXGlhJypeZXW9mb5jZJjP79lE+/zEze8XMcmb22VLcE0BtYrF944qedlph9CudVpq1X2hgRZcvM4tI+p6kGyTNlnS7mc0+4rJ3Jf2tpF8Wez8Atcvlcgr37JHMFOno8B0HHgxd+8XoFxpVKUa+Fkna5Jzb7JzLSHpA0k1DL3DOveOcWyspLMH9ANSo/O7dknMKxoyRxeO+48CD6GmnKXL66YXRr+XLfccBvChF+Zoo6b0hj7f1P3fSzOxOM1tpZis7OztLEA1ANWHKEZKUHBj9WrZMLpXynAaovKpacO+cu885t9A5t3Ds2LG+4wAosZDyBUnRKVMUmTqV0S80rFKUr+2SJg95PKn/OQD4AN7piAGMfqGRlaJ8rZB0hpmdbmZxSbdJeqQErwugjrh8fnCDVfb4QnTqVEWmTJFSKUa/0HCKLl/OuZykuyU9Lul1SQ8659ab2T+a2aclyczON7Ntkm6VdK+ZrS/2vgBqS9jZKeXzCkaNkiUSvuOgCgyMfmWWLZNLpz2nASqnJGd7OOcek/TYEc/9w5CPV6gwHQmgQbHYHkeKTJ2qyGmnKf/uu0ovX67kxz7mOxJQEVW14B5A/aJ84UhmpuQVV0iSMkuXMvqFhkH5AlARlC8czcDol0ullH75Zd9xgIqgfAEoOxeGg+90DChfGMLM3l/7tXQpu96jIVC+AJRduG+flMvJRoxQ0NTkOw6qTOT00wujX3196vn5zylgqHuULwBlx5QjPoqZqfkzn1EwcqTyO3ao52c/o4ChrlG+AJTdYPlify8cQzB8uFruuKNQwHbuVPfPfqaQAoY6RfkCUHaMfOFEBMOHq+Vv/1bBqFEKd+5Uz09/qrC313csoOQoXwDKyjlH+cIJC4YNK4yAjRqlcNcuChjqEuULQFmF+/dLmYysrU1Ba6vvOKgBwbBhhRGw0aMV7t6tnp/8RGFPj+9YQMlQvgCUFaNeOBVBW1thBGz0aIV79hRGwChgqBOULwBlld+xQxLlCycvaGsrjICNGVMoYIyAoU5QvgCU1cDmqpQvnIqgtbUwAjZ2rMLOzkIB6+72HQsoCuULQNl8YLH9hAme06BWUcBQbyhfAMrGHTwopVKy1lYFbW2+46CGBS0thQLW3q5w7171/PjHCg8f9h0LOCWULwBlw3ovlFLQ0qKWL31JQUeHwn37CiNgFDDUIMoXgLLhnY4otQ8VsB//WOGhQ75jASeF8gWgbChfKIegublQwMaNU7h/v7p/9CPltm71HQs4YZQvAGXBzvYop4ECFpk4Ua6rSz0//rH6nnxSLpfzHQ04LsoXgLJwXV1yfX2y5mbZsGG+46AOBU1Navnyl5W47DLJTJmXXlL3D3+o/O7dvqMBH4nyBaAsho56mZnnNKhXFokoedVVavnylxWMHKlw9251/+AHSr/0klwY+o4HHBXlC0BZMOWISopOnqzWb35T8XPPlfJ5pZ58snAk0cGDvqMBH0L5AlAWlC9UmsXjarrxRjXffruspUX5rVt1+J57lHn1VTnnfMcDBlG+AJScc+79Pb7Y2R4VFjvzTLXedZeiZ50lpdPq+8Mf1Pub3yjs7fUdDZBE+QJQBu7wYbneXlkyKRs+3HccNKCgpUXNn/ucmm66SYrHlXv9dXX/278p+9ZbvqMBlC8ApTd01IvF9vDFzBQ/5xy13XWXIqedJtfTo95f/lJ9jz4ql8n4jocGRvkCUHID672CceM8JwGkYMQItdxxh5LXXCNFIsqsWqXue+9VduNGuXzedzw0oKjvAADqz+Bie9Z7oUpYEChxySWKzpih3t/9TuGePer99a9lra2Kz5+v2IIFiowe7TsmGgTlC0DJ8U5HVKtIR4dav/51ZVasUOaVVxTu3av0iy8q/eKLikyZoviCBYrNni2LxXxHRR2jfAEoqfDwYbnubimRUDBypO84wIdYNKrERRcpfuGFym/bpswrryi7fr3yW7eqb+tW9f35z4rPmaP4uecqYJNglAHlC0BJsbM9aoWZKTp5sqKTJ8tdf72y69Yps3q18tu3K7NqlTKrVino6CiMhs2bp6CpyXdk1AnKF4CSYsoRtcgSCcXPO0/x885Tfs+ewmjY2rUKd+9W6i9/UerJJxWbNUvxc89VZOpU/mKBolC+AJQU5Qu1LtLerqbrr1fymmuUfeMNZVevVu7tt5Vdt07ZdesUmTxZyWuuUfS003xHRY2ifAEoKcoX6oVFo4qffbbiZ5+t8OBBZVavVmblSuXfe08999+v6JlnKnn11Yq0t/uOihpD+QJQMmFPj9yhQ1I8roC37aOOBCNGKHnllUpcfLHSL72k9NKlyr35prrfekux+fOVvOIKBZzmgBNE+QJQMoOjXuPGsSYGdckSCSWvvFLx889X+vnnlVm1Stk1a5R97TXFFy1S4tJLFTQ3+46JKscO9wBKZvBYIaYcUeeC1lY1feITav3WtxSbM0fK55VZulSH//VflVqyRC6b9R0RVYzyBaBk8rt2SaJ8oXFERo1S8y23qPXrX1d02jQpnVb66ad1+LvfVWbVKrkw9B0RVYjyBaBkhh6oDTSSyIQJavniF9XyxS8qMn683OHD6nv0UXX/278p+/rrcs75jogqQvkCUBJhb69cV5cUi7HYHg0rOm2aWr7+dTXdcouCkSMV7tun3gcfVM/99yu/f7/veKgSlC8AJfGBxfYBP1rQuMxM8Tlz1Pqtbyn5iU/IWlqUf+89dd97rzJr1jAKBsoXgNII2d8L+ACLRJQ4/3y13X13YVF+JqO+hx9W30MPyaVSvuPBI8oXgJJgc1Xg6CyZVNNnPqOmm2+W4nFl16/X4XvuUe7dd31HgyeULwAlQfkCjs3MFJ8/X63f+IYiEybIdXWp58c/VuqZZ3hHZAOifAEomuvrU3jggBSNKhg71nccoGpFRo1Sy1e+osSll0rOKf388+q5//7Cnx80DMoXgKIN7u/V0cFie+A4LBJR8uqr1XLHHbK2NuW3bdPhe+9V5rXXfEdDhfBTEkDRmHIETl506lS1fvObis6aJaXT6vvd79T7+9/LpdO+o6HMKF8Aikb5Ak5N0Nys5ltvVdOnPiXFYsquXavue+9Vbtu2iuZwzrH2rII4WBtA0QbLFzvbAyfNzBQ/7zxFpkxR70MPKdy1Sz0/+pESV1yhxKWXlnQq36XTCvfvV37vXoX79inct0/5/l+Vzys2c6Zi8+YpOmOGLBIp2X3xQZQvAEVx6XThB3ckwmJ7oAiRMWPU+tWvKvX008osXar0M88o/fzzsuZmWVNT4dfmZgUDHw95buDjoLlZisUUdnUVylV/yRooWO7w4Y/MkN2wQdkNG2RNTYqdfbZi8+YpMmmSzKxCvwuNgfIFoCgfWGzP35SBolg0qqbrrlN0+nT1Pfqo3MGDcocPH7c0nbBIRMGoUQpGj1Zk9GgFY8YoGD26cCRYNqvsa68ps3atws5OZVauVGblSgUjRyo2d26hiHF0WElQvgAUZfAw7XHjPCcB6kds+nTF/u7v5LJZud5eub4+ud7ewhmq/R8Pfd719RU+19srZTKyYcPeL1j9JSsyerRs+PCPnMZMXHqp4pdconD3bmXWrlV23TqFBw4o/fzzSj//vCITJyo2b55iZ5+toKWlgr8j9YXyBaAorPcCysdiMdnw4dLw4Sf8Nc65oqYJzUyRcePUNG6cktdco/w77xSK2OuvK799u/Lbtyv1l78oOmNGoYjNnCmLxU75fo2ooctXftcuBR0dzGUDRRgoXwHvdASqQin/n2ZBoOi0aYpOmyb3yU8q+8Ybyq5dq9ymTcq99ZZyb72lVGurktddp9icOfz/9AQ1bPkKDx1S9w9+oGDUKCUuukixefNk0Yb97QBOictkFO7dKwWBIu3tvuMAKCOLxRSfM0fxOXMU9vQou369MqtXK9y1S32/+50yr7yipk98QhHeeHNcDbvPV7h/v6y1VeHever74x91+F/+RaklSxT29fmOBtSMgcX2QXs7f3kBGkjQ0qLEokVqvfNONX3607LmZuXfeUfd99yjvieflMtkfEesaiUpX2Z2vZm9YWabzOzbR/l8wswq3b1DAAAgAElEQVR+3f/55WY2tRT3LUZ06lS1/cf/qKa/+isFHR1yPT1KP/20Dv+f/6O+P/+Zc7aAE8DmqkBjMzPFFyxQ6913K37eeVIYKvPSSzr8f/+vsuvXyzlXsnuFhw8ru2mTwq6ukr6uD0X/VdXMIpK+J+laSdskrTCzR5xzG4Zc9lVJB5xzM8zsNkn/S9Lni713sSwSUXzePMXmzlV+yxally5VbtMmZV5+WZkVKxSbNUvxiy9WdOJE31GBqkT5AiBJQVOTmj71KcUWLFDqsceU37FDvb/9raLTpil5ww2KjBlzSq8bdnUV9h57/XXl33tv8HlraVFk4kRFJkwY/DVobi7Vt1N2pRj5WiRpk3Nus3MuI+kBSTcdcc1Nkn7S//FvJV1tVbIqz8wU9C8obPnrv1brN7+p2x99VCP++3/XI7//vXp++EN133+/vv+P/ygz05133jn4tTt27JCZacIR7/I677zzZGZatWrV4HPf+c53ZGb6zne+M/jcqlWrZGY677zzPvD1EyZMkJlpR/9b+CXpzjvvlJnpvvvuG3zuj3/8o8xMN95444e+pyN/e2+88UaZmf74xz8OPnfffffxPfE9FfU93fLf/puk98tXPXxP9fjvie+J76lS31N04kS1fPWrWtzWphHf+Y5u+R//Q93f/75SixfLZTIn9D2FBw7oe//lv8jM9JWrrlLqiSeUf+897ezt1YjvfEdn/e//LdfTo9ybbyr97LM6/2MfU6SlRc/95/+s3t/+VumXXtI//P3ff+h7qialWKQxUdJ7Qx5vk3TBsa5xzuXMrEvSaEl7h15kZndKulOSTjvttBJEO3mRjo7B/Ypis2ZJsZjy776rzMqVkqT83r1yuRzrWwAVFtzLTJGODt9RAFQJCwLFzjhDkhQMHy6FodIvvKDMa6995Ndl1q/X4R07FO7apdzrrxeeDALFZs9WdPZstbW0SP/0T7LWVrX+h/9Q2PZixw7Z/fcXrj18WNn16wtvBFizRpKUXrFCvQ8/rMiECYrNmaOgqals3/fJsGLnTc3ss5Kud859rf/xFyVd4Jy7e8g16/qv2db/+O3+a/Ye7TUlaeHChW5lf+HxyaXTyqxerfSyZXJdXZIKw52JSy9V/Pzz2dEbDSu3bZt6/v3fFbS3q+2uu3zHAVClcu+9p77HHlPY/wad6IwZSt5wg5TPDx5nFO7Z8/4XxOOKnXlmoXTNmHFCe4i5MFTY2Tm4D1lux47Caw45LLzt7/++UAbLxMxWOecWnsi1pRi+2S5p8pDHk/qfO9o128wsKmm4pH0luHfZWSKhxIUXKr5okbLr1yu9dKnCnTuVevxxZVauVPLaaxU980z2NkHDGdzZns1VAXyE6OTJav3615VZuVKpp59WbtMmdX/3ux+8KJFQ7KyzFJs1S9Hp0096dsmCoDBz1dEhnXuuJMlls8rv2qX8jh0KOztlw4aV6lsqWinK1wpJZ5jZ6SqUrNskfeGIax6RdIekpZI+K+lpV2NvVbAgUHzuXMXmzFHuzTeVevJJhfv2qfeBBxQ5/XQ1XXcdx6ugoQwutue/ewDHYUGgxKJFis2erdRTTyn76quypiZFZ84sjHBNm1bymSSLxRSdPFnRyZOPf3GFFV2++tdw3S3pcUkRST9yzq03s3+UtNI594ikf5f0MzPbJGm/CgWtJpmZYjNnKjpjhjIrVij93HPKb9mi7nvvVWzBAiWvukpBa6vvmEDZcawQgJMVtLaq+eab5a6/XorFGnbpTklWjTvnHpP02BHP/cOQj1OSbi3FvaqFRSJKXHihYvPnK/3cc8qsWKHs6tXKrl+vxKWXKnHhhZx1hbrlcjmFnZ0stgdwSiyZ9B3Bq4bd4b5UgqYmNV1/vVrvukvRM8+UMpnCZq3f+54y69bV/EZwwNHkd++WwlDBmDGyeNx3HACoKZSvEomMGaOW229Xyxe/WNgxv6tLfQ89pJ4f/Ui5bdt8xwNKis1VAeDUUb5KLDptWuGsqxtvlLW0KN//dvzehx5SePCg73hASVC+AODUsVNoGVgQKH7uuYqdfbbSL7yg9NKlyq5bp+zrrytx4YVKXHaZLJHwHRM4ZSHlCwBOGSNfZWSJhJJXX622u+9WbM4cKZ9X+sUXdfhf/1XpFSvk8nnfEYGT5vJ55fs3RGSbCQA4eZSvCghGjFDzLbeo5atfVWTyZLneXqUee0zd99yj7BtvsCgfNSXcs0fK5xWMHs0ILgCcAspXBUUnTVLLl7+s5s99TsGoUQr37lXvAw+o56c/HVxDA1Q71nsBQHFY81VhZlY4PuHMMwubtD7/vPLvvKPu++5TbN68wiatZTx7CijW4LFClC8AOCWUL08GNmmNz5+v1JIlyrz8srJr1yq7YUNhUf6llzKlg6qU7z8cl/IFAKeGaUfPrKlJTdddp7ZvfUuxs8+WcjmlX3hBh7/7XaVXrpQbciI74JvL5ylfAFAkyleVCEaOVPNnP/v+ovyeHqX+9Cd1f//7ym3e7DseIEkK9+4tLLYfNarhjwcBgFNF+aoyg4vyb71VwciRCvfuVc8vfqHspk2+owGs9wKAEqB8VSEzU2z2bLV+61uKX3CBFIbqffBBjimCd4NTjuzvBQCnjPJVxSwSUfLjH1fsnHOkbFa9v/yl8p2dvmOhgQ2Ur4CRLwA4ZZSvKmdmarrxRkXPPFOur089P/+5wq4u37HQgJxz7498dXR4TgMAtYvyVQMsCNT82c8WFuIfOlQoYL29vmOhwbiDB6VMRtbaqqC11XccAKhZlK8aYbGYWm6/XUF7e2Fn/F/9Si6T8R0LDYT1XgBQGpSvGmJNTWr567+WDR+u/LZt6v3NbzicGxXDlCMAlAblq8YEw4ap5W/+RtbcrNymTep75BEO5kZFDC62Z+QLAIpC+apBkTFj1PyFL0ixmLJr1yr1xBMUMJRdfvduSUw7AkCxKF81Kjpxopo//3kpCJRZtkyZl17yHQl1LOzrk+vqkmIxBaNG+Y4DADWN8lXDYtOnq+mv/kqSlHrqKWVWr/acCPUqHLLeywJ+bABAMfgpWuPic+Yoef31kqS+P/5R2Tfe8JwI9YjF9gBQOpSvOpC44AIlLrtMck69v/2tcu++6zsS6szAei8W2wNA8ShfdSJx5ZWKn3uulMup51e/GvyfJVAK7PEFAKVD+aoTZqbkJz+p6FlnSamUen7xC4UHDviOhTrgcjmF/WeKRtrbPacBgNpH+aojFgRqvuUWRaZMkTt8WN0/+IGyGzf6joUaF3Z2SmGoYPRoWTzuOw4A1DzKV52xaFQtt92m6BlnyPX1qffXv1bfn/4kl836joYaxf5eAFBalK86ZMmkmm+/XcmPf7ywD9jKler+4Q+V37PHdzTUIHa2B4DSonzVKTNT4sIL1fq1rykYNUrhnj3q/sEPlFm1it3wcVJYbA8ApUX5qnOR8ePV+o1vKHbOOVIup75HH1Xvb38r19fnOxpqgHOO8gUAJUb5agAWj6v5ppvU9JnPSPG4chs26PC997IfGI7LdXVJ6bSspUVBa6vvOABQFyhfDSQ+d65av/ENRSZMkOvqUs+Pf6zUc8/JhaHvaKhSjHoBQOlRvhpMZNQotXzlK0pcconknNLPPquen/5U4aFDvqOhCg0utudYIQAoGcpXA7JIRMlrrlHz3/yNrKVF+a1b1X3PPewJhg9h5AsASo/y1cBi06er9a67FJ0x4/09wR57TC6f9x0NVYI9vgCg9ChfDS5oaVHzF76g5HXXFfYEW7FCmWXLfMdCFXCplNzBg1I0qmD0aN9xAKBuUL5Q2BPsoovU/PnPS5JSL7zAVhR4f8qxo0MW8KMCAEqFn6gYFD3jDEVOP11KpZR+8UXfceDZ0PIFACgdyhcGmZmSV18tSUovX847IBvcwHovjhUCgNKifOEDohMnKjprlpTLKf3cc77jwCPe6QgA5UH5wockr7pKMlNm9Wrl9+3zHQceuHxeYf9B7JH2ds9pAKC+UL7wIZExYwpnQTqn9NNP+44DD8K9e6UwVDBqlCyR8B0HAOoK5QtHlbziCikaVXbDBuV27PAdBxXGlCMAlA/lC0cVDBum+KJFkqT04sWe06DSBo8VonwBQMlRvnBMiUsvlRIJ5TZvVm7zZt9xUEGMfAFA+VC+cExBU1PhAG5JqaeeknPOcyJUgnNOIeULAMqG8oWPlLjwQllrq/I7dyq3YYPvOKgAd+iQXCola26Wtbb6jgMAdYfyhY9ksZgSl18uSUo9/TSHbjeAoVOOZuY5DQDUH8oXjiu+YIGCUaMU7t+v7Jo1vuOgzAYX23OsEACUBeULx2WRiBJXXSVJSj37rFw26zkRymngWCHWewFAeVC+cEJis2crGD9errtb6eXLfcdBGbHYHgDKi/KFE/KBQ7dffFGur89zIpSDS6UUHjggRSIKxozxHQcA6hLlCycsNn26IqefLqVSSr3wgu84KIPBKceODlnAjwcAKAd+uuKkDIx+ZV5+WeGhQ57ToNRYbA8A5VdU+TKzUWb2pJm91f/ryGNc9xczO2hmjxZzP/gXnThR0dmzpVxOqeee8x0HJcbO9gBQfsWOfH1b0