SynapseModel/STDP_Reward.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING    c:\\users\\valentin\\appdata\\local\\programs\\python\\python37-32\\lib\\site-packages\\sympy\\matrices\\matrices.py:2479: SymPyDeprecationWarning: \n",
      "\n",
      "Dot product of non row/column vectors has been deprecated since SymPy\n",
      "1.2. Use * to take matrix products instead. See\n",
      "https://github.com/sympy/sympy/issues/13815 for more info.\n",
      "\n",
      "  useinstead=\"* to take matrix products\").warn()\n",
      " [py.warnings]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABQgAAAKACAYAAADQJJVVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzs3XmczXX///Hna8zY0lgyyNIMSiUVmpKlrJeSkDahkLr86ip0LS2XukLlQl0tqm8bhdCqkkTShYpKxlKIkkhSiFCTfV6/P845c81yZsxY5pxxHvfb7dxmzvu8P5/P6/N5mz6fXue9mLsLAAAAAAAAQGyKi3QAAAAAAAAAACKHBCEAAAAAAAAQw0gQAgAAAAAAADGMBCEAAAAAAAAQw0gQAgAAAAAAADGMBCEAAAAAAAAQw0gQAgAAAAAAADGMBCEAAAAAAAAQw0gQAgAAAAAAADEsPtIBFBeVK1f2lJSUSIcBAAAAAAAAFMiiRYt+cfekg9UjQVhAKSkpSktLi3QYAAAAAAAAQIGY2fcFqccQYwAAAAAAACCGkSAEAAAAAAAAYhgJQgAAAAAAACCGkSAEAAAAAAAAYhgJQgAAAAAAACCGkSAEAAAAAAAAYhgJQgAAAAAAACCGkSAEAAAAAAAAYhgJQgAAAAAAACCGkSAEAAAAAAAAYhgJQgAAAAAAACCGkSAEAAAAAAAAYhgJQgAxZd26dTIz9enTJ9Kh5DJkyBCZmebOnVtkx+zTp4/MTOvWrSuyYwIAgOKtVatWMrNIh1FsHKnnrXHjxsnMNG7cuCMSV1ExM7Vq1SrSYQA4CBKEAIq9VatWqX///mrQoIHKly+vkiVLqnr16urYsaOef/557d69O9IhHlMWLlyonj17Kjk5WaVKlVJiYqLq1q2rTp066cEHH1R6erokKSUlRWZW4NeQIUMk/S9RGnrFxcUpMTFRycnJuuSSSzRy5Ej9+OOPYWMLJYCzvuLj41W1alV17NhRM2bMKKrLBAAoxnLeS0qVKqWkpCQ1btxYN954o2bMmKEDBw5EOkzgkETiS+nCOHDggEaPHq2WLVuqUqVKSkhIUJUqVXTWWWfpxhtv1NSpUyVJc+fOLdSzZtYkbc7n1ISEBJ1wwgk688wzdd111+n111/X3r17w8aX81nVzFSmTBnVq1dPt9xyizZs2FBUlwo4ouIjHQAAHI777rtPQ4cOVUZGhs4//3z17t1b5cqV06ZNmzR37lzdeOONevrpp5WWlhbpUA/q1ltv1TXXXKOTTjop0qHkaeLEierdu7fcXW3atFHXrl1VokQJrV27VmlpaZo2bZouv/xynXzyybrtttu0ffv2bNuPGzdO33//vXr37q2UlJRsn+X8Zrlly5aZZenp6frpp580f/58zZgxQ4MHD9aQIUN01113hY2zfPnyuu222yRJu3fv1hdffKHp06dr+vTpGjVqlAYMGHBErgcA4Ng2ePBgSYGExfbt27VixQpNmDBBzz//vFJTUzVp0iTVq1cvwlEWvRdffFF//PFHpMPAMejAgQO69NJL9d5776lChQrq2LGjatasqW3btmnNmjV66aWXtGrVKnXu3FkpKSmZf6Mh27dv16hRo7I9C2ZVoUKFbO8HDhyoChUqKCMjQzt37tTXX3+tt956SxMnTtQpp5yiCRMmqEmTJmFjzfqs+ssvv+j999/XU089pddee02fffaZ6tate2QuClBU3J1XAV7nnHOOA4guw4YNc0leq1Yt/+yzz8LWeeedd7xVq1aZ79euXeuSvHfv3kUUZXTr3bu3S/K1a9cetG56eronJiZ6iRIl/IMPPghbZ/78+f7rr7/muY+WLVu6JJ8zZ06edQYPHuySfPDgwbk+y8jI8MmTJ3ulSpVckg8fPjzb56H2TU5OzrXtCy+84JL8uOOO8/T09DyPDwCAJA/8r1JuP//8s1911VWZzyCbNm0q4uhQ3BTmeSs/Y8eOdUk+duzYw9pP6Fkrv+exI0mSt2zZskB1J0yY4JL87LPP9u3bt+f6PD093WfPnp3n9vk9C2aVnJycZ5ts377d+/fv75K8fPnyvnLlymyf5/WsunfvXm/btq1L8j59+uR7fKAoSUrzAuS9GGIMoFhat26dhgwZooSEBE2fPj3Pb/ZC30AezDfffKO77rpLqampSkpKUqlSpZScnKx+/fqFHSbg7ho/fryaNWumpKQklS5dWrVq1dJFF12kV199NVvdL7/8Ut27d1dKSkq2IUq33Xab9u3bl1kvv+Eeq1atUt++fTP3UaVKFV1wwQV6+umns9WbMmWKrr32WtWrV0/HHXecypUrp3POOUePP/64MjIyDnod8rN8+XLt3LlTDRo0UNu2bcPWadasWa5vZo8kM9MVV1yhyZMnS5KGDh2qn376qUDb9unTR8cdd5zS09O1YsWKoxYjAODYVrVqVb3yyitq1aqVfvjhB/373//OVWf16tXq1auXatSokTn1Sa9evbR69epcdbPe/8ePH69GjRqpTJkyqlKlivr27auff/451zaLFi3SwIEDdfbZZ6tSpUoqXbq0TjnlFP3973/Xr7/+mqt+1rnrZs2apQsuuEDlypVTUlKSrr/++swe/0uWLNGll16qihUrqly5curcuXPYefPCzUEYGu45ZMgQLV26VB07dlSFChVUtmxZtWzZUp988knY67l//3499dRTOv/885WYmKiyZcuqUaNGevLJJwv17BKKad++fbrvvvtUt25dlS5dWqeddppGjx6dWe+ZZ57RmWeeqTJlyqhmzZoaPHhwnsd57bXXdOGFF6p8+fIqU6aMzjzzTA0fPlx79uwJW/+DDz7QBRdcoOOOO06VKlXSZZddplWrVoWtm/V6hZOSkpJrtEVe5syZo379+ql+/fpKTExUmTJl1KBBAw0dOjTXVDspKSkaOnSoJKl169bZhslm9ccff2j48OFq2LBh5jNl06ZN9fLLL4eNYe/evbr//vtVt25dlSpVSrVr19Y999yT57XKS+jfSZ8+fVS+fPlcn5ctW1atW7cu1D4Lq3z58nr88cfVq1cv7dixI88RKzklJCSoX79+kqTPP//8aIYIHBUMMQZQLI0dO1b79u3TNddcowYNGuRbt1SpUgfd35tvvqlnnnlGrVu3VrNmzVSyZEmtWLFCY8aM0TvvvKO0tDTVqFEjs/7dd9+t4cOHq3bt2rr66qtVvnx5/fTTT1q4cKFef/11devWTVIgOdikSROZmTp37qzatWtr586d+vbbb/XUU0/pgQceUEJCQr6xvfvuu7rqqqu0Z88eXXzxxerevbu2b9+uL774Qg8++KBuvvnmzLp33XWX4uLi1KRJE9WoUUM7duzQ7NmzNXDgQC1cuFATJkw46LXIywknnCBJ2rhxo9LT03Xccccd8r4OV+vWrdWiRQvNmzdPb775pm655ZZCbc/E6gCAwxEXF6d77rlHc+fO1csvv6xHH300896ycOFCtWvXTr/99ps6d+6s+vXra9WqVZo0aZLefvtt/fe//1VqamqufT766KN6//331a1bN1188cWaN2+exo4dq7lz52rBggVKSkrKrDt69Gi99dZbatmypdq1a6cDBw5o8eLFeuSRRzRjxgwtWLBAxx9/fK5jTJ06VdOmTdOll16qm266SZ988onGjRuntWvXasSIEWrbtq0uuOAC3XDDDVq2bJneeecdrVmzRsuWLVNcXMH6lqSlpenBBx9U06ZNdeONN2r9+vV644031LZtWy1dulSnnnpqZt19+/apU6dOmjlzpk499VT16NFDpUuX1pw5c9S/f38tWLCg0M8u11xzjRYsWKBLLrlECQkJmjx5svr166eEhAR9+eWXGj9+vC699FK1bdtWU6dO1X333aeyZcvqzjvvzLafQYMGafjw4apcubJ69OihcuXKacaMGRo0aJBmzpypWbNmZXuGmzx5srp166aSJUuqW7duOvHEEzVv3jw1bdpUZ511VqHOobBGjhypVatWqVmzZurYsaN2796t+fPna8iQIZo7d64++OADlShRQpJ02223acqUKfrwww/DTvkiBYbptmnTRkuWLFHjxo3Vt29fZWRkaObMmerRo4dWrFihBx54ILO+u+vqq6/W22+/rbp16+rWW2/V3r179cILL2jZsmWFOpfQ8+Y333xz6BfkCBk8eLBefPFFTZs2TTt37lRiYuJBtwl01uJZE8VUQboZ8mKIMRBt2rRp45J89OjRhdouryHGGzZs8N27d+eqP3PmTI+Li/ObbropW3mlSpW8Ro0aYYeqbtmyJfP3v/3tby7Jp0yZkqvetm3b/MCBA5nvww332LJliycmJnpCQoLPnTs31z5++OGHbO+//fbbXHUOHDjgvXr1ckm5hmIXZshLRkaGn3vuuZnDPp588klfvHix79mz56DbhhzuEOOs7rnnHpfkvXr1yizLb1jJ888/nznE+I8//ihwzACA2KN8hhiH7N692+Pj412Sf/fdd+4euFeedtppLsknTpyYrf4rr7zikvzUU08Ne/9PSEjwxYsXZ9vmtttuc0net2/fbOXr1q3z/fv354ppzJgxLslHjBiRrTw0NLVEiRLZnicOHDjg7dq1c0lesWLFXDH37ds37HNM6H6e1Zw5czKvW84hsM8884xL8ptvvjlbeejcb7311mzns3///jyPnZdQTKmpqdmmO1mzZo0nJCR4hQoVPCUlxTds2JD52a+//uonnHCCV65c2fft25dZ/sknn2QOIf/pp58yy/ft2+eXXnqpS/Jhw4Zllv/2229eqVIlj4+P94ULF2aLK9SGOZ+3Qtcrr+ed5OTkXM8zeQ0xXrNmjWdkZOTaR+hZ6ZVXXslWfrAhxqHnw5EjR2Yr37Vrl1900UVuZr5kyZLM8kmTJrkkP//8833Xrl2Z5Vu3bvU6deoUaojx4sWLPSEhwc3Mr732Wn/jjTd83bp1BdrW/cgMMc6qZs2aLinbsOa8nlX37duX+f8oOf9mgUgSQ4wBHMtCw0pr1qx5RPZXo0aNsD0N27dvrzPOOEMzZ87M9VlCQkLmt7FZVa5cOVdZmTJlcpVVrFjxoN/Gjx8/Xjt37tTNN9+sli1b5vo85/mHmww5Li5OAwcOlKSw51FQZqbJkyerVatW+uKLL3TrrbeqcePGKleunJo0aaKRI0dq586dh7z/wgr16NyyZUuuz7Zv364hQ4ZkLmRyySWX6IYbbpAk/fvf/w7bHgAAFEapUqUyezuF7kWffPKJVq1apaZNm6pnz57Z6nfr1k0tWrTQ119/rXnz5uXa33XXXadGjRplKxsyZIjKly+vl156KdtQzeTk5LDPIH379lViYmKe9/vu3btne56Ii4vTddddJ0lq0KBBrph79eolSVq6dGn4ixBG8+bN1adPn1xxxcfHZxt2mZGRoSeffFLVqlXTo48+mu18SpQooYcfflhmpkmTJhX42JI0YsSIbNOd1KlTRy1atND27dv1r3/9K9uIkAoVKqhTp0765Zdf9OOPP2aWv/DCC5Kke+65R9WqVcssj4+P18MPP6y4uDiNGTMms/ztt9/Wtm3b1KNHj1y9Q0NteDTVqVMnbI+10CIdhXn+27p1qyZOnKjU1FTdcccd2T4rXbq0Ro4cKXfXSy+9lFk+duxYSYFnrNKlS2eWV6pUSf/6178KdS6NGjXSxIkTVbVqVU2cOFFXXHGFUlJSdMIJJ6hr16565513CrW/w5Xf8+bcuXMznzf79++v+vXra/bs2apcubLuvvvuIo0TOBIYYgygWPIj3H3f3TVp0iSNGzdOX3zxhX799VcdOHAg8/OSJUtmq9+zZ0898cQTOuOMM3TVVVepZcuWatq0aa4HwG7dumnUqFG67LLLdOWVV6pdu3Zq3rx5gVc1++yzzyRJHTp0KFD9rVu36qGHHtL06dP13XffKT09PdvnWR9+D8VJJ52kOXPmaOXKlZo1a5bS0tL0+eefZ76eeuopzZ07V7Vr1z6s4xREfv8GduzYkTm/TokSJVSpUiV16NBBt956qy655JKjHhsAIDbkvBctXrxYktSmTZuw9du0aaN58+ZpyZIluvDCC7N9Fu6LwPLly6thw4b68MMPtXLlSjVs2FBSYGjus88+q1deeUVfffWVduzYkW0evbzu9+GGNlevXl2SdM455+T6LJQcCTcfc17CHSMhIUFVq1bNNj/iN998o61bt+qUU07JNlw1qzJlymjlypUFPnZexy/oOSYnJ0vKvx3r1aunmjVrau3atdq+fbsqVKiQWf9gbXi0pKena9SoUXrrrbf0zTff6Lfffsv8tykV7vlv4cKFOnDgQJ7zI4bmz87aLosXL1ZcXJxatGiRq35old/CuPrqq9W1a1fNmTMn8+9l3rx5mjJliqZMmaJevXplzqt5tOX3vPnhhx9mtmvJkiVVq1Yt3XTTTRo0aJBq1ap11GMDjjQShACKperVq2vVqlWFemDNz9/+9jc99thjOvHEE3XRRRepRo0amb3Mxo0bp++//z5b/UcffVR169bVCy+8oBEjRmjEiBGKj4/XJZdcoocfflgnn3yyJOm8887Txx9/rGHDhmny5MmZ8+iceuqpGjx4sLp3755vXKFJw7N+251f3XPPPVdr167Veeedp169eqlSpUqKj4/X9u3bNWrUqEJPFJ2X008/Xaeffnrm+9AiKp9++qn++te/asqUKUfkOPnZuHGjJGWbkykkOTk57KTqAAAcKbt379a2bdsk/e9etGPHDknSiSeeGHabUHno/p5V1apVw24T6sEW2rcU+ALyrbfeUp06ddSlSxdVq1YtcyTEY489luf9PlxPtvj4+IN+lnVRtYPJa7Gy+Pj4bF++bt26VVJgQZfQl3rh/P777wU+tnRkzrEg7bh+/Xrt2LFDFSpUyKx/sDY8Gvbt26c2bdro888/V4MGDdStWzclJSVlzo84dOjQQj3/hdpl4cKFWrhwYZ71srbLjh07VKlSpbDzah/quSckJKh9+/Zq3769JOnAgQN644031LdvX7344ovq2rWrLrvsskPad2Hk97w5ePDgPBeZAYojEoQAiqUWLVpo9uzZ+u9//5s5dPRQbd68WY8//rgaNGigTz75JNek3uFWaytRooQGDhyogQMHavPmzZo3b55eeeUVvf7661qxYoVWrFiR+aDetGlTTZs2TXv27NGiRYv03nvv6YknnlCPHj2UlJSkdu3a5Rlb6CH7xx9/1JlnnpnveYwZM0Zr164N+7Dy6aefatSoUQW5HIfktNNO04QJE3TyySdr9uzZR+04Wc2ZM0eS8lzBGgCAo2nevHnav3+/qlatmrnQQygBFW7lYel/U6SES1Rt2rQp7DahfYW2SUtL01tvvaV27dpp+vTp2ZIyGRkZevDBBw/thIpY6Hy6du2qN998M8LRZJe1HcON+sjZjqGfB2vDrELTzOzfvz/sNjt27CjQ0OS3335bn3/+uXr37q1x48blijO/5Gs4oWP+9a9/1SOPPFLgbbZt26Z9+/blShLm9bdQWCVKlNDVV1+tZcuW6YEHHtDs2bOPeoLw22+/1YYNGxQfHx+29ylwrGEOQgDF0vXXX6+EhAS98cYb+uqrr/Kte7BvTb/77jtlZGSoffv2uZKDGzZs0HfffZfv9lWqVNHll1+u1157TW3atNGaNWu0fPnyXPVKlSqlZs2a6b777tPjjz8uKfBQl5/zzz9fkjRjxox860mBhxhJuuKKK3J9djSHtYSErl3WIS1Hy+zZszV//nyVKVNGXbt2PerHAwAgq4yMDA0bNkyS1KNHj8zy0ByCc+fODbtdqLxx48a5Pgt3r96xY4eWLl2q0qVLZ/bcD93vO3funCsZ8/nnn2vXrl2FO5kIOe2001ShQgV99tlnheqhWBTya8dQ0qh27dqZX+SG2jO/NsypYsWKkqQffvgh7DHC9TIN51Ce/0LzPWbt0Rly3nnnKS4uTh9//HGBji8Fzj8jIyPs3Jp5/S0cqqJ83rzvvvskSZ06dQq7KjhwrCFBCKBYSklJ0ZAhQ7R371517NhRaWlpYeu99957B52/L/St/7x587I9KP3+++/685//nOub3T179ui///1vrgeTffv2ZQ41Klu2rCTp448/zjYkKCT0DXOoXl569+6txMREPf300/roo49yfZ51iHXoPHI+iC1ZskTDhw/P9zgFsXbtWj3++ONhz8fdM/9HKeecSkeSu+vNN9/UVVddJSkwbOZoDtsBACCnzZs365prrtHcuXN10kknadCgQZmfNW/eXKeeeqrmzZunyZMnZ9tu8uTJ+uijj1SvXr2wc7VNmDBBS5YsyVY2ZMgQ7dixQ927d88cmZDX/X7z5s265ZZbjsAZFo34+Hj1799fP/30kwYMGBA2sfnTTz8d9Ivgo6Fv376SpAceeCDb4hQHDhzQP/7xD2VkZGQbwdKlSxdVrFhRL730Uq5n0lAb5nTaaacpMTFRb7/9tjZv3pxZvmvXLg0YMKDAseb17+G7777TnXfeGXab0OI669evz/VZlSpV1LNnT6Wlpen+++8P28NxzZo1Wrt2beb766+/XpJ09913a/fu3Znl27Zty3N+yby8/PLLmjVrVrY5NUN+/vlnjR49WtLRfd7cuXOnBgwYoAkTJqhChQoaMWLEUTsWEE0YYgyg2Bo0aJD279+voUOH6txzz1WzZs2UmpqqcuXKadOmTfroo4+0evXqsJNVZ1WtWjVdc801euWVV9SwYUO1b99eO3bs0KxZs1S6dGk1bNgw2ze/u3btUrt27ZSSkqImTZooOTlZu3fv1qxZs7Ry5Up17tw581v+hx9+WO+//75atWqlOnXqqFy5clqxYoVmzJihihUrql+/fvnGVrlyZb300ku68sor1bp1a3Xo0EFnnXWWdu7cqS+//FI//PBD5gNar1699NBDD+m2227TnDlzdMopp2j16tWaNm2aLr/8cr366quHdb137NihgQMH6vbbb1fz5s3VoEEDHX/88dq8ebNmz56t7777TlWqVNHDDz98WMcJCa0MJwWu+caNGzV//nytXbtWpUqV0siRI3X77bcfkWMBABBO6D6UkZGh7du3a8WKFZo3b5727t2r8847T5MmTVLlypUz65uZxo8frz/96U/q1q2bunTpotNOO01ff/21pkyZouOPP14vvvhi5vDSrDp06KDmzZvr6quv1oknnqh58+Zp3rx5SklJyZagOPfcc9W8eXO9+eabatasmVq0aKFNmzZpxowZOvXUUzMX5CgO/vWvf+mLL77QM888o3feeUdt2rRRjRo1tHnzZq1evVrz58/XsGHDVL9+/SKNq1mzZrrjjjv04IMPqkGDBrryyit13HHHacaMGVq+fLlatGiR7RmkXLlyeu6559StWzddcMEF6tatW2YbLl++XBdeeGGuL3oTEhI0cOBA3X///WrUqJG6du2q/fv3a9asWapevXqB27FTp046+eST9cgjj2jZsmVq1KiR1q9fr2nTpqljx45hk4CtW7dWXFyc/vnPf2r58uWZvRnvueceSdKTTz6p1atX695779WECRPUokULVa1aVRs3btTKlSu1cOFCvfzyy5mL0nXv3l2vvvqqpk6dqgYNGqhLly7at2+fJk+erHPPPVdr1qwp8LVfsGCBRo0apWrVqqlFixaZx1i7dq3effdd7dq1S126dNGVV15Z4H3m57HHHlOFChXk7tq5c6e+/vprffTRR0pPT1e9evU0ceJE1atX74gcC4h67s6rAK9zzjnHAUSnr776ym+99VY/44wz/Pjjj/eEhASvVq2aX3zxxT5mzBjfvXt3Zt21a9e6JO/du3e2faSnp/ugQYO8bt26XqpUKa9Zs6b/5S9/8V9++cVbtmzpgf9cBuzdu9dHjhzpF198sdeqVctLlSrllStX9iZNmvjTTz/te/bsyaw7c+ZM79Onj59++umemJjoZcuW9Xr16nn//v193bp12WIYPHiwS/I5c+bkOsfly5f7dddd59WrV/eEhASvUqWKX3jhhf7ss89mq7dixQrv1KmTJyUledmyZb1x48Y+evToPM+7d+/eLsnXrl170Ou8e/duf+utt/zmm2/2Ro0aeZUqVTw+Pt4TExO9cePGPmjQIN+8eXO++whdy3DnmPM6hF5m5uXKlfOTTjrJO3To4CNGjPANGzaE3TZ0nsnJyQc9HwAA8pL1PiTJS5Ys6SeccII3btzYb7zxRp8xY4YfOHAgz+1XrVrl1157rVerVs3j4+O9WrVq3rNnT1+1alWuulnv/2PHjvWzzz7bS5cu7ZUrV/Y+ffr4xo0bc22zdetWv/nmmz05OdlLlSrlderU8X/+85+enp7uycnJue6DY8eOdUk+duzYXPuaM2eOS/LBgwfn+iyv54ecz0YH24+7h43L3T0jI8NffPFFb9OmjVesWNETEhK8evXq3rx5cx82bJivX78+7P5yChdTSH7PO/k9f7388svevHlzL1eunJcqVcrr16/vDzzwgO/atSvscd5//31v3ry5lylTxitUqOCdO3f2lStX5nn8jIwMHz58uNepU8cTEhK8Vq1afvvttxe6HdevX+89evTw6tWre+nSpb1+/fo+cuRI37dvn0vyli1b5op1woQJmf/WQv/Os9qzZ48/8cQT3rRpU09MTPSSJUt6rVq1vE2bNv7oo4/6L7/8kqv+0KFDvXbt2l6yZElPTk72QYMG+e7du/OMIZz169f7k08+6ZdddpnXq1cv27N9hw4dfMKECfn+7RX0WTA5OTnb33h8fLxXrFjRGzRo4Ndee62//vrr2Z7pswr9m8nr3zoQbSSleQHyXuZFMHb/WJCamup5DWEEAAAAgEMxZMgQDR06VHPmzFGrVq0iHQ4A4BhjZovcPf9hdWIOQgAAAAAAACCmkSAEAAAAAAAAYhgJQgAAAAAAACCGMQdhATEHIQAAAAAAAIqTgs5BGF8UwcQCM4t0CAAAAFHrcL6U5jkLAAAgb0ei8x9DjAEAAAAAAIAYRoIQAAAAAAAAiGEkCAEAAAAAAIAYRoIQAAAAAAAAiGEsUnKEsBo0AADA0cFzFgAAwNFFD0IAAAAAAAAghpGCYpFMAAAgAElEQVQgBAAAAAAAAGIYCUIAAAAAAAAghpEgBAAAAAAAAGIYCUIAAAAAAAAghpEgBAAAAAAAAGIYCUIAAAAAAAAghpEgBAAAAAAAAGIYCUIAAAAAAAAghpEgBAAAAAAAAGIYCUIAAAAAAAAghpEgBAAAAAAAAGIYCUIAAAAAAAAghkU0QWhmB8xsqZktN7N3zKxChOJIMbPlkTg2AAAAELUmTZJSUqS4uMDPSZMiHVFsoz2iC+0B4BgS6R6Eu9y9obs3kLRN0i1FcVAzK1EUxwEAAACKrUmTpH79pO+/l9wDP/v1IwkSKbRHdKE9ABxjIp0gzOpTSTVCb8zsdjNbaGZfmtnQYNkdZjYg+PujZjY7+HtbM5sY/P1pM0szsxWh7YLl68zsXjObJ+kqMzvHzL4ws09VRIlJAAAAoNi4+27pjz+yl/3xR6AcRY/2iC60B4BjTFQkCIM9+tpKmhp8317SKZLOk9RQ0jlmdqGkjyRdENwsVVI5M0uQ1ELSx8Hyu909VdJZklqa2VlZDrXb3Vu4+yuSxkoa4O5N84mrXzDZmLZly5YjdboAAABA9Fu/vnDlOLpoj+hCewA4xkQ6QVjGzJZK2iqpkqRZwfL2wdcSSYslnaZAwnCRAsnC4yXtUaDXYaoCScNQgvBqM1sc3PYMSfWzHO9VSTKz8pIquPuHwfIJ4YJz9+fcPdXdU5OSko7A6QIAAADFxEknFa4cRxftEV1oDwDHmEgnCHe5e0NJyZJK6n9DfU3S8OD8hA3d/WR3f97d90laJ+l6SZ8okBRsLamupJVmVlvSPyS1dfezJL0rqXSW46Vn2b8f3VMDAAAAirFhw6SyZbOXlS0bKEfRoz2iC+0B4BgT6QShJMndd0gaIOkfwSHDMyX1NbNykmRmNcysSrD6RwokAT9SIEF4k6Sl7u6SEhVIAu4ws6qSOuRxvO3BOi2CRT2PzpkBAAAAxVTPntJzz0nJyZJZ4OdzzwXKUfRoj+hCewA4xlggrxahg5v97u7lsrx/R9Jr7j7BzAZKujH40e+SrnX3NWbWVtJ7CgwRTjezbyQ94+6PBPcxTlITSd8pMAx5qruPM7N1klLd/ZdgvXMkvSDpDwUSklcGV1MOKzU11dPS0o7k6QMAAAAAAABHjZktCq7VkX+9SCYIixMShAAAAAAAAChOCpogjIohxgAAAAAAAAAigwQhAAAAAAAAEMNIEAIAAAAAAAAxjAQhAAAAAAAAEMNIEAIAAAAAAAAxjAQhAAAAAAAAEMNIEAIAAAAAAAAxjAQhAAAAAAAAEMNIEAIAAAAAAAAxjAQhAAAAAAAAEMNIEAIAAAAAAAAxjAQhAAAAAAAAEMNIEAIAAAAAAAAxjAQhAAAAAAAAEMNIEAIAAAAAAAAxjAQhAAAAAAAAEMNIEAIAAAAAAAAxjAQhAAAAAAAAEMNIEAIAAAAAAAAx7KAJQjM7YGZLzWy5mb1uZmULexAzu+1QtgMOx6RJUkqKFBcX+DlpUqQjOrZxvaMD7QAAKArcb4oW1zs60A4AjmXm7vlXMPvd3csFf58kaZG7P1Kog5itk5Tq7r8caqCFZWbx7r7/SO0vNTXV09LSjtTucJRNmiT16yf98cf/ysqWlZ57TurZM3JxHau43tGBdgAAFAXuN0WL6x0daAcAxZWZLXL31IPWK2SC8CZJZ7n7X8zsb5L6BquNcffHzOw4Sa9JqimphKT7JVWV9B9JX0v6xd1b59h/H0mdJZWVVFfSW+5+R/Cz9pKGSiolaY2k693996wJRzNLlfQfd29lZkMkVZeUIumXYHxPS0qVtF/S39x9Tn7HzAsJwuIlJUX6/vvc5cnJ0rp1RR3NsY/rHR1oBwBAUeB+U7S43tGBdgBQXBU0QRhfiB3GS+og6T0zO0fS9ZKaSDJJC8zsQ0l1JG10947Bbcq7+45gMrF1Pj0IG0pqJGmPpK/N7AlJuyTdI6mdu6eb2Z2S/ibpvoOEeo6kFu6+y8z+LknufqaZnSbpfTOrl9cx3f2HHOfcT1I/STrppJMKcpkQJdavL1w5Dg/XOzrQDgCAosD9pmhxvaMD7QDgWFeQRUrKmNlSSWmS1kt6XlILBXrdpbv775LelHSBpGWS2pnZSDO7wN13FDCO/7r7DnffLekrScmSzpdUX9L84PF7B8sPZqq77wr+3kLSBEly91WSvpcUShCGO2Y27v6cu6e6e2pSUlIBTwXRIK98Lnneo4PrHR1oBwBAUeB+U7S43tGBdgBwrCtIgnCXuzcMvvq7+14Feg3m4u7fKNCDb5mk4WZ2b846ZtY1uOjJ0uDwYCnQiy/kgAI9G03SrCzHru/uNwTr7M8Se+kch0jPerh8zivcMXGMGDYsMCdIVmXLBspx5HG9owPtAAAoCtxvihbXOzrQDgCOdQVJEIbzkaTLzKxscN7BrpI+NrPqkv5w94kKzDvYOFj/N0nHS5K7v5Ul6ZffpH6fSWpuZidLUvBYod5/6xRIRErSFQeJs2dw+3qSTlJgLkQc43r2DEwYnJwsmQV+MoHw0cP1jg60AwCgKHC/KVpc7+hAOwA41hVqkZIc5eEWKblI0kOSMiTtk3Szu6eZWX9Jt0j6KY9FSlLd/dbg+2kKLDoy18zaSBqpwCIlknSPu081swsUGOq8SdKC4PahRUp+d/f/BPdVWtIzCiQTcy5SEvaYeV0HFikBAAAAAABAcXLEVjFGAAlCAAAAAAAAFCcFTRAe6hBjAAAAAAAAAMcAEoQAAAAAAABADCNBCAAAAAAAAMQwEoQAAAAAAABADCNBCAAAAAAAAMQwEoQAAAAAAABADCNBCAAAAAAAAMQwEoQAAAAAAABADCNBCAAAAAAAAMQwEoQAAAAAAABADCNBCAAAAAAAAMQwEoQAAAAAAABADCNBCAAAAAAAAMQwEoQAAAAAAABADCNBCAAAAAAAAMQwEoQAAAAAAABADCNBCAAAAAAAAMQwEoQAAAAAAABADIs/lI3M7ICkZcHtV0rq7e5/HMnAAAAAAAAAABx9h9qDcJe7N3T3BpL2Srop64cWENHeiWZ2SMlPRL9Jk6SUFCkuLvBz0qRIRxTbaI/oRLsAAA4F94/oQntEJ9oFwLHoSCTxPpZ0spmlmNlKM3tK0mJJtcysvZl9amaLzex1MyuXc2Mza2Vmc81sspmtMrNJZmbBz84xsw/NbJGZzTSzE4Plc80sNfh7ZTNbF/y9T/A470h6P5iofMjMlpvZMjPrdrBjIrpNmiT16yd9/73kHvjZrx835UihPaIT7QIAOBTcP6IL7RGdaBcAxypz98JvZPa7u5cL9tJ7Q9J7kmZI+k5SM3f/zMwqS3pTUgd3TzezOyWVcvf7cuyrlaS3JZ0haaOk+ZJul7RA0oeSurj7lmBy7yJ372tmcyX9w93TgsdJc/cUM+sj6QFJZ7n7NjO7QoHejRdLqixpoaQmkk4Nd0x3n5fXOaempnpaWlqhrxWOrJSUwE04p+Rkad26oo4GtEd0ol0AAIeC+0d0oT2iE+0CoLgxs0Xunnqweoc6DLeMmS0N/v6xpOclVZf0vbt/Fiw/X1J9SfODnfNKSvo0j/197u4bgoEvlZQiabukBpJmBbcvIemnAsQ2y923BX9vIelldz8gaZOZfSjpXEk78zhmtgShmfWT1E+STjrppAIcGkfb+vWFK8fRRXtEJ9oFAHAouH9EF9ojOtEuAI5VhzsHYUN37+/ue4Pl6VnqmALJulC9+u5+g5k1MbOlwVfnYN09WbY7oEDi0iStyLL9me7ePlhnf5bYS+eILWcMeQl3zGzc/Tl3T3X31KSkpHx2haKSV56W/G1k0B7RiXYBABwK7h/RhfaITrQLgGPV0VxI5DNJzc3sZEkys7JmVs/dF2RJ+k3NZ/uvJSWZWdPg9glmdkbws3