4-network-tests.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Network tests notebook\n",
    "This notebook is used to compare the inferred networks to the ground truth networks through various tests.\n",
    "\n",
    "Copyright (C) 2020  Andrew Sonta, Rishee K. Jain\n",
    "\n",
    "This program is free software: you can redistribute it and/or modify\n",
    "it under the terms of the GNU Affero General Public License as published\n",
    "by the Free Software Foundation, either version 3 of the License, or\n",
    "(at your option) any later version.\n",
    "\n",
    "This program is distributed in the hope that it will be useful,\n",
    "but WITHOUT ANY WARRANTY; without even the implied warranty of\n",
    "MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the\n",
    "GNU Affero General Public License for more details.\n",
    "\n",
    "You should have received a copy of the GNU Affero General Public License\n",
    "along with this program.  If not, see <https://www.gnu.org/licenses/>."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/site-packages/rpy2/robjects/pandas2ri.py:17: FutureWarning: pandas.core.index is deprecated and will be removed in a future version.  The public classes are available in the top-level namespace.\n",
      "  from pandas.core.index import Index as PandasIndex\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib\n",
    "import matplotlib.dates as mdates\n",
    "import matplotlib.pyplot as plt\n",
    "import datetime\n",
    "%load_ext rpy2.ipython\n",
    "import networkx as nx\n",
    "import src.__init__ as init\n",
    "timestep = init.TIME_STEP\n",
    "timesteps = int(1440/timestep)\n",
    "\n",
    "# Set fonts for plotting\n",
    "\n",
    "# font_path = 'MyriadPro-Regular.otf'\n",
    "# prop = fm.FontProperties(fname=font_path)\n",
    "# matplotlib.rcParams['font.family'] = prop.get_name()\n",
    "\n",
    "font = {'family' : 'Arial',\n",
    "        'size'   : 18}\n",
    "matplotlib.rc('font', **font)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%R\n",
    "glasso_adj = array(as.matrix(read.csv('data/graphs/glasso-adj.csv')), c(18,18))\n",
    "influence_adj = array(as.matrix(read.csv('data/graphs/influence-adj.csv')), c(18,18))\n",
    "interaction_adj = array(as.matrix(read.csv('data/graphs/interaction-adj.csv')), c(18,18))\n",
    "info_adj = array(as.matrix(read.csv('data/graphs/info-adj.csv')), c(18,18))\n",
    "prob_adj = array(as.matrix(read.csv('data/graphs/prob-adj.csv')), c(18,18))\n",
    "support_adj = array(as.matrix(read.csv('data/graphs/support-adj.csv')), c(18,18))\n",
    "socialnorm_adj = array(as.matrix(read.csv('data/graphs/socialnorm-adj.csv')), c(18,18))\n",
    "org_adj = array(as.matrix(read.csv('data/graphs/org-adj.csv')), c(18,18))\n",
    "ground_adj = array(as.matrix(read.csv('data/graphs/ground-adj.csv')), c(18,18))\n",
    "graphs = list(glasso_adj, influence_adj, interaction_adj, info_adj, prob_adj, support_adj, socialnorm_adj, org_adj, ground_adj)\n",
    "graphs[is.na(graphs)] <- 0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Graph correlation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1] 0.4531696\n",
       "\n",
       "QAP Test Results\n",
       "\n",
       "Estimated p-values:\n",
       "\tp(f(perm) >= f(d)): 0.002 \n",
       "\tp(f(perm) <= f(d)): 0.998 \n",
       "\n",
       "Test Diagnostics:\n",
       "\tTest Value (f(d)): 0.4531696 \n",
       "\tReplications: 10000 \n",
       "\tDistribution Summary:\n",
       "\t\tMin:\t -0.5159509 \n",
       "\t\t1stQ:\t -0.1195987 \n",
       "\t\tMed:\t -0.005944072 \n",
       "\t\tMean:\t -0.003388017 \n",
       "\t\t3rdQ:\t 0.1078308 \n",
       "\t\tMax:\t 0.55125 \n",
       "\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAEGWlDQ1BrQ0dDb2xvclNwYWNlR2VuZXJpY1JHQgAAOI2NVV1oHFUUPrtzZyMkzlNsNIV0qD8NJQ2TVjShtLp/3d02bpZJNtoi6GT27s6Yyc44M7v9oU9FUHwx6psUxL+3gCAo9Q/bPrQvlQol2tQgKD60+INQ6Ium65k7M5lpurHeZe58853vnnvuuWfvBei5qliWkRQBFpquLRcy4nOHj4g9K5CEh6AXBqFXUR0rXalMAjZPC3e1W99Dwntf2dXd/p+tt0YdFSBxH2Kz5qgLiI8B8KdVy3YBevqRHz/qWh72Yui3MUDEL3q44WPXw3M+fo1pZuQs4tOIBVVTaoiXEI/MxfhGDPsxsNZfoE1q66ro5aJim3XdoLFw72H+n23BaIXzbcOnz5mfPoTvYVz7KzUl5+FRxEuqkp9G/Ajia219thzg25abkRE/BpDc3pqvphHvRFys2weqvp+krbWKIX7nhDbzLOItiM8358pTwdirqpPFnMF2xLc1WvLyOwTAibpbmvHHcvttU57y5+XqNZrLe3lE/Pq8eUj2fXKfOe3pfOjzhJYtB/yll5SDFcSDiH+hRkH25+L+sdxKEAMZahrlSX8ukqMOWy/jXW2m6M9LDBc31B9LFuv6gVKg/0Szi3KAr1kGq1GMjU/aLbnq6/lRxc4XfJ98hTargX++DbMJBSiYMIe9Ck1YAxFkKEAG3xbYaKmDDgYyFK0UGYpfoWYXG+fAPPI6tJnNwb7ClP7IyF+D+bjOtCpkhz6CFrIa/I6sFtNl8auFXGMTP34sNwI/JhkgEtmDz14ySfaRcTIBInmKPE32kxyyE2Tv+thKbEVePDfW/byMM1Kmm0XdObS7oGD/MypMXFPXrCwOtoYjyyn7BV29/MZfsVzpLDdRtuIZnbpXzvlf+ev8MvYr/Gqk4H/kV/G3csdazLuyTMPsbFhzd1UabQbjFvDRmcWJxR3zcfHkVw9GfpbJmeev9F08WW8uDkaslwX6avlWGU6NRKz0g/SHtCy9J30o/ca9zX3Kfc19zn3BXQKRO8ud477hLnAfc1/G9mrzGlrfexZ5GLdn6ZZrrEohI2wVHhZywjbhUWEy8icMCGNCUdiBlq3r+xafL549HQ5jH+an+1y+LlYBifuxAvRN/lVVVOlwlCkdVm9NOL5BE4wkQ2SMlDZU97hX86EilU/lUmkQUztTE6mx1EEPh7OmdqBtAvv8HdWpbrJS6tJj3n0CWdM6busNzRV3S9KTYhqvNiqWmuroiKgYhshMjmhTh9ptWhsF7970j/SbMrsPE1suR5z7DMC+P/Hs+y7ijrQAlhyAgccjbhjPygfeBTjzhNqy28EdkUh8C+DU9+z2v/oyeH791OncxHOs5y2AtTc7nb/f73TWPkD/qwBnjX8BoJ98VQNcC+8AAAA4ZVhJZk1NACoAAAAIAAGHaQAEAAAAAQAAABoAAAAAAAKgAgAEAAAAAQAAAeCgAwAEAAAAAQAAAeAAAAAApZ9jSgAAQABJREFUeAHtnQeYFMXWhg85CkgWlCBZgiTJkjMIShYlqJgQAxiuOQfMijmgCErOSM5IRiQjQbICKjnn/vur+8/c2dmZ3dnZ6enq7q+eZ3Z7uqurTr3V06fCOVVpDDMIAwmQAAmQAAmQQFwJpI1rbsyMBEiABEiABEhAEaAC5oNAAiRAAiRAAjYQoAK2ATqzJAESIAESIAEqYD4DJEACJEACJGADASpgG6AzSxIgARIgARKgAuYzQAIkQAIkQAI2EKACtgE6syQBEiABEiABKmA+AyRAAiRAAiRgAwEqYBugM0sSIAESIAESoALmM0ACJEACJEACNhCgArYBOrMkARIgARIgASpgPgMkQAIkQAIkYAMBKmAboDNLEiABEiABEqAC5jNAAiRAAiRAAjYQoAK2ATqzJAESIAESIAEqYD4DJEACJEACJGADASpgG6AzSxIgARIgARKgAuYzQAIkQAIkQAI2EKACtgE6syQBEiABEiABKmA+AyRAAiRAAiRgAwEqYBugM0sSIAESIAESoALmM0ACJEACJEACNhCgArYBOrMkARIgARIggfREQAI+Av369ZMTJ074vib436ZNG+natWuCc6G+HD16VNKkSSO5cuWSdevWyccffyzfffddqKipPnfx4kX5+++/5dprr40qrV9//VUGDx4sX3zxRcj7H3roITl58qT/WqZMmaRkyZLSvXt3ue666/znrTyYPn26LF26VF577TWVDeoH5c6TJ0/Ms921a5f0799f/vzzT3n22WelQ4cOifJA3t9//71MmjRJDhw4IHnz5hU8G/fdd59kyZIlUfzLly/LAw88ICVKlJCnn346wXWUBc9cYMiQIYNiC8alS5cOvOQ/njZtmowcOdL/HQfp0qWTggULSqdOnaRatWoJrkX7ZcuWLfLmm2/KDz/8IIZhSO/evRWXsmXLRpuk/76dO3fK9ddfr77j93Hu3Dnp27ev/zoPvEGAPWBv1HNEpRw9erTs2LFDrrrqqkQfKJ/kwunTp6VMmTKCFzkCXrDr169P7raorzdt2lSgoKINe/bskVGjRoW9fcyYMQl4QPl8++23Uq5cOZk7d27Y+2J54Z9//pGtW7eqJC9duqT4+r7HMh+kBaX7xx9/yCOPPCI33nhjouSPHz8u9erVk+eee06qVKkijz32mPr+zjvvSPXq1ZVCDr4J9TNhwgR58cUX5eDBgwkunz9/XoYNGyZotPmeufTp06v4yH/VqlUJ4vu+bNq0ScaPHy/Fixf3f6655hpZsGCB1KxZU+bNm+eLmqr/aNxBvitXrigFvHbtWjl16lSq0sTNaEA0b97cn86+fftk9+7d/u888BABs2XHQAKKQL58+Yz33nsvahpHjhwxzJ+O8dtvv0WdRkpuNHsQxtdff52SWxLEHTt2rHH11VcnOBf4JRQPUwkbZs/QyJ8/v2G+jAOjW35s9pIU3yVLlliSV8OGDQ1TuYZN2+wBGkWLFjX++uuvBHEOHz5smI0So0WLFgnO48ttt91mmD07o1KlSsbrr7+e4LrZuFDlMZV0gvMXLlxQ6XXp0iXBed8XU+EbBQoU8H31/zd724bZEDBuueUW/7nUHJgKXclnNnxSk0yie81RF8Ux0QWe8BwB9oA91NiKRVHRIzNfjGoIFMOgt956q7/1jh4pwp133ilDhw4VU1FI3bp11bl///1XbrrpJpk4caKUL19ecubMqYYL9+/fL7fffrsassbQ3pw5c1R8/Fm+fLk0a9ZMcufOra43aNBATOWurmM40FQEaojwmWeeUecwJIq0MCxqKkjp06ePoNfmC+g5NW7cWHLkyKFkWb16te9SxP/RQ/vkk0/k0KFDMnnyZP99n376qeqdIu3atWur3pjvIoZeBw0apLigLEWKFBGzoeO7LL///rsqJ5hgGPXuu++WY8eOqes//vijug9ffHzvueceMRseqqyfffaZPx0cYMj/0UcfTXDO9wU9LUwj+Pj07NlTUC8IqEfwwHAoernBAeWFLBgKL1SoUILLKBN6wTNnzpTNmzf7ryHtn3/+WdUzngnIjN5kcgHD0FWrVlVD4cnFDbyeNm1axTZwGgVsW7durZ4fTFWgLjCSgYApEjxfmILANXDBMxOql4uhdDy/vucPvff//Oc/argc0y2tWrVSoyVIF9eQD4bds2fPLmajRV544QXVi164cKFiiGcV6e3du1feeOMNNaqAexHADNcwKoBh+I8++ui/F8y/48aNE0yN4B7IjN/gXXfdJWfOnFFxMGWC5wfPv9m4VD3tDRs2+O/ngWYEPNfkYIHDEkCPz5yvM8zh1UQfc3hZ3detWzfDnPMzzDkswxyuNEyloD64aL5cVI9hxIgRhjl8Z0ydOtXImDGjus+cV1TXzKFLY/HixSqu+eIyzBeIYb4ADfPFbXTu3NkwhxVVfPOFb5gvINUjM19S6jrywv0I5jCsYb6ADPPFZphD3gZ6TTfccIPRrl07dQ3pNWnSRH0Q33wxGeYwpWG+rAxzuM8whwENU3GkuAeMtBDMl6vxxBNPqOMPP/xQlWPKlCmG2aAwTGVrmPOhhvmCV9c7duxoZM6cWY0umArBePfddxWLFStWqOvmcKthzoUaYIR70Jt8+OGH1TXExXUEc0hW3WfOwRrmC9wwlaFhvqDVNfwxlZtRrFgxw1Si/nO+A/MFbWDEAL1DyGC+lA30eCtXrmygh2c2TlTv8d577zXMoVbfbf7/KJv56lIy+k8GHKB3bs7DGl999ZX/7Pvvv696epAL8uK62WjxXw/XAzaHqg1TyRumovHHDTxADxjPKu73fVCnGA0xlbdhKiwVHdcQ78EHH1R1Dn4VKlRQzzgi4Dk0lbZRp04d9Txv27ZN8TAbcer+wB4wRj5Q/vnz56trZiPIMO0BDIxGIG88u2YD0kAv/Mknn1TXVq5cqUYLzMaoygfPnNmwMl566SX1LK5Zs8YAN1PpG2bDSKVrNkANs5GnyoDnwWywGvidvPXWW+r6559/rn5Td9xxh3pWTGVtmA03w5yrVtcHDBhg1KpVS/0GzCkWA79X/C4Y9CSAVhkDCSgCeFmZc71K8UH5BX7wckIwe6GGaehimPN26jsUjjnPq46Dh6ADFTCGLfECM413VFz8gbI0eyf+77Nnz1ZxoBCQLl6keKH5wk8//aReRr7vgUPQ5jyjUnJm78d32cALCHlCqZm9N3Wv2TvxX4dyS+kQtO9mNASgrBCgLHwvfd91NFJMwyT1FbzQeAgMGMLGUCQCuD/11FMGlCTC9u3bVeMGx4EKOHgIGuWDAvEpcjSAsmXLphobuDcwDB8+XF0L5ANFB6UIxYAAhYwGTagwZMgQxfLs2bOhLqtzgYoCJ8yRDsOc+/XHBxOzp+j/7lPAYOlryJkjJkbWrFlVQyJ4qNt3IxQw6jX4A4UIXlD4CIiHBp3vO85BwUFJmyMjSgEjDTRIfAFKF0zRAAyngNGYQxqBjQk0OFFW8DXnp1XavjTxHw0jc7RCnQoegg5UwGAEpRkY8GyhoYpyoLGKxh2eBV9Ag8E0WlNfcYx6hDwIKAcaHgx6EqAVtPkLZPgfAQxtPf744/87EXQEQx0MQZtKQxm8mD0qwZBocgGW0QgwnPEFs1eohqJ9332GXuZPRWBUYypoMVv+aojWbAAIhtLMHrUveoL/uIb7zHnIBOfxBcOi+GCIO/B+GA5FG2A4hKFHs9EhGEb/8ssvExh07TIN0QKHvwPLjTxNRSkwqkIw50bF7PGqYWoMs7dv31569eqlriX1B0PZZi9fzIaJ1KhRQw37m71tNewZfB+GYmE8hmFNXzDnUcVUDIovhlCTChjOREB5g4egcR7Duhj+BBMEs1EgGPLHdAMsqxEwtPvLL7+oKQvk6wvm/LCUKlVKfUX94Plr2bJlSKtqREI94/mDgR+OYa2MYVgM/WP43fes4ZmAvL5pENxrKi4lq9nIwVc1HRFo1QzjL1PRqTRVhBB/YKiG8poNB/9VDPm+8sor6jueW9O+QNUpjBohJ54RpJtcQD3hWQgMZo9WzHl2MRss6jR+G77fCk5gmNv3rIGdqcTV7wfD+DjGsDqDngSogPWsF22lgvUmXiazZs0Ss8eq5qfMYbEkX1iBhYG7SCRh48aNghcPFAzyNHubao7N9zIPlQbmT33uOoHXoQSggDGPFxh8L+rAc5Ecw00HFtSBChzy4eUdGDCn6wuYnwwMgXnDhQfuM7COxRw45g/hZoP5wuQC5sLBBA0nWG1jjj1UQP7B5Uc8NAIwr51cQF2g7mBhjPlcBNwLRQT3I7O3qNKHEkTAXDJctjBP6QtwD4JSNIepVcPKd97s8SmF6/ue3H+wQ3lQ3whQSJiPRyMEDQDMt/sC6v7VV1/1ffX/h8L3Ndr8J80DX6MI89C+48DrOPbVZTBPfAcjzK2bQ9WqYYr6gVywPYgkhKonnxy+egp8doLTBGPMK8NKH79Pc+RC8UAjBY0EBr0IJHwr6CUbpdGQAJSFObeljHZgAASFgR88egWxDHCJwgsD/qYwOkHP0JxrS9CLCHwRQfnBKAu9PChtfNCzQc8U7lF4MUGpo+fqCzDySmlAL8YcLlayofcPAyT4BMPwx5cv/kMhRWLkhZ4LFBCMrqDY8MKEq9OiRYvCKgD0+nzBtDJWRj8YmUAvtWHDhr5LCf6bc72qR+ozusJF+KKi7ipWrJggbqgvqAsYuMGdyNcTg5EdesNvv/22mPOeKm/UAwyC0IBAw8CcB07w6dGjh1LO5px9qGwiOhdY774b0It++eWXBcZwYIeAMqOM6AH76gYjAODrSwO9dvTUfQH3mkPgYs4V+04l+o+GBUZv0Mv3BRhV4Rzq3LSBUGVGYxCjRaiXwB6wL2/fvYH/ITOencCA7+AMg6vkAhpv5ny9Mgr74IMPxJxnVs88/jPoRyD5pq9+MlMiCwnAAhP+j8EBw51YNAEWnlhYAS9WvFjQmylqWnnC/xeKAT0A9IhTO+yFlxx6mvDTxcsQLXoofHMOUilh9BTwMsVQH3pAsHTFywtKEb1B9GCef/55JS/kwwfKGQoLQ4UYIjbnC4OLmeA7yuPjgWMMoeJlCKWFhoFv4QlYuGLYEL0v9ADh+4qhczROkgvoJUNJYAjVNM5R/NBoQI/J1+PxpYHhWQw9fvPNN+oUFAtkgAKHJS+swcO93Nu2bas4YngUigE9V7ys69evH1Zp+/L1/Yf1tzlXqzjjWYCFLobLkQ7yxYIhCKgT1FOohVswXQFLbVjzIq1YBtMoTil+jEZg2BdMTQM5VeewWIayNY2UBMo6cCjeNGhSzwIaUahHDONCCYcLuIahbvhB4znEcDh62aYxl7LeRq8fDRDT+EkpPzQMEM9nXY28obDxG0JvOTBATlihoywYFYFCx/OMckQSUA/w40bwyYHfKayqGfQjkM58OF7WTyxKZAcBDFPhRYFeYvAHL3/0IjDHCgVtWuIqJYSWOY7hwgGFgVY6ehFQLHgBYQ4OLxIMo6GXjBc2XHUQ4BYDxe4bykXPCYs1mBal6iUPZY7eChZdQLpQMujlYrEFKLti5jwiemGQFYoXL3ykidWt4BIDZYhjzJGhXEjXNwSKHgkUHgJccEIF8IAMSB89VPR+oTTgThM4/4feNXrb6PnAVQflQCPEN+wImSAr4vkC5rTxwsZqSFCKkAvDtpAbQ5a4H8oV7j+QAcPweLli2NXXS8P9CCgf6gD3o0ceKqD84IOGBxQ4hrfBDEx98+K4hjoL1/tDD6+3OaQKuTBcjmFfKHLTyliVDQoJzwPqEHWERlFwwLwz5jPRoMGwNuZiEc83nBwcP9R3MMH94BYYUEYoGrDF84JyoAeKBTTwHJlWz0oZm0ZN6llFQwrc4BKGXjwaEGhYYIQDrFGPeNaRBr6jp4w5VYwG4Lfgawz5hr/xXIAR6h2NLzx7mPqAQvfVPeqsmPks4PlDfmhE4bkCFzyv4IffGHqxkA3x0JjzDfvj94Sh7kC2+E347m9ojoBgdAH3oiGExgKO0QBl0I9AGvNB/t94ln7yUSISIIFkCKAnhRc2eud2BihTNLQw0uCEgMYbeptQtL7RDCfITRndQ4BD0O6pS5bEYwTQE8ZQLno46H3bHXyWzHbLwfxJwCkEaITllJqinCQQRAC9NgyjwtgGQ6MMKSOAOVIMN8NegIEE7CDAIWg7qDNPEiABEiABzxNgD9jzjwABkAAJkAAJ2EGACtgO6syTBEiABEjA8wSogD3/CBAACZAACZCAHQSogO2gzjxJgARIgAQ8T4AK2POPAAGQAAmQAAnYQYAK2A7qzJMESIAESMDzBKiAPf8IEAAJkAAJkIAdBKiA7aDOPEmABEiABDxPgArY848AAZAACZAACdhBgArYDurMkwRIgARIwPMEqIA9/wgQAAmQAAmQgB0EqIDtoM48SYAESIAEPE+ACtjzjwABkAAJkAAJ2EGACtgO6syTBEiABEjA8wSogD3/CBAACZAACZCAHQSogO2gzjxJgARIgAQ8T4AK2POPAAGQAAmQAAnYQYAK2A7qzJMESIAESMDzBKiAPf8IEAAJkAAJkIAdBNLbkaldeY4dO1YuXbpkV/bMlwRIgARIQDMC+fPnl8aNG9siVRrDDLbkHOdMx40bJ++//7706tUrzjkzOxIgARLQh8DOnTtl+/bt0qJFC32EslGSQYMGyU8//SSVK1eOuxSe6QGj59uzZ0+5//774w6ZGZIACZCALgQWLlwo8+bN47vw/ytk27ZtcuXKFVuqh3PAtmBnpiRAAiRAAl4nQAXs9SeA5ScBEvAUgeuuu05q167tqTLrWljPDEHrWgGUiwRIgATiSeD6668XfBjsJ8AesP11QAlIgARIgAQ8SIAK2IOVziKTAAmQAAnYT4AK2P46oAQkQAIkEDcCf/31lyxfvjxu+TGj8ASogMOz4RUSIAEScB2BP/74Q6ZPn+66cjmxQFTATqw1ykwCJEACJOB4AlTAjq9CFoAESIAESMCJBKiAnVhrlJkESIAEoiRAP+AowVlwG/2ALYDKJEnAR+DUqVOyZ88eOXDggBw+fFiOHz8u586dEyzBniVLFsmXL5/yySxbtqxkypTJdxv/k4BlBOgHbBnaFCdMBZxiZLyBBMITOHr0qLIwXb16tWzatElOnjwpxYoVk8KFCytlmytXLsmcObOkSZNGzpw5I3/++acsWLBAduzYIVWrVpUOHTrYsih8+BLxCgmQgFUEqICtIst0PUVg2bJlMn78eNm6data5q9GjRpy9913S6FChSLigF4xFsj/6KOPBNuj9e/fX6655pqI7mUkEiABZxKgAnZmvVFqTQhs3rxZPv30UzWk3LVrVxk4cKCkS5cuxdKhV9y6dWtp1aqVTJgwQfr27SvPPvus3HTTTSlOizeQQFIE4Ae8b98+qVWrVlLReC0OBKiA4wCZWbiPwOXLl+WHH35Q/pT9+vWTBg0axKSQGJrGMDSGratVqyaLFi2Sm2++OSZpMxESAAH4AWO0hQrY/ueBCtj+OqAEDiOAudsXX3xRMmTIIN9//71kz5495iXAfPCWLVvk5ZdflgsXLkiTJk1ingcTJAESsJcA3ZDs5c/cHUYA1suPPfaYFC1aVN566y1LlK8PSZkyZeSdd95RQ9zYNJyBBEjAXQSogN1VnyyNhQTQE+3WrZukTZtWHn74YQtz+l/S8Nl84okn5JVXXpGzZ8/+7wKPSCBKAvQDjhKcBbdRAVsAlUm6kwB6vBUqVJAvv/wyrgWsW7euclH6+uuv45ovM3MnAfgBt2zZ0p2Fc1ipqIAdVmEU1x4Co0ePln/++UdZJtshwYMPPii//PKL/P7773ZkzzxJgAQsIEAFbAFUJukuAlB6I0aMUMPA0bgYxYJG1qxZ5aGHHpJPPvkkFskxDRIgAQ0IUAFrUAkUQV8C58+flzfffFOefPJJyZs3r62CNmrUSOW/cOFCW+Vg5s4mwP2A9ak/KmB96oKSaEjgm2++kYoVK0qdOnW0kO6ee+5Rrk+wxmYggWgIcD/gaKhZcw8VsDVcmaoLCGBZSfQ2MfSrS8DiHDly5FALdOgiE+UgARKIjgAVcHTceJfLCVy5ckXeffddgfFTtmzZtCptly5dZNSoUVrJRGFIgARSToAKOOXMeIcHCEyZMkWwc1Hjxo21K229evUE2xxityUGEkgpAfoBp5SYdfGpgK1jy5QdSuD06dNqned4LbYRDaZ27drJxIkTo7mV93icAP2A9XkAqID1qQtKogkBbLKADRCw3KSuATsnrVy5Uk6cOKGriJSLBEggGQJUwMkA4mVvETh48KDMnj1bYG2sc4BfcO3atWXWrFk6i0nZSIAEkiBABZwEHF7yHoFvv/1WOnXqpCyNdS99mzZt1HaIustJ+fQiQD9gfeqDClifuqAkNhOAf+SaNWuUArZZlIiyh38yNmjYvn17RPEZiQRAgH7A+jwHVMD61AUlsZnA4MGDpVevXpIpUyabJYk8++bNm3MYOnJcjEkCWhGgAtaqOiiMXQQ2bNgge/fuFQzrOilAAc+fP99JIlNWEiCB/ydABcxHgQRMAlhysnfv3mLXZgvRVkKhQoUkd+7csnbt2miT4H0eI0A/YH0qnApYn7qgJDYRWL16tZw8eVKaNWtmkwSpy7ZJkyayYMGC1CXCuz1DgH7A+lQ1FbA+dUFJbCIAy2f0fp0aGjRooNas5gYNTq1Byu1VAlTAXq15llsRWL58uVy6dEmgxJwaChYsqLZKxDw2AwmQgHMIUAE7p64oqQUEvv/+e7n77rstSDm+STZs2FAWL14c30yZmyMJ0A9Yn2qjAtanLihJnAlAYWHYFitKOT2gDEuWLHF6MSh/HAjQDzgOkCPMggo4QlCM5j4CWPNZ9yUnI6UOwxpsobhr165Ib2E8EiABmwlQAdtcAczeHgILFy6U9OnTS82aNe0RwIJcsYHEsmXLLEiZSZIACVhBgArYCqpMU3sCQ4cOdU3v1wcbjQkqYB8N/g9HgH7A4cjE/3z6+GfJHEnAXgJYOSpz5sxSvXp1ewWJce433nijGoI+fvy45MyZM8apMzm3EMB0BT4M9hNgD9j+OqAEcSaA3q8bLJ+DsWFIvVq1avLrr78GX+J3EiABDQlQAWtYKRTJOgLo/ebIkUMpKutysS/lGjVqCHybGUiABPQnQAWsfx1RwhgRgMsR/H579uwZoxT1SwbzwCtXrlTuVfpJR4l0IEA/YB1q4b8yUAHrUxeUxGIC6P1i4wIM07o15M2bV62KxT2C3VrDqS8X/YBTzzBWKVABx4ok09GaAHq/Q4YMcXXv11cBMC5bsWKF7yv/kwAJaEqACljTiqFYsSXg6/1WrVo1tglrmBrmgTEMzUACJKA3ASpgveuH0sWIAFa9cvKORynBAHeknTt3ytmzZ1NyG+N6hAD9gPWpaCpgfeqCklhEYN68ecryuXLlyhbloFeycEcqX768YJ9jBhIIJsD9gIOJ2Pddm4U4tm7dKj/99JP8/vvvcubMGcmaNauUKVNGunXrJhUqVLCPEHN2PIFhw4bJww8/7PhypKQAmAeGP3C9evVSchvjkgAJxJGAFj3giRMnCrZTO3HihDRu3Fg6d+6s/kMRN2/eXEaOHBlHJMzKTQSw5nOWLFnEC3O/gfWGeWD2gAOJ8JgE9COgRQ/4jTfekOnTp0uoIcIHHnhAOnbsqHrC+uGjRLoTwKhKnz59dBcz5vIVK1ZMTp06JX///bcUKFAg5ukzQecSgB/wvn37pFatWs4thEsk16IHfO7cOSlYsGBIpHny5JHz58+HvMaTJJAUgaVLl6rL6A16MfiGob1YdpY5PAH6AYdnE+8rWvSAsSdrixYt1Dxd2bJl1fwvlDIWExg0aJD06tUr3lyYnwsIoPfr5lWvkquim266SfkDt2nTJrmovE4CJGADAS0U8GOPPSZFixZVCyW89957cvr0acmWLZsywnrqqaeka9euNqBhlk4msGbNGuWG42UjJMx7f/bZZ06uRspOAq4moIUCBuHbbrtNfVxNm4WLG4Eff/zR83YDWJYyV65cgiHHkiVLxo09M9KbAP2A9akfbRRwOCSwYl23bp088sgj4aL4z2/atEm2bdvm/x54AF/Q7NmzB57isUsJ4BmAoUmzZs1cWsLIi4V54N9++40KOHJkro/J/YD1qWLtFTDmgk+ePBkRMWxEDuu+UAF+xnBHYXA/Acz9YtoiTZo07i9sMiWsUqWKTJ48Wbp06ZJMTF4mARKINwHtFTCMs/CJJNSpU0fwCRV27dolBw8eDHWJ51xEYP/+/bJhwwZ57rnnXFSq6IsCBfzmm2/KpUuXBCtkMZAACehDQAs3JOxU89133wmsoT///HO5ePGin9DmzZvltdde83/nAQkkRQCLtrRr104yZsyYVDTPXIMxY/HixdUKc54pNAuaJAHuB5wknrhe1EIBv/vuu/LRRx8J5iawbCDm7rCIAAJ6rbNnz44rFGbmTAJYSQ02Ax06dHBmASySGtbQq1atsih1Jus0AvQD1qfGtFDAw4cPl59//lkNGy5ZskRKlSqlVr/CsBnn8fR5WHSXZPz48WpJ0xw5cugualzlgwKGIRYDCZCAXgS0UMAYcsZQGULatGnlq6++ksyZM0vfvn0Fw9MMJJAcATTWpkyZQp/xEKCwmcmOHTu4PWEINjxFAnYS0EIBY6UerHblGyaDEoYlKxZTePrpp+3kw7wdQmDmzJlq4ZZChQo5ROL4iZkhQwapWLGirF27Nn6ZMidtCdAPWJ+q0UIBv/7664JNxCdMmOAnA5/duXPnSunSpWm96afCg3AExo0bR1ebcHDM89johLsjJQHIQ5e4H7A+la2FXwIsVrEjUnDAXB5WNLpy5UrwJX4nAT8BKBbYCoTaTcsfyeMH1apVk4EDB3qcAotPAnoR0KIHnBwSDEkzkEA4AjC+wh7SDOEJlClTRo4cOSLHjh0LH4lXSIAE4kqAmi2uuJlZrAlg4Q0sQdq4ceNYJ+269DBC4LOzcF3hWKCICdAPOGJUlkekArYcMTOwksDEiRPllltu4cIbEUDGMDTngSMA5fIo9APWp4KpgPWpC0qSQgJYJ3zGjBlq5asU3urJ6NiYgZbQnqx6FlpTAlTAmlYMxUqeAHa4wtrf+fLlSz4yY4jPRWvv3r2kQQIkoAEBKmANKoEiREdgzJgx0rJly+hu9uhdHIb2aMUHFJt+wAEwbD6kAra5Aph9dASw4xECXY9Sxu+mm27iPHDKkLkuNv2A9alSKmB96oKSpIAAFm3BrkcMKSOAdaGxwtzly5dTdiNjkwAJxJwAFXDMkTJBqwkcP35cfv31V2nVqpXVWbkufSxuc+211wq2+WQgARKwlwAVsL38mXsUBLDpQsOGDdWGHVHc7vlbatSowd2RPPwU0A9Yn8qnAtanLihJhASwdeWtt94aYWxGCyYAQ6wVK1YEn+Z3jxCgH7A+FU0FrE9dUJIICCxfvlxy584tMCRhiI4Atifcs2ePnDp1KroEeBcJkEBMCFABxwQjE4kXAfR+27dvH6/sXJlP+vTppVKlSrSGdmXtslBOIkAF7KTa8ris//77r6xfv14aNWrkcRKpL37NmjWVIVvqU2IKTiNAP2B9akyL7Qj1wUFJdCYwbdo0ad68Odd9jkElYVnK4cOHxyAlJuE0Api+4RSOHrXGHrAe9UApkiGAPaGnTp0qbdq0SSYmL0dCAK5IGIrevXt3JNEZhwRIwAICVMAWQGWSsSeAbfTy588vxYsXj33iHk0Rw9DcntCjlc9ia0GACliLaqAQyRGA7y97v8lRStl1+APDqpzBWwToB6xPfVMB61MXlCQMgaNHjyrjqyZNmoSJwdPREMCylFgR6+zZs9HcznscSoB+wPpUHBWwPnVBScIQgPEVVr7KmDFjmBg8HQ2BTJkyScWKFblHcDTweA8JxIAAFXAMIDIJawlAAXP42RrGmAdesmSJNYkzVRIggSQJUAEniYcX7Sawdu1ayZYtm5QpU8ZuUVyZf+3atWXZsmWuLBsLFZoA/YBDc7HjLP2A7aDOPCMmANcj7noUMa4URyxUqJBgh6Tt27dLqVKlUnw/b3AeAfoB61Nn7AHrUxeUJIjA6dOnlZUuFt9gsI4ArKGXLl1qXQZMmQRIICQBKuCQWHhSBwJz5syRm266SQ1B6yCPW2WoW7cu3ZHcWrksl9YEqIC1rh5vCzd9+nQaX8XhEcDuSPv375dDhw7FITdmYTcB+gHbXQP/y58K+H8seKQRgZ07dwr8f+GrymAtgbRp00qtWrU4DG0tZm1Spx+wNlUhVMD61AUlCSAwY8YMad26taRJkybgLA+tIoBhaLojWUWX6ZJAaAJUwKG58KyNBC5fviyzZs2SFi1a2CiFt7KGIdaGDRvkzJkz3io4S0sCNhKgArYRPrMOTQAWudh0oWDBgqEj8GzMCWTOnFmqVKlCn+CYk9UvQfoB61MnVMD61AUl+X8CM2fOpO+vDU9D/fr1ZfHixTbkzCzjSQB+wC1btoxnlswrDAEq4DBgeNoeAjC8wupXDRo0sEcAD+dap04dWblypVy4cMHDFFh0EogfASrg+LFmThEQmDt3rtp4ARsFMMSXwFVXXSXlypWTFStWxDdj5kYCHiVABezRite12Nh4gStf2Vc7jRo1kkWLFtknAHO2nAD9gC1HHHEGVMARo2JEqwnAP/H8+fNSqVIlq7Ni+mEI1KtXTxlicRg6DCAXnKYfsD6VSAWsT114XhL4/jZr1szzHOwEkDNnToFP8OrVq+0Ug3mTgCcIUAF7opr1LyR8f+fNm0frZw2qCiMQ8MNmIAESsJYAFbC1fJl6hASWL18u8E8sUKBAhHcwmlUEYIG+atUqOXv2rFVZMF0bCdAP2Eb4QVlTAQcB4Vd7CMD3lytf2cM+ONfs2bNL5cqV6RMcDMYl3+kHrE9FUgHrUxeeleTUqVOyZs0aady4sWcZ6FbwJk2ayOzZs3UTi/KQgKsIUAG7qjqdWRj4/tasWVOwHCKDHgRgDb1161Y5duyYHgJRChJwIQEqYBdWqtOKNGfOHBpfaVZpGTJkEChhGmNpVjExEId+wDGAGKMkqIBjBJLJREfgzz//VJvBYyMABr0IYE4ejSMGdxGgH7A+9UkFrE9deFISvOCx8hU2hWfQiwDckTA/v337dr0EozQk4BICfOu5pCKdWgwuvqF3zbVq1UqwPCgDCZBA7AlQAceeKVOMkMD69eslW7ZsArcIBj0JYNu6+fPny8WLF/UUkFKlmAD9gFOMzLIbqIAtQ8uEkyMANxf6/iZHyd7r+fLlk7Jly8qCBQvsFYS5x4wA/YBjhjLVCVEBpxohE4iGAHpUeKnD35RBbwJt27aVqVOn6i0kpSMBBxKgAnZgpblBZCw9ib1n8+TJ44biuLoMtWvXln379snevXtdXU4WjgTiTYAKON7EmZ8igKUnufKVMx6GdOnSCXrBkydPdobAlDJJAvQDThJPXC9SAccVNzMDAbi2rFu3TrDoP4MzCEABY1GOc+fOOUNgShmWAP2Aw6KJ+wUtFTAWZxg2bJiMHz9evazjToUZWkoA2w7WqFFDsmTJYmk+TDx2BGCMhcVSsGwoAwmQQGwIaKGA27Rp4/c1hGUsrC4///xzeeutt9Tx5s2bY1NapqIFASy+0bRpUy1koRCRE7j11ltl7Nixkd/AmCRAAkkS0EIBX7lyRQzDUII++eSTMnjwYFm2bJnak3TgwIEyYMCAJAvBi84h8O+//yqDnlq1ajlHaEqqCKAHnCZNGlm7di2JOJgA/YD1qTwtFHAgjgMHDkjHjh39p+644w61K4v/BA8cTQAjHPXr11cvckcXxKPC47c5ZswYj5beHcWmH7A+9aiNAh4xYoSa8w2eZxo3bpwUKVJEH2KUJFUEsKoSh59ThdDWm7FwypYtW9QGGrYKwsxJwAUEtFDATz/9tFqOcMiQIYL53tdee02hfeONN+Tee+8VDEMzOJ/A7t275fjx41KxYkXnF8ajJUifPr1ySRo9erRHCbDYJBA7Auljl1T0KcEdJdAlxefqcPvtt8tDDz0kuXLlij5x3qkNAVjQcuUrbaojakE6dOggd955p/Tp00eyZ88edTq80R4C8APGwiq0w7CHf2CuWvSAAwXCcebMmdUpzFXAOIsr8AQTcuZ3uB9x+NmZdRcodc6cOdU8/oQJEwJP89ghBOgHrE9FadEDTgrHDz/8oOaGFy1alFQ0de3rr7+W4cOHh4yHh6548eIhr/Gk9QQwb4gVlUqUKGF9ZszBcgIYnXrssccE/zEszUACJJByAtr/cvr16ycPPvhgRCW77777BJ9QoX///nLw4MFQl3guDgRofBUHyHHM4tprr5XSpUsL9nPGKlkMJEACKSeg5RB0YDHQus6UKVPgKR47kADmfxs1auRAySlyOALo/Y4aNSrcZZ7XlAD9gPWpGO0VsD6oKEm0BDZs2KAM6fDDZ3APAVizYz54yZIl7imUB0pCP2B9KlmLIeiHH35Y/v7777BUsG3dK6+8EvY6L+hNAPv+cucjvesoWum6desmP/74o9StWzfaJHgfCXiWgBYKGGvMwrWhd+/eUqlSpUSVUbBgwUTneMIZBGDFDgU8aNAgZwhMKVNEoF69egLjx/Xr14f87aYoMUYmAY8R0EIBwzf0p59+khdeeEE+/PBDSZuWI+NueQ43btwoefPmlcKFC7ulSCxHEIGuXbsq74NQjeegqPyqAQH6AWtQCf8vgjaaDpaUWPWKlsr6PByxkAS935tvvjkWSTENTQm0bNlSduzYIVjpjEF/AvQD1qeOtFHAQNK3b18pVKiQPnQoSaoJLFy4UBo2bJjqdJiAvgTg333bbbfJyJEj9RWSkpGAhgS0UsAa8qFIqSCwadMmufrqqwU+owzuJgA7jqVLl8rhw4fdXVCWjgRiSIAKOIYwmVRCAli9jNaxCZm49VvWrFkFOyWNHTvWrUV0TbnoB6xPVVIB61MXrpMEw8+Bm2y4roAsUAICXbp0kenTp8vZs2cTnOcXvQjQD1if+qAC1qcuXCUJDD2wghnX33ZVtSZZmHz58knVqlVl2rRpScbjRRIggf8SoALmk2AJAQ4/W4JV+0TRC+YuSdpXEwXUhAAVsCYV4TYxsDxh/fr13VYslicZAmXLllXLUy5evDiZmLxsFwH4AS9fvtyu7JlvAAEq4AAYPIwNAfzAjx8/LngZM3iPQOfOndUWot4ruTNKTD9gfeqJClifunCNJOj90PrZNdWZ4oJg4ZW9e/fKrl27UnwvbyABLxGgAvZSbceprBx+jhNoTbPBwhzwC544caKmElIsEtCDABWwHvXgGimOHTumej433nija8rEgqScAJaWxR7Qp0+fTvnNvMNSAvQDthRvihKnAk4RLkZOjgBWQ7rpppskfXot9vlITlxet4hArly5pHbt2jJr1iyLcmCy0RKgH3C05GJ/HxVw7Jl6OkUoYGxRx0ACt9xyi0yaNIkgSIAEwhCgAg4DhqdTTuDChQvy22+/SY0aNVJ+M+9wHQFsT4itRTds2OC6srFAJBALAlTAsaDINBSB1atXS5kyZSR79uwkQgKKQJs2bWTy5MmkoREB+gHrUxlUwPrUheMlWbFiBYefHV+LsS0A9grGog+nTp2KbcJMLWoC9AOOGl3MbwyrgC9fvhzzzJiguwn88ssvUqtWLXcXkqVLEYFs2bKpZ2L27Nkpuo+RScALBMIqYOxkU7lyZfnwww/ln3/+8QILljEVBNCqxtBz4cKFU5EKb3UjAQxDY5ckBhIggYQEwirgxo0by2effSZbtmyRChUqSPv27dUi6zC0YSCBYAKwfmbvN5gKv4MAGvInT54UNNIY7CdAP2D768AnQVgFjAhYTvCrr76Sffv2yZ133inPP/+8FCpUSB599FH+mHwE+V8RWLZsmfL7JA4SCEWgdevWMmPGjFCXeC7OBOgHHGfgSWSXpALGfRs3bpSXXnpJHn/8ccmaNau8+uqrao9X7HQDoxsGEsDqV7CshNsJAwmEItCsWTOZM2eO0LYkFB2e8yqBsMsVbd26Vbp16yb79+9XvV9sso2haF/YuXOnwOimZs2avlP871EC6P1Wr15d+Xx6FAGLnQyBggULSrFixZRFNDfqSAYWL3uGQFgFDAf6l19+WWBAEbisINZ2hWXjiy++KDly5PAMKBY0PAEoYOyAw0ACSRGASxKWpqQCToqS9dcwWoVpRdpsWM86uRzCDkGjgmBYE6h8MXyUL18+gSFW3rx5JWPGjMmlz+suJ4BnYt26dfwxu7yeY1E8TFv9+uuv9AmOBcxUpEE/4FTAi/GtiXrAsHpu0qSJnD17Vs6fPy9Dhw71Z3nu3DkpWrQoFa+fCA/Wrl0rRYoUkauuuoowSCBJArAhQa9r0aJFAqMsBhLwOoFECrhs2bKq54thRfSAH3nkET8j7PNJP08/Dh6YBPCccCiLj0KkBODeOGbMGCrgSIExnqsJJBqCRs+3QIEC0rlzZ3n99dclf/78/k+ePHkEvWAGEvARWLlypdSpU8f3lf9JIEkCMNqEAeehQ4eSjMeL1hGgH7B1bFOacqIe8IMPPqhWNLrtttukadOmidKDARbXdU2ExZMnYCeABlnx4sU9WX4WOuUEYFPSsGFDWbBggXTq1CnlCfCOVBOAHzA+DPYTSKSAhwwZ4pfKMAz/MQ9IIJgAFtnn8HMwFX5PjgAU8Ndff00FnBwoXnc9gURD0IElXrVqlaxfv14ZZN1///3SsWNH2bRpU2AUHnuYABZiqV27tocJsOjREKhataoagj5w4EA0t/MeEnANgbAKePfu3cpQAhsxoLUKZQz/vS5durim8CxI9ATOnDkjmzdvFrxMGUggpQTgkjR37tyU3sb4MSDA/YBjADFGSYRVwJijue+++9Q88KRJk6Rfv34yYMAAyZw5Mw0oYgTfycmgQYalJzNlyuTkYlB2mwg0atRI5s+fb1Pu3s6WfsD61H9YBYxFNmBgg3V+lyxZIljLFQtwoEcMfz4GbxPg8LO36z+1pa9YsaLaIQm9MQYS8CqBsAoYy8ZNmDBBqlSpIjjGLkg1atRQFoxUwF59XP5bbhjnwUec87/efg5SW3oMQ8+bNy+1yfB+EnAsgbAKOHfu3IJVjkaMGCHjxo0TLMIxcOBA+eGHHxxbWAoeGwK///67wCccPuIMJBAtAQxDY0MXhvgSoB9wfHknlVsiN6TAyMePH5cdO3YIXri+8NNPP0mPHj18X/nfgwTgfsQF9T1Y8TEucvny5eXw4cNqK0uusBdjuEkkRz/gJODE+VJYBYxFFrD9YLVq1QS9YV+A0Q0VsI+GN/8vXrxYnnjiCW8WnqWOKYEGDRqoRTnuuOOOmKbLxEjACQTCKmBsng2/3++++84J5aCMcSJw8OBBOXr0qNxwww1xypHZuJkAFPDnn38uVMBurmWWLRyBsHPAMLpKkyZNuPt43qMEuPevRyveomLDlQ2Nur///tuiHJhsMAH6AQcTse972B5w9erV5amnnlJDjejtpE37X10NYywOQdtXYXbnDKMZbNTBQAKxIIBGPqyh8VxxbehYEE0+DfgBw/qcy8gmz8rqGGEV8P79+6VkyZKya9cu9fEJwjlgHwnv/cfqV9u2bVN2Ad4rPUtsFYGbb75Zhg0bRgVsFWCmqy2BsAoYjvJwP0K4ePGiZMiQQdtCULD4EMCCLJUrVxYs0sJAArEiAEPP1157TY4cOZLA4DNW6TMdEtCVQNg5YAg8Y8YMgasAXrh79uxRi3KsXr1a17JQLosJwPoZvRUGEoglAUxrYU9pLO7CYD0B+gFbzzjSHMIqYChc7IA0aNAg/4pHDz30kHTv3j3StBnPRQQwCvLbb7+pF6WLisWiaEIAfuWLFi3SRBp3iwE/YKxuyGA/gbAKGJP0vXr1kiZNmqgeMIwl+vTpI9myZZN///3XfskpQVwJrFy5UkqXLi1XXXVVXPNlZt4gULNmTdm4caOcPn3aGwVmKUnAJBBWAefKlSuB8RVonTp1SrkL8CXsvWcHvRNYqzKQgBUEYNyJuWCsssZAAl4hEFYBt27dWq0FjSFnrIqFPYHr1asnHTp0UFsSegUQyyly5coV9WLEogkMJGAVAQ5DW0U2Ybr0A07Iw85vYRUwWqQwuoERFixfsSb0I488Ih9//LGd8jJvGwj8+uuvUqRIEcGoCAMJWEUADXwYecLegME6AtwP2Dq2KU05rBsSEsJewNgHGL1ezP/BWpHBewQWLlwo2LmGgQSsJJA9e3b1nlm1ahWN/awEzbS1IRCyBwwDLKx+VaxYMfXixaYMBQsWlPfee08bwSlIfAhcvnxZrVLE+d/48PZ6LnBz4xaFXn8KvFP+RAp48+bN0q5dO7n33nsF2xHCKhErIA0dOlQtms4haO88HCgprJ+LFy8uefPm9VbBWVpbCEABwxDLMAxb8vdCpvQD1qeWEylg7Pd71113Sf/+/SVHjhxKUswHt2rVSr755hv59ttv9ZGeklhOAMPPjRs3tjwfZkACIICGXv78+WXdunUEYhEB+gFbBDaKZBMpYKwBjWUoQwUYY2GBDgZvELhw4YIyxGvYsKE3CsxSakEAvWAYgDKQgNsJJFLAcDlJnz60bRaWpMScoBVh4sSJ0r59eylTpoxgiAT/b7nlFhk1apQV2THNCAhgaUDM/+fMmTOC2IxCArEhAAWMdccZSMDtBEJq2i1btsj8+fMTlR1zwVaEjz76SAYPHiwDBgxQVpBZs2aVs2fPyvbt2+Xtt99WLlDPPvusFVkzzSQIzJw5U62ElkQUXiKBmBMoWrSo6gTg91+qVKmYp+/1BOEHjLUduB2h/U9CIgWcJ08e1esM1/NE7zTWAcoXGz8ULlw4QdJYoL1t27ZqLWoq4ARoLP8CA7wNGzbIyy+/bHlezIAEggn4rKGpgIPJpP479wNOPcNYpZBoCPqDDz5Q87yY6w31Qe841iFz5sxy9OjRkMliizIYgTHEl8DcuXPVzkdkH1/uzO2/BLDqGueB+TS4nUCiHrAdBX7uuefUgh+dOnWSsmXLCoagz507J2ipjRgxQt5//307xPJ0ntOmTZN+/fp5mgELbx8B2ICcPHlS/vzzT7n22mvtE4Q5k4CFBLRQwLfeequUK1dO4AKFRf/he4xdl/AjnDVrljIEspABkw4igIYP5uBh9c5AAnYRwOIv6AV369bNLhFcmS/9gPWpVi0UMHBA2b766qv6kPGwJD///DP3C/Vw/etSdKwN/dVXX1EBx7hC4AeMD4P9BLRRwOFQjB07VvWKBw0aFC6K//ycOXPCzhstW7bMv7CI/wYeJCKAhfBhAf/dd98lusYTJBBPAjfeeKMcOHBA7T+eL1++eGbNvEggLgS0V8BoqUXqe1yiRAnJkCFDSHCbNm0SLCzBkDQBNGLg+wtreAYSsJNA2rRplSEg1obGhjAMJOA2Ator4KpVqwo+kQSsWYxPqICFPg4ePBjqEs8FEJg6dar06NEj4AwPScA+AnBHgm0IFXDs6oB+wLFjmdqUErkhpTZB3u9cAlj44NChQ1KjRg3nFoKSu4pAtWrVZNeuXQJ3RIbYEOB+wLHhGItUtOgBY/cTWN2GC9gIvkqVKuEu83yMCEyaNEktB5omTZoYpchkSCB1BLAsLhbkgTU0dmljIAE3EdBCAWOXJRj9YO4x1MIPUL6Iw2AdgRMnTsiCBQtk+PDh1mXClEkgCgJYlGPMmDFUwFGw4y16E9BCAWMpSvS64PvL/YbteWCw8AZedL4tKO2RgrmSQGICmBIZOHCgHDt2TDAaxpA6AvQDTh2/WN6tzRzwJ598otyNsPINQ3wJwMp83LhxNHSJL3bmFiEBDEPXrl075AYxESbBaAEEuB9wAAybD7VRwFmyZJE1a9Zw2TkbHgi4eWC5P7hxMZCAjgSwJzWmSBhIwE0EtFHAboLqtLKMHj1asA43AwnoSgDD0NgchtbQutYQ5YqGABVwNNRcdA8WKDl16pTUrVvXRaViUdxGIF26dOoZnTdvntuKFvfywA8YnicM9hOgAra/DmyVAPs+d+zY0VYZmDkJREKgSZMmnAeOBFQycegHnAygOF6mAo4jbN2ywjq7GzdulFatWukmGuUhgUQE4I6IZxY9OAYScAMBKmA31GKUZUDvt23btpIxY8YoU+BtJBA/AnBVRC949uzZ8cuUOZGAhQSogC2Eq3PS2Owc82kcfta5lihbMIHmzZsLNgxhiJ4A/YCjZxfrO6mAY03UIemNHz9eLbyRM2dOh0hMMUlApFSpUmrHM0ydMERHgH7A0XGz4i4qYCuoap4m9vzFus/dunXTXFKKRwKJCbRo0UJmzZqV+ALPkIDDCFABO6zCYiHujBkzpFy5clK4cOFYJMc0SCCuBJo1a6YW5eD+3nHFzswsIEAFbAFU3ZPEwvbdu3fXXUzKRwIhCeTJk0fKly+vdkgKGYEnkyRAP+Ak8cT1IhVwXHHbn9mSJUske/bs6gVmvzSUgASiIwDXuSlTpkR3s8fvoh+wPg8AFbA+dREXScaOHStdunSJS17MhASsIoCV27A0JfyCGUjAqQSogJ1ac1HIvX37drWIQf369aO4m7eQgD4EsDRly5YtlTGhPlJREhJIGQEq4JTxcnRsbDnYuXNnSZuW1e7oiqTwikC7du2UNTS202SInAD9gCNnZXVMvomtJqxJ+thFBvO/rVu31kQiikECqSNQsGBBKVmypFpQJnUpeetu+gHrU99UwPrUhaWSTJ48WeA/mS1bNkvzYeIkEE8Ct912G4eh4wmcecWUABVwTHHqmdilS5eUxSheVgwk4CYCtWrVkkOHDsnWrVvdVCyWxSMEqIA9UNHz58+XEiVKcOEND9S114qIDRo6deoksG9giIwA/YAj4xSPWFTA8aBscx5Y97lDhw42S8HsScAaAvAJxgbz6AkzJE+AfsDJM4pXDCrgeJG2KR/82I4fPy4YqmMgATcSgF0DlDAamgwk4CQCVMBOqq0oZMXQHNw1GEjAzQSwrebUqVPl9OnTbi4my+YyAlTALqvQwOLgZQTXo7Zt2wae5jEJuI5A/vz5pU6dOgJrf4akCdAPOGk+8bxKBRxP2nHOa/r06VKzZk219nOcs2Z2JBB3ArfffruMHj1azp8/H/e8nZQh/YD1qS0qYH3qIuaS/Pzzz3LrrbfGPF0mSAI6EihSpIhUqVKFmzToWDmUKSQBKuCQWJx/cv369QIXDWzbxkACXiHQo0cPGT58uHCvYK/UuLPLSQXs7PoLKz16vzS+CouHF1xKoHjx4lK5cmXOBSdRv/QDTgJOnC9RAccZeDyyO3XqlCxdulSaNWsWj+yYBwloReCuu+6SESNGcC44TK3QDzgMGBtOUwHbAN3qLGfPni3YLzV79uxWZ8X0SUA7ArDyrV69uowcOVI72SgQCQQSoAIOpOGSYww/Y2ECBhLwKoG7775bJkyYICdOnPAqApbbAQSogB1QSSkRcdu2bcoABfNgDCTgVQIFChSQpk2bytChQ72KIGy56QccFk3cL1ABxx25tRmi99uyZUtrM2HqJOAAArCInjt3rhw8eNAB0sZPRPoBx491cjlRASdHyEHX4XqxcOFCDj87qM4oqnUEcubMKZ07d5ZvvvnGukyYMgmkggAVcCrg6XYrlO8NN9wguXPn1k00ykMCthDAVoUbNmyQzZs325I/MyWBpAhQASdFx2HXZsyYweFnh9UZxbWWQMaMGeX++++Xzz77zNqMHJQ6/YD1qSwqYH3qIlWS/PPPP7Jz506pV69eqtLhzSTgNgJNmjSRK1euyPz5891WtKjKQz/gqACvpS4AACd0SURBVLBZchMVsCVY458oNl5o3LixpEuXLv6ZM0cS0JzAQw89JF988QWXqNS8nrwmHhWwS2p81qxZNL5ySV2yGLEnUKFCBalYsaLaLSn2qTNFEoiOABVwdNy0umvdunWSOXNmKVmypFZyURgS0IkA5oLHjBkjR44c0UmsuMtCP+C4Iw+bIRVwWDTOuTBz5kxp3bq1cwSmpCRgA4H8+fPLLbfcIl9//bUNueuTJf2A9akLKmB96iIqSc6dOyeLFi0SGJowkAAJJE0Ai3OsXr1aYIjEQAJ2E6ACtrsGUpn/L7/8orZfy5UrVypT4u0k4H4CmTJlEqwTTbck99e1E0pIBeyEWkpCRlg/N2/ePIkYvEQCJBBIAEu1njx5UtB49WKgH7A+tU4FrE9dpFgS+P7u2LFD6tSpk+J7eQMJeJVAmjRppF+/fmou+PLly57DQD9gfaqcClifukixJNj3F3O/6dOnT/G9vIEEvEwAu4XBGnjSpElexsCy20yACtjmCkhN9tOmTZMWLVqkJgneSwKeJQC3pGHDhsmpU6c8y4AFt5cAFbC9/KPOfdOmTQKDkjJlykSdBm8kAS8TKFq0qDRo0ECGDx/uKQz0A9anuqmA9amLFEmCjReaNWuWonsYmQRIICGBXr16CUaS/v3334QXXPyNfsD6VC4VsD51EbEkFy9eVL6/HH6OGBkjkkBIAldffbW0b99eBg8eHPI6T5KAlQSogK2ka1HacJ/A0DP3/bUIMJP1FIHbb79dVq1aJbt37/ZUuVlY+wlQAdtfBymWgPv+phgZbyCBsASwjvodd9whX331Vdg4brpAP2B9apMKWJ+6iEiSw4cPy7Zt26R+/foRxWckEiCB5AlgGHrPnj2ycePG5CM7PAb9gPWpQCpgfeoiIkmw8QKUL31/I8LFSCQQEQHso40lKr/55puI4jMSCcSCgBYKGC1PLA3nC/PmzZMnnnhCXnrpJdm6davvNP+bBLjvLx8DErCGABa1OXHihKxYscKaDJgqCQQR0EIBP/DAA8qqF7J9++230qlTJ+Uc/+eff0q9evVkzpw5QWJ78+vmzZvlypUrUq5cOW8CYKlJwEICWKKyT58+8t1331mYi/1J0w/Y/jrwSaCFAvYJg/8DBw6UqVOnypdffqlcA7BU3AsvvBAYxbPHGH5u1aqVZ8vPgpOA1QTq1q0rGI5euHCh1VnZlj79gG1Dnyhj7RTw+fPnpXr16n5Ba9WqJQcPHvR/9+rBhQsXZP78+dz5yKsPAMsdNwL33HOPavxjtImBBKwkoI0CfvPNN9WcL1pnviEgbDb/2muvSaVKlaxk4Ii04ftbvnx5yZMnjyPkpZAk4FQC1apVk7x588rcuXOdWgTK7RACWihg+N8NGDBAsDXYVVddJTDCQsDQ87hx4+Sjjz5yCE7rxOTGC9axZcokEEzA1wt243aF9AMOrm37vmuxj12RIkUEn44dOyYg8corr8i7776b4JwXv/z9998C3z0YpDGQAAlYTwCjTcWLF1f2KO3atbM+wzjmgHcJOjmY3mOwl4AWPeBwCLJmzaoWnfD1iMPF853HtmJQVqE+p0+flkuXLvmiOuo/XI+w8QJ9fx1VbRTW4QTQC8Z2hVh7nYEErCCgRQ84qYItXbpUYP3buHHjpKKpa1OmTAm7wfaaNWvkmmuuSTYN3SIYhiE///yzvPXWW7qJRnlIwNUESpYsKTfccINMmDBBunTp4uqysnD2ENBeAffu3VvwiSRgUXV8QoX+/fs70poaDQfs2ALjNAYSIIH4EkAvGO8ODENjzWg3BPoB61OL2ihgDPPAzeb333+XM2fOCIafseNPo0aN1Mbz+iCLryTo/bZu3Tq+mTI3EiABRQC2KXCLHD16tPTs2dMVVNCYZ4Nej6rUYg4YC6CXLl1ann32WVm5cqXs2LFD/X/xxRcFw0C//fabHrTiLAWW5/z111/V/G+cs2Z2JEAC/08Aa0SPHz9erc5HKCQQSwJa9IDhgoQ5zm7duiUqG+Z1H374YVmyZEmia24/gW0H69SpI1myZHF7UVk+EtCWQIECBZQNyvDhw+W+++7TVk4K5jwCWvSAsRnDLbfcEpJey5Yt5cCBAyGvuf0kfH/DcXF72Vk+EtCJwJ133qmMIQ8dOqSTWFHJQj/gqLBZcpMWCrhp06Zy//33y5YtW9RmAygprH8xFI3zDRo0sKTwOie6YcMGxQD+iAwkQAL2EsidO7cyxBo6dKi9gsQgd+4HHAOIMUpCCwX8/vvvC4Z5sNBEhgwZlAEWfF5r1Kgh2bJlk08//TRGxXVOMjC+Yu/XOfVFSd1PoHv37rJ48WLBLm0MJBALAlooYJj3QwljeAeLaMASGv8PHz4sn3zyiVLCsSisU9KA8RXmvFu0aOEUkSknCbieADwzunbtKlg6l4EEYkFACwUcWBAsgl60aFG1GHrgeS8dw/gK26Jlz57dS8VmWUlAewJYLnf79u2CvbmdGugHrE/NaaeA9UFjnySTJ0+WW2+91T4BmDMJkEBIApgaw+IcTp4Wgw8wjFsZ7CdABWx/HSSQYPXq1WrhkXLlyiU4zy8kQAJ6EMC67FhXftGiRXoIRCkcS4AKWLOqmzRpUqJdoTQTkeKQgOcJYG2CL7/8Um2h6nkYBBA1ASrgqNHF/kYYnq1bty6ijSdinztTJAESiJRAxYoV1XKOY8aMifQWbeLRD1ibqhAqYH3qQrDqV6tWrTy99rVG1UFRSCBJAn379pVRo0bJ8ePHk4yn20X6AetTI1TAmtTFhQsXlAKm8ZUmFUIxSCAZAoUKFVKugoMHD04mJi+TQGgCVMChucT97OzZs6VSpUpSsGDBuOfNDEmABKIjgB2SsDgHXJMYSCClBKiAU0rMoviYS4KPIQMJkIBzCGBxDiyXiwWDnBLoB6xPTVEBa1AX2HIwY8aMUrlyZQ2koQgkQAIpIYAV6+CWhFEsJwT6AetTS1TAGtQFDDk6deqkgSQUgQRIIBoC/fv3l6+//lrOnDkTze28x6MEqIBtrvhdu3bJ3r17pUmTJjZLwuxJgASiJVCqVCmpXbu20CArWoLevI8K2OZ6HzFihHTu3FnSpUtnsyTMngRIIDUE7rvvPlmwYIHaRjU16Vh9L/2ArSYcefpUwJGzinnMgwcPyooVK6RNmzYxT5sJkgAJxJcANk+BQdYHH3wQ34xTmBv9gFMIzMLoVMAWwk0uacz9wu83S5YsyUXldRIgAQcQaN68uRrNwoYqDCSQHAEq4OQIWXQdex3PnTtXOnToYFEOTJYESMAOAk8++aR8//33cvToUTuyZ54OIkAFbFNljRw5Ug0958yZ0yYJmC0JkIAVBOBn265dO219g+kHbEWtR5cmFXB03FJ115EjR2TmzJnSpUuXVKXDm0mABPQk0KNHD7U61sqVK7UTkH7A+lQJFbANdTF8+HDV+7366qttyJ1ZkgAJWE0gffr08sQTT8i7774rZ8+etTo7pu9QAlTAca64Q4cOyZw5c+T222+Pc87MjgRIIJ4EbrzxRuUbjAU6GEggFAEq4FBULDz33XffqfmhHDlyWJgLkyYBEtCBwAMPPCBLly6VDRs26CCOkoF+wNpUBfcDjmdV7Nu3T/n9duvWLZ7ZMi8SIAGbCGCzhscff1wGDhwo2HJUh0A/YB1q4b8ysAccx7r48ssvpXv37oIfJQMJkIA3CNSoUUNtNfrFF194o8AsZcQEqIAjRpW6iL/99pvs3r1b2rdvn7qEeDcJkIDjCDz88MOCXc/Wrl3rONkpsHUEqICtY5sg5XHjxkm/fv0E1pEMJEAC3iKAUa8BAwbIW2+9JadOnbK18PQDthV/gsypgBPgsObLxIkT5fz588oi0pocmCoJkIDuBKpUqSINGzaU9957z1ZR6QdsK/4EmVMBJ8AR+y9Yjm7IkCHyyCOPxD5xpkgCJOAoAvfee6/ACnn69OmOkpvCWkOACtgarv5UYXjRtm1bKVKkiP8cD0iABLxJAFNQL7/8ssA3eM+ePd6EwFL7CVAB+1HE/gCGVxs3bpRevXrFPnGmSAIk4EgChQsXVvYgL7zwgpw7dy7uZaAfcNyRh82QCjgsmtRdwPJz77zzjvIBzJAhQ+oS490kQAKuItCkSRPBnPAbb7wR93LRDzjuyMNmSAUcFk3qLmCIqXr16lKtWrXUJcS7SYAEXEkAdiHHjx+XYcOGubJ8LFTyBKiAk2eU4hirV6+W5cuXS9++fVN8L28gARLwBoF06dLJq6++KtOmTZMFCxZ4o9AsZQICVMAJcKT+C3z8sOzcM888wxWvUo+TKZCAqwnkypVL3nzzTRk0aJBs2rQpLmWlH3BcMEeUCRVwRJgijwRH+xYtWqil5yK/izFJgAS8SqB48eLy4osvCoyysFqe1YF+wFYTjjx9KuDIWSUbc+zYsWpO56677ko2LiOQAAmQgI9A5cqV1UpZTz/9tMB7gsEbBLguYozqGe5Go0aNks8//1wwt8NAAiRAAikhUK9ePcmSJYsy3IQShpU0g7sJsAccg/o9dOiQvPTSS4LWa758+WKQIpMgARLwIgF4TSxevFjNC8NdyIpAP2ArqEaXJhVwdNz8d2GPT8zfdOnShS5Hfio8IAESiJZA3bp1pX///vLUU0/JmjVrok0m7H30Aw6LJu4XqIBTiRxGV2XLlpWuXbumMiXeTgIkQAL/JVCnTh3lovT666/LzJkzicWlBKiAU1Gx33zzjRw5coT+vqlgyFtJgARCE6hQoYJ88sknMmLECPn000/FMIzQEXnWsQSogKOsusmTJ6u5Giwlxz1+o4TI20iABJIkUKhQIcGGLgcOHJBHH31UNfiTvCGCi/QDjgBSnKJQAUcBeuHChapVin09s2fPHkUKvIUESIAEIiMAy2g09GvXri3YznDlypWR3RgmFv2Aw4Cx4TTdkFIIfd68ecrV6P3336fFcwrZMToJkED0BG6//Xa1wA/mhWGodf/99ws3eomepw53sgecglqYM2eOvPvuuwLDq6JFi6bgTkYlARIggdQTKF++vAwePFhOnjwpffr0ka1bt6Y+UaZgGwH2gCNED+Xbr18/tckC1m9lIAESIAE7CGTNmlWtNb906VJ59tln1dK3vXv3lowZM0YkDvyA9+3bJ7Vq1YooPiNZR4A94AjYYqcSGEJMmTJFqHwjAMYoJEAClhOAq9KQIUPkn3/+Ub3hdevWRZQn/YAjwhSXSOwBJ4N5woQJMnLkSPnggw847JwMK14mARKIL4GrrrpKnn/+eVmxYoWaGsNKWtgGNVu2bPEVhLlFRYA94DDY4HOHXu/EiROVLx7nfMOA4mkSIAHbCdSsWVP1hqF4e/bsKTNmzLBdJgqQPAEq4BCMTp8+rdZ13rlzp7J4zp8/f4hYPEUCJEAC+hDInDmz6v3CPXLatGnqeMuWLYkEpB9wIiS2naACDkK/efNmueeee6RkyZLK4plDOUGA+JUESEBrAthfeNCgQdKhQwe1xzB8iA8ePOiXmX7AfhS2H3AO+P+r4PLlyzJ06FD5+eef1SLoGNJhIAESIAGnEmjatKnUr19fbZP6wAMPSOPGjaV79+6SN29epxbJdXJr1wM+d+6cWm4N/+MV1q5dK3fffbfs2bNH+dhR+caLPPMhARKwkgBck3r06KE6F3Bfwuge1jLAu47BfgJaKOCLFy/KK6+8Ihg6wbJrefLkUf+LFCmi/NzOnz9vCant27crfzpYOD/44IPy8ssv083IEtJMlARIwE4COXLkUK5KP/30k0ARo8OBLQ9nz54tVr1f7SyvU/LWYggaG9nDN23cuHFSunRppXzRA4aCxPwFFiH/8ssvY8IUQ81LliyRSZMmyf79+9WQTJs2bSRtWi3aIjEpIxMhARIggVAEsHY9dln6+++/1QIes2bNko8//lhuuukmqVevnmD0j+vbhyJnzTktFPDUqVOVH1vOnDn9pYTxU+XKleXHH3+UcuXK+c+n5mDbtm3SsWNHKVGihLRt21YaNGhAxZsaoLyXBEjAkQTSpEmj1pPGmtLw+li8eLEsWrRIKWOMPOLdiw+WvsSoJIM1BLRQwNdcc41gWbVWrVolKiV6q7FyA4JS//bbb2mEkIgyT5AACXiVAN6LLVq0UJ8rV67Ixo0bBatqYR9iuDEVLlxYbrjhBqWM8R/fGWJDQAsFDKOAdu3aSalSpaRs2bJqjuLs2bNqWBoPA4amYxHw4NACMBYkmQYJkIBTCSTlB4ypuEqVKqkPjLegkDEVuGnTJrUNIjxFjh07pqYKy5Qpo/7DZfPaa691Kg5b5dZCAVevXl1VMlZvQYvrzJkzSlHCjB4tMyy3xkACJEACJJB6AvADxieSAIUMRYuPL5w6dUrtwoSdmLA3+nfffSeHDh1SS/XCkBZD2FDyhQoVUp9MmTL5buX/IAJaKGDIhGEQzM8ykAAJkAAJ6EsARlpYcxofX7hw4YJg5cDdu3ernZZmzpypjFyxAEj69OlVhwqjj1dffbXyNEGnCungg3c/5pnxwWpecJ2C0sZ/fHC/W4P2JYP1M3YhgqFWcgEGW+PHjw8Zbf369cr4Cg8KAwmQAAmQQGwJhOtZo8eMHvLRo0fl+PHjcuLECWX4BV9kGIBhuhEfuEPhAw8YvKd9H3iuQAnjkyFDBmU4i2P0zqG8cQ+OYVgW/MGa/gg4D2+bAgUKJCo07rUraK+A4SIEE/lIAnrQiB8qjBkzRlU2WlQMJEACJOBVAvHeDzh37tyCT2oClDDWi8Dn0qVLam4a56Bg8d93jDwwbw2Fi3MIOEa8ggULKgWuTgb8QXy7gnYKGK0ZtIYwHIFhCMwl4BNJ8A1jhIoLFyc7QYeSiedIgARIIN4EsObCvHnzpFatWvHOOur80qVLJ/hgiNpNwb6+dwBFtGpee+01NUQMwJgnwP9ixYqpvS6hlBlIgARIgARIwE0EtFDAzzzzjDJxHzVqlJojwNg/5gkwnwur6Mcee8xNzFkWEiABEiABEhAthqCxA9GKFSskcCUsTLZXrVpVrYQF528GEiABEiCB1BNIyg849akzhZQQ0KIHjJWwli9fHlJunM+XL1/IazxJAiRAAiSQMgLcDzhlvKyMrUUP+J133pH27dsrZ+/glbCwJFqsVsKyEiTTJgESIAESIIGUENBCAcPNCBslTJ8+3b8SVq5cuaRhw4bSsmVLwVZaDCRAAiRAAiTgJgJaKGAAxYoonTt3dhNbloUESIAEtCMQbz9g7QBoJJAWc8Aa8aAoJEACJOBqAvADxmgjg/0EqIDtrwNKQAIkQAIk4EECVMAerHQWmQRIgARIwH4Cacw1Mv+7WrX9slgqwdq1a9U60VWqVElxPvPnz3fdEmgphpDEDVg83bdYehLRPHsJS6BiNTcslcoQmgAW5cfC+gyhCfg2HMD6CKkNeB7x2sfSjm4JPjVWp06dFBcJuzjNnj1bsF98vINnFHBqwMIae8GCBalJwtX3vvjii9KsWTO5+eabXV3OaAuHfVM//PBDwc5eDKEJ8DcWmovvLJ6fEiVKSLt27Xyn+D+AwD///CMPP/ywYDVFJwUOQTuptigrCZAACZCAawhQAbumKlkQEiABEiABJxGgAnZSbVFWEiABEiAB1xCgAnZNVbIgJEACJEACTiJABeyk2qKsJEACJEACriFABeyaqmRBSIAESIAEnESAbkgR1NaBAwcEWyYyhCZw9OhR5eOaOXPm0BE8fvbixYty/PhxyZs3r8dJhC8+f2Ph2eAKnh/4AGfNmjXpiB69Ct/mQ4cOSf78+R1FgArYUdVFYUmABEiABNxCgEPQbqlJloMESIAESMBRBKiAHVVdFJYESIAESMAtBKiA3VKTLAcJkAAJkICjCFABO6q6KCwJkAAJkIBbCFABu6UmWQ4SIAESIAFHEaACdlR1UVgSIAESIAG3EKACdktNshwkQAIkQAKOIkAFHKK6sPn1rl27BAsoRBKwF+Xly5cjieqaOPv27ZNjx44lWR5sko14//77b5Lx3HQRTFDm5EKk8ZJLx2nX8ZvCb+vChQtJio7r2Cj99OnTScZz48W//vpLjhw5ElHRzpw5oxbpiCiySyKdOHFC9u7dm2xpfO/x5J61ZBOyMAIVcBDcuXPnSsmSJaVXr17q/8qVK4NiJPy6evVqufbaawUr+XghQHHUqVNHOnToIJUqVZLXXnstZLG3bNkiN9xwg3Tp0kXFb9++vZw7dy5kXLecfPXVV6VKlSpy6623St26dcO+GF9//XWpXLmyigeWyTVk3MJnxYoVUqpUKfXbuv7662XevHkhizZixAgpXry43HPPPSr+22+/HTKe205CmTZv3lzatm0r1apVk8cffzzJImL1pzZt2shbb72VZDw3Xfz444+lQoUK0rlzZ/VbQ+cnVBg1apR6du666y71LC1atChUNPvPmb0Uhv8ncOnSJcNUpsYvv/yizkycONGoWLFiWD5m69wwX7iGuUScYfZ6wsZz04UnnnjCeOCBB1SRzCUoFa9169YlKmLLli2N999/X50H10aNGhmDBg1KFM8tJ9asWWMULlzYMFvnqkj333+/8dRTTyUqHlgVKlTIMJcWVNcefPBBA0y9EMqXL29MnjxZFdV8IRpFihQxzJGjBEU/deqUcdVVVxkbN25U5/fv32/kyJHD2LZtW4J4bvzy5ptvGl27dlVFw7ulTJkyxoIFC8IW1WyYGDlz5jT+85//hI3jpgs7duwwzOVcDXNETRXr6aefNu69995ERcQzg3h//PGHujZ8+HD1nk4UUYMT7AEHtIHWr1+vvtWrV0/9b9eunezZs0d2794dEOt/h2ihopWeJUuW/510+dHUqVPl9ttvV6XMlSuXaq1PmTIlUal79+4tPXr0UOfTpUunesPmDyNRPLecABf0RkzloYp05513iqlsEhUP8Vq3bi2mUkkyXqIbHX4CvyEMG4IRws0336z+r127Vv33/TEba/Lpp5+KqazVKazBjjW0vTDC9PPPP0u3bt1UubHmM0aZQj1DiABuY8eOFVMBqfhe+DNjxgxp2LChf031cL8xvI/wnJUoUUIOHz6sesvLli3TEhEVcEC14AVx3XXX+c+kSZNGzN5KyPkGvEjxUnnooYf88b1wEMzI7PWF5GO25CVfvnwKCe4ZOXKkGnJ1K6NgLpiWwLnggPnhwGcsXLzg+5z+HeXGs5I27f9eOaHKbvbopGfPnv7ijh49Ws0X16hRw3/OrQdgZI4K+IsXig8uYirn7rvvlsGDB0vGjBn98d1+gN9TMJ+///47kT3B1q1blU0OpoOqV6+uhqI3bdqkJZ70WkoVJ6Gwiw8qEAG9WLS+seNIYMD3YGMszDs8+eSTMmfOnMCorjuGYdn27dv95UKLMphRKD7+G8wDNFIwr/XMM89IzZo1Ay+56jiYS/r06RWr4ELiWQp8xnCMe90eUG4wCQzJlR09mX79+gn+e2GnrWBG4fiYUxvSvXt3MafHVMM2kKmbj/E7CXyGfL8jnA9siGAuHb3lVatWSbFixeSLL76QO+64Q37//Xft8PyvOaqdaNYLNGbMGGncuLH6mPOacvXVVwss7AIDtgHD+cDwwgsvqCHEDz/8UClitEjfeOONBMoqML5Tj9FA8fHBf7TQgxmF4uMrL4Z9YIxkznEma1Diu8ep/8EFLHwBzxGG6INDcDzcEype8H1O/x783KA8SZUdxjZ9+/ZVL1I3N9wC6zWYUSg+GzZskG+//VYNyaMTAKPR+fPny5AhQwKTcuVxKD5QvMFTgJjeQaMfyhcBowUwCg38faoLGvxJ2CTVQKB4inDfffcJPr6A/STNiXtBCwpzMHCfwbly5cr5oqj/t912m2p9+k5iWA3Da25rpWPuLXjeFsM6v/32m5rTRflxjLmY4ADlC8vnYcOGSYsWLYIvu+571apVVVl9BYN1PM4FB5z7/vvv/afDxfNHcMkBfkOYj8NvClMTcC/C6MqNN96YqIQDBw5ULJcuXZpguD5RRJedwLNhGvOJzwYFzwas5ANDnjx55J133vGfgt0KrKF90z3+Cy48AJ9AexPwwfsIU4WBAfFMQ1r/KYxyZsuWTX38J3U50MAQTCsRTPcjo1OnTsbMmTMNU3EYpqGVku/kyZPqu2kMkkheWGmavcNE5914whzaMcyWpTFhwgQDVpvmsLQBi02E559/3vjpp58M84VgmO4mRq1atZSFJqw08Zk0aZIbkagygYHpOmOYLiGKTdGiRY1Zs2apa2YPxejTp486Nht3humCo9iBIViCqRdC//79DVjH47fVsWNHw3QR8RcbvzVTIRubN282zAatYQ4ZJnh2zB6MP65bD8xGq7KQN+e9jY8++kh5GJj+wKq4sHj+6quvEhX92Wef9YwVtDlEb8CSHu8ZUxEbpUuXNsxRTMXEHG72W5CbI5KG6aqkvAvwrDVo0MAwp8ASsdPhRLqXzaBLY0AHOWChihYThnVM1xkxTd2V4QhamegN1q5dO1FPF3MRGGrNlCmTDkWwVAb4SMMPD9aZaHlifgWtcgT0bmAkge9wgjfdKBQTcMEH1zCP7MaAZ8BsuAn8DU0XGvXcNGvWTBUVvT3MU8EgBPHgw4gWOuKZDRM1XOZGJsFlatq0qZhuRsp2AixeeeUVgYU8AgxsMNRsNnSVXz16dL7nBv/hc547d+7gJF31HUZXZqNV/bbOnj2rfluwAkfA6EGBAgUE/tOBAfzwm4LftNsDRhrx24E/OUbeHnnkEb9hJ6YBMXIJfpgnxhw5huuXLFki8GYZMGBAop6yDrzSoBWggyCUgQRIgARIgAS8RMDTRlheqmiWlQRIgARIQC8CVMB61QelIQESIAES8AgBKmCPVDSLSQIkQAIkoBcBKmC96oPSkAAJkAAJeIQAFbBHKprFJAESIAES0IsAFbBe9UFpSIAESIAEPEKACtgjFc1ikgAJkAAJ6EWACliv+qA0JEACJEACHiFABeyRimYxSYAESIAE9CJABaxXfVAaEiABEiABjxCgAvZIRbOYJEACJEACehGgAtarPigNCZAACZCARwhQAXukollMEiABEiABvQhQAetVH5SGBEiABEjAIwSogD1S0SwmCZAACZCAXgSogPWqD0pDAiRAAiTgEQJUwB6paBaTBEiABEhALwJUwHrVB6UhARIgARLwCIH0Hikni0kCWhD49ddf5a+//gopS9OmTSVbtmwhryV38sqVK5I2beL29MWLF2X69Onyxx9/SJ48eaR169aSL18+f3Lh7vNH+P8DX7wzZ87IwoULpVWrVsFREnxPafwEN/MLCXiEQOJfrEcKzmKSgB0E5s6dK99884363HnnnfL666/7vx8/fjwqkZo1ayb//PNPonuhLGvWrCnvvvuu/PnnnzJt2jQpU6aMLF68WMX9/fffpVu3bonuCz6xdetW6dKlizp9+PBhefzxx4OjJPie0vgJbuYXEvAQAfaAPVTZLKr9BP7zn/8IPggVKlSQF154Qdq1a5dIsGPHjkn27NklffqEP9GTJ09KunTpJGvWrP57li1b5j8OPJg8ebK6/5dffvGffuedd+Sll14SNAT27Nkj+/fv91/DweXLlwV558qVS+WDc4HxrrvuOtm8eTNO+8ORI0ckZ86cEcc/e/asoIccbW/fnzEPSMDhBNgDdngFUnx3EUCvtHLlylK/fn25/vrr5b333lMFPHXqlKCnW6tWLalYsaK0b99ezp8/Lw888IBAoTVu3Fh2796dAAYUOBQs0vSFvn37yvfffy+nT5+WBx98UNasWSMdO3ZUl5955hkpWbKkYCg8f/78MnjwYEEvGnmsW7dObrvtNtm7d68UL15cxV+xYoWULl1ayVW0aFH56KOPkox/7tw5Qa+/XLlyUqlSJZUvysBAAp4lYDCQAAnYQqB8+fLGpEmTEuRtKlfDVJDqnNkTNUqUKGGYPVhj2LBhhjl/q86bvVSjZ8+exurVq9V3sydpHDhwQB0H/jF7mcajjz5qZMiQwTAVpXHvvfca5jC0P4o5N2zUrVtXfd+4caNh9sgNUzGr71OmTDFMZayOZ86cadSuXVsd79q1y8idO7c6NhWy8e2336pjU/kr+ZBnuPhvvvmm0bJlS8OclzZQBrNBYQwZMkTdzz8k4EUCCce3PNsMYcFJwH4C+/btkw0bNqgh4E8++UQJVLBgQTEVmpqDRY+1a9euYioxef/99yVv3rxJCp0mTRrVK8WQ87x589SnR48e0qFDB/n6668T3Gs2BgQGYitXrlQGWwsWLFC92QSRgr7UqFFDnnvuOVm/fr2SaeLEiYI8wwUMe99zzz3+YXWUK6n44dLheRJwCwEOQbulJlkOxxPAsHCmTJlUOTAXi0+nTp3UkDCGnTEMjOHpoUOHCoZ8Z82alWSZEW/58uVy9dVXq+Hezz77TFatWiVmr1Up+cCbd+7cKaVKlZIPPvhAGWxBuSYXnn76aRkxYoRSoo899pgaWsb8cbiA4ewsWbL4L1P5+lHwwKMEqIA9WvEstn4EMLdqDhdLo0aNBAoNHxg4XbhwQSZMmKB6wpinnT9/vtx///1iDhOrQkCRmcN3iQoEYykYecEVyRcwlwwlmDlzZnXKd9+oUaOkevXqKh/cA1clGEohhEv/qaeeUsodc79btmxRDQbMKYeLD6MzNAB8AcZo5rC07yv/k4DnCHAI2nNVzgLrSgC934EDB8ott9wiffr0EXO+VRYtWiSPPPKIcjOC0sU5WA+PGTNG9T5RFgxT33333fLpp5+KOWfsLx4UXPPmzZWhVMOGDZWxFnrNH374oVLA11xzjaxdu1Ypelhiw13p7bffVgZaUPJwi0IvHOljmBlyDBgwwJ9+4cKF1dB47969lXEWlDrcnnbs2BEyPhQ2GhcwGoMl9+jRo/0uUf5EeUACHiKQBhPfHiovi0oC2hBYsmSJUo6BC2NAuE2bNgmuwbWnTZs2yh0J57GAx+zZs5VSbNKkiRQrVgynlZvQ0qVL1VB1cFroxUKJo4cKBQ8F6LsP98InGD7EmBdG7xWWzXA1wjwz5oFhdQ2FD3kOHjyoFvKA2xOsrhEwZ4z7MMwNWX2uReHio0c/depU1btu27atWhxEJcQ/JOBBAlTAHqx0FpkESIAESMB+ApwDtr8OKAEJkAAJkIAHCVABe7DSWWQSIAESIAH7CVAB218HlIAESIAESMCDBKiAPVjpLDIJkAAJkID9BKiA7a8DSkACJEACJOBBAlTAHqx0FpkESIAESMB+AlTA9tcBJSABEiABEvAgASpgD1Y6i0wCJEACJGA/ASpg++uAEpAACZAACXiQABWwByudRSYBEiABErCfABWw/XVACUiABEiABDxIgArYg5XOIpMACZAACdhPgArY/jqgBCRAAiRAAh4kQAXswUpnkUmABEiABOwnQAVsfx1QAhIgARIgAQ8SoAL2YKWzyCRAAiRAAvYToAK2vw4oAQmQAAmQgAcJ/B/hNyacBqWcZQAAAABJRU5ErkJggg==\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "library(sna)\n",
    "library(showtext)\n",
    "\n",
    "# Use Myriad Pro\n",
    "# library(showtext)\n",
    "# font_add(\"MyriadPro\",\n",
    "#     regular = \"MyriadPro-Regular.otf\",\n",
    "#     bold = \"MyriadPro-Bold.otf\")\n",
    "# showtext_auto()\n",
    "\n",
    "# Use Arial\n",
    "library(showtext)\n",
    "font_add(\"Arial\", \"Arial.ttf\")\n",
    "showtext_auto()\n",
    "\n",
    "# Create a fully connected graph as a null model for baseline and test\n",
    "full_adj <- matrix(rep(1, 18*18), 18,18)\n",
    "length <- dim(full_adj)[1] * dim(full_adj)[2]\n",
    "noise <- matrix(runif(length, -0.01, 0.01), dim(full_adj)[1])\n",
    "full_adj <- full_adj + noise\n",
    "diag(full_adj) <- 0\n",
    "\n",
    "graph.Cor <- qaptest(list(ground_adj, interaction_adj), gcor, g1=1, g2=2, reps=10000)\n",
    "print(gcor(ground_adj, interaction_adj), diag=FALSE, mode='graph')\n",
    "print(summary(graph.Cor))\n",
    "\n",
    "#png(width=1600, height=1600, res=300, file='cor-interaction.png')\n",
    "plot(graph.Cor, xlim=c(-0.4, 0.6), family='Arial')\n",
    "\n",
    "# Calculate graph correlations for all models\n",
    "gcors <- gcor(graphs)\n",
    "rownames(gcors) <- c('graphical\\n lasso','influence\\n model','interaction\\n model', 'communi-\\n cation','advice','trust','social','organi-\\n zational','overall\\n ground')\n",
    "colnames(gcors) <- c('graphical\\n lasso','influence\\n model','interaction\\n model', 'communi-\\n cation','advice','trust','social','organi-\\n zational','overall\\n ground')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/site-packages/rpy2/rinterface/__init__.py:146: RRuntimeWarning: corrplot 0.84 loaded\n",
      "\n",
      "  warnings.warn(x, RRuntimeWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "gcors_upper = gcors\n",
    "UT = upper.tri(gcors_upper)\n",
    "gcors_upper[UT] = gcors[UT]\n",
    "diag(gcors_upper) = NA\n",
    "\n",
    "library(corrplot)\n",
    "#pdf(width=3.5, height=3.5, file='cor-talk.pdf')\n",
    "#corrplot.mixed(gcors, lower='number', lower.col='black', upper='shade', tl.col='black', tl.cex=.35, cl.cex=0.6, number.cex=0.6, family='MyriadPro')\n",
    "corrplot(gcors_upper, type='upper', method='shade', tl.col='black', tl.pos='d', tl.srt=90, family='Arial', addCoef.col = 'black', tl.cex=.35, cl.cex=0.5, number.cex=0.5, family='Arial')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## OLS Network Model test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "OLS Network Model\n",
       "\n",
       "Coefficients:\n",
       "            Estimate  Pr(<=b) Pr(>=b) Pr(>=|b|)\n",
       "(intercept) 0.8415642 0.999   0.001   0.001    \n",
       "x1          4.8199532 0.735   0.265   0.498    \n",
       "\n",
       "Residual standard error: 0.6064 on 304 degrees of freedom\n",
       "F-statistic: 2.115 on 1 and 304 degrees of freedom, p-value: 0.1469 \n",
       "Multiple R-squared: 0.006908 \tAdjusted R-squared: 0.003641 \n",
       "\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "# Choose any two adjacency matrices to compare\n",
    "nl<-netlm((ground_adj), # Dependent variable/network\n",
    "          list(influence_adj), # List the independent variables/networks\n",
    "          reps=1000, # select the number of permutations\n",
    "          mode='digraph',\n",
    "          nullhyp='qap',\n",
    "          intercept=TRUE)\n",
    "\n",
    "#Examine the results\n",
    "print(nl)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Mantel test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "$z.stat\n",
      "[1] 15.86033\n",
      "\n",
      "$p\n",
      "[1] 0.00609939\n",
      "\n",
      "$alternative\n",
      "[1] \"two.sided\"\n",
      "\n",