mLn3BmSFvc/Ppp/lvTFIu+FKpG88krJTNnVq5Xfu9d3HJQQ73QEgPIrtnzdJOkn/R//RNLNR7vIObdY0uEi74UqERkzRrEFCyTnlHrmGd9xUCIun1e4Z4+kwpovAEB5FFu+OpxzO/s/3iWpqJ/YZnanma00s5WdnZ1FRkM5JS+/XIpGlduwQbnt233HQQmE+/ZJ+byCkSNliYTvOABQt45bvszsKTNbd5R/bhp6nSuculzUycvOufuccwudcwvHjh1bzEuhzIJhwxRftEiSlFq8mEO368DgYnumHAGgrKLHu8A5d82xPmdmu81svHNup5mNl7SnpOlQ1ZKXXqrMK68ov2WLcps3KzZ9uu9IKAKL7QGgMoqddnxE0h39H98h6eEiXw81xJqalLzkEkmMftUDyhcAVEax5et/SrrWzN6SdE3/Y5nZQjP74cBFZrZE0m8kXW1m28zs40XeF1UifsEFsrY2hTt3Kvvqq77j4BQ5594/VojF9gBQVseddvwozrl9kq4+yvMrJX1tyOPLirkPqpfFYkpefbX6/vAHpRYvVmzWLBZr1yB3+LBcX5+sqUk2bJjvOABQ19jhHkWLzZunyMSJhUO3lyzxHQenYOhiezPznAYA6hvlC0UzMyWvv16SlF62TPn9+z0nwsnKM+UIABVD+UJJRCdNUmz+fCmfV+qJJ3zHwUliZ3sAqBzKF0omefXVUjyu3BtvKPv2277j4CSEvNMRACqG8oWSCdralLis8N6K1OOPy4Wh50Q4ES6dVrh/vxSJKBgzxnccAKh7lC+UVOLCCxWMHKmws1OZFSt8x8EJGJhyDMaOlUUintMAQP2jfKGkLBpV8rrrJEmpZ59V2NvrORGOh81VAaCyKF8ouejMmYpOmyalUko/84zvODgOyhcAVBblCyVnZkp+/OOSmTKrVg1Oa6E6hbzTEQAqivKFsoi0tyu+cKHknPr+8hfOfaxSLgyV37NHEnt8AUClUL5QNokrr5Q1NSn/zjvKbdzoOw6OIty3T8rlZCNGyJJJ33EAoCFQvlA2QVOTEldcIUnqe+IJuVzObyB8COu9AKDyKF8oq/jChQra2+UOHlR66VLfcXAEyhcAVB7lC2VlQaCmj39ckpReskThoUOeE2GokDMdAaDiKF8ou+i0aYqedZaUzSq1eLHvOOjnnGPkCwA8oHyhIpquu06KRJRdu1a5997zHQeSXHe3XG+vlEzKhg/3HQcAGgblCxURjBypxEUXSZJSbD1RFYaOepmZ5zQA0DgoX6iYxGWXyVpbld+xQ9lXX/Udp+HlWe8FAF5QvlAxFo8rec01kqTU4sVy6bTnRI2Nne0BwA/KFyoqNm+eIhMnynV3K71kie84DY3F9gDgB+ULFWVmSl5/vSQpvWyZ8vv3e07UmFwmU9jdPggUjB3rOw4ANBTKFyouOmmSYvPnS/m8Uk895TtOQxoY9QrGjpVFIp7TAEBjoXzBi+TVV0uRiHKvvz54sDMqJ/fuu5IKRRgAUFmUL3gRtLUpvmCBJCn94oue0zSe/DvvSJIiU6d6zQEAjYjyBW8SF18smSn72msKDx70HadhuDAc3Og2OmWK5zQA0HgoX/AmGDlSsblzJeeUfukl33EaRn7nTimTUTB6tIK2Nt9xAKDhUL7gVeKSSyRJmdWrFXZ3e07TGAanHBn1AgAvKF/wKtLerujMmVIup8yyZb7jNITc1q2SpCjrvQDAC8oXvEtceqkkKb1ypVwq5TlNfXNh+H75YuQLALygfMG76KRJhXfdpdNKr1jhO05dy+/aVVjvNXKkgmHDfMcBgIZE+UJVGBj9yixbJpfNek5Tv9hiAgD8o3yhKkSnTVNk/Hi53l5lVq/2HaduMeUIAP5RvlAVzOz9tV8vvSSXz3tOVH8+sN6LkS8A8IbyhaoRnTVLwejRcl1dyq5b5ztO3Ql375bSadmIEQqGD/cdBwAaFuULVeMDo18vvCDnnOdE9SXXv96LUS8A8IvyhaoSmztXNmyYwr17lXvjDd9x6grrvQCgOlC+UFUsEimc+SgpvWQJo18l4pxTnvVeAFAVKF+oOvFzz5U1Nyu/Y4fyW7b4jlMXwt275VIp2fDhCkaM8B0HABoa5QtVx2IxxS+4QJKUeuEFz2nqA+u9AKB6UL5QlRLnny/F48pv2aLc9u2+49Q81nsBQPWgfKEqWVOT4gsXSiq88xGnjvVeAFBdKF+oWomLLpIiEeU2blS+s9N3nJoV7tkj19cnGzZMxnovAPCO8oWqFbS2Kn7OOZKk9Isvek5Tu4ZOOZqZ5zQAAMoXqlrikkskM2Vfe03hwYO+49QkFtsDQHWhfKGqBSNHKjZnjhSGSr/0ku84NWfoeq8Ii+0BoCpQvlD1Bo4cyqxerbCnx3Oa2hJ2dsr19sra2hSMGuU7DgBAlC/UgEh7u6JnninlcsosW+Y7Tk1hvRcAVB/KF2rC4IHbK1bIpVKe09QOphwBoPpQvlATopMnKzJ1qpROK71ype84NcE5x2J7AKhClC/UjMG1X8uWyWWzntNUv3DfPrmeHllLi4LRo33HAQD0o3yhZkSnTVMwfrxcT48yr77qO07VGzrqxXovAKgelC/UDDMr7HovKfPKK57TVD/WewFAdaJ8oabEZs2SJZMKd+5Uftcu33GqFuu9AKB6Ub5QUywaVWzuXEmFfb9wdOH+/XLd3bLmZgVjxviOAwAYgvKFmhNfsECSlF27Vi6X85ymOuX7R70irPcCgKpTVPkys1Fm9qSZvdX/68ijXHOOmS01s/VmttbMPl/MPYHI+PEKxo2TS6WU3bjRd5yqNHRzVQBAdSl25OvbkhY7586QtLj/8ZF6JX3JOXe2pOsl/YuZjSjyvmhwg6NfTD1+COu9AKC6FVu+bpL0k/6PfyLp5iMvcM696Zx7q//jHZL2SBpb5H3R4GJz50qRiHKbNys8eNB3nKoSHjggd/iwrKlJwVj+qAFAtSm2fHU453b2f7xLUsdHXWxmiyTFJb19jM/faWYrzWxlZ2dnkdFQz4KmJsVmzZIkZdas8ZymurDeCwCq23HLl5k9ZWbrjvLPTUOvc845Se4jXme8pJ9J+rJzLjzaNc65+5xzC51zC8fyN3Ycx8DUY2bNGhX+84PEei8AqHbR413gnLvmWJ8zs91mNt45t7O/XO05xnXDJP1J0n91zi075bTAEJHTT5cNHy7X1aX8li2KTpvmO5J3H1jvRfkCgKpU7LTjI5Lu6P/4DkkPH3mBmcUl/V7ST51zvy3yfsAgM1P8nHMksefXAHfwoNyhQ7JkUkHHR64CAAB4Umz5+p+SrjWztyRd0/9YZrbQzH7Yf83nJH1M0t+a2Zr+f84p8r6AJA2Wr+zrr8v19XlO49/AqFdkyhTWewFAlTrutONHcc7tk3T1UZ5fKelr/R//XNLPi7kPcCzBiBGKTpum3ObNyrz2mhKLFvmO5NXgei+2mACAqsUO96h5sYGF90w9stgeAGoA5Qs1L3bWWYXDtnftUn7nzuN/QZ0KDx6UO3hQSiRY7wUAVYzyhZpn0ahi8+ZJauzRr6GjXhbwRxsAqhU/oVEXBo8beu21hj1smy0mAKA2UL5QFyLjxikYP76hD9vOs9geAGoC5Qt1Y3DbiQacegy7uhQeOFBY7zVunO84AICPQPlC3Yg38GHbg+u9TjuN9V4AUOX4KY26YQ182DbrvQCgdlC+UFc+cNh2eNTz2+vSwHqvCOu9AKDqUb5QVyKnny4bMUKuq0u5LVt8x6mI8PBhhfv3S/G4IuPH+44DADgOyhfqytDDthtl4f3glCPrvQCgJvCTGnVnsHxt3KiwAQ7bzg85TBsAUP0oX6g7wfDhik6fLuXzyq5d6ztO2XGYNgDUFsoX6tLQw7adc57TlE9+926F+/ZJySTrvQCgRlC+UJdiM2fKmpoU7t6tcNcu33HKJvPqq5Kk+NlnyyIRz2kAACeC8oW6ZNGoYnPnSpIyr7ziOU15uDBU9rXXJEmx/nVuAIDqR/lC3Rrc82vdOrls1nOa0stt3izX3a1g1ChFJk70HQcAcIIoX6hbkXHjCuug6vSw7Wz/lGNs/nyZmec0AIATRflCXRu68L6euCGFMj5vnuc0AICTQflCXYvPnStFo8pv2aLwwAHfcUomu2GDlMspMnWqghEjfMcBAJwEyhfqmiWTdXnYdqZ//zJGvQCg9lC+UPfq7bDt8MCBwkHa0ahis2f7jgMAOEmUL9S9yNSpCkaOlDt0SNn1633HKdrAqFds1ixZIuE5DQDgZFG+UPfMTInLLpMkpRYvlsvlPCc6dc65D7zLEQBQeyhfaAix+fMVtLfLdXUps3y57zinLL9tm8IDB2StrYqefrrvOACAU0D5QkOwIFDyuuskSaklSxT29npOdGoGR73mzZMF/PEFgFrET280jNj06YpOny6l00o/95zvOCfN5XLK9K9ZizPlCAA1i/KFhpK87jrJTJmVK5Xft893nJOSe+MNKZVSMH68Iu3tvuMAAE4R5QsNJdLeXjiEOgyVeuop33FOCnt7AUB9oHyh4SSvvFKKxZTbuFG5rVt9xzkhYXe3cm+9JQWBYnPn+o4DACgC5QsNJ2hrU+LiiyVJqSeekHPOc6Ljy65bJzmn6IwZClpafMcBABSB8oWGlLj4Yllrq/I7dhSKTZXL9L/LkYX2AFD7KF9oSBaPK3nVVZKqf+PV/O7dCnftkpJJRc8803ccAECRKF9oWLWy8ergqNfZZ8uiUc9pAADFonyhYdXCxqsuDJV97TVJHCcEAPWC8oWGVu0br+Y2b5br7lYwapQikyb5jgMAKAHKFxpe8tprq3bj1Wz/3l6x+fNlZp7TAABKgfKFhhfp6KjKjVddOq3s669LYmNVAKgnlC9A1bnxanbDBimXU2TKFAUjRviOAwAoEcoXoOrceJW9vQCgPlG+gH7VtPFqePCg8lu3StGoYrNne80CACgtyhfQz+LxwvSj/G+8OjDqFZs1S5ZIeMsBACg9yhcwROycc7xvvOqce/9djiy0B4C6Q/kChqiGjVfz27Yp3L9f1tqq6LRpFb8/AKC8KF/AEXxvvJodmHKcN08W8EcUAOoNP9mBo/C18arL5ZRZv14S73IEgHpF+QKO4gMbrz7+eMW2nsi9+aaUSikYN06R9vaK3BMAUFmUL+AYkldeKcXjyr31ltJPP12Re7K3FwDUP8oXcAxBW5uab71VMlP6hReUWb26rPcLe3qU27RJMlNs7tyy3gsA4A/lC/gIsRkzlPzEJyRJfY8+quzbb5ftXtk1a6QwVPSMMxS0tJTtPgAAvyhfwHEkFi5U/OKLpTBU729+o/yePSW/R2btWqUWL5YkxRcsKPnrAwCqB+ULOAHJa64pHPOTTqvnF79QePhwyV47vXy5+n7/e8k5JS69VNGZM0v22gCA6kP5Ak6Amanp5psVmTRJ7tAh9f7qV3KZTFGv6ZxT6tlnlfrLXyQVtrdIXn21zKwUkQEAVYryBZwgi8XUfNttCkaOVH7nTvU+9JBcGJ7SaznnlPrLXwqbuJqp6dOfVuLii0ucGABQjShfwEkIWlrU/IUvyJJJ5d58U6nHHz/p13D5vPr+8AdlXn5ZikTUfOutrPMCgAZSVPkys1Fm9qSZvdX/68ijXDPFzF4xszVmtt7MvlnMPQHfImPGqPm226RIRJmXX1Z62bIT/lqXzar3wQcLB2fHYmr5whcUmzWrjGkBANWm2JGvb0ta7Jw7Q9Li/sdH2inpIufcOZIukPRtM5tQ5H0Br6JTpqjp05+WJKUef1zZjRuP+zUulVLPL36h3Jtvypqa1HLHHRycDQANqNjydZOkn/R//BNJNx95gXMu45xL9z9MlOCeQFWIz5unxJVXSpJ6H3pIue3bj3lt2NOj7p/8RPmtW2VtbWr58pcVnTixUlEBAFWk2CLU4Zzb2f/xLkkdR7vIzCab2VpJ70n6X865Hce47k4zW2lmKzs7O4uMBpRf4rLLCmdA5nLq/dWvFB48+KFrwoMH1XP//Qp37VIwapRav/IVRcaO/f/bu78Yqc46jOPfh92FpYt1F4sFSlshYGNDGxI3TS/axFhaqkHxglRa0Q1eGPWaRBriTRMTNSZG40XTNNlSDaGIUTY12tDWGmNslRoK9GLLtmj4J13RSl0pFPh5cV7M7O4sO7Iz75nZeT7JyZ455z2z73n23Tm/OXNmpoTemplZM5i2+JL0vKTDVaYNle2i+Obhqt8+HBHHIuJOYCUwIKlqkRYRT0REf0T0L/LByVqAJOavX0/H8uXE2BhjO3cS7733v/WXRkf59+Agl8+cYc7ixfRs2cKc3t4Se2xmZmWbtviKiLURsU/+sPwAAAb9SURBVLrKtBc4LWkJQPp51Y/+Tme8DgP31qPzZs1AHR30PPQQcxYt4vLoKGO7dxOXLnHx5EnGBgeJs2fpuOUWFgwMMGfBgrK7a2ZmJZvpy45DwECaHwD2TmwgaZmk+Wm+D7gHGJ7h7zVrKurupueRR1BPD5eOHuU/u3YxtmMHce4cnStX0rN5M+ruLrubZmbWBGZafH0buF/SEWBtuo2kfklPpjYfA16R9BrwW+B7EXFohr/XrOnM6e3luocfhs5OLo6MwIULdK1ezXWbNqGurrK7Z2ZmTULFpVrNp7+/P/bv3192N8z+b+8PD3NuaIiuO+6ge906f12QmVkbkPRqRPTX0raz0Z0xazddt91G59atLrrMzKwqf+aWWQO48DIzs6m4+DIzMzPLyMWXmZmZWUYuvszMzMwycvFlZmZmlpGLLzMzM7OMXHyZmZmZZeTiy8zMzCwjF19mZmZmGbn4MjMzM8vIxZeZmZlZRi6+zMzMzDJy8WVmZmaWkYsvMzMzs4xcfJmZmZll5OLLzMzMLCMXX2ZmZmYZufgyMzMzy0gRUXYfqpI0Cvy17H40wA3A38vuRBNxHpM5k/Gcx2TOZDznMZkzGS9HHrdGxKJaGjZt8TVbSdofEf1l96NZOI/JnMl4zmMyZzKe85jMmYzXbHn4ZUczMzOzjFx8mZmZmWXk4iu/J8ruQJNxHpM5k/Gcx2TOZDznMZkzGa+p8vA1X2ZmZmYZ+cyXmZmZWUYuvszMzMwycvHVAJIWSton6Uj62Velza2S/izpgKTXJX21Yt1LkobTugOSPpx3D+qrDnl8XNIhSSOSfihJefeg/mrMZI2kP6Q8Dkr6fMW6pyQdrRgja/LuQX3VIY/lkl5JY+QZSXPz7kH91ZJJavdrSe9IenbC8rYbI6ndVHm08xgZSG2OSBqoWD4rjjWSHkz7MSJpW5X189LffCSNgY9UrHs0LR+WtC5bpyPCU50n4LvAtjS/DfhOlTZzgXlpfgHwF2Bpuv0S0F/2fjRRHn8E7gYE/Ar4VNn7lCmTjwKr0vxS4BTQm24/BWwsez+aKI/dwKY0/zjwtbL3KUcmad19wGeAZycsb7sxMk0ebTlGgIXAW+lnX5rvS+ta/lgDdABvAivSceQ14PYJbb4OPJ7mNwHPpPnbU/t5wPJ0Px05+u0zX42xAdiR5ncAn5vYICIuRMT5dHMes/ss5DXnIWkJcH1EvBzFf8vT1bZvQbVk8kZEHEnzJ4G3gZo+PbkFXXMe6UzoJ4E9V9u+BU2bCUBEvAC8m6tTJbrmPNp8jKwD9kXEPyLin8A+4MFM/cvhLmAkIt6KiAvALopcKlXmtAe4L42JDcCuiDgfEUeBkXR/DTebD/hlujEiTqX5vwE3Vmsk6WZJB4FjFM9YTlasHkyngb85C15mm0keNwHHK5odT8taXU2ZXCHpLopndW9WLP5Wevnt+5LmNaifucwkjw8B70TExbS6LcfIFNp2jEzQzmPkJorH1Csm7nurH2um279xbdIY+BfFmKhl24bozPFLZiNJzwOLq6zaXnkjIkJS1c/ziIhjwJ2SlgK/kLQnIk4DX4iIE5I+APwM+CLFGZ+m1ag86t/TfOqRSbqfJcCPgYGIuJwWP0rxYDuX4vNrvgE8Vo9+N0qj8mjN40WhXplMoW3HyGzS4Exa7lgzW7j4ukYRsXaqdZJOS1oSEafSgeLtae7rpKTDwL3Anog4kZa/K2knxWnQpv6HaGAevweWVaxeBpyoR58brR6ZSLoe+CWwPSJerrjvK892z0saBLbWsesN0cA8zgC9kjrTs9q2GiNXue+2HCNTaOcxcgL4RMXtZRTXetGKx5oqTgA3V9yu9re90ua4pE7ggxRjopZtG8IvOzbGEHDlHSUDwN6JDSQtkzQ/zfcB9wDDkjol3ZCWdwHrgcNZet0415xHOoCclXR3OiX+pWrbt6BaMpkL/Bx4OiL2TFi3JP0UxXUe7TBGquaRrgX8DbDxatu3oGkzuZp2HCNTafMx8hzwgKS+9Nj6APDcLDrW/AlYpeLdrHMpLqgfmtCmMqeNwItpTAwBm9K7IZcDqyje4NV4Oa7qb7eJ4rXkF4AjwPPAwrS8H3gyzd8PHKR4p8VB4CtpeQ/walr2OvADMr37ohnzqGh3mOL6nh+RvpmhlacaM9kMvA8cqJjWpHUvAodSLj8BFpS9TyXnsYLiQXME+CnpnbOtPNWSSbr9O2AUOEdxzcq6dh0j0+TRzmPky2m/R4AtadmsOdYAnwbeSMeI7WnZY8Bn03x3+puPpDGwomLb7Wm7YTK+k95fL2RmZmaWkV92NDMzM8vIxZeZmZlZRi6+zMzMzDJy8WVmZmaWkYsvMzMzs4xcfJmZmZll5OLLzMzMLKP/AutYV0B3k+lQAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "011914_cell_2\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x576 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x576 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x576 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "i = 0;\n",