WSzgn+fmU++/hIUjczK2FmSZIulPT5YZwTImzYMKls2exlZcsGylH0aI/oRLsAAA4F94/oQntEJ9oFwLHqqCUI3X2LpD6SXjazLxVIGJ5WiO33KpD8G2lmX0haKqlZ8OP/SLrZzD5RYG7BvLwl6UtJX0iaLekOd/+5kKeCKNKzp/Tcc4E5PswCP597LlCOokd7RCfaBQBwKLh/RBfaIzrRLgCOVYe0SEksYpESAAAAAAAAFCcFXaTkaA4xBgAAAAAAABDlSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDzN0jHUOxYGZbJH0f6TgAAAAAAACAAkp296SDVSJBCAAAAAAAAMQwhhgDAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDDSBACAAAAAAAAMYwEIQAAAAAAABDD4iMdQF7MLE7S2ZKqS9olaYW7b4psVAAAAAAAAMCxxdw90jFkY2Z1Jd0pqZ2k1ZK2SCotqZ6kPyQ9K2m8u2dELEgAAAAAAADgGBGNCcKXJT0t6WPPEZyZVZHUQ9Kv7j4+EvEBAAAAAAAAx5KoSxCGmFkpd99zsLKiUrlyZU9JSYnEoQEAAAAAAIBCW7Ro0S/unnSwelE7B6GkTyU1LkBZkUhJSVFaWlokDg0AAAAAAAAUmpl9X5B6UZcgNLNqkmpIKmNmjSRZ8KNESWUjFhgAAAAAAABwDIq6BKGkiyT1kVRT0sP6X4Jwp6RBEYoJAAAAAAAAOCZFXYIwuPjIeDO7wt3fiHQ8AAAAAAAAwLEsLtIB5GRm15qZ5ZUcNLO6ZtaiqOMCAAAAACCbuXOle+6JdBQAcNiirgehpBMkLTWzRZIWSdoiqbSkkyW1lPSLpLsiFx4AAAAAREjjxtLJJ0uvvRbpSCBJvXtL69dLPXpI9etHOhoAOGRR14PQ3UcpsFLxy5KSJLUNvv9R0nXufoW7r45giAAAAAAQGUuWSK+/Lu3ZE+lIIElNmwZ+TpgQ2TgA4DBFYw9CufsBSbOCLwAAAABAVtOmSVdcEekoULVq4OfEidKwYVJc1PXBAYACiboEoZk9Icnz+tzdBxRhOAAAAAAQPWrUkH78UXrxRRKE0WTDhsB8hG3aRDoSADgk0fj1RpoCcw+WVmBo8ergq6GkAxGMCwAAAAAiq0SJwM/p06UtWyIbCwJKlZISEwNJWwAopqIuQeju4919vKRTJLV29yfc/QkF5iJsGNnoAAAAACDCUlOl/full1+OdCSQpDJlpKuukiZPltLTIx0NABySqEsQZlFd0vFZ3pcLlh2UmV1sZl+b2bdmlmvFYzPrY2ZbzGxp8HXjEYoZAAAAAI6uBg2kRo3osRZNevUKJAenTIl0JABwSKI5QThC0hIzG2dm4yQtlvTvg21kZiUk/Z+kDpLqS+puZuHWm3/V3RsGX2OOYNwAAAAAcHT17i0tWiStWBHpSCBJLVpIKSkkbQEUW1GbIHT3sZKaSHor+GoaHHp8MOdJ+tbdv3P3vZJekdTl6EUKAAAAHKPcpalTpX37Ih0JcurePTAf4YQJkY4EUmD14muvlT74ILCIDAAUM1GbIJQkd//Z3d8Ovn4u4GY1JP2Q5f2GYFlOV5jZl2Y22cxqhduRmfUzszQzS9vCBMAAAACINUuWSF26SDcyI0/UqVJF6tAh0GNt//5IRwNJ6tNHysiQxo6NdCQAUGhRnSA8RBamzHO8f0dSirufJekDSWF7Jrr7c+6e6u6pSUlJRzhMAAAAoJh48cVAb0JElxtvlH76SXr33UhHAkmqW1dq00Z6/vlAohAAipFjMUG4QVLWHoE1JW3MWsHdt7r7nuDb0ZLOKaLYAAAAgOLp008jHQFy6thROvFEafToSEeCkD//WVq3LjDUGACKkahOEJpZCzO7Pvh7kpnVLsBmCyWdYma1zaykpGskTc2x3xOzvO0saeWRihkAAAA4Jj33XKQjQE7x8dINN0gzZkjr10c6GkhS167SCSeQtAVQ7ERtgtDMBku6U9I/g0UJkiYebDt33y/pVkkzFUj8vebuK8zsPjPrHKw2wMxWmNkXkgZI6nOk4wcAAACOGdWrS6++Kv36a6QjQU433BAY/v3CC5GOBJJUqlRghekpU6RNmyIdDQAUWNQmCCV1VaB3X7okuftGSccXZEN3n+7u9dy9rrsPC5bd6+5Tg7//093PcPez3b21u686SucAAAAAFH/9+km7d0uTJkU6EuSUkiK1bx+Y947FSqLDjTcG2mJ82KnuASAqRXOCcK+7u4ILjJjZcRGOBwAAAIhNjRpJqamBYcYsVhJ9+vWTNmyQ3nsv0pFAkk4/XWrRQhozhr8XAMVGNCcIXzOzZyVVMLM/K7DaMBM5AAAAAJHQr5+0bJm0YEGkI0FOnTpJVasyT2Q06ddPWr1amjMn0pEAQIFEbYLQ3f8jabKkNySdKuled38islEBAAAAMeqaa6Ry5aSnnop0JMgpIUHq21d6913p++8jHQ0k6corA4uVPPlkpCMBgAKJ2gShJLn7LHe/3d3/4e6zIh0PAAAAELOOP17q0yewWAmLL0Sfm26SzEjgRosyZaQ//1l6+22StgCKhahNEJrZ5Wa22sx2mNlOM/vNzHZGOi4AAAAgZt16q7R3r/Tss5GOBDmddJLUtas0erT0xx+RjgaSdPPNgZ8kbQEUA1GbIJT0oKTO7l7e3RPd/Xh3T4x0UAAAAEDMOvVU6eKLpaefDiQKEV0GDJB+/ZXVpotSfouQhJK2Y8ZIu3YVXUwAcAiiOUG4yd1XRjoIAAAAAFkMGCD9/LM0eXKkI4lN+SWkWrSQGjaUHn+c1XOjRf/+0rZt0ksvRToSAMhXNCcI08zsVTPrHhxufLmZXR7poAAAAICYdtFFUr16gSQUIsMs7/KBA6Xly6W5c4s0pJiWV3tI0oUXSmeeSdIWQNSL5gRhoqQ/JLWX1Cn4ujSiEQEAAACxLi4u0CtqwYLAC9HlmmukypWlUaMiHQmkQPJwwADpyy+ljz6KdDQAkKeoTRC6+/VhXn0jHRcAAAAQ83r3Dqxq/NhjkY4EOZUuLf2//ydNnSqtWRPpaCBJPXpIJ5wgPfxwpCMBgDxFXYLQzO4I/nzCzB7P+Yp0fAAAAEDMO/54qV8/6bXXpLVrIx0NcrrlFikhQfrPfyIdCSSpbNlAr9t33pG++irS0QBAWFGXIJQUWpgkTdKiMC8AAAAAkfbXv0olStArKhqdeGKgl+fYsdKmTZGOBlIgaVumjPTQQ5GOBADCiroEobu/E/w5PtyrIPsws4vN7Gsz+9bM7sqn3pVm5maWeqTiBwAAAGJCjRrStddKL7wgbdkS6WiQ0z/+Ie3dy2Iy0aJyZemGG6RJk6QNGyIdDQDkEnUJwhAzSzWzt8xssZl9GXoVYLsSkv5PUgdJ9SV1N7P6YeodL2mAJGZWBgAAAA7F7bdLu3dLTzwR6UiQU7160uWXS089Je3cGeloIEl//7uUkcECMgCiUtQmCCVNkjRW0hX63yrGnQqw3XmSvnX379x9r6RXJHUJU+9+SQ9K2n1kwgUAAABizOmnS126SE8+Kf3+e6SjQU533ilt3y4991ykI4EkpaRIV18tPftsoF0AIIpEc4Jwi7tPdfe17v596FWA7WpI+iHL+w3Bskxm1khSLXeflt+OzKyfmaWZWdoWhk0AAAAg1rgfvM6dd0q//iqNHn3040HhnHuu1Lq19Oij0p49kY4GknTHHdJvvwV6dgJAFInmBOFgMxtjZt3N7PLQqwDbWZiyzCcbM4uT9Kikvx9sR+7+nLununtqUlJSwSMHAAAAjiUW7hE76PzzpZYtA4uV7GZwTtS56y5p40ZpfIGmc8fR1rCh1KGD9Mgj9LoFEFWiOUF4vaSGki7W/4YXX1qA7TZIqpXlfU1JG7O8P15SA0lzzWydpPMlTWWhEgAAAOAQ/etf0o8/SmPGRDoS5PSnP0lNmkjDhgUWLUHkDR4sbd0q/d//RToSAMgUzQnCs4O993q7+/XBV98CbLdQ0ilmVtvMSkq6RtLU0IfuvsPdK7t7irunSPpMUmd3TzsqZwEAAAAc69q0kS64QBo+nF6E0cZMGjpUWr8+sOI0Iq9Jk0AvwoceCgw3BoAoEM0Jws/CrT58MO6+X9KtkmZKWinpNXdfYWb3mVnnIx0kAAAAEPNCSaiNG1kQIxq1bx8YCv7vfzMXYbQYMoRehACiSjQnCFtIWmpmX5vZl2a2zMy+LMiG7j7d3eu5e113HxYsu9fdp4ap24regwAAAMBhat06MBfh8OHSrl2RjgZZhRK4P/xAL8Jocd550iWXSP/5D70IAUSFaE4QXizpFEnt9b/5BztFNCIAAAAAeRs6VPr5Z+mZZyIdCXL605+kZs3oRRhNQnMRPvlkpCMBgOhNELr79+7+vaRdCqxCHHoBAAAAiEYtWwbmIxw5UkpPj3Q0yCrUi3DDBoaBR4vzzpM6dgzMRfjrr5GOBkCMi9oEoZl1NrPVktZK+lDSOkkzIhoUAAAAgPw98IC0aZP0yCORjgQ5tW0rtWol3X+/tHNnpKOBFOjRuX27NGJEpCMBEOOiNkEo6X5J50v6xt1rS2oraX5kQwIAAACQr6ZNpcsvlx58UNq8OdLRICuzQLts2RLotYbIO+ss6brrpFGjAnNEAkCERHOCcJ+7b5UUZ2Zx7j5HUsNIBwUAAADgIEILldx3X6QjQU7nnitdc02gh+fGjZGOBlLg78RduvfeSEcCIIZFc4Jwu5mVk/SxpElmNkrS/gjHBAAAAOBg6tWT/t//k559Vvrmm0hHg5yGDZP27ZOGDIl0JMWfH4Fp8pOTpf79pfHjpWXLDn9/AHAIojlB2EWBBUpuk/SepDViFWMAAACgeLj3Xql0aWnQoEhHcmw5EgmpOnWkv/xFev556auvDn9/sc7s8PcxaJCUmCjdddfh7wsADkHUJgjdPV1SkqRLJG2T9FpwyDEAAACAaFe1qnTHHdIbb0jzmUr8iDoSCal77pHKlZNuv/3w94XDV6lSIEk4fbo0c2akowEQg6I2QWhmN0r6XNLlkq6U9JmZ9Y1sVAAAAAAK7G9/k2rWDAyfPHAg0tEgq8qVA708p0+Xpk2LdDSQpIEDpZNPDvzcuzfS0QCIMVGbIJR0u6RG7t7H3XtLOkfSnRGOCQAAAEBBHXec9PDD0pIl0ujRkY4GOQ0YIJ1+unTbbdLu3ZGOBqVKSY89Jn39tfTEE5GOBkCMieYE4QZJv2V5/5sk1n0HAAAAipOrrpJat5buvlvayoxBUSUhQXr8cWnNmsCqxoi8jh0Dr6FDpZ9/jnQ0AGJINCcIf5S0wMz+P3t3Hm91We7//3UxigKKgoqggohjailqnZyHnCVLf2LHtBwwTzZpk1ZO5ZyW35MNppaZ5pQZnVQy545DgpXmlDhCamLOE+P1++Ne+7DYbmDjXvBZe+/X8/G4H+szf64FbcI393BiRJwA3A1MiYijI+LoimuTJEmS1B4RpTfUq6+Wee/UXHbaCT72sbKy8VT7YzSF730PZsxwwRJJS1UzB4SPA9cCLct0/RZ4DhhQawsUEbtGxKMRMSUi3vWnakR8JiIeiIi/RsSfImKDRhcvSZIkqWbDDcs8hD/5CUyeXHU1au2cc2DuXDjmmKorEcDo0WX+zosvdoEfSUtNr6oLWJDMPAkgIgaU3XyjPfdFRE/gPGBnyjDleyNiQmY+VHfZZZn549r1ewPnALs2sn5JkiRJdU48EX71KzjiCLj7bujVtP8p0v2suWYZAv6tb8Hvf1+GuKpa3/hG+Xk5/PAyh2ffvlVXJKmLa9oehBHxvoj4C/B34MGImBwRG7bj1i2AKZn5RGbOBC4HxtZfkJmv1e0ux7xeipIkSZKWhOWXL0ONJ0+Gc8+tuhq19tWvlp6eRx4Jr7++6Ou1ZPXvDz/6ETz8MJx+etXVSOoGmjYgBM4Hjs7MNTNzTeAYoD1Lnw1j/sVMptWOzSciPhsRjwNnAp9v60ERMT4iJkXEpOnTpy/2F5AkSZI6tWzwv6Pvuy/svXfpqfb44419tjqmT5+y0vS0aaX3mqq3225wwAFw6qklKJSkJaiZA8LlMvOWlp3MvJXS229Roo1j7/qbTWael5mjgK8Bbc6WnJnnZ+aYzBwzZMiQ9lUtSZIkdTXR1l+x3+NzfvjDsnruEUc0PoBUx3zoQ3DUUfCDH8Bdd1VdjQC+//3Sm/Dww8s8kZK0hDRzQPhERHwrIkbU2jeBJ9tx3zRg9br94cCzC7n+cuCjHahTkiRJUnsNGwZnnAE33QQ//3nV1ai1U06B4cPhsMPKSrqq1sorw9lnl8VKfvSjqquR1IU1c0B4CDAEuAb4TW370+24715gdESMjIg+wDhgQv0FETG6bncP4LGGVCxJkiRp0caPh623hi99CZ55pupqVG/AAPjxj+Ghh+CEE6quRgAHHwwf+UiZJ/Ix/9NV0pLRtAFhZr6cmZ/PzE0z8wOZ+YXMfLkd980GjgImAg8DV2bmgxFxcm3FYoCjIuLBiPgrcDRw8BL7IpIkSZLm16MH/OxnMGcOfOpTDp1sNrvvXoa0nnkm3HFH1dUoAi66qKxk/MlPwuzZVVckqQvqVXUBrUXE71jIqsKZufeCztVdcx1wXatjx9dtf6EjNUqSJEnqoFGjyvxqhx1WPo8+uuqKVO+cc8ow8IMOgiuXCdMAACAASURBVL/9DQYOrLqi7m3YsDLEeNy4sqrxN9ucRl+S3rNm7EH4XeBsynyDb1NWLv4p8Abw9wrrkiRJktRIhxwCY8fCscfCAw9UXY3q9e8Pl1xShoB/8YtVVyOA/fcvqxqfdBJMnlx1NZK6mKYLCDPztsy8DfhAZu6fmb+rtU8AW1VdnyRJkqQGiYDzz4cVVoADD4R33qm6ItX7j/8o4e3Pfga//nXV1QjgvPNglVXgP/8T3nij6mokdSFNFxDWGRIRa7XsRMRIykIlkiRJkrqKlVcu86vdf39ZtETN5fjjYfPN4dBD4Yknqq5GgwbBL39ZFis58kjIBc7OJUmLpZkDwi8Bt0bErRFxK3AL4NyBkiRJUlezxx7wla+U1XMvu6zqalSvTx+44orS23O//ezl2Qy22w5OPLEEhRddVHU1krqIpg0IM/MGYDQlFPwCsG5m/qHaqiRJkiQtEaecAlttBePHwyOPVF2N6o0cCRdfDPfdB8ccU3U1AjjuONhpJzjqqNL7VpI6qGkDQoDMnJGZf6u1GVXXI0mSJGkJ6d0bLr8c+vWDffeFt96quiLV23tv+PKX4Yc/LD0KVVQ1xLdnT7j00jLkeL/94PXXq6lDUpfR1AGhJEmSpG5k2LASejz0EBx2mPOrtaXKX5NTTy0Llxx6qKtO14uo5r0rrwy/+hU8/nhZ5Gfu3GrqkNQlNGVAGMXqVdchSZIkaSn7yEfgO98pwcfpp1ddTXOqKpDq3RuuugqWX770KJw+vZo6NM+228L3vgcTJpQFZSTpPWrKgDAzE7i26jokSZIkVeDYY+GAA8o8a7/9bdXVqN5qq8G118Lzz5eh4DNnVl2Rjjqq9Lg95RSHf0t6z5oyIKy5OyI2r7oISZIkSUtZBFx4IWy+Ofznfzqctdlsvnn5/bn9dvjc5xwKXrUIOO88+PCH4dOfLovJSNJiauaAcHvgroh4PCLuj4gHIsLlmSRJkqTuoF8/+M1vYOBA2GsvePbZqitSvU98ovT0PP98OOecqqtRnz7w61