      "[1] 0.5066379\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAEGWlDQ1BrQ0dDb2xvclNwYWNlR2VuZXJpY1JHQgAAOI2NVV1oHFUUPrtzZyMkzlNsNIV0qD8NJQ2TVjShtLp/3d02bpZJNtoi6GT27s6Yyc44M7v9oU9FUHwx6psUxL+3gCAo9Q/bPrQvlQol2tQgKD60+INQ6Ium65k7M5lpurHeZe58853vnnvuuWfvBei5qliWkRQBFpquLRcy4nOHj4g9K5CEh6AXBqFXUR0rXalMAjZPC3e1W99Dwntf2dXd/p+tt0YdFSBxH2Kz5qgLiI8B8KdVy3YBevqRHz/qWh72Yui3MUDEL3q44WPXw3M+fo1pZuQs4tOIBVVTaoiXEI/MxfhGDPsxsNZfoE1q66ro5aJim3XdoLFw72H+n23BaIXzbcOnz5mfPoTvYVz7KzUl5+FRxEuqkp9G/Ajia219thzg25abkRE/BpDc3pqvphHvRFys2weqvp+krbWKIX7nhDbzLOItiM8358pTwdirqpPFnMF2xLc1WvLyOwTAibpbmvHHcvttU57y5+XqNZrLe3lE/Pq8eUj2fXKfOe3pfOjzhJYtB/yll5SDFcSDiH+hRkH25+L+sdxKEAMZahrlSX8ukqMOWy/jXW2m6M9LDBc31B9LFuv6gVKg/0Szi3KAr1kGq1GMjU/aLbnq6/lRxc4XfJ98hTargX++DbMJBSiYMIe9Ck1YAxFkKEAG3xbYaKmDDgYyFK0UGYpfoWYXG+fAPPI6tJnNwb7ClP7IyF+D+bjOtCpkhz6CFrIa/I6sFtNl8auFXGMTP34sNwI/JhkgEtmDz14ySfaRcTIBInmKPE32kxyyE2Tv+thKbEVePDfW/byMM1Kmm0XdObS7oGD/MypMXFPXrCwOtoYjyyn7BV29/MZfsVzpLDdRtuIZnbpXzvlf+ev8MvYr/Gqk4H/kV/G3csdazLuyTMPsbFhzd1UabQbjFvDRmcWJxR3zcfHkVw9GfpbJmeev9F08WW8uDkaslwX6avlWGU6NRKz0g/SHtCy9J30o/ca9zX3Kfc19zn3BXQKRO8ud477hLnAfc1/G9mrzGlrfexZ5GLdn6ZZrrEohI2wVHhZywjbhUWEy8icMCGNCUdiBlq3r+xafL549HQ5jH+an+1y+LlYBifuxAvRN/lVVVOlwlCkdVm9NOL5BE4wkQ2SMlDZU97hX86EilU/lUmkQUztTE6mx1EEPh7OmdqBtAvv8HdWpbrJS6tJj3n0CWdM6busNzRV3S9KTYhqvNiqWmuroiKgYhshMjmhTh9ptWhsF7970j/SbMrsPE1suR5z7DMC+P/Hs+y7ijrQAlhyAgccjbhjPygfeBTjzhNqy28EdkUh8C+DU9+z2v/oyeH791OncxHOs5y2AtTc7nb/f73TWPkD/qwBnjX8BoJ98VQNcC+8AAAA4ZVhJZk1NACoAAAAIAAGHaQAEAAAAAQAAABoAAAAAAAKgAgAEAAAAAQAAAeCgAwAEAAAAAQAAAeAAAAAApZ9jSgAAQABJREFUeAHtnQW8VMX7xt9Ld3d3d3d3KAqCGCAICIIgAsoPUbETLFJRUAkFBCRE6e7u7gbprv2fZ/yfde+9u/fu3q0Tz3w+yz17zpyJ7xz2OfPOOzMRDi0IAwmQAAmQAAmQQEgJxAtpbsyMBEiABEiABEhAEaAA80EgARIgARIggTAQoACHATqzJAESIAESIAEKMJ8BEiABEiABEggDAQpwGKAzSxIgARIgARKgAPMZIAESIAESIIEwEKAAhwE6syQBEiABEiABCjCfARIgARIgARIIAwEKcBigM0sSIAESIAESoADzGSABEiABEiCBMBCgAIcBOrMkARIgARIgAQownwESIAESIAESCAMBCnAYoDNLEiABEiABEqAA8xkgARIgARIggTAQoACHATqzJAESIAESIAEKMJ8BEiABEiABEggDAQpwGKAzSxIgARIgARKgAPMZIAESIAESIIEwEKAAhwE6syQBEiABEiABCjCfARIgARIgARIIAwEKcBigM0sSIAESIAESoADzGSABEiABEiCBMBCgAIcBOrP0jcCqVavk6NGjvt1kwNjXrl2Tv//+W+7fv+916c6fPy/Tpk2TOXPmyK1bt7y+z1PEpUuXyunTpz1dNt35kydPyvLly53l3rZtm+zatcv5nQckYGQCFGAjtw7Lpgh0795diVCwcYwdO1ZKly4dtGwOHDggTZo0katXr3qVx9mzZ6Vo0aLy8ccfy0cffSRXrlzx6r6YIrVv314WLFgQUxSVz0svvWQKoZ43b548//zzzvp88MEH8tVXXzm/h+PgwoUL0q1bN7l48WI4smeeJiJAATZRY7GowSVQpEgRadq0qTOTffv2Sa9evZzfQ32A3rLD4ZD169fL6tWrJVu2bAEvQt++faP1GPv166fEIxj5BbwCcUzw0aNH6mXIl9snT54sP/74Y6y3ZMyYUW7fvi19+vSJNS4j2JsABdje7W/Y2kP8Dh48qMoHEYoa8AMHcyNMtK7h3r17TlMtzNY7d+6UBw8euEZRx7i2devWSL3KSpUqyZtvvumMO27cODl8+LDqsSI/mJCjmo+RH857CjAbo5yXL19WYuou3pkzZ2T79u1y9+7dSJdxT+rUqeXGjRvqB12/iJ7xli1b5Ny5c/op9Rf3I65rQLk9ma4vXbok6PWjx4ZeOUTpyJEjMn78eBk0aJBKBulFvf/69euRyuOany/HYKmnffz4cdmxY0c0vlGZ41nQy+ptXqgT2trV8vDXX3/JmjVr1DnURw9oTzwzu3fvjvTc4BkaM2aMwOSNdBAPAeXZv3+/ar+bN2/qySh+EOy9e/c6z/GABKIR0B4gBhIwDAHNzOqoXr26I1myZA6tB+Zo1KiRo0SJEo7PP//cWcZZs2Y50qZN68iRI4cjceLEDq0X57ym/Ug6atas6Wjbtq0jS5YsjlSpUjly5crlOHTokIqj/Ug66tSp40iZMqUjZ86cjvjx4zs0EVLXRowY4ciXL586fvvtt1UZkD7SeeWVVxy1a9d2aL0aZ144wPn69etHOqd/0Uy9qpyZM2d2ZMiQwaH1pvEm4dAET0V5+PCho2vXro6kSZM6smbN6tB6To7Fixera//73//UvSgf8h84cKA6j7wQP3fu3KrsmqlYz04xKlmypPM7DjTzrOOZZ55xnkNamsA6Tp065ShUqJAqT/r06VUe2suG4/3333cULlzYGV8z56qy//PPP+qcJvyOBAkSOMvpjKgdbNy40aG9tET7/Pzzz67RnMe//PKLo2LFio4OHTo4wEh72VBtvmfPHmcctNOUKVOc37UXJ1VmPc53332n2leP0KZNG0eXLl3UV0281bOEZwDPCsqNPGfPnu1AnSMiIlS9K1SooOJPnTpVPRd47tKkSePInj27QxNidU0bOlD3ozxgiHS0FycHeKdLl061X6JEiRzz58/Xi+IoU6aMQ3uRcX7nAQlEJYA3OAYSMAyBAQMGOPLnz+/Qf/C///579YOrCzDEK0WKFA6td6HKjB9IiCREGQE/yBC5YcOGqe8Q3GLFiimhxIlJkyapH3ucRzh27JhD60WrY1cBxglt7NnRsmVLdQ3/QFwglBBOBK3HqH6kdQFXJ///H613p8REfznQeqJKqF0FePTo0Uo8tB6tuuutt95SP+5a70p9//bbb50vBHraWq/NofUa1VfNRK3qqo0tq+9g5K0A4wawRHk2bNig7sc/DRo0cLiKOuqKFyJtTFPVt0qVKo6XX37ZGd/1YMKECQ4IYNTPs88+6xrNeQwRQ/4ffvihOgdGZcuWdbz44ovOOP4IMNoLL1lIF0HrCTufKzwHEHzXoDmnOXRh16wJDs0i4hRzxMOL4Keffuq8BeUuV66cQ+sdq3N4FvXnCideffVVR9WqVZ3xeUACUQnQBK39AjAYh8Aff/whHTt2FK1XoQql/RgrM6xeQk1oRftRlqefflqdgpOS1ouSuXPn6lFEE2Tn2K3Wkxat56pMq4iAdOHMtHLlShVf6x0Lxuy8CZqwKHPzkiVLVPR169Yp823r1q2j3a71BpUTE8ZYEZIkSSKacEWKN3HiRNHESbTenzqv9QQF5uVNmzZFiuf6RRNA0XrA6lTdunXV30B6NcOcqvWAnVnGixdPjXtq4io9evRQ5dNEyHnd9QB10XqR0T64113QeqCC9F0Z1atXz9lW7u7x5ZxmJRHtRU6ZmnFfnjx5nM+Vu3S0lyuBHwCC1puVatWqCYYHPAU8SydOnFBDDIiDZxHPmx7AETwZSMATgQSeLvA8CYSDgGYaFc286swaP9KuzkAYz8OPIn649aC9VYpmLlRfEV8z90rChAn1y+oY45sIjRs3Fq2nKU888YRovUUZPHiwtGjRwhk3pgOt5y0QYa33JJopWHlmw2lLM1dGuw1jhSij1gNzXtPMoM5jHKAuK1askE8++STS+Zi8nbVesRqjhehqJtVI9wXiC/KOWh/NVC2vvfaawMN4+vTpAg7uwsKFC50vNq7X0RauY+uu15CX/kKB86iT3lau8eJy/Nhjj0n//v2lefPmovVUVVvDC91T0Hq/6tmA0xueKbDAy5ungJdDjBVrvVypU6eOvPPOO0q09fioW0xtqcfjX/sS+O9XzL4MWHMDEcCPdVRHJ1cnKvRM0NOA44z+gaPQTz/95KwFRDimADFAzwU/xk8++aRyRIopvuu1F154QYkQnHB+//130cZXXS87j1EPCIlr2V2PERF1gajp9cBf1KVhw4bOdFwPUEdtbFiGDh2qetdRe1cQjaghqmNX1OtRv6NMUe9BXWfOnKlehLTx2Ki3OL/jPte66MdRHcOcN2gHsbWVa1wcRy1b1Ouu35H2u+++K3DwgpBqwwmimb1doziP0TbgDqc3OIPh+ejUqZPzursDtPE333wjmn+BFCxYUGrVqhVpihfKCmsMAwl4IhD4V2hPOfE8CXhBAL3SzZs3C3oXCPCC1Rxv1DH+gaey5iikhEo33Tov+nAA8+GQIUNUbxqLXGiOO9HuRg82am8MP+To2QwfPlxN1cGPuh7gcYweHH6YUQ8EeCvDRI4Ab2jXgLqgt5U8eXLX0x6PtbFa1ZNDbwsBnuKuAWZueE67Bnj/ak5Grqecx7r4aeO8znPosWvj4s7vOABvhLVr16p6zZgxQ1kQ1EmXf9DTxCeQAQLm2osET18DLCLaeK3yYMZQBeYNo21d6w2LAqwvmlOcc8gDHsx4IdGDu+cB12DZwPOAF6I///zT+QIF4Y9q9dDT4l8SAAEKMJ8DQxHo2bOnoJeJXq7mkSyjRo1SY756IdFLgXDhh753795K7PCjDDMyeiCxhfHaFBtMu9GcfQTTcDDX1p34Ip0CBQrIb7/9psaXNa9ZwfgrRAtj1BgHbdWqVSTzafHixVVPCGmi7OhhY0ELmCYxdWXkyJGRivfGG2+I5uCkxoYx9gnxxNQYT3NNNWcyQfnRC4UwfPnll5HGHGvUqKHmnqKHr3mCqx56VEF2LQDGSFEvzVlJiRyEGi8LGNvWA16GUFeMe0OcIWQYC8aLiD5Or8cNxl+06RdffKHG6dGTRp29DZoDn+qRlypVSr0sLVq0SDTHKHW75uinXuIQB/UCf/DQHPHUFCu0IczLmiezMzs8D/BRKF++vMB3QPOmVu2PZxViixckPLt6wAsLnlUGEvBEIL7WCxji6SLPk0CoCWiepuoHEeOs+AGDgGH8DiY+jEVCANu1a6fm/2KJRogFeqSPP/64aNNNlJCg1+raE8N8WfSCIFDopcJRCGOZmHsLE/Lrr7+uekQQK6SvL8aBVbEwDxjxMU6J+xEwJg1RwgpVKJcesCQiyohxZgT0jjGX+ddff1XlwjxSzFvG+DN6dtp0FvXioE09Uj/mMGXC8QcfhKjlwUsDRAhssKoW8ke98CMPRzI4EUEk4Ki2VFtyEiIJQcCYbeXKlVWaKA9EDePsqCt66jAvY7nPZs2aSaZMmdSYNFYfQ88c4gxx0k3tEGkwgclW7+WrhOPwD6wbMNNirFYP4AMzMMqOAEcz1BXthWvoaYIBygpnPH1OML4joCebJ08eJZKYY4yXFfTYMbcXLODwhXqjDdEGmKt7VLOw4JnCWC7mB2N1LTyHeJHBs6S3O+qONkYvFy9bYI8XNPBD2yFtvJwhoFx4mdSmszkdu9QF/kMCLgQi4Bbt8p2HJEACsRDADy5MlXAGC4YjVCzZB/Uyfg7wwoPePXruDHEj8Nlnnym/BIwnw3TNQALuCPDJcEeF50jAAwH0vtD77dy5s+XEF1VG7xBmWZi6MSWKwXcCmPqkzUdXHCm+vvOz0x3sAduptVlXvwhgPBDjgjD5wgwJkzcDCZAACcSVAAU4ruR4n+0IYFwPY8gYu2UgARIgAX8JUID9Jcj7SYAESIAESCAOBDgGHAdovIUESIAESIAE/CVAAfaXIO8nARIgARIggTgQoADHARpvIQESIAESIAF/CVCA/SXI+0mABEiABEggDgQowHGAxltIgARIgARIwF8CFGB/CfJ+EiABEiABEogDAQpwHKDxFhIgARIgARLwlwAF2F+CvJ8ESIAESIAE4kCAAhwHaLyFBEiABEiABPwlQAH2lyDvJwESIAESIIE4EKAAxwEabyEBEiABEiABfwlQgP0lyPtJgARIgARIIA4EKMBxgMZbSIAESIAESMBfAhRgfwnyfhIgARIgARKIAwEKcByg8RYSIAESIAES8JcABdhfgryfBEiABEiABOJAgAIcB2i8hQRIgARIgAT8JWBYAb5z547gw0ACJEACJEACViSQwAiV+vnnn+XZZ5+V+PHjy40bN+S5556TP//8UxwOh9SsWVN++eUXyZ49u99FnTZtmjx48MDvdJgACZAACZCANQhkypRJ6tWrF5bKRGgi5whLzi6Z5s6dW44dO6bOfPLJJ7J//34ZMWKEEuRPP/1Utm3bJhBPf8Lvv/8uQ4cOlY4dO/qTDO8lARIgAdMS+OOPP6RMmTKC31yGfwl88803MnHiRMUl1EwM0QNOnjy5nD59WrJly6aE+Pnnn5ekSZMqFq+++mpAwKDn26FDB3nppZdCzZj5kQAJkIAhCBw+fFhatWollSpVMkR5jFAIdPgePXoUlqIYYgwYvdIXXnhBiXDLli1l7ty5Thjjx4+X0qVLO7/zgARIgARIgASsQMAQPeABAwbIzZs3pVixYpI1a1Y5c+aMLF68WJ1LlSqVTJ8+3QqsWQcSIAESIAEScBIwhADHixdP3nvvPRk0aJAa74UAIxQoUEBKlCjhLCwPSIAESIAESMAqBAwhwDrMJEmSSOXKlfWv6u+tW7fk7t27kjZt2kjn+YUESIAESIAEzEzAUALsDmSXLl1k7dq1AueB2MJ3330nkyZNchvtyJEjUqRIEenevbvb6zxJAiRAAiRAAqEkYHgBHjJkiPzzzz9eMenWrZvg4y707dtXzp496+4Sz5EACZAACZBAyAkYSoCxCMehQ4cEZmdMTcqXL58UKlQo5FCYIQmQAAmQAAkEm4AhBBg90x49esi8efMEq5JgLBgifOHCBWnSpIl8//336nywYTB9EiABEiABEggVAUPMA8Yc4LJlyypT8/Hjx9VKWCdPnlQCjB7wyy+/HCoezIcESIAESIAEQkLAED3gzZs3q7WfMR3JNWAO8ODBg6VcuXKup3lMAiRAAiRAAqYnYAgBLlq0qHz++efSp08fZX7WqWJxjs8++0zKly+vn+JfEiABjcDly5fVYjWwFKVLl05q164tuXLlIhsSIAETEYjc5QxTwceNGyd//fWXpE6dWvLkyaOmC2Gx8AwZMihz9MiRI8NUMmZLAsYjgI1JOnXqJEePHpW8efOqefLw8v/111+NV1iWiARIwCMBQ/SA4e28ZMkSuXTpkprve/v2bUmRIoUULFhQ/fVYel4gARsRwII0kydPli1btijHxIwZMzpr365dO8F3bOn51FNPOc/zgARIwLgEDCHAOh6Y0vBhIAESiEzg/v370r9/f8mZM6d89dVXEhERESlCypQp1Tx3CDHWU69Ro0ak6/xCAiRgPAKGMEEbDwtLRALGIQDxHTZsmLIQvf7669HEVy9pmjRpBHubQqCvX7+un+ZfEiABgxKgABu0YVgsEgAB7GP92muvycaNG9Wm4bFRKVy4sNSsWVN+/vnn2KLyOgmQQJgJUIDD3ADMngQ8EXj48KGahoc58lOnTvUULdp5OGj9/fffcu7cuWjXeIIESMA4BCjAxmkLloQEIhH44IMPBD3azp07Rzof2xfMn2/durVMmDAhtqi8TgIkEEYCFOAwwmfWJOCJwCeffCKYitehQwdPUWI8DwFetmyZGjeOMSIvkgAJhI0ABThs6JkxCbgngDHfzJkzS/v27dW0IvexYj6LaXwNGjSQGTNmxByRV0mABMJGgAIcNvTMmASiE2jWrJkS32eeeUYSJ04cPYIPZ9q0aSNz585Vjlw+3MaoJEACISJAAQ4RaGZDArERGDVqlBQoUEAGDBjgt/gir2zZskn+/PmVKTq2vHmdBEgg9AQowKFnzhxJIBIBh8Ohpg0dOHBArYkedVOSSJF9/PL444+rXrCPtzE6CZBACAhQgEMAmVmQgCcC2Pd6+PDhaglWbEjir9k5aj5Vq1ZVaZ85cybqJX4nARIIMwEKcJgbgNnblwBEEXtdJ0iQQIYMGRJnh6uYCGJt6MaNG6t5wTHF4zUSIIHQE6AAh545cyQBWbdunfTs2VOwdnOPHj2CSgQCPH/+/KDmwcRJgAR8J2CozRh8Lz7vIAFzEXj06JHayWjx4sXy0Ucfqa03g10D7DaWJEkS2bFjh5QsWTLY2TF9EiABLwmwB+wlKEYjAX8JnD59WpmcT5w4IT/88ENIxFcvc5MmTWThwoX6V/4lARIwAAEKsAEagUWwPoEFCxZIr169BPN8scQkFsoIZahbt67acxvrSzOQAAkYgwBN0MZoB5bCogSwleC4ceNkzZo18uWXX6rlJcNR1YwZM6o5xuvXrxd4RjOQAAmEnwB7wOFvA5bAogSuXbsm/fv3V7sSjRkzJmziq+OtU6eOLFq0SP/KvyRAAmEmQAEOcwMwe2sSuHfvnvTp00fKlCkjb731liRKlCjsFYUAoweMsjGQAAmEnwAFOPxtwBJYkEDz5s3VvF7szWuUgG0KixYtKqtWrTJKkVgOErA1AQqwrZuflQ80ATg5oedbv359GTt2bKCT9zu9evXqCaZAMZAACYSfAAU4/G3AEliIwIgRIyRNmjQycOBAQ9aqVq1asnXrVrl9+7Yhy8dCkYCdCFCA7dTarGtQCaxevVqtcGVU8UXlkyZNqsally9fHlQWTJwESCB2AhTg2BkxBgnESuD69esybNgwGTx4sBK5WG8IYwSYx5cuXRrGEjBrEiABEKAA8zkggQAQGDlypDRs2FA5OQUguaAmUa1aNbUs5dWrV4OaDxMnARKImQAFOGY+vEoCsRLYuXOnbN68WYzk8RxToTElCiJMM3RMlHiNBIJPgAIcfMbMwcIEHA6HfP3112pHIyPM9fUWNZamxPKYDCRAAuEjQAEOH3vmbAECELFkyZIJFrkwU6hUqZIcP35cLly4YKZis6wkYCkCFGBLNScrE0oCmPOLXY2CvZ9vMOoUP358qV27NnvBwYDLNEnASwIUYC9BMRoJRCUwe/ZsKVSoUEi3FYxaBn++w2mMa0P7Q5D3koB/BCjA/vHj3TYlgF2OJk6cKB07djQtgRIlSsjNmzfl4MGDpq0DC04CZiZAATZz67HsYSPw119/qZ5vgQIFwlaGQGTcpEkTmqEDAZJpkEAcCFCA4wCNt9ibADyfJ02aJM8995zpQTRq1EgJMMazGUiABEJLgAIcWt7MzQIEMG6aJUsWKVy4sOlrky1bNsmZM6ds3LjR9HVhBUjAbAQowGZrMZY37ASmTJkizz77bNjLEagCwAw9b968QCXHdEiABLwkQAH2EhSjkQAIYMUrmGsrVKhgGSCYw4weMJemtEyTsiImIUABNklDsZjGIDB16lRp3bq1MQoToFJgh6SaNWtySlKAeDIZEvCWAAXYW1KMZ3sCp0+fln379gkcl6wWYIaeM2eO1arF+pCAoQkYUoAfPXok69evF+yvevfuXUMDZOHsQ2DGjBnSokULSZAggeUqXbp0abl37556wbBc5VghEjAoAUMI8JtvvimnTp1SiCC4WCKvWbNmytEF8yz3799vUHwsll0I3L59W/7++2957LHHLFvl5s2by6xZsyxbP1aMBIxGwBAC/OeffzodQH7//XeJiIgQmPuOHDkiAwcOFAg0AwmEk8DChQulfPnykiFDhnAWI6h5N23aVFasWCF37twJaj5MnARI4F8ChhBg18bAsnhwctG3dsNiB1u3bnWNwmMSCDmBmTNnWrr3C6Bp0qSRMmXK0Bkr5E8XM7QrAUMIMFYWGj16tHz33Xdy+PBhNRalN8j169clceLE+lf+JYGQE9izZ49g7eeyZcuGPO9QZ9iyZUuBRYqBBEgg+AQMIcCvv/66pE2bVvV0b926pXaY0asOR6zGjRvrX/mXBEJO4I8//lDOVyHPOAwZVqxYUf755x81/BOG7JklCdiKgCHcOZ955hmP0Nu2bSv4MJBAOAjghRAvgT179gxH9mHJE2PBcMbq06dPWPJnpiRgFwKG6AHHBHvNmjUyffr0mKI4r924cUPOnTvn9oNt1x48eOCMywMS8IYA1n3GqlcpU6b0Jrol4mCq1eLFi/n/xRKtyUoYmYAhesAxARo7dqysXbtWnnzyyZiiqWvYIB3mQndhy5YtkjVrVneXeI4EPBLA4hRdu3b1eN2KF9KnT6+2Wly6dKk0aNDAilVknUjAEAQML8A//PCDYGEOb0L79u0FH3ehb9++cvbsWXeXeI4E3BI4evSomh5npXWf3VbUzUn0guH5TQF2A4enSCBABAwjwFeuXFE7suzdu1cw7pY8eXLljIXxKDhoMZBAqAnMnTvXtg6A1apVky+++EK9tGLrRQYSIIHAEzDEGPCCBQuU2E6YMEEgxPHjx5dLly7JTz/9JAULFpQlS5YEvuZMkQRiIACrC8Z/8QJox4D/g1jzGqt/MZAACQSHgCF6wPC2RG8DUyCiBqxANGjQIIEzFgMJhIrAunXrJEeOHGLn3h8E+J133pGOHTuGCjvzIQFbETBED/jy5ctSvHhxt+BLlCghuM5AAqEkMH/+fMEOQXYOsD4lS5ZMdu7caWcMrDsJBI2AIQS4VatWaqGDadOmCbyVsfLQpk2