    "for cell in rf.keys():\n",
    "    if i<2:\n",
    "        print(cell)\n",
    "        r = rf[cell] # 38 x 38 x 38\n",
    "        #plt.plot(r)\n",
    "        vis.spatial(r)\n",
    "        vis.temporal(time, r)\n",
    "    i = i+1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[0;31mSignature:\u001b[0m \u001b[0mvis\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtemporal\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtime\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfilt\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
       "\u001b[0;31mDocstring:\u001b[0m\n",
       "Plot the temporal component of a full linear filter.\n",
       "\n",
       "If the given linear filter is 1D, it is assumed to be a temporal filter,\n",
       "and is plotted directly. If the filter is 2 or 3D, it is decomposed into\n",
       "its spatial and temporal components, and the temporal component is plotted.\n",
       "\n",
       "Parameters\n",
       "----------\n",
       "time : array_like\n",
       "    A time vector to plot against.\n",
       "\n",
       "filt : array_like\n",
       "    The full filter to plot. May be than 1D, but must match in size along\n",
       "    the first dimension with the ``time`` input.\n",
       "\n",
       "ax : matplotlib Axes object, optional\n",
       "    the axes on which to plot the data; defaults to creating a new figure\n",
       "\n",
       "Returns\n",
       "-------\n",
       "fig : matplotlib.figure.Figure\n",
       "    The figure onto which the temoral STA is plotted.\n",
       "\n",
       "ax : matplotlib Axes object\n",
       "    Axes into which the temporal STA is plotted\n",
       "\u001b[0;31mFile:\u001b[0m      /usr/local/lib/python3.7/site-packages/pyret-0.6.0-py3.7.egg/pyret/visualizations.py\n",
       "\u001b[0;31mType:\u001b[0m      function\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "vis.temporal?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "vis.play_sta?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "resp = d['boxes']['membrane_potential']\n",
    "\n",
    "cell = 0;\n",
    "\n",
    "a = resp[cell]\n",
    "plt.plot(a)\n",
    "plt.plot(rf[cell])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "#\n",
    "with h5py.File(filename, 'r') as f:\n",
    "    print('Keys :', list(f.keys())) # contents of the dataset\n",
    "                          # 'boxes', 'expts', 'flashes', 'lines' \n",
    "    for key in f.keys():\n",
    "  \n",
    "        print(key)\n",
    "    \n",
    "        dset = f[key]\n",
    "        \n",
    "        if key == 'expts': # cell names # sam\n",
    "            data[i][key] = np.array(dset)\n",
    "            \n",
    "        else:\n",
    "            data[i][key] = dict()\n",
    "            \n",
    "            print(list(dset.keys()))\n",
    "            #\n",
    "            for kk in dset.keys():\n",
    "                \n",
    "                data[i][key][kk] = np.array(dset[kk])\n",
    "    #\n",
    "    #dset = f['exp1']\n",
    "    #print(list(dset.keys()))\n",
    "    #print(list(dset.attrs.keys()))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data[0]['lines'].keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(data[0]['lines']['rfs'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(data[0]['expts'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "f = h5py.File(filename, 'r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "f['lines'].keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "f['expts']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "is_dataset = isinstance(item, h5py.Dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def h5py_dataset_iterator(g, prefix=''):\n",
    "    for key in g.keys():\n",
    "        item = g[key]\n",
    "        path = '{}/{}'.format(prefix, key)\n",
    "        if isinstance(item, h5py.Dataset): # test for dataset\n",