/D4MGw557w9NNVVySpk2nmgHA3YBSwA7AXsGftU5IkSVJ3MGwY/O538OKLsPvu8OqrVVeket/5TllB98tfhssuq7oarbIKXH89vP027Lor/PvfVVckqRNp2oAwM5/OzKeBt4Gsa4sUEbtGxKMRMSUivt7G+aMj4qFaz8SbImLNxlYvSZIkqSE22wyuuQYefBD22QdmzKi6IrXo0QN+8YuyUManPgV//GPVFWnDDcu8nU8+WRaSefvtqiuS1Ek0bUAYEXtHxGPAk8BtwFPA9e24rydwHqUH4gbAARGxQavL/gKMycyNgauBMxtYuiRJkqRG+shH4Gc/g1tugYMOgrlzq65ILZZZpixast56JcB1/rvqbbMN/PKXcNddZSj47NlVVySpE2jagBD4NvBB4B+ZORLYEfjfdty3BTAlM5/IzJnA5cDY+gsy85bMfKu2ezcwvHFlS5IkSV1EMy0+ceCBcNZZcOWVcOSRhoTNZIUVytDWFVcsQ1sfeqjqirTvvnDuuSW8PfhgmDOn6ookNblmDghnZea/gR4R0SMzbwHe3477hgFT6/an1Y4tyKEsoGdiRIyPiEkRMWn69OntrVuSJEnSknDMMfMWxnD13OYybFgZYtyrF+ywAzzySNUV6XOfg1NPLfNDHnaYobqkhepVdQEL8UpE9AduBy6NiBeA9vSNjjaOtfk3h4g4EBgDbNvW+cw8HzgfYMyYMf7tQ5IkSd1TtPVX7ApEwCmnlCGTZ50FvXvD977XPPV1d6NHw803lzkJd9gBbrutHFN1jj22zNt50kllpeMf/9ifF0ltauaAcCzwDvAl4D+B5YGT23HfNGD1uv3hwLOtL4qInYBvANtmpjMdS5IkSZ1BBJxxBsyaBd//PvTsCd/9rqFHs1hvvRISbrcdbL99mTfSkLBaJ5xQQsLTTy8Ly5x3XvmUpDpNGxBm5pt1uxcvxq33AqMjYiTwT2Ac8In6CyLiA8BPgF0z84WO1ipJkiRpKYqAc84p86qdcw68+WYJPXr2rLoyQVlJ96abYMcdYeut4Q9/gI03rrqq7iuiDDWeOxfOPBPeeAMuuqj0wJWkmqYLCCPiddoeEhxAZubAhd2fmbMj4ihgItATuCgzH4yIk4FJmTkBOAvoD1wV5V8an8nMvRv5PSRJkiQtQRFlEYb+/eG00+CVV+AXvyjDKFW9jTeG22+HnXcuQ46vuw4+9KGqq+q+IkoPwuWXh298A15/HS6/vKxCLUk0YUCYmQMa8IzrgOtaHTu+bnunjr5DkiRJUsVaekYNGgRf/Sq89hpcfTUsu2zVlQlg/fXhT38qIeFOO8FvfgMf+UjVVXVfEXDccTBwYFnAZM89y+/JgA7/J7ikLqBpJx6IiDXaalXXJUmSJKnJfOUr8NOfwsSJZXGMf/2r6orUYsQIuOMOWHtt2GMP+NnPqq5IRx0FF18Mt94KW20F06ZVXZGkJtC0ASHw+7p2E/AEcH2lFUmSJElqTocdBr/+NTzwAGy5Jfz971VXpBarrlpWNN5uOzjkkNKLbe7cqqvq3g46CH7/e3jyyfLzct99VVckqWJNGxBm5kZ1bTSwBfCnquuSJEmS1KQ++tHSW23WLPiP/4Abbqi6IrVYYYUyD+H48WXOyP33h7ffrrqq7m2XXeDOO6FXr7KYzO9+V3VFkirUtAFha5l5H7B51XVIkiRJamKbbgr33AOjRpUhrd/9LmRbayBqqevdG378Yzj77NLbc6ut4Kmnqq6qe3vf++Duu2GDDWDsWDj5ZHt3St1U0waEEXF0XftyRFwGTK+6LkmSJElNbvjw0pPwYx8r8xPuuy+8+mrVVQnKQhlHHw2//S08/ngJdK93JqlKDR1ahoAfeCCccALstRe8/HLVVUlaypo2IAQG1LW+lLkIx1ZakSRJkqTOoX9/uPJKOOecEkZtvnmZn1DNYa+9YPJkWGON0tPzxBM7Z8+1rtI7ddlly8IlP/wh3HgjbLYZ/OUvVVclaSlq2oAwM0+qa6dk5qWZ+U7VdUmSJEnqJCLgS1+CW26B118vizH86EedO9TpzLW3NmpUmQPvoIPgpJNgp51g6tSqq1p8EVVX0BgRcOSRcPvtMHMmfPCDJWDvjMGtpMXWdAFhRPwuIiYsqFVdnyRJkqROZuutS2+obbaB//qv0nvtX/+quqr3rqsEUlB6rv3sZ3DhhfDnP8PGG8MVV1RdVff2wQ/CX/8Ku+0GxxxTFjP55z+rrkrSEtZ0ASHwXeDshTRJkiRJWjyrrlrmuvt//w9uugk22ggm2P+gKUTAIYfA3/4G660H48bBJz8Jr7xSdWXd1+DB8JvfwPnnl16eG29chux3pR6skubTdAFhZt62sFZ1fZIkSZI6qQj43OfK3HfDhpVVWw84oHP3JuxKRo0qi8uceCL86ldlZd1rrqm6qu4rAg4/vPS+XWst2H9/2GcfexNKXVTTBYQR8UBE3N9GeyAi7q+6PkmSJEmd3AYbwD33wMknlwBq/fXLMFd7R1WvV6+yku4998Aqq8DHP24oVbV11oG77oIzz4SJE8vPz/nnOzeh1MU0XUAI7Ans1UZrOS5JkiRJHdOnD3zrW2VY6/veV4a47rgj/P3vVVcmKKvo/vnPcMYZcMMNJcQ991yYNavqyrqnXr3gK18pK4FvthkccUSZ0/O++6quTFKDNF1AmJlPt7TaodG17ReAl9rzjIjYNSIejYgpEfH1Ns5vExH3RcTsiNi3geVLkiRJ6kzWWw9uvRV+8pOyMMMmm8BnPwsvvlh1ZerdG7761RLafuhD8MUvlrnwrr++6sq6r7XXLnN4Xngh/OMfMGZMGYb8wgtVVyapg5ouIGwREYcDVwM/qR0aDlzbjvt6AucBuwEbAAdExAatLnsG+BRwWaPqlSRJktRJ9egB48fDY4+VVY5/8hMYPRq+/32YObPq6jRqVOlFOGECzJ4Nu+9e2sMPV11Z99SyqMw//gFf+hL8/Ofl5+Wss+Dtt6uuTtJ71LQBIfBZ4MPAawCZ+Riwcjvu2wKYkplPZOZM4HJgbP0FmflUZt4POGmCJEmStCDdbU6+lVaC//7vMux4iy1K+LHOOmV+wtmzq66ue4uAvfaCBx8sQdT//m8ZGn7wwfDEE1VX1z2tsAKcfXbp4fnhD5fenqNHl4DdoeBSp9PMAeGMWsAHQET0AtrzN5RhwNS6/Wm1Y4stIsZHxKSImDR9+vT38ghJkiRJnc2GG5YeazfcAEOGlN5SG2xQVtZ1YYZq9ekDX/4yTJlSAtwrr4R114XPfAamTau6uu5p3XXhuuvglltgzTXL78V668EllxisS51IMweEt0XEcUC/iNgZuAr4XTvuizaOvad/+szM8zNzTGaOGTJkyHt5hCRJktT5RVt/xe7iImCXXcpCGddeC8ssA5/4RAkPL7oIZsyousLubcgQ+O53S1A4fnz5PRk1Cg47DB59tOrquqfttoM//Ql+/3sYOBAOOqiEhz/+sUOPpU6gmQPCrwPTgQeAI4DrgG+2475pwOp1+8OBZxtenSRJkqSuLwLGji0LmFx+eQkKDz0URo6EM8+EV1+tusLubdgwOO+8Mh/eIYfApZeWFY/32Qfuuqvq6rqfiDI/5OTJcM01Jcg98kgYMQJOPRVeeaXqCiUtQDMHhP2AizJzv8zcF7iodmxR7gVGR8TIiOgDjAMmLME6JUmSJHV1PXrA/vvDfffBH/5QehJ+7Wuw+urwuc+VufFUnREj4Ec/gqefhm98A267Df7jP8rqx5dcAu+8U3WF3UuPHvNC2ltvhU03Lb8vw4fDEUeUeT4lNZVmDghvYv5AsB/wx0XdlJmzgaOAicDDwJWZ+WBEnBwRewNExOYRMQ3YD/hJRPj/5pIkSZIWLQJ23hluvLGEhXvvDeefXxbM2HbbMk+hw4+rs/LK8O1vwzPPwLnnwksvlaGuw4eXRTQef7zqCruXiPJzcf31pRfuuHElsH3/+8vCJpde6s+L1CSaOSBcJjPfaNmpbS/bnhsz87rMXCczR2XmKbVjx2fmhNr2vZk5PDOXy8yVMnPDJfINJEmSJHVdH/gA/PKXZXGMM88sn5/4ROlV+IUvwKRJ3W8l6GbRvz98/vPwyCPwxz+W+fHOOQfWXht22KGsTP3aa1VX2b1ssglccAH885/l92L6dDjwQBg6FP7rv+Duu/15kSrUzAHhmxGxactORGwGOLOpJEmSpOYyZAh85Svw2GNl5eNtt4Wf/AQ237ysfnzKKfDUU1VX2T1FwI47wtVXl+HHJ51UehcecgisuioccEBZgXfWrKor7T4GDSorUD/ySBmuv9tu8POfl+Hg665beoA+8UTVVUrdTjMHhF8EroqIOyLiDuAKytBhSZIkSWo+PXqUlY+vugqef74MPV55ZfjmN8uiJptvDqedVoIRLX3DhsHxx5cg96674NOfLgHVHnvAKqvApz4Fv/2tK+4uLT16lOH6l15afl4uumje79GoUaWH7re/DQ89ZM9CaSlo2oAwM+8F1gOOBP4LWD8zJ1dblSRJkiS1wworwOGHl8UynnwSTj8devaE444rq+yuv35ZtOGuu2DOnKqr7V4i4IMfLKsfP/dcCQX32qt8fvSjMHgw7LtvGT7+wgtVV9s9DBxYAttbbim9bc8+G5ZbDk44oSwItP76cOyxcMcd9vaUlpCmDQhr1gU2AD4AHBARB1VcjyRJkiQtnhEjyorHd98NU6fCD34Aq60GZ5xRVtodPBj226/MzzZ1atXVdi99+pSFZi6+uISBN94IBx8Md94Jn/xk6Vm42WYl2L3tNpg5s+qKu74114Sjj4Y//anMV/jDH5Z5Pc86C7bZpvy8fPzjpYfuM89UXa3UZUQ2aVfdiDgB2I4SEF4H7Ab8KTP3raKeMWPG5KRJk6p4tSRJklSNP/8ZttwSfv972H33qqvpel56qSygMXFimbvw2WfL8fXXL/MYbrMNbL11WYG3xbBh5ffipz+tpubuYu7cskr1xIml3Xln6enZv3/5vWlpF1wAv/lNWXBDS9arr8LNN5eflRtumBcOrrNO+VlpaWuuWW2dUpOJiMmZOWaR1zVxQPgAsAnwl8zcJCJWAS7IzL2qqMeAUJIkSd2OAeHSk1nmWps4scyLd+ed8Prr5dyIESUo3HprGD8eDj20BFNaelrCqYkT4dZb4dFH550bPNiAcGnLLHN53nBDGZZ8xx3wyivl3BprlKBwq63KvJ8bbQS9e1dbr1Sh9gaEvZZGMe/R25k5NyJmR8RA4AVgraqLkiRJkqSGiyhzrW24YRleOXs23H9/CT7uuKMEU5dcUq7t27faWruj5ZeHffYpDcqiGnfcUYYdr7ZatbV1RxHz5vL80pdKj8+//x1uv720P/yhzCEJ5efl/e8vYeGYMeVz3XXLnKCS/k8z9yD8IXAcMA44BngD+GtmfrqKeuxBKEmSpG7HHoTNI7Osvjt5cukZtfrqVVckNa/MsjjQvffOa5Mnw5tvlvPLLAMbbFB6F260EbzvfeVz6NASPkpdSKfuQRgRAZyWma8AP46IG4CBmXl/xaVJkiRJ0tIXUeZaW2edqiuRml8ErLVWafvvX47NmVOGJd97LzzwQGkTJ5YFalqsuCKst175ORs9el5be+0y/6TUhTVlQJiZGRHXApvV9p+qtiJJkiRJktRp9ew5bxh/vRdfLMOTH3igfD76aBmi/POfz3/d0KElLBwxosxzuMYaZUGUlu1ll11a30RaIpoyIKy5OyI2z8x7qy5EkiRJkiR1QYMHw3bblVbvjTdgypQytL++3XIL/POfZd7D1s9ZY42y6vjQobDqqqXVb6+yShneLDWhZg4ItweOiIingTeBoHQu3LjasiRJkiRJUpfWv39Z3OT973/3udmzS0j4zDPz2tNPl88nn4S77lrwytYrrFBCw8GDy5DmlVaa1+r367ddmEhLQTMHhLu91xsjYlfgXKAncEFmnt7qfF/gF5QhzP8G9ncYsyRJkiRJWqRevcrw4jXXXPA1s2aVkPD550t77rl5288/X4Y2P/FEmRPx3/+GGTMW/Ky+fWHgwPa3AQPKkOcFtX79oEePxv+6qFNr5oDwO5n5yfoDEXEJ8MkFXN9yTU/gPGBnYBpwb0RMyMyH6i47FHg5M9eOiHHAGcD+Da1ekiRJkiR1T717w2qrldYeb71VgsL69tJL5fO1197dpk6dt/3qqyWQXBzLLNN2eNi3b2l9+szbbmt/Ydf06VO+f69eC2/tucYgc6lp5oBwvplDa8HfZu24bwtgSmY+UbvvcmAsUB8QjgVOrG1fDfwgIiIzc4FPff55OOusdhcvSZIkdXrPPFN1BZLUPbQEdKuv/t7unzFjXlj45pslcFxYW9A1M2aU+RdnzICZM8tnS2vZnzmzsd99YSLmDwx79iytR4+223s91557I+bV1Lq1dXxxrl3Sz2iHpgsII+JY4DigX0S81nIYmAn8tB2PGAZMrdufBmy5oGsyc3ZEvAqsBLzYqpbxwHioJZNf/erifBVJkiSp8+vdG4YNq7oKSdLC9O0LQ4aUtqRllpBwQQHijBkwZ06Zq7F1mzWr7ePtabNmlefOndt26+i5mTMXfH7OnHnfvXVr6/jiXLskn7GQfnCtNV1AmJmnAadFxGmZeex7eERb8WjrX5H2XENmng+cDzBm002TO+54D+VIkiRJnVivXk6QL0maJ2LesOIBA6quRovSzl6ETRcQ1plSv1MbYvzNzDxpEfdNA+r75A4Hnl3ANdMiohewPPDSQp/aowcst1w7ypYkSZIkSZI6j2ae7XHHiLguIoZGxEbA3UB7oul7gdERMTIi+gDjgAmtrpkAHFzb3he4eaHzD0qSJEmSJEldVNP2IMzMT0TE/sADwFvAAZn5v+24b3ZEHAVMBHoCF2XmgxFxMjApMycAFwKXRMQUSs/BcUvsi0iSJEmSJElNrGkDwogYDXwB+DWwPvDJiPhLZr61qHsz8zrgulbHjq/bfgfYr7EVS5IkSZIkSZ1PNOvI2oh4BPhsZt4UEQEcDRySmRtWVM904Okq3i1JkiRJkiS9B2tm5iKXt27mgHBgZr7W6tjozHysqpokSZIkSZKkrqbpFimJiK8CZOZrEdF6GPCnKyhJkiRJkiRJ6rKaLiBk/gVDjm11btelWYgkSZIkSZLU1TVjQBgL2G5rX5IkSZIkSVIHNGNAmAvYbmtfkiRJkiRJUgc03SIlETEHeJPSW7Af8FbLKWCZzOxdVW2SJEmSJElSV9N0AaEkSZIkSZKkpacZhxhLkiRJkiRJWkoMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRuzIBQkiRJkiRJ6sYMCCVJkiRJkqRurFfVBXQWgwcPzhEjRlRdhiRJkiRJktQukydPfjEzhyzqOgPCdhoxYgSTJk2qugxJkiRJkiSpXSLi6fZc5xBjSZIkSZIkqRszIJQkSZIkSZK6MQNCSZIkSWpyjz0Gjz5adRWSpK7KgFCSJEmSmtxee8F668H228MVV8DMmVVXJEnqSgwIJUmSJKnJ/etf5fPpp2HcOFh9dTjuOHjyyWrrkiR1DQaEkiRJktTkBg2CT34SpkyBG26AD30IzjgDRo2C3XeHCRNgzpyqq5QkdVYGhJIkSZLUSfToAbvsAtdeW3oTHn88/O1vMHYsjBwJ3/42PPts1VVKkjobA0JJkiRJ6oSGD4cTT4SnnoJrroH11y+B4RprwD77wP/8D8yeXXWVkqTOoFMEhBGxa0Q8GhFTIuLrbZzvGxFX1M7fExEj6s4dWzv+aETsUnd8hYi4OiIeiYiHI+JDS+fbSJIkSVLj9O5dAsGJE8tqx0cfDXfeWRY2WWMNOPbYclySpAVp+oAwInoC5wG7ARsAB0TEBq0uOxR4OTPXBr4HnFG7dwNgHLAhsCvww9rzAM4FbsjM9YBNgIeX9HeRJEmSpCVp7bXhzDNh2jT4zW9gzBg46yxYZx3Ydlu4+GJ4882qq5QkNZumDwiBLYApmflEZs4ELgfGtrpmLHBxbftqYMeIiNrxyzNzRmY+CUwBtoiIgcA2wIUAmTkzM19ZCt9FkiRJkpa43r3hox8ti5c88wycdho89xx86lMwdCgccQT8+c+QWXWlkqRm0BkCwmHA1Lr9abVjbV6TmbOBV4GVFnLvWsB04GcR8ZeIuCAilmv94ogYHxGTImLS9OnTG/V9JEmSJGmpWW01+PrX4dFH4fbb4WMfg1/+ErbcEjbaCM4+u4SHkqTuqzMEhNHGsdb/zrWgaxZ0vBewKfCjzPwA8CbwrrkNM/P8zByTmWOGDBmyeFVLkiRJUhOJgK23hp//vASC558P/fvDl79cFjzZdVe49FKHIEtSd9QZAsJpwOp1+8OBZxd0TUT0ApYHXlrIvdOAaZl5T+341ZTAUJIkSZK6vIED4fDD4e674ZFH4LjjyueBB8Kqq5ahyDfdBHPmVF2pJGlp6AwB4b3A6IgYGRF9KIuOTGh1zQTg4Nr2vsDNmZm14+NqqxyPBEYDf87M54GpEbFu7Z4dgYeW9BeRJEmSpGaz7rrw7W/DE0/AbbfB/vuXBU522gnWXLMMT37wwaqrlCQtSQ0NCCNiUERsGBFrRURDnl2bU/AoYCJlpeErM/PBiDg5IvauXXYhsFJETAGOpjZcODMfBK6khH83AJ/NzJZ/A/sccGlE3A+8Hzi1EfVKkiRJUmfUowdssw1ccAE8/zxccQW8//3w3e/C+94Hm24K3/sePNt6PJckqdOL7OCyVRGxPPBZ4ACgD2Xxj2WAVYC7gR9m5i0drLNyY8aMyUmTJlVdhiRJkqRuaK21YKut4Be/WPrvfuEFuPxyuOQSmDSpzGW47bYwbhx8/OMwePDSr0mS1D4RMTkzxyzqukb08ruaslLw1pm5bmZuVVvYY3XgDGBsRBzagPdIkiRJkpaylVeGz38e7r23zFN44omlh+FnPlPmK9xtN7j4Ynj11aorlSS9Vx3uQdhd2INQkiRJUlVGjiwrEFfRg7AtmXD//aVn4eWXw1NPQd++sPvupWfhnnvCsstWXaUkaWn2IGx54U3tOSZJkiRJWnwRVVcwTwRssgmcdlpZ3OTuu0uPwrvvLoucrLwyfOITcO218PbbVVcrSVqUDgeEEbFMRKwIDK4tUrJirY0AVuvo8yVJkiRJzSsCttwSvv99mDoVbrkFDjwQ/vAH2GcfGDKkhIZXXglvvFF1tZKktjSiB+ERwGRgvdpnS/stcF4Dni9JkiRJ6gR69oTttoMf/xiee66EhAceCLfeWkLCIUPgox8tC5688krV1UqSWnQ4IMzMczNzJPDlzFwrM0fW2iaZ+YMG1ChJkiRJ6mR694addy5h4bPPwm23wfjxMHkyHHRQGYa8225wwQXw4otVVytJ3VsjhhhvBZCZ/72A8wMj4n0dfY8kSZIkqXPq2RO22QbOPReefrrMVfiFL8Cjj8Lhh8Mqq8AOO8APflDOS5KWrkYMMf54RNwZEcdHxB4RsUVEbBMRh0TEJcD/AP0a8B5JkiRJUifXo0eZs/Css+Dxx+G+++DYY8uQ5M99DkaMgA98AE44ofQ2zKy6Yknq+iIb8KdtRAwC9gU+DAwF3gYeBn6fmX/q8AuawJgxY3LSpElVlyFJkiSpGxo5svTAu/jiqitZsv7xD5gwobT//V+Ye//ubgAAIABJREFUOxeGDYO99oKxY2H77aFv36qrlKTOIyImZ+aYRV7XiICwOzAglCRJklSV7hIQ1nvxRfj970tYOHEivPkm9O8Pu+xSwsLdd4eVVqq6Sklqbu0NCBsxxLjlhadGxAp1+4Mi4juNer4kSZIkqfsYPBgOPhh+/esSFl53XVkR+a675i1yss02cPrp8Le/ORRZkjqiYQEhsFtm/t9C9Zn5MrB7A58vSZIkSeqGllmmrHj8ox/B1Klw773wjW/A66+X+Qvf/35YfXU47DC45hp47bWqK5akzqWRAWHPiPi/2SAioh/g7BCSJEmSpIbp0QPGjIGTT4a//AWefRYuugg+9CG4+mr4+MfL0OPtt4czz4S//93ehZK0KI0MCH8J3BQRh0bEIcCNQDeaIUOSJEmStLQNHQqf/jRcdRVMnw633QZf/jK89BJ87Wuw0Uawxhowfjxce23pdShJml9DFymJiF2BnYAA/pCZExv28Iq5SIkkSZKkqnTHRUoaYdo0uOEGuP56uPHGEg726gVbbgk771zaFluUY5LUFbmKcYMZEEqSJEmqigFhx82cCXfeCX/4QwkLJ08uQ48HDizDkVsCw9GjIaLqaiWpMZb6KsZLUkTsGhGPRsSUiPh6G+f7RsQVtfP3RMSIunPH1o4/GhG7tLqvZ0T8JSL+Z8l/C0mSJElSVfr0ge22g1NPLYucTJ8OV14J++9fVkE+6ihYd10YMaIsdnLFFWX1ZEnqDpq+I3VE9ATOA3YGpgH3RsSEzHyo7rJDgZczc+2IGAecAewfERsA44ANgdWAP0bEOpk5p3bfF4CHgYFL6etIkiRJ0mJz4FfjrbQS7LdfaZnw+OOlZ+GNN5bFTi68sPQk/MAHYKedYIcdYKutYLnlqq5ckhqvoT0II6JfRKzbyGcCWwBTMvOJzJwJXA6MbXXNWOYtiHI1sGNERO345Zk5IzOfBKbUnkdEDAf2AC5ocL2SJEmS1HAOe11yImDtteHII+Gaa0rPwbvugpNOKoHg974Hu+4KgwaVkPD440uvQ0nqKhoWEEbEXsBfgRtq+++PiAkNePQwYGrd/rTasTavyczZwKvASou49/vAV4G5C3pxRIyPiEkRMWn69Okd+Q6SJEmSpE6iVy/44AfhW9+C22+Hl1+GiRPhmGNg1iw45ZQSGEpSV9HIHoQnUnrnvQKQmX8FRjTguW39O1nrDvYLuqbN4xGxJ/BCZk5e2Isz8/zMHJOZY4YMGdK+aiVJkiRJXcpyy8FHPgKnnQb33AOf/Sy8807VVUlS4zQyIJydma828HktpgGr1+0PB55d0DUR0QtYHnhpIfd+GNg7Ip6iDFneISJ+uQRqlyRJkiR1MQ73ltTVNDIg/HtEfALoGRGjI+K/gTsb8Nx7gdERMTIi+lAWHWk9dHkCcHBte1/g5szM2vFxtVWORwKjgT9n5rGZOTwzR9Sed3NmHtiAWiVJkiRJkqROpZEB4ecoqwXPAC6jzAP4xY4+tDan4FHARMqKw1dm5oMRcXJE7F277EJgpYiYAhwNfL1274PAlcBDlLkRP1u3grEkSZIkSZLU7fVq1IMy8y3gG7XWUJl5HXBdq2PH122/A+y3gHtPAU5ZyLNvBW5tRJ2SJEmSJElSZ9PIVYxvjIgV6vYHRcTERj1fkiRJkiRJUuM1cojx4Mx8pWUnM18GVm7g8yVJkiRJkiQ1WCMDwrkRsUbLTkSsCWQDny9JkiRJkiSpwRo2ByFl7sE/RcRttf1tgPENfL4kSZIkSZKkBmvkIiU3RMSmwAeBAL6UmS826vmSJEmSJEmSGq+RPQgB5gAvAMsAG0QEmXl7g98hSZIkSZIkqUEaFhBGxGHAF4DhwF8pPQnvAnZo1DskSZIkSZIkNVYjFyn5ArA58HRmbg98AJjewOdLkiRJUreULv8oSVqCGhkQvpOZ7wBERN/MfARYt4HPlyRJkqRuK6LqCiRJXVUj5yCcFhErANcCN0bEy8CzDXy+JEmSJEmSpAZr5CrG+9Q2T4yIW4DlgRsa9XxJkiRJkpqFw74ldSUNCQgjogdwf2a+DyAzb2vEcyVJkiRJkiQtWQ2ZgzAz5wJ/i4g1GvE8SZIkSZKalfNBSupqGjkH4VDgwYj4M/Bmy8HM3LuB75AkSZIkSZLUQI0MCE9q4LMkSZIkSZIkLQWNDAh3z8yv1R+IiDMA5yOUJEmSJEmSmlRD5iCs2bmNY7s18PmSJEmSJEmSGqzDAWFEHBkRDwDrRsT9de1J4P6OlwgRsWtEPBoRUyLi622c7xsRV9TO3xMRI+rOHVs7/mhE7FI7tnpE3BIRD0fEgxHxhUbUKUmSJEmSJHU2jRhifBlwPXAaUB/evZ6ZL3X04RHREziP0kNxGnBvREzIzIfqLjsUeDkz146IccAZwP4RsQEwDtgQWA34Y0SsA8wGjsnM+yJiADA5Im5s9UxJkiRJkiSpy+twD8LMfDUzn8rMAzLz6brW4XCwZgtgSmY+kZkzgcuBsa2uGQtcXNu+GtgxIqJ2/PLMnJGZTwJTgC0y87nMvK9W/+vAw8CwBtUrSZIkSZIkdRqNnINwSRkGTK3bn8a7w7z/uyYzZwOvAiu1597acOQPAPe0fnFEjI+ISRExafr06R36EpIkSZIkSVIz6gwBYbRxLNt5zULvjYj+wK+BL2bma++6MPP8zByTmWOGDBmyGCVLkiRJkiRJnUPDAsKIOCoiBjXqeXWmAavX7Q8Hnl3QNRHRC1geeGlh90ZEb0o4eGlmXrME6pYkSZKkhsjWXSQkSWqgRvYgXJWygMiVtVWH2+q9917cC4yOiJER0Yey6MiEVtdMAA6ube8L3JyZWTs+rrbK8UhgNPDnWm0XAg9n5jkNqlOSJEmSJEnqdBoWEGbmNykB3IXAp4DHIuLUiBjVwefOBo4CJlIWE7kyMx+MiJMjYu/aZRcCK0XEFOBoaqspZ+aDwJXAQ8ANwGczcw7wYeCTwA4R8dda270jdUqSJEnSktSwLhiSJLXSq5EPy8yMiOeB54HZwCDg6oi4MTO/2oHnXgdc1+rY8XXb7wD7LeDeU4BTWh37E23PTyhJkiRJkiR1Kw0LCCPi85Rhvi8CFwBfycxZEdEDeAx4zwGhJEmSJEnNxHkhJXUljexBOBj4WGY+XX8wM+dGxJ4NfI8kSZIkSZKkBmnkIiUjW4eDEXEJQGY+3MD3SJIkSZJUGeeDlNTVNDIg3LB+JyJ6Aps18PmSJEmSJEmSGqzDAWFEHBsRrwMbR8RrtfY68ALw2w5XKEmSJEmSJGmJ6XBAmJmnZeYA4KzMHFhrAzJzpcw8tgE1SpIkSZIkSVpCOrxISUSsl5mPAFdFxKatz2fmfR19hyRJkiRJkqQloxGrGB8DHA6c3ca5BHZowDskSZIkSZIkLQEdDggz8/Da5/YdL0eSJElSs3vzTdhjD3jjDRgwoLT+/Utr2V7YsfpzvXtX/W0kSVIjhhh/bGHnM/Oajr5DkiRJUvOYOhVuu61sb701TJtWwsLXXy+fb7zR/mf17bvoQLH1seWWg2WXnf+z9bFejRgrJUlSN9GI/9vcayHnEjAglCRJkrqQzPL5q1/BuHHvPj93Lrz11vyhYX142J5jzz03/7kZMxavxj592hckvpdjyy5bej5GdPzXUpKkZtCIIcafbkQhkiRJkrqGHj3m9fxbddXGPHPWrHmB4ZtvlvbWW/N/tmf7tddK+Nj6/Ny5i/8d+/Urbdll52231RZ1vj3XtISykiQtCY0YYnxgZv4yIo5u63xmntPRd0iSJElqHi1h1dLsQde7NwwaVFqjZcLMme0LGFs+33574e2VV+bff+ut8jl79nuvs0ePxn1nSZLqNWKI8XK1zwENeJYkSZIkLVURZS7Evn1hxRWX7Ltmz150uFgfKLa0d96B/fZbsrVJkrqvRgwx/knt86SOlyNJkiSp2VXRg7Cr6NVr3srPkiQ1i4Z1Uo+ItSLidxExPSJeiIjfRsRajXq+JEmSJEmSpMZr5CwWlwFXAkOB1YCrgF814sERsWtEPBoRUyLi622c7xsRV9TO3xMRI+rOHVs7/mhE7NLeZ0qSJElqmz0IJUnqWhoZEEZmXpKZs2vtl0CH19qKiJ7AecBuwAbAARGxQavLDgVezsy1ge8BZ9Tu3QAYB2wI7Ar8MCJ6tvOZkiRJkiS1yZWlJXUljVjFuGUa31tqPfEupwSD+wO/7+jzgS2AKZn5RO19lwNjgYfqrhkLnFjbvhr4QURE7fjlmTkDeDIiptSeRzueOZ9//xsuuaQB30aSJEnq5KZOLZ/2IFR35f/2JXU1jVjFeDIlEGz5I/KIunMJfLuDzx8GTK3bnwZsuaBrMnN2RLwKrFQ7fnere4fVthf1TCJiPDC+7G3GQQe9168gSZIkdT1DhlRdgSRJaoRGrGI8shGFLERb/zbTujP3gq5Z0PG2hla/q4N4Zp4PnA+w0UZj8tprF16oJEmS1F0sswwMG7bo6yRJUvNrxBDjHTLz5oj4WFvnM/OaDr5iGrB63f5w4NkFXDMtInoBywMvLeLeRT1zPn37wqhRi127JEmSJEmS1NQaMcR4W+BmYK82ziXQ0YDwXmB0RIwE/klZdOQTra6ZABwM3AXsC9ycmRkRE4DLIuIcysrKo4E/U3oWLuqZkiRJkiRJUpfXiCHGJ9Q+P93xctp8/uyIOAqYCPQELsrMByPiZGBSZk4ALgQuqS1C8hIl8KN23ZWUxUdmA5/NzDkAbT1zSdQvSZIkSZIkNbNG9CAEICKObuPwq8DkzPxrR56dmdcB17U6dnzd9jvAfgu49xTglPY8U5IkSZIkSepuGhYQAmNq7Xe1/T0ow4M/ExFXZeaZDXyXJEmSJEmS1OXNng1vvVXa228vXmuvRgaEKwGbZuYbABFxAnA1sA0wGTAglCRJkiRJUpeQCTNnwptvzgvwFtUW59qW62fNem/1RbT/2kYGhGsAM+v2ZwFrZubbETGjge+RJEmSJEmS2mXOnBK0vfFGafXb9W1Rx9sK8ObOXfx6+vWDZZd9dxswAFZdte1zyy47775+/drXeveGHj3aV1MjA8LLgLsj4re1/b2AX0XEcpRFQiRJkiRJkqQFyixDY197DV5/vXy2tNdfbzvAW1S4tzhDbXv0gP7952/LLQeDB5fPlrCufntRrf7aZZZpf2i3NDUsIMzMb0fEdcBWQACfycxJtdP/2aj3SJIkSZIkqbnMmjV/kNc62GvPuZb9OXPa985ll50/xOvfv/TCGzr03cfbCv3aOrbMMos3NLer6HBAGBEDM/O1iFgReLLWWs6tmJkvdfQdkiRJkiRJWjLmzCnh3KuvwiuvlFa/3Xq/ZfvVV+eFeu+80753DRhQ2sCBpQ0YAKusMv/+grZbWv/+JRzs2XPJ/rp0J43oQXgZsCdlIZKsOx61/bUa8A5JkiRJkiS1Ye7cEta99BK8/PKCg70FBX2vvbbodwwYAMsvDyusUNrQobDeevMCvIWFey37/fs35/BaNSAgzMw9a58jO16OJEmSJElS9zRrVgn4Xnpp4a31NS+/XObuW5AePeYP95ZfHkaNmrfdcrz1fsv2wIHQq5GrWKjpNGKI8aYLO5+Z93X0HZIkSZIkNZOFhTFSZumZ9+KL89r06fDvfy88+HvjjQU/MwIGDYIVVyxt0KAS8rXs1x8fNGj+oK9//+45r57arxH579kLOZfADg14hyRJkiRJTcGgpft56635g7764G9Bxxa00Ebv3vMHequvDptsMn/A1zr0W3HFEvg5PFdLSiOGGG/fiEIkSZIkSZKWhnfegX/9q7QXXph/uz7sa9l+++