bZPLkyYJrWI6SgQRCRQA+CNiNq1atWqHK0rD51KtXTzBExEACJBB4AoYwQX/77bcycuRIGT58uBw6dEgtBp8iRQopWrSo9O7dW2JaqCPwSJii3QksX75cDYfgGbR7gBka07AwTBQvniHe1+3eJKy/hQgYQoCxvyqEFh8GEgg3AfgdYE1kBlG7P2EsfMOGDVK5cmUiIQESCCABvtIGECaTMj8BeOFj/+nq1aubvzIBqgHmAmNRDgYSIIHAEqAAB5YnUzM5AUx5q1q1qsAqw/AvgZo1a8rKlSu5NCUfCBIIMAEKcICBMjlzE1i2bJnUr1/f3JUIcOnTpUsn+fPnl82bNwc4ZSZHAvYmQAG2d/uz9i4EsPgL9qMuV66cy1keggB6wXg5YSABEggcAQpw4FgyJZMTgPdzjRo1aH520461a9dW2zI6HA43V3mKBEggLgQowHGhxnssSQCORujpMUQnkCFDBsmUKZNs3749+kWeIQESiBMBCnCcsPEmqxGA9/PBgwfdLodqtbrGtT7wDF+1alVcb+d9JEACUQhQgKMA4Vd7Eli9ejW9n2NpepjnuSZ7LJB4mQR8IEAB9gEWo1qXAMZ/Ofc35vbNly+f3L17V44fPx5zRF4lARLwigAF2CtMjGRlAnfu3JFdu3apHrCV6xmIuuElBTtFMZAACfhPgALsP0OmYHICa9euFey6lThxYpPXJPjFr1SpEs3QwcfMHGxCgAJsk4ZmNT0TwLgmxjcZYidQvnx5tVQnrAYMJEAC/hGgAPvHj3ebnMCjR49Uj65KlSomr0loip8oUSK1Sxm2C2UgARLwjwAF2D9+vNvkBDD2mz17dkmfPr3JaxK64sMMjf2SGUiABPwjQAH2jx/vNjkBTD/iNnu+NSIEmD1g35gxNgm4I0ABdkeF52xDAA5Y2P2IwXsCuXPnFowBnz592vubGJMESCAaAQpwNCQ8YRcCFy5ckGvXrknhwoXtUuWA1bNChQqycePGgKXHhEjAjgQowHZsddZZEYD3c8WKFUkjDgRgtuf2hHEAx1tIwIUABdgFBg/tRQCORPR+jlublylTRo0Dc3ekuPHjXSQAAhRgPge2JPDgwQPZunWrYF4rg+8E0qZNK5kzZ1Zzgn2/m3eQAAmAAAWYz4EtCezYsUPgTJQyZUpb1j8QldZ7wYFIi2mQgB0JUIDt2Oqss2zYsIHmZz+fA1gPOA7sJ0TebmsCFGBbN799K4/xX5qf/Wv/smXLyp49ewTmfAYSIAHfCVCAfWfGO0xO4MqVK4IpSEWLFjV5TcJb/CRJkigz/s6dO8NbEOZOAiYlQAE2acOx2HEnAPMzxi8jIiLingjvVATKlSsnW7ZsIQ0SIIE4EKAAxwEabzE3ASwggYUkGPwnADM+vMkZSIAEfCdAAfadGe8wOQGsY4z1jBn8J1C8eHFuT+g/RqZgUwIUYJs2vF2rfezYMcGWepjDyuA/AX17QuwqxUACJOAbAQqwb7wY2+QE0Pul93NgG7FUqVLcHSmwSJmaTQhQgG3S0KzmvwQwb5Xjv4F9GjAdiePAgWXK1OxBgAJsj3ZmLTUCjx49UkIBD2iGwBEoUaKEwLSPLQoZSIAEvCdAAfaeFWOanMDevXsle/bskjp1apPXxFjFjx8/vhQpUoS9YGM1C0tjAgIUYBM0EosYGAIwP7P3GxiWUVMBV84HjkqF30kgZgIU4Jj58KqFCEAgMF7JEHgCcGzbtm1b4BNmiiRgYQIUYAs3Lqv2HwGsVwwTNHvA/zEJ5FHhwoXVOPCNGzcCmSzTIgFLE6AAW7p5WTmdANYrxvaDWL+YIfAEMA6M6Uj0hg48W6ZoXQIUYOu2LWvmQgDmZ6xbzBA8ApjeRQEOHl+mbD0ChhZgTBthIIFAEMD4JMd/A0HScxqlS5fmghye8fAKCUQjYFgBPnz4sGB+IQMJ+Evg3r17avy3ZMmS/ibF+2MgULBgQbXNI7Z7ZCABEoidQILYowQ/Rvv27eXEiRORMsKk/qNHj0qNGjUEb9YjRoyIdJ1fSMBbAjt27JBChQqpNaC9vYfxfCeA7R1h5sd0r3r16vmeAO8gAZsRMIQA58yZU/2nHThwoOTKlUs1walTp+Stt96Sd999V9KkSWOzZmF1A0kA5meO/waSqOe0YObHetsUYM+MeIUEdAKGMEF/9tlnMnr0aBk6dKgyFeI/L3q+yZMnl/r163PxfL21+DdOBNAjgxWFIfgE6IgVfMbMwToEDCHAwFm3bl1Zu3atbN++XRo2bKjmFFoHM2sSLgJ3794V+hOEjj6sWRhzP336dOgyZU4kYFIChjBB6+xSpEghY8aMkfnz50uXLl0kceLE+iX+JYE4EcALHcZ/EyZMGKf7eZPvBLAqFszQ2bJl8/1m3kECNiJgmB6wK/NGjRqpdWUnT57seprHJOAzAcxL5epXPmPz6waYoWH2ZyABEoiZgKF6wK5FTZUqlRq3GzZsmOzevVvGjh3retntMUzYnhaEhyMOe9RusVn6JNq9W7dulq6j0SoHh7evv/5aHA6HwDOagQRIwD0BwwqwXtyUKVN67QUNp6306dPrt0b6iyUI48UzZIc/Ujn5JXAEMJXt0KFDUrRo0cAlypRiJZAuXTrJnDmz7N+/X7BGNAMJkIB7AoYX4K5du7ovuZuzWGjB02ILa9askbNnz7q5i6esSgDzfyEAHP8NfQtjHHj9+vUU4NCjZ44mImCYLuH06dOlbdu2akH3AgUKKPPzU089JdOmTTMRThbVSATggMX5v+FpkUqVKikBDk/uzJUEzEHAED3gd955R+bMmSM9evSQnj17qh1rbt26Jfv27ZMhQ4bIwYMHBYt0MJCALwTgCOSLBcWXtBk3ZgKYd43pX/h/nCxZspgj8yoJ2JSAIQT4hx9+kF27dknq1KkjNQPmBjdv3lzgFU0BjoSGX2IhgPFfCEDx4sVjicnLwSCQIEECNRy0YcMGqV27djCyYJokYHoChjBBwzv5zJkzbmFiSUo4YjGQgC8EsP9vsWLFOP7rC7QAx61cuTLN0AFmyuSsRcAQPeD33ntPqlevLo0bN5b8+fNL0qRJlekKXpTLli2TiRMnWos6axN0AjA/Y4N4hvARqFq1qkyaNCl8BWDOJGBwAoboAT/77LOyceNGqVKlily7dk3tgoSxIyxJiZ5MgwYNDI6RxTMaAS7AEf4WyZIli3qZPnDgQPgLwxKQgAEJGKIHDC558+aV3r17GxARi2Q2Ahj/PXbsGPeTNkDDoReMBXKwVzADCZBAZAKG6AFHLhK/kYB/BPT1n+PHj+9fQrzbbwLVqlWTlStX+p0OEyABKxKgAFuxVW1eJyxHin1pGcJPAOPwWADn8uXL4S8MS0ACBiNAATZYg7A4/hNAD5gC7D/HQKSAtaCxKMeKFSsCkRzTIAFLEaAAW6o5WRk472H+L9d/Ns6zUKNGDZqhjdMcLImBCFCADdQYLIr/BOA1j8U3sBAEgzEIYHYDdjS7ffu2MQrEUpCAQQhQgA3SECxGYAhgI3ju/xsYloFKBQvtYGnK1atXBypJpkMCliBAAbZEM7ISOgEswMENGHQaxvlbp04dmqGN0xwsiUEIUIAN0hAshv8Ebty4oZY05fiv/ywDnQLmA2N7wrt37wY6aaZHAqYlQAE2bdOx4FEJYPpRiRIlBJ63DMYikCJFCrU5AxblYCABEviXAAWYT4JlCMD8zOlHxm1O7Iq0ePFi4xaQJSOBEBOgAIcYOLMLHoFt27ZJ+fLlg5cBU/aLQK1atQROcvfu3fMrHd5MAlYhQAG2SkvavB6XLl2SixcvSoECBWxOwrjVT548uTJDc1EO47YRSxZaAhTg0PJmbkEiQPNzkMAGONkmTZoIxuoZSIAERCjAfAosQQA/6jQ/G78psSzl0qVL5fr168YvLEtIAkEmQAEOMmAmHxoC2E+a839Dw9qfXJImTar2/V6+fLk/yfBeErAEAQqwJZrR3pU4ffq0mnqUI0cOe4MwSe3r1asnCxYsMElpWUwSCB4BCnDw2DLlEBFA75fTj0IEOwDZYG3oY8eOyYULFwKQGpMgAfMSoACbt+1Y8v8ngKktHP81z+MQL148wZzg+fPnm6fQLCkJBIEABTgIUJlkaAnAAatChQqhzZS5+UWgcePGsmjRIr/S4M0kYHYCFGCzt6DNy79nzx7JmDGjpEmTxuYkzFV9rNeNdaH3799vroKztCQQQAIU4ADCZFKhJ4DxX/Z+Q889EDliTjCdsQJBkmmYlQAF2Kwtx3IrAjA/V6xYkTRMSKBBgwaycOFCefjwoQlLzyKTgP8EKMD+M2QKYSJw+/ZtgQm6VKlSYSoBs/WHQNasWSVXrlyyYcMGf5LhvSRgWgIUYNM2HQuOzRdKliwpiRIlIgyTEkAvmN7QJm08FttvAhRgvxEygXARWLduHacfhQt+gPKtX7++6gHDmsFAAnYjQAG2W4tbqL4wXXL819wNmixZMrWE6JIlS8xdEZaeBOJAgAIcB2i8JfwEzp07J3fu3JF8+fKFvzAsgV8E6A3tFz7ebGICFGATN56diw7zM6cfWeMJwA5Jhw4d4tKU1mhO1sIHAh4F+MqVKz4kw6gkEFoCMD9Xrlw5tJkyt6AQiB8/vtStW5crYwWFLhM1MgGPAjx27FjJmzevDBo0SHbu3GnkOrBsNiOAeaPc/9dajQ5vaC7KYa02ZW1iJ+BRgPv37y9Tp04Vh8MhTz75pJru8fHHH8uRI0diT5UxSCCIBHbs2KHmj6ZKlSqIuTDpUBLAdDJ4QvP3JZTUmVe4CXgUYBQMY2wQXazXOm7cOFm2bJnkz59fqlWrJuPHj5dHjx6Fu/zM34YEsPwkzc/Wa/g6deqolbGsVzPWiATcE4hRgGHqg1moc+fO0rx5c0GPY86cOfLee+/JhAkT5PXXX3efKs+SQBAJrF27lg5YQeQbrqQbNmwoS5cuDVf2zJcEQk7AowDPnDlTcuTIIW+++ab6sdu7d69MmTJFmjVrJhiv+fXXX+Wvv/4KeYGZob0JXLp0SS5evCjFihWzNwgL1h4+JwkTJpTdu3dbsHasEglEJ5Ag+ql/z6RMmVJmz54dradx4MABKViwoKRIkULefvttT7fzPAkEhcD69evVwg0RERFBSZ+JhpcAesFYlIMvWOFtB+YeGgIee8Aw8x09ejRaKVq3bi03b96UJEmSSNu2baNd5wkSCCYBzP/FvFEGaxLAOPDixYuV86c1a8hakcB/BNz2gN9//331nwBTPbZu3eqMjbnBp0+f5uL3TiI8CCUBeORv2rRJ+vTpE8psmVcICWTPnl0yZMgg8HTnLlchBM+swkLAbQ+4Vq1aqjB3796VW7duOT+pU6dWO5dgnIaBBEJNYNeuXZI5c2ZJkyZNqLNmfiEkgF7wokWLQpgjsyKB8BBw2wOuXbu2mmaEtXYLFy4c9JJh/t8333wjWN8XHtdw/urSpYtyxoDD12effaZM3kEvCDMwNAGsflWlShVDl5GF859AvXr15OWXX5a+ffv6nxhTIAEDE4gmwJhmBNMPxBc9DnxcA5aNe/zxx11P+X38yiuvCLysq1atqhb9wLJ0hQoVkpdeekk+//xzgUn8ww8/9DsfJmBuAvBL6Nmzp7krwdLHSgBWDpihMfxVpkyZWOMzAgmYlUA0AZ44caJygLh+/bpgOcqoAc5XgRZgTGeC0MPEXbNmTfXmi8XZEbDwB1biogBHbQl7fYf/walTp6R48eL2qrhNa4teMBb+oQDb9AGwSbWjCfDkyZOdVX/mmWecx8E8QK86UaJEKgv8h8uUKZMzuyxZssi1a9ec33lgTwJY/Qp7/+JZYbA+AfihwNrRu3dv4ZQz67e3XWvo1gkLMDAZXl8c/cGDB2pBDgjy5s2bA84KY86TJk1S6cL8NHLkSGce+/btk2zZsjm/88CeBLj9oL3aPWvWrJIxY0ZuBGOvZrddbT0KMAQRUwEQvv76a7VEHOZftm/fPuCQsLRl9erVVbqJEyeWsmXLOvO4evWqvPHGG87vPLAnAfSA6YBlr7avUaOGMkPbq9asrZ0IeBRgLPmXL18+uX//vvJQ/uGHH+TVV19VpmKIYiBDnjx5pEiRIm6ThDDXr1/f7TWetAcBWEEw9Sh9+vT2qDBrqQhgOtKKFStIgwQsSyDaGLBeU4y3DRs2TG26gJ6vLpCXL18W9FJDFTp16iQwP2IJzNgCXhJ+++03t9HwI547d26313jS2ASw/CRXvzJ2GwWjdDlz5pRkyZLJnj17pGjRosHIgmmSQFgJeBTgjh07qv05sfC9vvIQvFCfe+65kM7JxY5LFy5c8AoSxLpDhw5u4/br10/Onz/v9hpPGpsAzM94HhnsRwCzItALpgDbr+3tUGOPAhwvXjw1Gd4VApaJ++STT1xPBfQY5u7jx4+rtaax2QPegH1Zjg5lxsddgPcsvSndkTH2OazEButHiRIljF1Qli4oBOCg+e6770q3bt2Ckj4TJYFwEvAowJgHjAUysPgBvKD1kDRpUqdzln7O37///POP6mVPnz5d9a6RBzZ8wFKYmAP87bffSrp06fzNhvebkAC87kuXLu2cpmbCKrDIfhDAOgD37t2TI0eOCLYrZCABKxHwKMDDhw9XS0NiMQ4svqEHTz1M/Xpc/sJ0DBPTiRMnIjnaYGlKeEj36tXLOU0pLunzHvMSwPhv+fLlzVsBltxvAnDGWrlyJQXYb5JMwGgEPArwmTNn1FKQmAoQ7IBe9syZM6OZjzEn+KOPPuIPcLAbwMDpY/wXVhAG+xLATAisFf/888/bFwJrbkkC7gdMtariRw+iiHHZYIeCBQvKqFGj5OHDh5GyQt4jRoyQkiVLRjrPL/YggJdADH9gmhqDfQng/z+GqWARYyABKxHwKMB44LExA8ZesSMSTMT4uC6SESgQmD40fvx4NdcTzjYVKlRQa/6mTZtWeUBCnBnsR4DmZ/u1uacaoxe8fPlyT5d5ngRMScCjCRqrDs2ePTtapYIxBow5xthqDmPA2IQB2xPCCxrnsRwdgz0JwPyM8T8GEsB0JKxT/9RTTxEGCViGgEcBxpQjfLAiFj4FChQIeqUx7QgfBhJwOByybds2GTBgAGGQgJQrV045ZGJXLKyKxkACViDg0QSNXmibNm3Uvpx4+BFOnjwpzZo1s0K9WQeDE8DqR9gVK1WqVAYvKYsXCgIJEiRQq6GtWrUqFNkxDxIICQGPAoxxVyw5iaUnsTk2Qo4cOWTLli2CxREYSCCYBDD/F74ADCSgE8AWhatXr9a/8i8JmJ6ARwHGxHd4QqdOnTpSJWEaxIeBBIJJAOO/uuUlmPkwbfMQqFy5smzdulX5iJin1CwpCXgm4FGAGzZsKJ9++ql648T0IGxm0L9/f7UkYPLkyT2nyCsk4CcBrICG5w0rYDGQgE4AK+RhaVp4xzOQgBUIeBTgxx57TLp37y5dunRRY7/YHQmbMWCfYAYSCCYB7EMND/hQ7roVzPow7cARwMJAnI4UOJ5MKbwEPHpBo1idO3dWn/AWkbnbjcCmTZuCMt/cbhytWF8I8JgxY9SiPdhghYEEzEzAowDjRxDzgM+ePas8UTERvkWLFsKH3szNbY6ywwELG4EwkEBUAvBJwb7e+H3iHtFR6fC72Qi4NUFjD14sgLB7927BuAv24+3Zs6dgMjx2SWIggWARwC5YGOrg/q/BImz+dOkNbf42ZA3+JRBNgDG+MmXKFNm7d6/6++WXX8q4cePUClVYlQq7EzGQQLAIYJpb8eLFaWkJFmALpIuOAH6nOBvDAo1p8ypEE+C5c+cq8x9WwXINcIj5/PPP1QYNrud5TAKBJIBpJpz/G0ii1ksrS5Ysao36Xbt2Wa9yrJGtCEQTYCw7iTEWd6FQoULKHO3uGs+RQCAIYGyP048CQdLaaaAXzFWxrN3GdqidWyesJUuWeBzrffTokR24sI5hIIBV1y5evCjYnpKBBGIiAAEePHiw2rM8pni8RgJGJhBNgLH35t9//y1Hjx51W+4GDRq4Pc+TJOAvAZifsd1lRESEv0nxfosTyJcvn2BnNqzYlzdvXovXltWzKoFoAty7d2/Bh4EEQk0A04/KlCkT6myZn0kJYGrksmXLKMAmbT8WWyTaGDChkEC4CMADGj1gBhLwhgCmSnIc2BtSjGNUAhRgo7aMzcqFsV/sskVzos0a3o/qYq74P//8I6dPn/YjFd5KAuEjQAEOH3vm7EIA5md6P7sA4aFXBGrXri0rV670Ki4jkYDRCFCAjdYiNi0Pzc82bXg/q421oTEOzEACZiRAATZjq1mwzNu3b6cDlgXbNdhVgs/AyZMnBesXMJCA2QhQgM3WYhYs77lz59Qm67ly5bJg7VilYBLAVKQqVarIihUrgpkN0yaBoBCgAAcFKxP1hQDMz+XKlfPlFsYlAScBfW1o5wkekIBJCFCATdJQVi4mBJjzf63cwsGtW+XKlWX//v1y48aN4GbE1EkgwAQowAEGyuR8JwAPaM7/9Z0b7/iXQMKECZUFhd7QfCLMRoACbLYWs1h5MYczfvz4EnX3LYtVk9UJMgFORwoyYCYfFAIU4KBgZaLeEuD8X29JMV5MBOCIhaGM27dvxxSN10jAUAQowIZqDvsVhvN/7dfmwahxsmTJpFSpUrJ+/fpgJM80SSAoBCjAQcHKRL0lsG3bNilfvry30RmPBDwSgDc0F+XwiIcXDEiAAmzARrFLkY4dOyaJEiWSjBkz2qXKrGcQCWBVrI0bN8rDhw+DmAuTJoHAEaAAB44lU/KRAL2ffQTG6DESSJUqleTPn59m6Bgp8aKRCFCAjdQaNisLzM8VKlSwWa1Z3WASqFWrlixfvjyYWTBtEggYAQpwwFAyIV8IOBwO2bRpExfg8AUa48ZKAAK8Zs0aefToUaxxGYEEwk2AAhzuFrBp/gcOHJAMGTJI2rRpbUqA1Q4GgfTp00u2bNnUlKRgpM80SSCQBCjAgaTJtLwmgOlH9H72Ghcj+kCgTp063JzBB16MGj4CFODwsbd1zjA/c/lJWz8CQas8pyMFDS0TDjABCnCAgTK52Algmsju3bspwLGjYow4EMiaNavAFL1r16443M1bSCB0BCjAoWPNnP6fwM6dOwV7/2L1IgYSCAYBOGMtXbo0GEkzTRIIGAEKcMBQMiFvCcD8zPFfb2kxXlwIYHOGFStWxOVW3kMCISNgeAG+deuWHDx4MGRAmFHwCcABi/N/g8/Zzjnkzp1brbK2b98+O2Ng3Q1OwBACfPz4cWnVqpWUKFFCunbtKidOnHBi27Fjh3Tu3Nn5nQfmJnDz5k31QlW8eHFzV4SlNzwBekMbvolsX0BDCHCXLl3UEnLDhw9XY4NVq1aVxYsXq8bhhHprPaPo/ZYsWVISJEhgrYqxNoYjgHHgJUuWGK5cLBAJ6AQM8SuIXu5ff/0l8eLFE7y1PvHEE/L444/LZ599Jjly5NDLyr8WIID1nytVqmSBmrAKRidQoEABiYiIkMOHD0u+fPmMXlyWz4YEDNEDhsju3bvXiR+maLy5vvPOOzJ+/HjneR6YnwD2ay1Xrpz5K8IamIIAe8GmaCbbFtIQAtyvXz+ZMGFCpEbANBVMI4DHLIM1CJw7d05u377N3og1mtMUtYA3NPcINkVT2bKQhjBBP/3009KuXbtoDYC1gvGf58KFC9Gu8YT5CKxbt04qVqxovoKzxKYlULhwYXnw4IEcOXJE8ubNa9p6sODWJGAIAQZajNW4CzBNnz9/Xjlnubvueu7evXuCj7uA83TockcmdOewWTpMggwkEEoC+qIcFOBQUmde3hAwjAB7KuyXX34pa9eulf3793uK4jw/efJkmTZtmvO764G++pLrOR6HjgBefrD/L4YbGEgglATq1asnH374oXTq1CmU2TIvEoiVgOEFeNy4ccqEFGtNtAgdO3ZUH3dx+/btK2fPnnV3iedCQADr8mbJkkVSp04dgtyYBQn8R6BQoUJy9+5dekP/h4RHBiFgGAHGileLFi1S3tA4Tp48ueA/Tv369dWxQXixGHEkAPNz5cqV43g3byMB/wjgdwROnZyO5B9H3h1YAobwgl6+fLlaiOOLL75QAgynK+yW8+mnn6r/MGvWrAlsrZlayAnAAYvrP4ccOzP8fwJ169ZVL/gEQgJGImCIHnCvXr1kypQpgn08o4bZs2fLgAEDZOXKlVEv8btJCFy9elVOnjyplho1SZFZTIsRwKIc8ePHV74ksKwxkIARCBiiBwwvZ0/TU6pUqcJpSEZ4Uvwog777EX4AGUggXATgjLVw4cJwZc98SSAaAUMIcLNmzaR169ZqOcoDBw6ozRiwiwl6v23btpWnnnoqWsF5wjwEMITA5SfN015WLWmDBg3UOLBV68d6mY+AIUzQo0aNkmHDhsngwYOVpyJWS0qRIoUULVpUnnvuObVDkvnQssQ6AThgde/eXf/KvyQQFgJY8jZt2rSCDUHKli0bljIwUxJwJWAIAU6cOLH873//Ux/XwvHY/AT27NmjfvTSp09v/sqwBqYn0LBhQ+WMRQE2fVNaogKGMEFbgiQr4ZYAvJ85/cgtGp4MAwGMA2N5W08r5oWhSMzSxgQowDZu/FBUfcOGDQJHOgYSMAKBdOnSSfHixYVTG43QGiwDBZjPQNAIYPrRsWPHOP0oaISZcFwIwBkL+48zkEC4CVCAw90CFs4fvd8KFSqo+ZcWriarZjIC2JwBC/1cvnzZZCVnca1GgAJstRY1UH04/chAjcGiOAkkSpRIatSoIfPnz3ee4wEJhIMABTgc1G2QJ3Y/4vivDRrapFVs2rSp/P333yYtPYttFQIUYKu0pMHqARNf1qxZBU4vDCRgNAIlSpRQntCYJsdAAuEiQAEOF3mL5wvzM6cfWbyRTV69Fi1ayLx580xeCxbfzAQowGZuPQOXff369VK1alUDl5BFszuBJk2aqKUpOSfY7k9C+OpPAQ4fe8vmjO0k8SlSpIhl68iKmZ9AmjRppEyZMrJgwQLzV4Y1MCUBCrApm83YhUbvF4tvREREGLugLJ3tCbRs2VLmzp1rew4EEB4CFODwcLd0rqtXr+buR5ZuYetUDvPUL126JNiFjYEEQk2AAhxq4hbPD+NpW7dupQBbvJ2tUj1YaR577DG19alV6sR6mIcABdg8bWWKkm7btk0KFy6stpM0RYFZSNsTwH7kixcvllu3btmeBQGElgAFOLS8LZ8bzc+Wb2LLVRDOWPDY58Iclmtaw1eIAmz4JjJXAbH9YLVq1cxVaJbW9gRghp45c6btORBAaAlQgEPL29K5HT16VLAEZa5cuSxdT1bOegRKliwpCRIkkM2bN1uvcqyRYQlQgA3bNOYrGFa/4uIb5ms3lvhfAk888YT88ccfxEECISNAAQ4ZautnhPm/ND9bv52tWsOGDRsqD/7z589btYqsl8EIUIAN1iBmLc6NGzdk3759amUhs9aB5bY3gcSJEwuWp5w1a5a9QbD2ISNAAQ4ZamtnhN5v+fLlJWHChNauKGtnaQIwQ//555/y4MEDS9eTlTMGAQqwMdrB9KVYtWqVWn7S9BVhBWxNIEuWLGoe+8KFC23NgZUPDQEKcGg4WzoXeD5v3LhRqlevbul6snL2IPDkk0/KjBkz7FFZ1jKsBCjAYcVvjcy3b98u2bJlEyxowEACZidQsWJFuXnzpuzatcvsVWH5DU6AAmzwBjJD8bD6Fb2fzdBSLKO3BFq3bs1esLewGC/OBCjAcUbHG3UCa9eupflZh8G/liDQtGlT2bBhg9opyRIVYiUMSYACbMhmMU+hjh8/Lnfu3JF8+fKZp9AsKQnEQiBJkiRSv359Lk8ZCyde9o8ABdg/fra/G6tf1axZ0/YcCMB6BNq0aSNz586Vhw8fWq9yrJEhCFCADdEM5i3EypUrufykeZuPJY+BABwL8+fPL4sWLYohFi+RQNwJUIDjzs72d167dk1ggi5btqztWRCANQmgF8xdkqzZtkaoFQXYCK1g0jKg94vVr+LHj2/SGrDYJBAzgUqVKgmWWd29e3fMEXmVBOJAgAIcB2i85V8CWP2Ki2/wabA6gf6S+hYAACVfSURBVFatWsm0adOsXk3WLwwEKMBhgG6FLO/fvy+nT5+WunXrWqE6rAMJeCTQrFkztU/wxYsXPcbhBRKICwEKcFyo8R6B93PGjBklXjw+QnwcrE0AU5IaNGjAhTms3cxhqR1/PcOC3fyZYvUrTj8yfzuyBt4RgDPWvHnz5N69e97dwFgk4AUBCrAXkBglMgHMi+T4b2Qm/GZtAtglqUSJEjJ//nxrV5S1CykBCnBIcVsjsy1btqiVr9KlS2eNCrEWJOAFAawPPWXKFC9iMgoJeEeAAuwdJ8ZyIbB06VKpUaOGyxkekoD1CZQuXVqSJUsm69evt35lWcOQEKAAhwSzdTJxOBzKAat27drWqRRrQgJeEmjbti17wV6yYrTYCVCAY2fEGC4EYH7OlCmT+ric5iEJ2IIAXjyx+tuBAwdsUV9WMrgEEgQ3ee9Sf/HFF2X//v0eI8P0M3z4cI/XeSF0BJYvXy516tQJXYbMiQQMRACrvum94DfffNNAJWNRzEjAEAKcPXt2OXTokLz99tvy6NGjaBzTpk0b7RxPhJ4A2gbjv6NHjw595syRBAxCoGXLljJhwgQ5e/aswDuagQTiSsAQAjxkyBBp3ry5HDt2TDp16hTXuvC+IBOA+Rk7xPBHJ8igmbyhCSROnFgee+wxmTRpkrz22muGLisLZ2wChhgDxmpKeJifeOIJY9OyeenQ+61Xr57NKbD6JCCChTmWLVsmly9fJg4SiDMBQwgwSg8zc5o0aeJcEd4YXAIPHjwQjv8GlzFTNw+BVKlSSZMmTeTXX381T6FZUsMRMIQJOiYqo0aNkr1798rXX38dUzR1DSbSnTt3uo23Z88ebpvnlox3Jzdu3KgW38iQIYN3NzAWCVicQLt27eSFF16Q9u3bs/Ng8bYOVvUM0wP2VEHsunP37l1PlyOdxxKJiO/ug2sMcSeAJfhofo47P95pPQJYCa5Ro0bsBVuvaUNWI8P3gHv37u01jAoVKgg+7sKOHTuU16K7azwXM4E7d+7Ihg0b6HASMyZetSGBZ599VvWCMTWJS7Pa8AHws8qG6QGjh9W5c2epVq2alClTRm30Do/ov/76y88q8nZ/CcD5CnOxU6RI4W9SvJ8ELEUAvivYL/jnn3+2VL1YmdAQMIQAf/rpp9K3b18pVaqUYHL70KFDZeDAgVK8eHHp1auXV+O/ocFlz1wWLlwojRs3tmflWWsSiIUAesF4ST1z5kwsMXmZBCITMIQJGqtcbd26VdKnTx+pdJjwjh1IWrRoIX369Il0jV9CQ+DChQtq2b2qVauGJkPmQgImIwDLEEzQ33//vVpMyGTFZ3HDSMAQPeAECRLI1atX3WK4cuWKJE2a1O01ngw+Ad35Cm3EQAIk4J7AU089Jbt27VIzNtzH4FkSiE7AEL+qMDtXqVJFWrVqJfnz51eCe+vWLbU+9Ny5c+XHH3+MXnKeCQkBjMEPHjw4JHkxExIwK4GECRNKly5d1Jr1XLferK0Y+nIbogeMBxdjKHny5FGiu3btWjl8+LCULFlSMP8Uy1QyhJ7A9u3bBT3fwoULhz5z5kgCJiPQsGFDtZY9rEYMJOANAUP0gFHQYsWKqY83hWac0BCYN28eX35Cg5q5WITAq6++qhxJa9asyaEzi7RpMKthiB5wMCvItONG4ObNm7JixQrBWz0DCZCAdwQKFSqkplCOHTvWuxsYy9YEKMC2bn7PlV+0aJFUrlxZUqdO7TkSr5AACUQj0LVrV7Vu+r59+6Jd4wkScCVAAXalwWMngVmzZqnpX84TPCABEvCKQPLkydX6BVjfgEvgeoXMtpEowLZtes8Vx8YVWE+7bNmyniPxCgmQgEcCtWvXluzZs3OFLI+EeAEEKMB8DqIRmDlzJnu/0ajwBAn4RqB///4yZ84coSnaN252ik0BtlNre1HXa9euyZo1a6Rp06ZexGYUEiABTwTgP4Eldj/88EOvd3TzlBbPW5MABdia7RrnWmHqEaZQcOOFOCPkjSTgJFCjRg21nsG3337rPMcDEtAJUIB1EvyrFhH4/fff5YknniANEiCBABHAlqrYDnXJkiUBSpHJWIUABdgqLRmAeqxatUo5jhQoUCAAqTEJEiABEEicOLEMGTJE0As+ceIEoZCAkwAF2ImCB1OnTlW7T5EECZBAYAnkzZtXunfvLm+99ZbcuXMnsIkzNdMSoACbtukCW3BMPcLWg9WrVw9swkyNBEhAEWjUqJGa2vfRRx+RCAkoAhRgPgiKAHq/7dq1k4iICBIhARIIEoFXXnlFbb06fvz4IOXAZM1EgAJsptYKUllPnTolmzZt4tSjIPFlsiSgE4gXL56alrRgwQI6ZelQbPyXAmzjxter/ttvv6mxXziLMJAACQSXAKb4ffzxx8opa+fOncHNjKkbmgAF2NDNE/zCXbp0SZYtW8apR8FHzRxIwEkgV65c8vbbbyunLHpGO7HY7oACbLsmj1zhiRMnKtNzypQpI1/gNxIggaASKFOmjGCO8IABA+TixYtBzYuJG5MABdiY7RKSUqH3i20H27dvH5L8mAkJkEBkAnXr1lXOj6+99prcuHEj8kV+szwBCrDlm9hzBdH7bdasGff89YyIV0gg6ASw8ly9evWkX79+cvv27aDnxwyMQ4ACbJy2CGlJzp07x95vSIkzMxLwTOCFF16QKlWqyE8//aS2AvUck1esRIACbKXW9KEuY8eOlSeffFI49usDNEYlgSAS6NSpk9y6dUsmT56s/gYxKyZtEAIUYIM0RCiLceDAAdm2bZsaewplvsyLBEggZgIYCz5//rxwTDhmTla5SgG2Skv6UI/hw4cL3rY579cHaIxKAiEi0L9/fylVqpQyR9+9ezdEuTKbcBCgAIeDehjzXLx4sdocvGnTpmEsBbMmARKIicDLL7+s/p/++uuvdMyKCZTJr1GATd6AvhT/5s2bMmrUKOnbt68vtzEuCZBAGAjADH358mWBvwbGhhmsR4ACbL029Vij0aNHS61ataRw4cIe4/ACCZCAcQi8+uqrkjBhQsHf69evG6dgLElACFCAA4LR+Ils2LBBNm/eLC+99JLxC8sSkgAJOAlgH2FsEwqzNBbPYbAOAQqwddrSY03w5vzZZ5/JwIEDJVGiRB7j8QIJkIAxCXTs2FEef/xx6dWrl2D3MgZrEKAAW6MdY6zF+++/r9Z7LlmyZIzxeJEESMC4BNq0aSNYsAPrR+/fv9+4BWXJvCZAAfYalTkj/vjjj/Lw4UM17cicNWCpSYAEdAKNGjWS119/Xd544w1Zt26dfpp/TUqAAmzShvOm2Nj0G5stvPPOOxIREeHNLYxDAiRgcAKVK1eWTz/9VL744guZPn26wUvL4sVEgAIcEx0TX1u1apWMGTNG/UdNlSqViWvCopMACUQlUKhQIRk5cqT8/fffaulKWLkYzEeAAmy+Nou1xN98840MGjRIPvnkE8mRI0es8RmBBEjAfAQyZswoWNXuyJEj0q1bNzpnma8JhQJswkaLqchYOQemqVmzZkmBAgViisprJEACJieAOcJ42cbGKvCQnjt3rslrZK/iU4At1N5YMWfOnDmyevVq7vFroXZlVUggNgLNmzeXb7/9VpmksdLdsWPHYruF1w1AgAJsgEbwtwhnz56VV155RQ4fPixY7YpmZ3+J8n4SMB8B/L/H8FO9evXUcrNfffWVWsrSfDWxT4kpwCZva3hBYvynbt268tFHH0mKFClMXiMWnwRIwB8CLVu2lF9++UWSJk2q5g1jnBhbHDIYj0AC4xWJJfKGwNatW2XEiBGSJk0a+e677yRLlize3MY4JEACNiCQPHlytexsu3btZOrUqeq4TJkycu7cORvU3jxVpACbp61USQ8dOqR2Rzl58qT06NFDqlWrZrIasLgkQAKhIoAX9K5du0qHDh0E6wL8/vvvMmTIEHnmmWekYcOGkjlz5lAVhfm4IWAIAcYcNoxdwoN37969austvMFhrtvTTz+tFiGPF8/e1nJMNRg/frzs2rVL/Wdq0aKF2J2Jm+eZp0iABNwQSJw4seA3Y8WKFVKxYkU1ZQmbPOTMmVPq1KkjNWvWFExrYggtAUMIMLz2ILzYLABTZzB2gf0v9+3bp6bUoLeHOa12DGAwYcIE2bNnj3prffvttyV+/Ph2RME6kwAJBIBArly5BOtKY03pTZs2ybJly9RvTLp06aRKlSpSoUIFKV68uNoGMQDZMYkYCBhCgDFGAQ9eCK9rKFKkiHogsIet3QR448aNMmnSJDlz5owSXpiNKLyuTwePSYAE/CGA5Wkhtvj069dPdu/eLevXr1eWNmz2gM4QhLhYsWLqkz59en+y471uCBhCgDFOsW3bNiW2Ucu4ZcsWyZQpU9TTlvx+584dNU4zbdo0tW1g+/btlXcz13G2ZHOzUiRgKAK60KJQ9+/fV4K8Y8cOmT9/vnz99dficDjUsCCGBvEpWLCgbX6bg9VQhhDgoUOHSrNmzQReevnz55ckSZIoEzTewjD2OWPGjGDV3xDpwqMZDhIYnylfvrx6Gy1VqpQhysZCkAAJ2I8AVtgqXbq0+ui1v3TpkhoWxO/yvHnzBPOM7969q8QYveW8efOqXjPGlbnvuE4t5r+GEGCIL0zQeNPC39u3bwvGKZ544glp0KCBEuSYq2GuqxcvXpTt27cLzMxYtSpr1qzKI/Gll14SbpxgrrZkaUnALgQwRly1alX10et8/fp1tTcxfrc3b94sU6ZMkdOnT0uGDBkkd+7c6nccC4Tgkz17dqEZWyf3719DCDCKAjN027ZtI5dO+/bo0SN58OCBJEhgmKJGK2PUE/fu3ZMbN26oVWjw1oiVqjCWe/ToUTl48KCqU8mSJaVcuXJqn156H0YlyO8kQAJmIJAyZUpltYPlzjWcOnVKLYd54sQJNXMDnSucg2BjzQJMf8LQIoQaogxxhwagA4IPrKB26EUbXtVefPFFWbt2rfICdm1gd8fwFva0PyZ6nDBvQxx9DXh4YHLBGAgCXgpwjOlT+IvxErwk4C+EFy7/WJEqderU6sHCA4aHDuu1wkyDh841xKVMrvfzmARIgAS8IYDfLPxOBfs3B50KfODg5RqQN1blwgeWQHzQKbly5Ypcu3ZN/X7iL/xh8JuK31IIMX4z8dsKR1RMv8TH1TcG37EiYNGiRV2z8+o4nNM5DS/AWGwCO314E1q3bq1Ezl1ceFrfvHkzTm9V1atXV96Aerp64+MvHgj0zvGQYNwkWbJkejT+JQESIAFDEcDvFX6nwtW7RL7ohODjS8CLw4ULF1TnB8d6JwgdIAgxzmXLli1O9UJa4QqGEmDAxJsRhBI9SLxBVapUyWs2mMYUdSqTfjN6o3EFnTZtWsGHgQRIgARIIPQE8OJgxeV2DbG81NWrVwWrsmAcAM5XZcuWVYP2+A7HJJgkGEiABEiABEjASgQMIcAY54UZF9Nx4NYOQcZfeAnjL7baYyABEiABEiABKxEwhAl6+fLlyksYZgY9wK4PpynMNYs6kK/H4V8SIAESIAESMCsBQ/SA8+TJo/avxBiwa8CY7U8//SRYkpKBBEiABEiABKxEwBA94LFjx6pdj15//XXJly+fcyUsbL2H5dEmTpxoJeasCwmQAAmQAAmIIQQYyy5iIXDs+APRxUpY8IKG+GI1FQYSIAESIAESsBoBQwiwDhWTqOMykVq/n39JgARIgARIwCwEDDEGbBZYLCcJkAAJkAAJBIoABThQJJkOCZAACZAACfhAgALsAyxGJQESIAESIIFAEYjQpv5EnvsTqJQNlg4W+cBmCFhlK5Bh6dKlasHwQKZph7Sw3Gjy5MntUNWA1VFfSB87xTB4TwCL+WDddqyBbPeATRiw6JE3GxDcunVLqlWr5nF5X6uwxFaK2I8d2yWGOthGgIMFtk6dOgIRZvCeAN756tevL4sXL/b+JsZUe0fPnTtXPvzwQ9LwgcDQoUOlcOHC0qJFCx/uYtQ2bdrImDFjuIdvEB8FmqCDCJdJkwAJkAAJkIAnAhRgT2R4ngRIgARIgASCSIACHES4TJoESIAESIAEPBGgAHsiw/MkQAIkQAIkEEQCFOAgwmXSJEACJEACJOCJAAXYExmeJwESIAESIIEgEuA0JD/hnjlzRrJmzepnKva7ndx8b3PMZ8XczLRp0/p+s43vuHr1qpoDnCxZMhtT8L3q58+flwwZMng1Z9j31HkHCFCA+RyQAAmQAAmQQBgI0AQdBujMkgRIgARIgAQowHwGSIAESIAESCAMBCjAYYDOLEmABEiABEiAAsxngARIgARIgATCQIACHAbozJIESIAESIAEKMB8BkiABEiABEggDAQowGGAzixJgARIgARIgAIch2fg8uXLgs3RXcOdO3dky5Ytcvr0adfTPHYhcPHiRZdv/x5iQY5NmzYJmDJEJ3D//n3BQhKewvHjx+X27dueLtv2PBYswcc14BnD4hL6B2wZIhO4du2auOPy6NEj2bVrl2zbti3ab1/kFPjNFwIUYB9oYSP5Dz74QK0Oc/jwYeedP//8sxQsWFD+97//SeXKlWXgwIHOazwQuXnzpjz99NNSpEgRJ44HDx5Is2bNpG7duvLWW29JoUKFZOTIkc7rPBA5evSoVKlSRZ599lm3OBYuXCh58uSRVatWub1u15PLly+X/Pnzy5dffhkJAf5vlilTRipUqKA+eGFm+I/Ad999J5kyZZIVK1b8d1I7Ak/8vvXu3VsGDx4s7l6kI93AL14TSOB1TEaU9u3bS5IkSSRnzpyRaJQtW1Y2btwomTNnlnPnzkmuXLmkX79+kjFjxkjx7PgFvRCIyOOPPy4QDD0kSJBAPvnkEylVqpQ6BRGBIL/88st6FFv/RW+jZcuW6nPo0KFoLNBTefXVV5WQRLto4xNTpkyRd999V2rXrh2NAoQDL85p0qSJds3uJwYMGKB6uPr/R50HnrO2bdvK9OnTpVq1avpp/g0QAfaAfQD5/vvvy/jx4wXmGNdQsmRJJb44B9GFuGDdXgYRrL/7+++/uxVW1//siRIlUnHJ7F8ChQsXVi8stWrVcovktddek169ekny5MndXrfryTp16iiLQI4cOSIhgMUFw0QU30hYnF+6d+8us2fPVmtmO09qBzNmzFBWPYqvK5XAHVOAfWAJM0xsYebMmVK8eHGJ+gMQ231Wvg7zMsz37sK0adPk888/lzfeeEN+/PFHd1FseQ4vcfny5XPLbd68eYKxX/xoQlgY/iMAE6o7kf3nn3/Ui3O9evUEn48//tjtWOd/KdnrCCb7+PHjR6s0/DPQqWjatKkaQuratWu0sfVoN/GE1wQowF6jij0izFvomWAshcE7AjDZwzQIqwLH5GJnduXKFTW88cMPP8QemTGcBDA8hGGOMWPGKD8OvCjTV8OJx+MBHNfWr1+vftPggHX9+nX58MMPPcbnBd8IcAzYN14eY0NIMIb5zTffKEcPjxF5IRKBnj17qu/w5MXY+WOPPSYlSpSIFIdf/iMwaNAgxQnmQgR4kc+aNUtSpUollSpV+i8ij6IRKF++vDoHS9Z7770nffr0kaFDh0aLxxP/EUiaNKl06NDB6ffy3HPPybBhw/6LwCO/CFCA/cL3781wVID44scRAsLgO4HEiRMrE1jUqSO+p2TtO+DFC0fAgwcPqorixeXUqVMCEyuD9wQwtStdunTe32DTmOXKlZN169Y5a3/hwgXuR+2k4f8BBdh/hmqayL1799S0EbxZI8AZxJMDTQCyNH0SR44cUWO/NWvWVE5rv/32m+rZYYoIg2cC3bp1i3Rxw4YN0qNHD2nQoEGk8/wSmcCyZctk/vz5yjqFl5WPPvpIvvjii8iR+C0agWeeeUb9P4XZOXfu3MrDfPTo0dHi8UTcCMQfooW43WrfuxImTKim1sA8g4AecIECBSRevHjOD+ZnwqTK8C+BiIgISZEihVStWlWdgPcu3qbh5AExhpclTFvwhmb4jwCeqSxZsijHvv/O/neEZxHT4Nw5Hv0Xy35HcCjC/0n9/2Dq1Kll//79snnzZuVE9Pbbb0vjxo3tByaWGkd9nmCZwjQk9ILx/xRrHfBlLxaIPlyO0LxT3bun+pAIo5IACZAACZAACfhGgF7QvvFibBIgARIgARIICAEKcEAwMhESIAESIAES8I0ABdg3XoxNAiRAAiRAAgEhQAEOCEYmQgIkQAIkQAK+EaAA+8aLsUmABEiABEggIAQowAHByERIgARIgARIwDcCFGDfeDE2CZAACZAACQSEAAU4IBiZCAmQAAmQAAn4RoAC7BsvxiYBEiABEiCBgBCgAAcEIxMhARIgARIgAd8IUIB948XYJEACJEACJBAQAhTggGBkIiRAAiRAAiTgGwEKsG+8GJsESIAESIAEAkKAAhwQjEyEBEiABEiABHwjQAH2jRdjkwAJkAAJkEBACFCAA4KRiZAACZAACZCAbwQowL7xYmwSIAESIAESCAgBCnBAMDIREiABEiABEvCNQPwhWvDtFsYmgeATOH36tKxZs0by588fKbP58+dL7ty5JV48/94dV65cKSlTppSkSZM60798+bJMmTJFdu3apfJNlCiR89qWLVvkjz/+kNu3b6v89Qv379+XmTNnqrJmyZJFpalfO3HihEydOlWOHz8uBQsW9KnMp06dklWrVsnBgwfV/ShLqlSp9KQD8vfll1+Whw8fSuHChb1KD2xu3LghOXLkiBb/7bfflqNHj0rZsmWjXfv555/lt99+k/r160e7FtcThw8flmnTpsmZM2ekQIECEhER4TYpvd3u3LkjuXLlcsa5deuWzJgxQzZt2iTZs2eXZMmSqWvgsXjxYlm4cKGcP39e8uXL52y3CxcuyKxZs2T9+vUqfoYMGZzp/fPPP6o8yC9r1qySIkUK5zUekIAnAv79inlKledJwE8CS5YskcaNG8vEiRMjpdSqVSvBj2dcAwS0c+fOUqdOHdm9e7czmStXrkj58uUFwgyRr1atmuBHGwHCg3whqL169ZLPP//ceV/Lli3lhx9+UKJdsWJFOXTokLqGtPF9z549MmbMGGnbtq3zHm8OIAAvvviiTJgwQUaOHCllypSRN99805tbvY6D+ly9etXr+Pv27VMM3N2AFyaIEAJEevTo0c5oly5dElwPVIBoVq1aVQ4cOCBffvmlak93af/666/OdsPLxtChQ1U0PD+VKlVSYrpx40b10gBBR2jatKl8++23cvLkSXn//fcF7Yuwd+9eKVeunKxevVrAoWbNmkpwce3s2bOqfVCuHTt2qOMjR47gEgMJxEzAwUACBiTwyy+/OCpUqODIli2b49q1a84Saj1WhyYazu++Hjz77LOOd955x1G5cmWHJrbO24cNG+Z46qmnnN+bN2/uGDdunPperFgxx99//62OtR9mR+rUqR3aj7hD+zF25MmTx/HgwQN1TfvBdrz00kvqWBN5x7vvvquOcV3rSTk2bNigvnvzz/jx4x3aC4gzqiYUDq0X7Hj06JE6p71IOLQesmPFihWqLHpElE97cXDs37/fsXTpUsf169f1S+rvuXPnHFoPz6GJpQN11Hqn6lgTSGc8TZjVOf2E1gtXaSKOZiXQTzs0YVVpaQLkQH21FxN1bcSIEY4WLVo4NNFy3Lx50wG2HTt2dGg9ShXfNS9nYj4coJ004VV33L1716H1OFVeUZMoUqSIY8GCBeo06pQmTRoHuP30008OTWid0VE+lB9Be3lwnsexZmlx6OV1vaaJtKNZs2YqLhiivnp45plnHMOHD9e/8i8JeCSQIGZ55lUSCB+BUqVKifYjKpqQyRdffOGxIOjV9u/fP9r1Tp06iSbikc6PHTtWkiRJInPmzIl0ft26dVKrVi3nubp164ombtKmTRvVi9WvwVwJE+PWrVsF91SvXl3ix4+v7sM93bp1U8e49vzzz6tjXMf9SC9qeZwZxnIA02jGjBmVqVUTYdUDhOlVEzjV+962bZtoAiPdu3cX7X+7Sg3xtm/frj7p0qUTTZBVT7xhw4Zy7NgxQc8UASbXr776SvX+8b1GjRpSu3Zt0YRKpV+8eHHRhFsGDBggVapUUVYA5NeoUSNlSYAZWBN6KVq0qOD8jz/+KBcvXlTtNmjQIFUelKN169bKfP3nn3+q3qfOFHkifP/994rrv9/+/Rfm8d69e7ueUtwHDhyozsE0D2sFLBeupnTtpc3ZU0VEmM0zZcqkyochATxXeqhXr56MGjVKfU2ePLl+