    "            yield (path, item)\n",
    "        elif isinstance(item, h5py.Group): # test for group (go down)\n",
    "            yield from h5py_dataset_iterator(item, path)\n",
    "\n",
    "with h5py.File('test.hdf5', 'r') as f:\n",
    "    for (path, dset) in h5py_dataset_iterator(f):\n",
    "        print(path, dset)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "#date = '17-11-08'\n",
    "#retina_num = 1\n",
    "#cell_num = 5\n",
    "\n",
    "### Rarely change\n",
    "fine_name_i_rf_line = date + '-r' + str(retina_num) + '-c' + str(cell_num) + '-1-line.h5'\n",
    "\n",
    "\n",
    "frame_num_rf_30m = 54000\n",
    "kernel_length_rf = 0.5 # in [s]\n",
    "upsample_factor_g_rf = 4\n",
    "\n",
    "#\n",
    "frame_num_bar_25r = 26100\n",
    "frame_time = 0.01666574578544061\n",
    "up_sampling = 16\n",
    "\n",
    "repeat_time = 23 # roughly takes 18[s] for each repeat (8 trials)\n",
    "sigma_fr = 0.01\n",
    "\n",
    "gain = 0.000152587890625\n",
    "#gain = np.double(f['data'].attrs['gain'])\n",
    "offset = 0\n",
    "threshold_intra_rf_line = 0.9\n",
    "sample_rate = 10000.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "order_l = 4 \n",
    "order_h = 3\n",
    "cutoff_l = 15      # desired cutoff frequency of the lowpass filter, [Hz]\n",
    "cutoff_h = 0.04\n",
    "\n",
    "def butter_lowpass(cutoff, fs, order=5):\n",
    "    nyq = 0.5 * fs\n",
    "    normal_cutoff = cutoff / nyq\n",
    "    b_l, a_l = butter(order, normal_cutoff, btype='low', analog=False)\n",
    "    return b_l, a_l\n",
    "\n",
    "def butter_lowpass_filter(data, cutoff, fs, order=5):\n",
    "    b_l, a_l = butter_lowpass(cutoff, fs, order=order)\n",
    "    y_l = filtfilt(b_l, a_l, data)\n",
    "    # filtfilt is to compensate delay introduced by this lowpass filter\n",
    "    return y_l\n",
    "\n",
    "def butter_highpass(cutoff, fs, order=5):\n",
    "    nyq = 0.5 * fs\n",
    "    normal_cutoff = cutoff / nyq\n",
    "    b_h, a_h = butter(order, normal_cutoff, btype='high', analog=False)\n",
    "    return b_h, a_h\n",
    "\n",
    "def butter_highpass_filter(data, cutoff, fs, order=5):\n",
    "    b_h, a_h = butter_highpass(cutoff, fs, order=order)\n",
    "    y_h = filtfilt(b_h, a_h, data)\n",
    "    return y_h"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "# Load file\n",
    "f_intra_rf_line = h5py.File(data_dir + fine_name_i_rf_line, 'r')\n",
    "pd_intra_rf_line = np.double(f_intra_rf_line['data'][0, :])\n",
    "pd_intra_rf_line = pd_intra_rf_line - np.min(pd_intra_rf_line)\n",
    "pd_intra_rf_line = pd_intra_rf_line / np.max(pd_intra_rf_line)\n",
    "maxtab_intra_rf_line, mintab_intra_rf_line = spk.peakdet(pd_intra_rf_line, threshold_intra_rf_line)\n",
    "print(maxtab_intra_rf_line, '\\n', len(maxtab_intra_rf_line))\n",
    "volt_intra_rf_line = (-np.double(f_intra_rf_line['data'][1, :]) * gain - offset) * 0.1 * 1000 # [mV]\n",
    "f_intra_rf_line.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Stim data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load stimulus\n",
    "stim_dir = '/Users/peterfish/Modules/1d_wn_amacrine_DL/'\n",
    "stim_filename = 'lineColor_w4.mat'\n",
    "f_stim_rf_line = sio.loadmat(stim_dir + stim_filename)\n",
    "stim_intra_rf_line = np.array(f_stim_rf_line['lineColor'])[0, :, :4500]\n",
    "## [0, :, :4500]??\n",
    "stim_intra_rf_line = np.reshape(stim_intra_rf_line, [64, 4500])       \n",
    "stim_intra_rf_line = np.einsum('ij->ji', stim_intra_rf_line)\n",
    "stim_intra_rf_line.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "imshow(np.vstack((stim_intra_rf_line[0, :], stim_intra_rf_line[0, :])), cmap='gray')\n",
    "# vstack?? 2 identical lines. (1 line is too thin..)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "# normalize the stimulus \n",
    "stim_intra_rf_line = zscore(stim_intra_rf_line)\n",
    "# upsampling\n",
    "stim_intra_rf_line_us = st.upsample(stim_intra_rf_line, up_sampling)[0]\n",
    "# zscore again?\n",
    "stim_intra_rf_line_us_input = zscore(stim_intra_rf_line_us)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = np.linspace(1, 10, 10)\n",
    "print(a)\n",
    "a[-3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Recording data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(files)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rcParams['figure.figsize'] = [5, 3]\n",