22n9OjB6y0EgwZAoMHw7rrwoc/XLYHD553vKWttJK9+dScGjHE+KuZeWZte7/MvKru3KmZeVxH3yFJkiRJkrQgmWWhjdahX+vwr2X79dfbfs6AAbDyyiXMW2012Hjj+QO+1qHfCivYq09dQyOGGI8DzqxtHwtcVXduV8CAUJIkSZIkLbZ33oHnn4fnnivt2WfnbbcOAmfMePf9ESXIW3llWGUVGDOmfK6yyrxjLdsrrwz9+i397yg1g0YEhLGA7bb2JUmSJElSN/fmm22Hfq2Pvfzyu+/t2bOEequuWkK9DTd8d+DXsj94sKvvSu3RiB+TXMB2W/uSJEmSJKmLmj27BHvTps3fWgd/bQ3x7dMHhg4tbd11YbvtyvZqq807PnRoCf169lzqX03q0hoREG4SEa9Regv2q21T21+mAc+XJEmSJEkVmzED/vnPd4d/9ceefx7mzp3/vn795gV9m2wCu+7advC34oou3iFVpRGrGJvbS5IkSZLUic2cWQK+Z54prXUIOG1aWdG3tYEDYfjw0t73vnnb9W2FFQz+pGbX1CPxI2JF4ApgBPAU8P9l5rtmIIiIg4Fv1na/k5kX145vBvwc6AdcB3whMzMizgL2AmYCjwOfzsxXluiXkSRJkiSpIm+8AU8/Pa8988z8+88+W1YCrrfiivNCvs03f3fwN2xYCQgldX5NHRACXwduyszTI+Lrtf2v1V9QCxFPAMZQ5jycHBETakHij4DxwN2UgHBX4HrgRuDYzJwdEWdQVl+e77mSJEmSJHUGmfDii/MHfq2DwJdemv+eXr1g9dVhzTVhp53K5xprzPscPhyWXbaa7yNp6Wv2gHAssF1t+2LgVt4d5O0C3JiZLwFExI3ArhFxKzAwM++qHf8F8FHg+sz8Q939dwP7LqH6JUmSJEnqsLffhqeegieeaLu99db81/fvX8K+NdeELbect93SVl3VhT4kzdPsAeEqmfkcQGY+FxErt3HNMGBq3f602rFhte3Wx1s7hDKM+V0iYjylByJrrLHGYhcvSZIkSVJ7zJ1bFvhYUAD43HPzX7/ccjByJKy1VukBOHLk/L0ABw1y3j9J7Vd5QBgRfwRWbePUN9r7iDaO5UKO17/7G8Bs4NK2HpyZ5wPnA4wZMybbukaSJEmSpPaYPbsM933ssXnt8cfhySdLe+ededdGlGG+a61VVv1da63525AhBoCSGqfygDAzd1rQuYj4V0QMrfUeHAq80MZl05g3DBlgOGUo8rTadv3xZ+uefTCwJ7BjZuupWCVJkiRJWnxz5sDUqfOHgC3tiSdKSNhiwAAYNQrWXx/22GP+AHCNNaBv3+q+h6TupfKAcBEmAAcDp9c+f9vGNROBUyNiUG3/I5QFSF6KiNcj4oPAPcBBwH8DRMSulLkMt83Mt9p4piRJkiRJC/TGG3Dzze8OAR9/HGbOnHfdssvC6NGw8cbw8Y+X7Za28sr2ApTUHJo9IDwduDIiDgWeAfYDiIgxwGcy87BaEPht4N7aPSe3LFgCHAn8HOhHWb34+trxHwB9gRuj/Gl8d2Z+Zil8H0mSJElSJ9enT/ncccfyucwypSfguuvCnnuW8G+ddcrn0KGGgJKaXzi6tn3GjBmTkyZNqroMSZL+//buPMyuqkzY/v1UZYYMmAQISUgCBGJACBBmQaaAChhbQUERBGxeW6BRHLG/7lY/7QYURBERGgcmmeU10IDIPEQwQWKAECBCICFhSCCQeah63j/qJBQhlRQktfepOvfvus519l57rb2eTV2HVD1nDZIkqWSzZsHttzdtBjJ8eNNagXV1ZUclSe8WEY9m5uh11av2EYSSJEmSJFWVLbaAE08sOwpJ2nD8jkOSJEmSJEmqYSYIJUmSJEmSpBpmglCSJEmSJEmqYW5S0koR8RrwQtlxSJIkSZIkSa00JDP7r6uSCUJJkiRJkiSphjnFWJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGmaCUJIkSZIkSaphJgglSZIkSZKkGtap7ADai379+uXQoUPLDkOSJEmSJElqlUcffXROZvZfVz0ThK00dOhQJk6cWHYYkiRJkiRJUqtExAutqecUY0mSJEmSJKmGmSCUJEmSJEmSapgJQkmSJEmSJKmGuQahpA3mxw/9mB/c/4Oyw1AzW22yFY+e/Cid6vzfvSRJkiRpzfyLUdIG87eX/0anuk6cOOrEskMRMGHWBB548QGWrFjCxl02LjscSZIkSVKVMkEoaYPadKNNOffQc8sOQ8BPxv+EB158oOwwJEmSJElVzgShJEmSJEmttKJxBdNen8a8JfMY3GswW/TcgogoOyxJWi8mCCVtMJlZdghaA38ukiRJ711mMn3edJ549Ymm12tN71PnTGVZw7JV9Qb3Gsxtn7+N7TfdvsRoJWn9mCCUtEEFfnsqSZKk9iUzeWn+S0ycNZEJL01gwqwJTJw1kTeWvLGqzpa9t2SHTXfgo1t/lB023YG+Pfoyfd50fnj/Dznw8gO55/h7GNl/ZIlPIUnvXyEJwogYBBwN7AtsASwGngD+F7gtMxuLiEOSaonJWkmSpDVrzEamvDaF+6bfx30v3MeDLz7I7AWzAaiPej602Yc4cuSRjN5iNDtutiMj+4+kV9dea7zXQcMOYv/L9ufAy5qShB/s/8EiH0WSNog2TxBGxG+BgcAtwNnAq0A3YFvgo8C/RcR3MvP+to5FUttKnMpajfy5SJIkwT9e/we3TbuNu5+/m/tfuJ+5i+cCMKjXIA4YdgB7DNyD3bbYjVGbj6J75+6tvu92/bbjnuPvYf/f7c8Blx3AXcfd5XRjSe1OESMIz83MJ9ZQ/gTwh4joAmxZQBySCuACzZIkSaoGS1Ys4f4X7ufWZ2/ltmm38czcZwAY1mcYR2x3BB8Z8hH2G7Ifw/oMW+/fYUf0G8E9x9/DQZcfxL6/3ZdbPncLew/ee0M8hiQVos0ThCuTgxFxemb+rPm1ZmXT2joOSao1JmslSVKtWbx8MbdPu53rp1zPzc/czIJlC+jWqRv7D92fU3c7lY8N/xjbfGCbNun7g/0/yEMnPsShVx7KwZcfzPVHXc9h2x7WJn1J0oZW5CYlxwM/W63si2sok9ROuVuuJEmSiraicQW3PXsbVz9x9aqkYN/ufTlmh2MYu91YDhh2AD069ygklmGbDOPBEx/kY1d9jLHXjOV/jvgfTtj5hEL6lqT1UcQahMcAnwOGRcS4Zpd6AnPbun9JqnUmbiVJUkf09Jyn+c1jv+HyyZfz8oKXVyUFjxp5FAcMO4BOdUWOh3nbphttyj3H38Onr/s0J447kSdfe5KzDz6b+rr6UuKRpNYo4v+Y44HZQD/g3Gbl84HJrblBRHyUppGG9cClmXnWate7ApcDu9KUdPxsZk6vXDsTOAloAP41M/9UKe8DXArsACRwYmb+5f09oqSV3DlXkiRJbWV5w3JumHIDF064kIdmPER91HP4todz4s4n8rFtPkbn+s5lhwhAr669uPVzt3LGn87g3L+cy5TXpnD1p6+md7feZYcmSWtURILwxcx8AdirpQoREdnCEJeIqAcuBMYAM4EJETEuM6c0q3YS8EZmbhMRR9O0W/JnI2IkcDSwPbAFcGdEbJuZDTQlHG/PzCMrG6UUM+Zc6sDcLbe6mKyVJEkdxdxFc7nk0Uu4cMKFvDT/JYZ/YDjnHHwOX9jpC2y+8eZlh7dGnes7c8HHL+BDm32IU249hT0u3YMbPnMDO2y6Q9mhSdK71BXQxz0RcVpEvGOn4ojoEhEHRsRlNK1P2JLdgWmZ+VxmLgOuAcauVmcscFnl+AbgoGhanX8scE1mLs3M52naDGX3iOgF7Af8GiAzl2XmvPV8TkmqSiZuJUlSezXzrZmcdutpDP7pYL5793cZ2X8k//u5/2XqqVP55j7frNrkYHMn73oydx13F/OWzGO3/9mNS/92qUvASKo6RYwg/ChwInB1RAwD5gHdaUpO3gH8NDMnraX9QGBGs/OZwB4t1cnMFRHxJtC3Uv7wam0HAouB14DfRsROwKPA6Zm58H09oaRV3DlXkiRJ62vGmzM468GzuPSxS2nMRo7b8Ti+ttfX2u3ou/2G7Mffv/x3jr3pWP755n/m7ufv5leH/4peXXuVHZokAQUkCDNzCfBL4JcR0ZmmtQgXv4cRe2vKNqz+dUtLdVoq7wTsApyWmY9ExM+A7wD//o6bRpwMnAyw5ZZbrn4fSavxm9DqYrJWkiS1N3MWzeEH9/2AX038FQAnjDqBM/c9k6F9hpYb2Aaw2cabcfvnb+esB8/iP+79D8bPGM9vxv6GA4cdWHZoklTIFGMAImJroC4zZwOjIuJfKxuFrMtMYHCz80HArJbqREQnoDfw+lrazgRmZuYjlfIbaEoYvkNmXpKZozNzdP/+/VsRqiRJkiTpvVqyYgk/fujHbPPzbfjlhF9ywqgTePa0Z7n4iIs7RHJwpfq6ev5tv3/jgRMeoEt9Fw66/CD+5ZZ/Yf7S+WWHJqnGFZYgBG4EGiJiG5rW/hsG/L4V7SYAwyNiWGUzkaOBcavVGcfb6xgeCdxd2fRkHHB0RHStTG8eDvw1M18GZkTEdpU2BwFTkLTe3Bij+jiyU5IkVavM5KanbuKDF36Qb935LT685YeZ/C+TufiIixnSZ0jZ4bWZvQfvzaQvT+KMPc/g4kcv5kMXfYhbnrml7LAk1bAiE4SNmbkC+BRwfmZ+DRiwrkaVNqcCfwKeAq7LzCcj4gcR8YlKtV8DfSNiGnAGTdOFycwngetoSv54sWZeAAAgAElEQVTdDpxS2cEY4DTgqoiYDIwC/msDPackSZIkaR1efPNFPnntJ/nUdZ+iZ5ee3PmFO7nlc7cwsv/IskMrRI/OPTj30HN54IQH6NG5B0dcfQRHXH0E/3j9H2WHJqkGFbFJyUrLI+IY4DjgiEpZ59Y0zMxbgVtXK/uPZsdLgKNaaPsj4EdrKJ8EjG5V5JJaxd1yq4ujOSVJUjVqaGzg54/8nH+/599Jkp+M+Qmn73k6neqK/PO0euyz5T5M+vIkfv7Iz/n+fd9n+19uzzf3/ibf2udb9Ozas+zwJNWIIkcQngDsBfwoM5+vTPm9ssD+JRXAjTGqj4lbSZJULZ5/43k+8ruPcMYdZ7D/0P2Z8pUpfH3vr9dscnClLvVd+Mbe32DqKVP51Ac/xQ8f+CFb/Xwrfvbwz1i6YmnZ4UmqAYUlCDNzSmb+a2ZeXTl/PjPPKqp/SZIkSVI5MpPfPvZbdvzVjjz+6uNc8U9XcPMxN3fodQbfj4G9BvL7T/+eR770CB/a9EN89U9fZbtfbMfvJv2O5Q3Lyw5PUgdW5AhCSR2cm2FUF0dzSpKkavDmkjf5zA2f4cRxJ7LrgF2Z/OXJHLvjsf6usha7D9ydu467izuOvYO+Pfpywh9PYPgFw/nFX3/BouWLyg5PUgdkglCSJEmS1CaeePUJdvuf3bjpqZs466CzuOu4uxw12EoRwZitxzDxnydy8zE3M7DXQE677TSGnj+UH97/Q15b+FrZIUrqQEwQStqg3Bij+jiyU5IkleH3j/+ePS7dg/nL5nPP8ffw7Q9/m/q6+rLDancigsO3PZyHTnyI+794P7sN3I1/v+ffGfTTQRz7h2P5y4y/+PuepPVW2EqwEbEt8E1gSPN+M/PAomKQ1LbcDEOSJEmN2ch37vwOPx7/Y/bdcl+uPfJaBvQcUHZYHcK+Q/Zl3yH78tRrT3HRxIu47O+XcdXjVzFq81GctPNJHL3D0fTr0a/sMCW1Q0WOILwe+Bvw/9GUKFz5kiS1AUdzSpKkoi1avojPXP8Zfjz+x3xl9Fe467i7TA62gQ/2/yA//9jPeemMl7j48IsBOO220xhw7gA+cfUnuP7J61myYknJUUpqT4rcS35FZl5UYH+SSuBi09XHkZ2SJKkIryx4hU9c8wkmvDSB8w45j6/u+VV/N2xjG3fZmJN3PZmTdz2Zya9M5srJV3LV41dx8zM306trLw4bfhifHPFJPrbNx+jZtWfZ4UqqYm2eIIyID1QOb46IrwA3AUtXXs/M19s6BknFcO0TSZKk2jR93nQOvvxgZs2fxR8++wc+OeKTZYdUc3bcbEfOGXMO/33Qf3Pv9Hu5+omrGff0OK5+4mq61Hfh4K0OZux2Yzlk60MY2mdo2eFKqjJFjCB8FEhYNdet+bTiBLYqIAZJqjl+Yy9Jkorw9JynOfiKg1mwbAF3H383ew7as+yQalp9XT0HbXUQB211EA2NDYyfMZ6bpt7ETVNv4tZnbwVgmw9sw5itxjBmqzEcMOwA+nTrU3LUksrW5gnCzBwGEBHdMvMdiyBERLe27l9SsVz3TpIkqXZMfmUyY64YQ2Zy7/H3stPmO5Udkpqpr6tftbHJuYecy1NznuLP//gzf37uz1z+98u5aOJF1EUdO222E/sM3oe9B+/NPlvuw+Beg/2yWaoxRa5BOB7YpRVlkqQNyKnfkiSpLUx+ZTIHXHYA3Tt1587j7mREvxFlh6S1iAhG9h/JyP4jOX3P01nWsIyHZz7MXc/dxYMzHuS3k37LLyb8AoCBPQey1+C92HXAruy8+c7sPGBnNt1o05KfQFJbKmINws2BgUD3iNiZt6ca9wJ6tHX/korjZhiSJEm14ek5TzPmijF079Sd+0+4n602ceWo9qZLfRf2G7If+w3ZD4AVjSuY/Mpkxs8Yz0MzHuLhmQ9zw5QbVtXfoucWTcnCzXdmZP+RjOg3gm37bstGXTYq6xEkbUBFjCA8FPgiMAg4r1n5fOC7BfQvqUBORageTveWJElt4fk3nuegyw8C4K7j7jI52EF0quvELgN2YZcBu3Dq7qcCMG/JPCa9PInHZj/GYy83vW6fdjsN2bCq3ZDeQxjRbwQj+o1gu77bsdUmWzG0z1CG9BlCt06uKia1F0WsQXgZcFlEfDozb2zr/iRJ7+TITkmStKHMnj+bg684mEXLF3HvF+9lu37blR2S2lCfbn3Yf+j+7D90/1VlS1cs5dnXn2XqnKk89dpTTJ3b9P7Aiw+waPmid7QfsPEAhm0yjKF9hjKszzAG9xrMFj23YEDPAWzRcws23WhTOtUVufKZpJYU+UkcEhFnrFb2JvBoZk4qMA5JbcS17iRJkjquBcsWcPjVh/PKgle4+/i72XGzHcsOSSXo2qkrO2y6AztsusM7yhuzkZfeeonp86Yzfd50np/3PM/Pe57p86bz0IsPcc0T19CYje9oUxd1bLbRZqsShptvtDn9N+pP3+596dejH317NL3369GPvt370rtbb+qirsjHlWpGkQnC0ZXXzZXzw4AJwJcj4vrMPKfAWCSpw3O6tyRJ2lAaGhs45sZjmPTyJMYdPY7dB+5edkiqMnVRx+DegxncezD7Dtn3XdeXNyznlYWvMHv+bGbNn8Ws+bOYveDt45lvzWTCSxOYu3guKxpXrLGP+qinb4++bNJtE3p17UXvbr3p1bVX03HXdx/37tabnl160qNzj3e9unXq5u/LUjNFJgj7Artk5gKAiPhP4AZgP+BRwASh1AG47p0kSVLHkpmcfvvp3PLMLVz48Qs5bNvDyg5J7VDn+s4M6jWIQb0GrbVeZvLW0reYs2gOcxfPbXpfNPcd528seYO3lr7FW0vfYvb82auO31r61ntaXqd7p+5rTB726NyDrp260qW+C13rm97fdbym683KOtd3plNdJ+qjnk51nZqO6+rfUbbyvLVl9VFPXdSZ2FSbKDJBuCWwrNn5cmBIZi6OiKUFxiGpjbjWXXVy6rckSVofF064kAsnXMg39voGX9ntK2WHow4uIujdrTe9u/Vma7Z+T20bs5GFyxby1tK3eHPpm6uShouXL2bR8kUtv1a883z+svnMXTyXpSuWsqxhGUsbmt6XNSxbVba8cXkb/RdYtyBWJQrroq7pmGbH61G+rjYrk5MrB4a83/MNcY8Ndd5S2Vp/Bu8hSfteBtG01X1bo8gE4e+BhyPij5XzI4CrI2IjYEqBcUiSJEmSWuHBFx/ka3/6GkdsewRnjzm77HCktaqLOnp27UnPrj0ZyMA27asxG1nesLzFBOKyhmU0ZAMrGlfQ0Nj0vqJxRavKVp6vqayhsYEkacxGGrORzGbH76WcluusrT28PQBh5QCR93q+rjqZud59tPa8pbK1eS8DY97LYI22uO97uWdhCcLM/P8j4jZgHyCAL2fmxMrlzxcVh6S25XD36uF0b0mStD5mzZ/FUdcfxbA+w7jin65wcwipmbqoo2unrnTt1JWe9Cw7HKlFcUrr/i4sej/xx4BZK/uNiC0z88WCY5DURpzKWp2c+i1Jkt6rZQ3LOOr6o5i/dD53fuFOenfrXXZIkqQ2VFiCMCJOA/4TeAVooGkUYQI7FhWDJEmSJGndvv3nbzN+xniuO/I6tt90+7LDkSS1sSJHEJ4ObJeZcwvsU1LBnNZaPZzuLUmS3o/bnr2N8x85n9N2P42jtj+q7HAkSQUochGJGcCbBfYnqWBOZZUkSWrfXl7wMsf/3+PZcbMdOWfMOWWHI0kqSJEJwueAeyPizIg4Y+WrNQ0j4qMR8XRETIuI76zheteIuLZy/ZGIGNrs2pmV8qcj4tDV2tVHxGMRcct6PpskVS3XhpQkSa3RmI0c/3+PZ/6y+Vz96avp1qlb2SFJkgpS5BTjFyuvLpVXq0REPXAhMAaYCUyIiHGZOaVZtZOANzJzm4g4Gjgb+GxEjASOBrYHtgDujIhtM7Oh0u504Cmg1/o9mqSVnNYqSZLUPl3wyAXc8Y87uOiwixjZf2TZ4UiSClRYgjAzvw8QERtl5sL30HR3YFpmPldpfw0wFmieIBwLfK9yfAPwi2jKUowFrsnMpcDzETGtcr+/RMQg4DDgR0CrRjJKUnviepCSJKm1pr0+jTPvOpPDtz2c/7Pr/yk7HElSwQqbYhwRe0XEFJpG7BERO0XEL1vRdCBN6xeuNLNStsY6mbmCprUO+66j7fnAt4DGtcR8ckRMjIiJr732WitClWqbU1mrk2tDSpKktWnMRk4adxJd6rvwq8N+5YwQSapBRa5BeD5wKDAXIDP/DuzXinZr+tdp9b92W6qzxvKIOBx4NTMfXVvHmXlJZo7OzNH9+/dvRaiSHLUmSZLUvlw04SLuf+F+zjv0PAb2Wn0shiSpFhSZICQzZ6xW1LDGiu80Exjc7HwQMKulOhHRCegNvL6WtvsAn4iI6cA1wIERcWXrnkKS2ge//ZckSesyfd50vn3ntzlk60M4YdQJZYcjSSpJkQnCGRGxN00j+LpExDeoTDdehwnA8IgYFhFdaNp0ZNxqdcYBx1eOjwTuzqa5juOAoyu7HA8DhgN/zcwzM3NQZg6t3O/uzDx2vZ9QqnFOZZUkSWo/MpNTbj2FiOCSwy/xy0VJqmFF7mL8ZeBnNK0BOBO4AzhlXY0yc0VEnAr8CagHfpOZT0bED4CJmTkO+DVwRWUTktdpSvpRqXcdTRuarABOabaDsSTVBNeGlCRJazLu6XHc+uytnHfIeQzpM6TscCRJJSokQRgR9cAXMvPz76d9Zt4K3Lpa2X80O14CHNVC2x/RtFNxS/e+F7j3/cQl6d385lmSJKn6LVq+iNNvP50dNt2BU3c/texwJEklK2SKcWXU3tgi+pJUHkeqVRc3jJEkSS357wf+mxfefIELP34hnes7lx2OJKlkRU4xfigifgFcCyxcWZiZfyswBkmqOa4NKUmSmnt27rOcM/4cjt3xWPYbsl/Z4UiSqkCRCcK9K+8/aFaWwIEFxiCpjTlqTZIkqbp948/foGt9V845+JyyQ5EkVYkiE4QnZeZzzQsiYqsC+5fUxhypVl1cD1KSJK3u/hfuZ9zT4/ivA/+LAT0HlB2OJKlKFLIGYcUNayi7vsD+JUmSJKlmNWYj37jjGwzqNYiv7vnVssORJFWRNh9BGBEjgO2B3hHxqWaXegHd2rp/ScVy1Fr1cfMYSZIEcN2T1zFh1gR+N/Z3dO/cvexwJElVpIgpxtsBhwN9gCOalc8H/rmA/iVJkiSppi1dsZQz7zqTnTbbiWN3PLbscCRJVabNE4SZ+UfgjxGxV2b+pa37k1QeR6pVFzeMkSRJK1008SKmz5vOHcfeQX1dfdnhSJKqTJFrEP5TRPSKiM4RcVdEzIkIv7qSOhiTUtXHzWMkSapti5Yv4qwHz+LAYQcyZusxZYcjSapCRSYID8nMt2iabjwT2Bb4ZoH9S5IkSVLNuWjCRbyy8BV+sP8Pyg5FklSlikwQdq68fxy4OjNfL7BvSQVwpFp1ccMYSZK0cNlCzn7obMZsNYZ9ttyn7HAkSVWqiE1KVro5IqYCi4GvRER/YEmB/UuSJElSTblwwoW8tug1vr//98sORZJUxQobQZiZ3wH2AkZn5nJgETC2qP4lFcNRa9XHzWMkSapNC5Yt4Mfjf8yhWx/KXoP3KjscSVIVK3IEIZn5RrPjhcDCIvuX1LZMREmSJFWPiyZcxJxFc/je/t8rOxRJUpUrcg1CSVKB3FFakqTatXTFUs5/5HwOHHYgew7as+xwJElVrtARhJI6PpNS1cfNYyRJqj2/f/z3zJo/i9+O/W3ZoUiS2oHCRhBGxD9FRO9m530i4pNF9S+p7ZmIkiRJKl9jNnLO+HMYtfkoxmw1puxwJEntQJFTjP8zM99ceZKZ84D/LLB/SaopbhgjSVJtuuWZW5g6Zyrf2vtb/j4gSWqVIhOEa+rLKc5SB+MvoZIkSeU6+6GzGdpnKEdtf1TZoUiS2okiE4QTI+K8iNg6IraKiJ8CjxbYv6Q25i7G1cmfiyRJtWP8jPGMnzGer+/1dTrVOR5DktQ6RSYITwOWAdcC1wNLgFMK7F+SJEmSOrQL/noBvbv25oRRJ5QdiiSpHSnsK6XMXAh8p6j+JJXDXYyrhz8LSZJqy6z5s7hhyg2ctvtpbNRlo7LDkSS1I22eIIyI8zPzqxFxM7x7i9PM/ERbxyBJtczdpSVJqg0XT7yYhsYGTtnNiVqSpPemiBGEV1Tef1JAX5JKZCJKkiSpHMsalnHxoxfz8eEfZ+sPbF12OJKkdqbN1yDMzJUbkYzKzPuav4BRrblHRHw0Ip6OiGkR8a5pyhHRNSKurVx/JCKGNrt2ZqX86Yg4tFI2OCLuiYinIuLJiDh9/Z9UEriLcTXxZyFJUu24YcoNvLLwFU7d/dSyQ5EktUNFblJy/BrKvriuRhFRD1wIfAwYCRwTESNXq3YS8EZmbgP8FDi70nYkcDSwPfBR4JeV+60Avp6ZHwT2BE5Zwz0lSZIkqV244K8XMPwDwzlk60PKDkWS1A4VsQbhMcDngGERMa7ZpZ7A3FbcYndgWmY+V7nfNcBYYEqzOmOB71WObwB+EU1DZ8YC12TmUuD5iJgG7J6ZfwFmA2Tm/Ih4Chi42j0lvUeZTjGuRv5cJEnq2B6b/RgPz3yY8w89n7oocgyIJKmjKGINwvE0JeP6Aec2K58PTG5F+4HAjGbnM4E9WqqTmSsi4k2gb6X84dXaDmzesDIdeWfgkdU7joiTgZMBttxyy1aEKkmSJEnFuvRvl9K1vivH7XRc2aFIktqpItYgfCEz783MvYCngd5AL2BWZq5oxS3WtIjW6sNhWqqz1rYRsTFwI/DVzHxrDbFfkpmjM3N0//79WxGqpFjjx05l8GchSVLHt3j5Yq56/CqOHHkkm3TfpOxwJEntVGHjzyPiJOCvwKeAI4GHI+LEVjSdCQxudj4ImNVSnYjoRFMS8vW1tY2IzjQlB6/KzD+81+eR9G7uYlyd/LlIktRx3fjUjby59E2+tMuXyg5FktSOFblAxbeAnTPzi5l5PLAr8O1WtJsADI+IYRHRhaZNR8atVmccb2+CciRwdzYtujUOOLqyy/EwYDjw18r6hL8GnsrM89b7ySRJkiSpBJf+7VK2+cA2fGTIR8oORZLUjhWZIJxJ07qDK83nnWsLrlFlGvKpwJ+Ap4DrMvPJiPhBRHyiUu3XQN/KJiRnAN+ptH0SuI6mzUduB07JzAZgH+ALwIERMany+viGeEip1jXl31UN/FlIktSxPTv3We574T5O2vkk/92XJK2XIjYpWekl4JGI+CNN6wCOpWk03xkAaxvJl5m3AreuVvYfzY6XAEe10PZHwI9WK3uQNa9PKGk9ZKafLEmSpIL85rHfUB/1HL/T8euuLEnSWhSZIPxH5bXSHyvvPQuMQZJqTtOKC5IkqSNZ0biC3/39dxy27WEM6Dmg7HAkSe1cYQnCzPx+UX1JKo8750qSJLW9u567i5cXvMwXd/pi2aFIkjqAwhKEEdGfpo1Ktge6rSzPzAOLikGSaonJWkmSOq4rH7+SPt368PHhLqUuSVp/RW5SchUwFRgGfB+YTtMOxZI6iMSprNXIn4skSR3LwmULuempm/jMyM/QtVPXssORJHUARSYI+2bmr4HlmXlfZp4I7Flg/5IK4A56kiRJbeuPT/+RhcsX8vkdP192KJKkDqLITUqWV95nR8RhwCxgUIH9S1JNMVkrSVLHdOXkK9my95Z8eMsPlx2KJKmDKDJB+MOI6A18HbgA6AV8rcD+JbUxd8uVJElqW68ufJU7/nEH39z7m9RFkRPCJEkdWZG7GN9SOXwTOKCofiWp1pm4lSSp47j2iWtpyAaO3fHYskORJHUgRe9i/M/A0Ob9VtYilNRBuHOuJEnSuzU0NrCsYdmq19KGpe84X1W+4t3lSxuWsnj5Yia/MpnrplzHqM1Hsf2m25f9SJKkDqTIKcZ/BB4A7gQaCuxXUkHcLbe6mKyVJOm9acxGFixbwFtL32rxNX/pfBYtX8TiFYtZtHzRqte7zpcvfse1FY0r1ju+jTpvxCFbH8K/7ftvG+BpJUl6W5EJwh6Z+e0C+5MkYeJWklR7MpO3lr7FKwtfYc6iOcxdNJe5i+e+47j5+bwl85qSf8vmt+r+Xeq70L1Td3p07kGPzj3o3vnt4/49+r9dXqnTvXN3unXqRpf6LnSt70qX+i7venXt1EJ5s/qbb7w59XX1bfxfT5JUi4pMEN4SER/PzFsL7FNSwdw5V5IktaVlDct48c0XmT5vOrPmz2L2/NlN7wtmM3vB7FVli1csXmP7TnWd6Nu9L3179KVv974M7zucTbptQu+uvenVtVeLr97dmq5v3GVjOtUV+WeUJEltr8h/2U4HvhsRS4HlQACZmb0KjEE15pm5z/CP1/9Rdhg1443Fb9Cjc4+yw1CFyVpJUnuUmby68FWmzpnKtNenMX3edKa/OZ3p86bz/BvPM2v+rHeNju/ZpScDeg5gwMYD2GPgHgzYeAADeg5gs402o1+PfvTr0W9VQrBX117+GylJ0mqK3MW4Z1F9SSvt/7v9mb1gdtlh1JSR/UeWHYIkSWoHMpPn5z3P5FcmM3XOVJ6e+zRT50xl6pypzFsyb1W9uqhjcK/BDO0zlIO3OpihfYYytM9QhvQewqBegxjQcwAbd9m4xCeRJKn9K3IX4xuA3wC3Z2ZjUf2qts1fNp9jdjiG0/c4vexQaoYJwuqT6RqEkqRyLW9YzpOvPcmklyfx2OzHmPTKJCa9PIm3lr61qs4WPbdgRL8RHLPDMYzoN4IR/UYw/APDGdRrEJ3rO5cYvSRJHV+RU4x/BZwAXBAR1wO/y8ypBfavGtSYjQzsOZA9Bu1RdihS4dzFWJJUltcWvsZfZv6F8TPGM37GeCbMmsCSFUsA6NG5BzttthPHfuhYRm0+ip0234kR/UbQq6srD0mSVJYipxjfCdwZEb2BY4A/R8QM4H+AKzNzeVGxqHY0ZqM7vUmSJLWxuYvmcvfzd3Pnc3dyz/R7ePb1ZwHoXNeZXQbswr+M/hd2H7g7O2++M9t8YBt/P5MkqcoUuv1WRPQFjgW+ADwGXAV8GDge2L/IWFQbGhobqIu6ssOQSvXwzIff11qc6zM1efXF46u5z/Xpt1b6XJ9+a6XP9em3Vvpcn35rpc/16besPqe8NoU7n7uTv83+G0nSq2svPjLkI3xply+x9+C92XXArnTv3P19xyZJkopR5BqEfwBGAFcAR2Tmyr9Wr42IiUXFodrSmI0mCFWzVk7VOnHciSVHIknqqDrXdWavwXvx/f2/z5itxzB6i9F0qit0DIIkSdoAivzX+xeZefeaLmTm6ALjUA0xQahadug2h/LIlx5ZtebT+/F+1zGMeP/rH7anPten3zL6XJ9+a6XP9enXn2l19rk+/fozXbfNNt7MHYQlSeoA2jxBGBG7ATNWJgcj4jjg08ALwPcy8/W2jkG1KTNJkvpwjRvVprqoY/eBu5cdhiRJkiSpyhUxtOpiYBlAROwHnAVcDrwJXFJA/6pRjdkI4AhCSZIkSZKktShiinF9s1GCnwUuycwbgRsjYlIB/atGmSCUJEmSJElatyIyJ/URsTIReRDQfB1CVzBWmzFBKEmSJEmStG5FZE6uBu6LiD8Ci4EHACJiG5qmGa9TRHw0Ip6OiGkR8Z01XO8aEddWrj8SEUObXTuzUv50RBza2nuq/VuZIKyvcw1CSZIkSZKklrT5CL7M/FFE3AUMAO7IzKxcqgNOW1f7iKgHLgTGADOBCRExLjOnNKt2EvBGZm4TEUcDZwOfjYiRwNHA9sAWwJ0RsW2lzbruqXauIRsARxBKkiRJkiStTSFTfDPz4TWUPdPK5rsD0zLzOYCIuAYYCzRP5o0Fvlc5vgH4RUREpfyazFwKPB8R0yr3oxX3fIfn3niOz1z/mVaGrGqwvHE5YIJQkiRJkiRpbdrDGoADgRnNzmcCe7RUJzNXRMSbQN9K+cOrtR1YOV7XPYmIk4GTATpv0ZknXn3i/T+FSrHjZjuy56A9yw5DkiRJkiSparWHBGGsoSxbWael8jUNKVv9nmTmJcAlAKNHj86Jp0xce6SSJEmSJElSO9Me5l7OBAY3Ox8EzGqpTmXH5N7A62tp25p7SpIkSZIkSR1ee0gQTgCGR8SwiOhC06Yj41arMw44vnJ8JHB3ZTOUccDRlV2OhwHDgb+28p6SJEmSJElSh1f1U4wrawqeCvwJqAd+k5lPRsQPgImZOQ74NXBFZROS12lK+FGpdx1Nm4+sAE7JbNradk33LPrZJEmSJEmSpLJF00A7rcvo0aNz4kTXIJQkSZIkSVL7EBGPZuboddYzQdg6ETEfeLrsOKQq1g+YU3YQUhXzMyK1zM+HtHZ+RqSW+fmQ1m67zOy5rkpVP8W4ijzdmoyrVKsiYqKfEallfkaklvn5kNbOz4jUMj8f0tpFRKumw7aHTUokSZIkSZIktREThJIkSZIkSVINM0HYepeUHYBU5fyMSGvnZ0RqmZ8Pae38jEgt8/MhrV2rPiNuUiJJkiRJkiTVMEcQSpIkSZIkSTXMBKEkSZIkSZJUw0wQtkJEfDQino6IaRHxnbLjkapJRPwmIl6NiCfKjkWqNhExOCLuiYinIuLJiDi97JikahIR3SLirxHx98pn5PtlxyRVm4ioj4jHIuKWsmORqk1ETI+IxyNiUkRMLDseqdpERJ+IuCEiplb+JtmrxbquQbh2EVEPPAOMAWYCE4BjMnNKqYFJVSIi9gMWAJdn5g5lxyNVk4gYAAzIzL9FRE/gUeCT/hsiNYmIADbKzAUR0Rl4EDg9Mx8uOTSpakTEGcBooFdmHl52PFI1iYjpwOjMnFN2LFI1iojLgAcy89KI6AL0yMx5a6rrCMJ12x2YlpnPZeYy4FbYVLgAAASfSURBVBpgbMkxSVUjM+8HXi87DqkaZebszPxb5Xg+8BQwsNyopOqRTRZUTjtXXn57LVVExCDgMODSsmORJLUvEdEL2A/4NUBmLmspOQgmCFtjIDCj2flM/ONOkvQeRcRQYGfgkXIjkapLZfrkJOBV4M+Z6WdEetv5wLeAxrIDkapUAndExKMRcXLZwUhVZivgNeC3laUqLo2IjVqqbIJw3WINZX6zLUlqtYjYGLgR+GpmvlV2PFI1ycyGzBwFDAJ2jwiXq5CAiDgceDUzHy07FqmK7ZOZuwAfA06pLH8kqUknYBfgoszcGVgItLivhgnCdZsJDG52PgiYVVIskqR2prKu2o3AVZn5h7LjkapVZcrLvcBHSw5Fqhb7AJ+orLF2DXBgRFxZbkhSdcnMWZX3V4GbaFoiTFKTmcDMZrMzbqApYbhGJgjXbQIwPCKGVRZ0PBoYV3JMkqR2oLIBw6+BpzLzvLLjkapNRPSPiD6V4+7AwcDUcqOSqkNmnpmZgzJzKE1/g9ydmceWHJZUNSJio8omcFSmTR4CPFFuVFL1yMyXgRkRsV2l6CCgxc0SOxUSVTuWmSsi4lTgT0A98JvMfLLksKSqERFXA/sD/SJiJvCfmfnrcqOSqsY+wBeAxytrrAF8NzNvLTEmqZoMAC6LiHqavri+LjNvKTkmSVL7sBlwU9P3sXQCfp+Zt5cbklR1TgOuqgx4ew44oaWKkelyepIkSZIkSVKtcoqxJEmSJEmSVMNMEEqSJEmSJEk1zAShJEmSJEmSVMNMEEqSJEmSJEk1zAShJEmSJEmSVMNMEEqSJOk9iYi+ETGp8no5Il5qdj6+jfrcOSIuXcv1/hFxe1v0LUmS1NF1KjsASZIktS+ZORcYBRAR3wMWZOZP2rjb7wI/XEtMr0XE7IjYJzMfauNYJEmSOhRHEEqSJGmDiYgFlff9I+K+iLguIp6JiLMi4vMR8deIeDwitq7U6x8RN0bEhMprnzXcsyewY2b+vXL+kWYjFh+rXAf4v8DnC3pUSZKkDsMEoSRJktrKTsDpwIeALwDbZubuwKXAaZU6PwN+mpm7AZ+uXFvdaOCJZuffAE7JzFHAvsDiSvnEyrkkSZLeA6cYS5Ikqa1MyMzZABHxD+COSvnjwAGV44OBkRGxsk2viOiZmfOb3WcA8Fqz84eA8yLiKuAPmTmzUv4qsMWGfwxJkqSOzQShJEmS2srSZseNzc4befv30Dpgr8xcTMsWA91WnmTmWRHxv8DHgYcj4uDMnFqps7b7SJIkaQ2cYixJkqQy3QGcuvIkIkatoc5TwDbN6mydmY9n5tk0TSseUbm0Le+ciixJkqRWMEEoSZKkMv0rMDoiJkfEFODLq1eojA7s3Wwzkq9GxBMR8XeaRgzeVik/APjfIoKWJEnqSCIzy45BkiRJWquI+BowPzPXtInJyjr3A2Mz843iIpMkSWr/HEEoSZKk9uAi3rmm4TtERH/gPJODkiRJ750jCCVJkiRJkqQa5ghCSZIkSZIkqYaZIJQkSZIkSZJqmAlCSZIkSZIkqYaZIJQkSZIkSZJqmAlCSZIkSZIkqYb9Pwcli4rU2VGkAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1296x648 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#!/usr/bin/env python\n",
    "# -*- coding: utf-8 -*-\n",
    "\n",
    "'''\n",
    "STDP modulated with reward\n",
    "\n",
    "Adapted from Fig. 1c of:\n",
    "Eugene M. Izhikevich \n",
    "Solving the distal reward problem through linkage of STDP and dopamine signaling. \n",
    "Cerebral cortex 17, no. 10 (2007): 2443-2452.\n",
    "\n",
    "Note:\n",
    "The variable \"mode\" can switch the behavior of the synapse from \"Classical STDP\" to \"Dopamine modulated STDP\".\n",
    "\n",
    "Author: Guillaume Dumas (Institut Pasteur)\n",
    "Date: 2018-08-24\n",
    "'''\n",
    "from brian2 import *\n",
    "\n",
    "# Parameters\n",
    "simulation_duration = 6 * second\n",
    "\n",
    "## Neurons\n",
    "taum = 10*ms\n",
    "Ee = 0*mV\n",
    "vt = -54*mV\n",
    "vr = -60*mV\n",
    "El = -74*mV\n",
    "taue = 5*ms\n",
    "\n",
    "## STDP\n",
    "taupre = 20*ms\n",
    "taupost = taupre\n",
    "gmax = .01\n",
    "dApre = .01\n",
    "dApost = -dApre * taupre / taupost * 1.05\n",
    "dApost *= gmax\n",
    "dApre *= gmax\n",
    "\n",
    "## Dopamine signaling\n",
    "tauc = 1000*ms\n",
    "taud = 200*ms\n",
    "taus = 1*ms\n",
    "epsilon_dopa = 5e-3\n",
    "\n",
    "# Setting the stage\n",
    "network = Network()\n",
    "\n",
    "## Stimuli section\n",
    "input_indices = array([0, 1, 0, 1, 1, 0, \n",
    "                       0, 1, 0, 1, 1, 0])\n",
    "input_times = array([ 500,  550, 1000, 1010, 1500, 1510, \n",
    "                     3500, 3550, 4000, 4010, 4500, 4510])*ms\n",
    "input = SpikeGeneratorGroup(2, input_indices, input_times)\n",
    "\n",
    "neurons = NeuronGroup(2, '''dv/dt = (ge * (Ee-vr) + El - v) / taum : volt\n",
    "                            dge/dt = -ge / taue : 1''',\n",
    "                      threshold='v>vt', reset='v = vr',\n",
    "                      method='linear')\n",
    "neurons.v = vr\n",
    "neurons_monitor = SpikeMonitor(neurons)\n",
    "\n",
    "synapse = Synapses(input, neurons, \n",
    "                   model='''s: volt''',\n",
    "                   on_pre='v += s')\n",
    "synapse.connect(i=[0, 1], j=[0, 1])\n",
    "synapse.s = 100. * mV\n",
    "\n",
    "## STDP section\n",
    "synapse_stdp = Synapses(neurons, neurons,\n",
    "                   model='''mode: 1\n",
    "                         dc/dt = -c / tauc : 1 (clock-driven)\n",
    "                         dd/dt = -d / taud : 1 (clock-driven)\n",
    "                         ds/dt = mode * c * d / taus : 1 (clock-driven)\n",
    "                         dApre/dt = -Apre / taupre : 1 (event-driven)\n",
    "                         dApost/dt = -Apost / taupost : 1 (event-driven)''',\n",
    "                   on_pre='''ge += s\n",
    "                          Apre += dApre\n",
    "                          c = clip(c + mode * Apost, -gmax, gmax)\n",
    "                          s = clip(s + (1-mode) * Apost, -gmax, gmax)\n",
    "                          ''',\n",
    "                   on_post='''Apost += dApost\n",
    "                          c = clip(c + mode * Apre, -gmax, gmax)\n",
    "                          s = clip(s + (1-mode) * Apre, -gmax, gmax)\n",
    "                          ''',\n",
    "                   method='euler'\n",
    "                   )\n",
    "synapse_stdp.connect(i=0, j=1)\n",
    "synapse_stdp.mode = 0\n",
    "synapse_stdp.s = 1e-10\n",
    "synapse_stdp.c = 1e-10\n",
    "synapse_stdp.d = 0\n",
    "synapse_stdp_monitor = StateMonitor(synapse_stdp, ['s', 'c', 'd'], record=[0])\n",
    "\n",
    "## Dopamine signaling section\n",
    "dopamine_indices = array([0, 0, 0])\n",
    "dopamine_times = array([3520, 4020, 4520])*ms\n",
    "dopamine = SpikeGeneratorGroup(1, dopamine_indices, dopamine_times)\n",
    "dopamine_monitor = SpikeMonitor(dopamine)\n",
    "reward = Synapses(dopamine, synapse_stdp, model='''''',\n",
    "                            on_pre='''d_post += epsilon_dopa''',\n",
    "                            method='exact')\n",
    "reward.connect(p=1.)\n",
    "\n",
    "# Simulation\n",
    "## Classical STDP\n",
    "synapse_stdp.mode = 0\n",
    "run(simulation_duration/2)\n",
    "## Dopamine modulated STDP\n",
    "synapse_stdp.mode = 1\n",
    "run(simulation_duration/2)\n",
    "\n",
    "# Visualisation\n",
    "dopamine_indices, dopamine_times = dopamine_monitor.it\n",
    "neurons_indices, neurons_times = neurons_monitor.it\n",
    "figure(figsize=(18,9))\n",
    "subplot(411)\n",
    "plot([0.05, 2.95], [2.7, 2.7], linewidth=5, color='k')\n",
    "text(1.5, 3, 'Classical STDP', horizontalalignment='center', fontsize=20)\n",
    "plot([3.05, 5.95], [2.7, 2.7], linewidth=5, color='k')\n",
    "text(4.5, 3, 'Dopamine modulated STDP', horizontalalignment='center', fontsize=20)\n",
    "plot(neurons_times, neurons_indices, 'ob')\n",
    "plot(dopamine_times, dopamine_indices + 2, 'or')\n",
    "xlim([0, simulation_duration/second])\n",
    "ylim([-0.5, 4])\n",
    "yticks([0, 1, 2], ['Pre-neuron', 'Post-neuron', 'Reward'])\n",
    "xticks([])\n",
    "subplot(412)\n",
    "plot(synapse_stdp_monitor.t/second, synapse_stdp_monitor.d.T/gmax, 'r-')\n",
    "xlim([0, simulation_duration/second])\n",
    "ylabel('Extracellular dopamine d(t)')\n",
    "xticks([])\n",
    "subplot(413)\n",
    "plot(synapse_stdp_monitor.t/second, synapse_stdp_monitor.c.T/gmax, 'b-')\n",
    "xlim([0, simulation_duration/second])\n",
    "ylabel('Eligibility trace c(t)')\n",
    "xticks([])\n",
    "subplot(414)\n",
    "plot(synapse_stdp_monitor.t/second, synapse_stdp_monitor.s.T/gmax, 'g-')\n",
    "xlim([0, simulation_duration/second])\n",
    "ylabel('Synaptic strength s(t)')\n",
    "xlabel('Time (s)')\n",
    "tight_layout()\n",
    "show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}