    "# plot style test\n",
    "print(plt.style.available)\n",
    "#plt.style.use(['classic'])\n",
    "plt.style.use(['seaborn-dark','dark_background'])\n",
    "# plt.style.use('seaborn-dark') # Working in dark theme jupyter lab. Figures for white presentation\n",
    "#\n",
    "plt.rcParams['font.size'] = 20\n",
    "#plt.rcParams['axes.labelsize'] = 40\n",
    "plt.rcParams['xtick.labelsize'] = 18\n",
    "plt.rcParams['ytick.labelsize'] = 18\n",
    "#\n",
    "a = np.linspace(1, 10, 10)\n",
    "plt.plot(a)\n",
    "plt.xlabel('xlabel')\n",
    "plt.ylabel('ylabel')\n",
    "plt.title('title')\n",
    "# etc\n",
    "#ax2.xaxis.grid()\n",
    "#ax2.set_xlim([0,n_data_res])\n",
    "#ax2.set_yticklabels([])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "keys = ['resp']\n",
    "values \n",
    "data = [dict(zip(keys, values)) for c in range(n_cell)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a list of ditionaries with just cell id? Then, I can add more fields in a flexible way?\n",
    "thisdic = { \n",
    "\"cell id\": 1,\n",
    "\"test\": \"asdf\"\n",
    "}\n",
    "thisdic[\"newkey\"] = 3\n",
    "print(thisdic)\n",
    "l = [thisdic for c in range(3)]\n",
    "print(l)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#l[1][\"newkey\"] = np.linspace(1, 10, 10)\n",
    "l[1][\"newkey\"] = 5\n",
    "print(l[1])\n",
    "print(l)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "listdic = [dict() for c in range(3)]\n",
    "print(listdic)\n",
    "listdic[0]['f'] = 1\n",
    "listdic[0]['g'] = 3\n",
    "listdic[1]['f'] = 'asdf'\n",
    "listdic[2]['f'] = np.linspace(1, 5, 5)\n",
    "print(listdic)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load all recording data files\n",
    "n_cell = len(files)\n",
    "data = [dict() for c in range(n_cell)]\n",
    "ignorebefore = 100\n",
    "# plot properties\n",
    "n_subplots = 6 \n",
    "\n",
    "for i in range(n_cell):\n",
    "    filename = data_dir+files[i]\n",
    "    with h5py.File(filename, 'r') as f:\n",
    "        #print(\"Keys: %s\" % f.keys())\n",
    "        fig = plt.figure(figsize=(18, 3))\n",
    "        #ax1 = plt.subplot(1, n_subplots, 1)\n",
    "        #ax2 = plt.subplot(1, n_subplots, 3)\n",
    "        ax1 = plt.subplot2grid((1, n_subplots), (0,0), colspan=1)\n",
    "        ax2 = plt.subplot2grid((1, n_subplots), (0,1), colspan=n_subplots-1)\n",
    "\n",
    "        # Get PD signal and center it.\n",
    "        pd_intra_rf_line = np.double(f['data'][0, :])\n",
    "        pd_intra_rf_line = pd_intra_rf_line - np.min(pd_intra_rf_line)\n",
    "        pd_intra_rf_line = pd_intra_rf_line / np.max(pd_intra_rf_line)\n",
    "        # Peak detection\n",
    "        maxtab_intra_rf_line, mintab_intra_rf_line = spk.peakdet(pd_intra_rf_line, threshold_intra_rf_line)\n",
    "        #print(maxtab_intra_rf_line, '\\n', len(maxtab_intra_rf_line))\n",
    "        \n",
    "        # Get Vol signal (2nd channel)\n",
    "        volt_intra_rf_line = (-np.double(f['data'][1, :]) * gain - offset) * 0.1 * 1000 # [mV]\n",
    "        # raw data plot\n",
    "        #ax2.plot(volt_intra_rf_line)\n",
    "        \n",
    "        # end-data floating error handling\n",
    "        volt_intra_rf_line[-10100:] = np.ones(volt_intra_rf_line[-10100:].shape[0]) * volt_intra_rf_line[-10100]\n",
    "        \n",
    "        # filtering\n",
    "        volt_intra_rf_line_hf = butter_highpass_filter(volt_intra_rf_line, cutoff_h, sample_rate, order_h)\n",
    "        volt_intra_rf_line_hf_lf = butter_lowpass_filter(volt_intra_rf_line_hf, cutoff_l, sample_rate, order_l)\n",
    "        \n",
    "        # down-sampling & resampling\n",
    "        volt_intra_rf_line_ds = st.downsample(volt_intra_rf_line_hf_lf[np.int(maxtab_intra_rf_line[0, 0]):\n",
    "                                                               np.int(maxtab_intra_rf_line[2, 0])], 10)[0]\n",
    "        volt_intra_rf_line_ds_rs = scipy.signal.resample(volt_intra_rf_line_ds, stim_intra_rf_line_us.shape[0])\n",
    "        # data centering\n",
    "        #volt_intra_rf_line_ds_rs_output = volt_intra_rf_line_ds_rs - np.mean(volt_intra_rf_line_ds_rs) # mean substraction by DL\n",
    "        volt_intra_rf_line_ds_rs_output = zscore(volt_intra_rf_line_ds_rs)\n",
    "        vol_resampled_centered = zscore(volt_intra_rf_line_ds_rs)\n",
    "        #\n",
    "        n_data_res = len(vol_resampled_centered)\n",
    "        #\n",
    "        print('File id: %d (%s)' % (i+1, files[i]))\n",
    "\n",
    "        # Plot the filtered, resampled trace.\n",
    "        ax2.plot(vol_resampled_centered)\n",
    "        ax2.set_xlim([0,n_data_res])\n",
    "        ax2.xaxis.grid()\n",
    "        ax2.set_yticklabels([])\n",
    "        \n",
    "        # ignore until 100 time points?\n",
    "        stim = stim_intra_rf_line_us_input[ignorebefore:n_data_res]\n",
    "        resp = vol_resampled_centered[ignorebefore:n_data_res]\n",
    "        data[i]['resp'] = resp\n",
    "        rc_line, rags_line = ft.revcorr(stim, resp, 300)\n",
    "        #\n",
    "        frame_line_500ms = np.int((0.5 / (rags_line.shape[0] * frame_time/8)) * rags_line.shape[0])\n",
    "        #\n",
    "        im = ax1.imshow(rc_line[-frame_line_500ms:].T, cmap='seismic', interpolation='nearest', \n",
    "                       vmin=1.6*np.min(rc_line), vmax=-1.6*np.min(rc_line), aspect='auto')\n",
    "        cb = plt.colorbar(im, ax=ax, fraction=0.03, pad=0.02)\n",
    "        cb.ax.set_yticklabels(cb.ax.get_yticklabels(), fontname='Arial', fontsize=12)\n",
    "        ax1.axis('off')\n",
    "        plt.tight_layout()\n",
    "        plt.show()\n",
    "        # save to npy file\n",
    "        #np.save('cell_id_%d__resp_' % i + files[i] + '.npy' , resp)\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save resampled trace as npy\n",
    "# Cell id from 1 (i+1) to ...\n",
    "for i in range(n_cell):\n",
    "    np.save( data_dir+'cell__%03d__resp_' % (i+1) + files[i] + '.npy' , data[i]['resp'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Temporal filters: Who has a biphasic filter?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "volt_intra_rf_line[-10100:] = np.ones(volt_intra_rf_line[-10100:].shape[0]) * volt_intra_rf_line[-10100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "volt_intra_rf_line_hf = butter_highpass_filter(volt_intra_rf_line, cutoff_h, sample_rate, order_h)\n",
    "volt_intra_rf_line_hf_lf = butter_lowpass_filter(volt_intra_rf_line_hf, cutoff_l, sample_rate, order_l)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "volt_intra_rf_line_ds = st.downsample(volt_intra_rf_line_hf_lf[np.int(maxtab_intra_rf_line[0, 0]):\n",
    "                                                               np.int(maxtab_intra_rf_line[2, 0])], 10)[0]\n",
    "volt_intra_rf_line_ds_rs = scipy.signal.resample(volt_intra_rf_line_ds, stim_intra_rf_line_us.shape[0])\n",
    "volt_intra_rf_line_ds_rs_output = volt_intra_rf_line_ds_rs - np.mean(volt_intra_rf_line_ds_rs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "# ignore until 100 time points?\n",
    "stim = stim_intra_rf_line_us_input[100:72000]\n",
    "resp = volt_intra_rf_line_ds_rs_output[100:72000]\n",
    "\n",
    "#rc_line, rags_line = ft.revcorr(stim_intra_rf_line_us_input[100:72000], volt_intra_rf_line_ds_rs_output[100+300-1:72000], 300)\n",
    "# For pyret 0.6.0 (Juyoung)\n",
    "rc_line, rags_line = ft.revcorr(stim_intra_rf_line_us_input[100:72000], volt_intra_rf_line_ds_rs_output[100:72000], 300)\n",
    "#rc_line = rc_line - np.mean(rc_line, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "frame_line_500ms = np.int((0.5 / (rags_line.shape[0] * frame_time/8)) * rags_line.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "fig = plt.figure(figsize=(6, 3))\n",
    "ax = fig.add_subplot(111)\n",
    "im = ax.imshow(rc_line[-frame_line_500ms:].T, cmap='seismic', interpolation='nearest', \n",
    "               vmin=1.6*np.min(rc_line), vmax=-1.6*np.min(rc_line), aspect='auto')\n",
    "cb = plt.colorbar(im, ax=ax, fraction=0.03, pad=0.02)\n",
    "cb.ax.set_yticklabels(cb.ax.get_yticklabels(), fontname='Arial', fontsize=12)\n",
    "ax.axis('off')\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "# save\n",
    "np.save(fine_name_i_rf_line+'_resp'+'.npy', resp);\n",
    "np.save(fine_name_i_rf_line+'_stim'+'.npy', stim);\n",
    "np.save('resp.npy', resp);\n",
    "np.save('stim.npy', stim);"
   ]
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}