WpnUtZctSZs2bbRrsHTowyPaC5rs3LlT8URbo1fdo0cPZzo8IAFPBCjAnsjwfNgJwAQMs2GJEiWUORY/8O5CwoQJReuBRLsEsYwaIL7uAsb3XMf0IHYwLeI8xgdd73O95u4epO8pPXd5uzsHEYWYtW/fXu7du6fGOjFmiYDxb5g79XFwmFMh7jCXYiwU5YMIIkCcYFJ/+umnBeO0H330kXTp0kWZ8TGWjgCx7dChg8oHJnPwhqAhYBy+dOnSAiFyDR988IESRpjF8XKgtw3ivvDCC8pUO2nSJHWL1gtVAo2x0wQJEihTOoYWogowXk6iji+nT5/eNVt17I4tXgJcA+JgHDZx4sTO03q7FSpUSLReryo3BHP27NmqfHpErfeqxnlhdp47d66z7TFM8b///U9gXka5dDM7hi4guBB1PIvgjJcYBhKIjQAFODZCvB42AhAhODbhR++VV15RjjHuCgPRQa80anAVzajXon735MTj6Tzuj+u1qHl7+g4x6tq1qxJGOICh16eZwpWIQZzRi0dPEy8K+ng1yoTxUT1oJnw1JovvmzdvFlgAEPBSgXFlBAgTetNwIEJvGM5SsCogDwgxeohRA9LSe6YQMbwEeAooE8bDIb4IKBMctqIG9DSjMoWjnLug9/LdXcO5qOm4xkNPHM5UEGLkCUuLaz6oC8q4aNEiZVnBCwx6xnieNLO9Yg7xBcu+ffsqJzm88EB4Ma7+zTffqHhRXyZcy8BjEgABCjCfA8MTwA89HJ3g9epOVCE+7pz5X3vttWi9LE+VhdCj16QHHGfOnFmZLSFG+Oge0/o13IMeoh708/juLr2SJUvqUWP9CwFBL0sXvyZNmgjMyNo4sjILo5f59ddfK7MrTJ6uwdV72/U8eqoQSz3ogojvMJ+vXbtW5s2bp3p2YArhgSf266+/rt/i/BtTWs5ILgeeyuQSRXlKu/LENYjje++95xpNvWzhxUPvwcNbGQ5SrgFtB/M86qE/M3r7gAGc2+CMhWM4a+kOZEgDAozPk08+qZyt0EOGBQHp6JYWmOLRY4cAf/bZZ6KNSyvTPO7Hy43mZ6CeWXxnIAFPBCjAnsjwvGEIwKwHz1R4L+tmV9fCoXeCqUD+BJhD0btETxsBvR94PuPHFD/uGD+Fhyy8Y9HLgakV+WJsWnOyUr073IMfZgSkh3sgbBCrZcuWqd6suhiHfzAGC0FBjw09X80BSI2pYpw3qvnVU/L6GC3GLnGfqxc4yompQkgLY5p44UCPDnHcmVMRBz1kmLgRMAYK0zUC2gh19jW88cYbXt0CD2SwhelXc8JSLwnwE0DA2DZ6nmgb9PDBHd70GPfVnKmUoOuZoG3Rdt99951TPPVr+Is6QLTdTSmC1UE/D69v/WUA92FYAqZ8BhKIjQAFODZCvB4WAlFNiDCLap7Lal5tMAr03HPPKZFHzxJze/FDjnMIGO+Ec1W7du0EDkQwBeMHHiIMAYIYYvwP05UgDAj9+vVTvSfMLcZYIgRLFysVwYt/4LikeW2r8sAUjLFb5APHL5QTY9wYC9adhGJLEqZ8WBNgPoZguppdMQ6MNPXpUnjpgPhi/rLe83dNHw5Z6BWCE8ZEwUwPmAv81ltvqV4geoaBDuDfoEED9TKEuuDlAeIPIcS83fHjxysHOLSb5pWu6oSxXNQfoouAMWCM6eI8LBOoC8Qcoo5nDeKKFyowhvUBDmIYBkCbw+EMbf3xxx+rtDBMALM2RBmmcbzIwFrDQAKxEYjQHph/XSZji8nrJBBCAuiJofcBE6QeYHaE4OAH0tV8ql/35S9MnegRwnNYD+hhoncJ8cdYH0RWD5j7iTmgGDeEAOoBPcm//vpLjQtCFDB2qAeYRtGrRh7oPftSZvTWMKcUAeVBr1X3usU5lB/Cjt4p0oWwwOyKcVz81csBb230yPTxSPRaIVp4GUBPFXXU46K3mDdvXueCFTB34zpeHhDwQoC66Ata6Ewg1jiPn5I8efKouHAKw9xavJzA1AvR0nvI6I1C/FzbVt3kwz/gA+cueDajh6tbRiC+yBPnEcAIrKK2G8aAYXbGy4I+Fo742vQtZeaHGOPlA3zBHz3l5cuXq7FrmKJxXueA+1AezF1HwPOJIQgGEoiNAAU4NkK8TgIkQAIkQAJBIMCVsIIAlUmSAAmQAAmQQGwEKMCxEeJ1EiABEiABEggCAQpwEKAySRIgARIgARKIjQAFODZCvE4CJEACJEACQSBAAQ4CVCZJAiRAAiRAArERoADHRojXSYAESIAESCAIBCjAQYDKJEmABEiABEggNgIU4NgI8ToJkAAJkAAJBIEABTgIUJkkCZAACZAACcRGgAIcGyFeJwESIAESIIEgEKAABwEqkyQBEiABEiCB2AhQgGMjxOskQAIkQAIkEAQCFOAgQGWSJEACJEACJBAbAQpwbIR4nQRIgARIgASCQIACHASoTJIESIAESIAEYiNAAY6NEK+TAAmQAAmQQBAIUICDAJVJkgAJkAAJkEBsBP4PAq8DbBHGovoAAAAASUVORK5CYII=\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "library(ade4)\n",
    "library(ape)\n",
    "## Choose any two adjacency matrices to compare\n",
    "print(mantel.test(interaction_adj, org_adj, nperm=10000, graph=TRUE, alternative='two.sided'))\n",
    "print(cor(as.vector(interaction_adj), as.vector(ground_adj), method='pearson'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## How does gcor vary over time?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "from importlib import reload\n",
    "import src.interaction_model as im\n",
    "im = reload(im)\n",
    "classified = pd.read_csv('data/classified-berkeley.csv', index_col=0)\n",
    "# Data segment 1: 08-09 to 10-09 (61 days) -- all occupants\n",
    "overlaps1, similarities1 = im.interaction_matrix(classified['2018-08-09':'2018-10-09'])\n",
    "# Data segment 2: 10-09 to 04-24 (197 days) -- occupant 10 gone\n",
    "overlaps2, similarities2 = im.interaction_matrix(classified['2018-10-09':'2019-04-24'])\n",
    "# Data segment 3: 04-24 to 06-17 (54 days) -- occupant 5 gone, occupant 10 gone\n",
    "overlaps3, similarities3 = im.interaction_matrix(classified['2019-04-24':'2019-06-17'])\n",
    "# Data segment 4: 06-17 to 08-09 (53 days) -- occupants 5 gone, occupant 10 in space 13, occupant 13 gone \n",
    "overlaps4, similarities4 = im.interaction_matrix(classified['2019-06-17':'2019-08-09'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Adjust similarity calculations for different data segments\n",
    "similarities1_2 = (similarities1 + similarities2)/(61+197)\n",
    "similarities1_2[9,:] = similarities1[9,:]/61\n",
    "similarities1_2[:,9] = similarities1[:,9]/61\n",
    "similarities1_3 = (similarities1_2*(61+197) + similarities3)/(61+197+54)\n",
    "similarities1_3[4,:] = similarities1_2[4,:]\n",
    "similarities1_3[:,4] = similarities1_2[:,4]\n",
    "similarities1_3[9,:] = similarities1_2[9,:]\n",
    "similarities1_3[:,9] = similarities1_2[:,9]\n",
    "similarities = (similarities1_3*(61+197+54) + similarities4)/365\n",
    "similarities[4,:] = similarities1_2[4,:]\n",
    "similarities[:,4] = similarities1_2[:,4]\n",
    "similarities[:,9] = (similarities1_3[:,9]*(61+197+54) + similarities4[:,12])/365\n",
    "similarities[9,:] = (similarities1_3[9,:]*(61+197+54) + similarities4[12,:])/365\n",
    "similarities[:,12] = similarities1_3[:,12]\n",
    "similarities[12,:] = similarities1_3[12,:]\n",
    "similarities = similarities[:18,:18]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [],
   "source": [
    "###### import src.interaction_model as im\n",
    "import dateutil\n",
    "classified = pd.read_csv('data/classified-berkeley.csv', index_col=0)\n",
    "classified = classified.iloc[:timesteps*364,:]\n",
    "aggsims = []\n",
    "batches = 52\n",
    "for i in range(1,batches+1):\n",
    "    data = classified.iloc[0:len(classified)-(batches-i)*(len(classified)//batches),:] # forward selection of data\n",
    "    last = dateutil.parser.parse(data.index[-1])\n",
    "    if last <= dateutil.parser.parse('2018-10-09 23:45:00'):\n",
    "        aggsims.append(im.interaction_matrix(data)[1][0:18,0:18]/(i*365/52))\n",
    "    elif last <= dateutil.parser.parse('2019-04-24 23:45:00'):\n",
    "        s1 = im.interaction_matrix(data['2018-08-09':'2018-10-09'])[1][0:18,0:18]\n",
    "        s2 = im.interaction_matrix(data['2018-10-09':])[1][0:18,0:18]\n",
    "        s1_2 = (s1 + s2)/(i/batches*365)\n",
    "        s1_2[9,:] = s1[9,:]/61\n",
    "        s1_2[:,9] = s1[:,9]/61\n",
    "        aggsims.append(s1_2)\n",
    "    elif last <= dateutil.parser.parse('2019-06-17 23:45:00'):\n",
    "        s1 = im.interaction_matrix(data['2018-08-09':'2018-10-09'])[1][0:18,0:18]\n",
    "        s2 = im.interaction_matrix(data['2018-10-09':'2019-04-24'])[1][0:18,0:18]\n",
    "        s3 = im.interaction_matrix(data['2019-04-24':])[1][0:18,0:18]\n",
    "        s1_2 = (s1 + s2)/(61+197)\n",
    "        s1_2[9,:] = s1[9,:]/61\n",
    "        s1_2[:,9] = s1[:,9]/61\n",
    "        s1_3 = (s1_2*(61+197) + s3)/(i/batches*365)\n",
    "        s1_3[4,:] = s1_2[4,:]\n",
    "        s1_3[:,4] = s1_2[:,4]\n",
    "        s1_3[9,:] = s1_2[9,:]\n",
    "        s1_3[:,9] = s1_2[:,9]\n",
    "        aggsims.append(s1_3)\n",
    "    else:\n",
    "        s1 = im.interaction_matrix(data['2018-08-09':'2018-10-09'])[1][0:18,0:18]\n",
    "        s2 = im.interaction_matrix(data['2018-10-09':'2019-04-24'])[1][0:18,0:18]\n",
    "        s3 = im.interaction_matrix(data['2019-04-24':'2019-06-17'])[1][0:18,0:18]\n",
    "        s4 = im.interaction_matrix(data['2019-06-17':])[1][0:18,0:18]\n",
    "        s1_2 = (s1 + s2)/(61+197)\n",
    "        s1_2[9,:] = s1[9,:]/61\n",
    "        s1_2[:,9] = s1[:,9]/61\n",
    "        s1_3 = (s1_2*(61+197) + s3)/(61+197+54)\n",
    "        s1_3[4,:] = s1_2[4,:]\n",
    "        s1_3[:,4] = s1_2[:,4]\n",
    "        s1_3[9,:] = s1_2[9,:]\n",
    "        s1_3[:,9] = s1_2[:,9]\n",
    "        s = (s1_3*(61+197+54) + s4)/(i/batches*365)\n",
    "        s[4,:] = s1_2[4,:]\n",
    "        s[:,4] = s1_2[:,4]\n",
    "        s[:,9] = (s1_3[:,9]*(61) + s4[:,12])/(i/batches*(61+53))\n",
    "        s[9,:] = (s1_3[9,:]*(61) + s4[12,:])/(i/batches*(61+53))\n",
    "        s[9,9] = 0\n",
    "        s[:,12] = s1_3[:,12]\n",
    "        s[12,:] = s1_3[12,:]\n",
    "        aggsims.append(s)\n",
    "        \n",
    "# build model on 10 weeks of data for different parts of the year\n",
    "aggsims_select = []\n",
    "\n",
    "start = 10\n",
    "for i in range(1,11):\n",
    "    data = classified.iloc[timesteps*start*7:timesteps*start*7+timesteps*i*7,:]\n",
    "    s2 = im.interaction_matrix(data)[1][0:18,0:18]\n",
    "    s2[9,:] = 0\n",
    "    s2[:,9] = 0\n",
    "    aggsims_select.append(s2)\n",
    "\n",
    "start = 24\n",
    "for i in range(1,11):\n",
    "    data = classified.iloc[timesteps*start*7:timesteps*start*7+timesteps*i*7,:]\n",
    "    s2 = im.interaction_matrix(data)[1][0:18,0:18]\n",
    "    s2[9,:] = 0\n",
    "    s2[:,9] = 0\n",
    "    aggsims_select.append(s2)\n",
    "\n",
    "start = 37\n",
    "for i in range(1,11):\n",
    "    data = classified.iloc[timesteps*start*7:timesteps*start*7+timesteps*i*7,:]\n",
    "    s3 = im.interaction_matrix(data)[1][0:18,0:18]\n",
    "    s3[9,:] = 0\n",
    "    s3[:,9] = 0\n",
    "    s3[4,:] = 0\n",
    "    s3[:,4] = 0\n",
    "    aggsims_select.append(s3)\n",
    "\n",
    "    #aggsims.append(im.interaction_matrix(classified.iloc[0:len(classified)-(50-i)*(len(classified)//50),:]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [],
   "source": [
    "# agginteraction_adjs = []\n",
    "# for aggoverlap in aggoverlaps:\n",
    "#     #totalaggoverlap = np.sum(aggoverlap[0][:,:,np.r_[0:96]], axis=2)\n",
    "#     #agginteraction_adj = totalaggoverlap[0:18,0:18]\n",
    "#     #agginteraction_adjs.append(agginteraction_adj)\n",
    "#     agginteraction_adjs.append(aggoverlap[1][0:18,0:18])\n",
    "# %Rpush agginteraction_adjs\n",
    "\n",
    "%Rpush aggsims\n",
    "%Rpush aggsims_rev\n",
    "%Rpush aggsims_select\n",
    "%Rpush batches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%R\n",
    "gcors = c()\n",
    "gcors_rev = c()\n",
    "gcors_select = c()\n",
    "gcors_s = c()\n",
    "gcors_o = c()\n",
    "mantels = c()\n",
    "library(ggplot2)\n",
    "for (i in 1:(batches)) {\n",
    "    # Choose aggsims or aggsims_rev for plotting\n",
    "    gcors <- c(gcors, gcor(aggsims[i], ground_adj))\n",
    "    #gcors_rev <- c(gcors_rev, gcor(aggsims_rev[i], ground_adj))\n",
    "    #gcors_s <- c(gcors_s, gcor(agginteraction_adjs[i], socialnorm_adj))\n",
    "    #gcors_o <- c(gcors_o, gcor(agginteraction_adjs[i], org_adj))\n",
    "    #mantels <- c(mantels, mantel.test(agginteraction_adjs[i], ground_adj, nperm=1000))\n",
    "}\n",
    "for (j in 1:(9*3)) {\n",
    "    aggsims_select_temp <- aggsims_select[[j]]\n",
    "    if (j <= 18)      {gcors_select <- c(gcors_select, gcor(aggsims_select_temp[-10,-10], ground_adj[-10,-10]))}\n",
    "    else if (j <= 27) {gcors_select <- c(gcors_select, gcor(aggsims_select_temp[c(-5,-10),c(-5,-10)], ground_adj[c(-5,-10),c(-5,-10)]))}\n",
    "}\n",
    "\n",
    "# Plotting using native R\n",
    "# plot(gcors, type='l', xlab='% data used to build interaction model', ylab='Correlation between interaction model and ground truth', xaxt='n')\n",
    "# axis(1, at=seq(from=0,to=batches,by=batches/10), labels=seq(from=0,to=100,by=10))\n",
    "# abline(v=batches*(61/365), col='darkblue')\n",
    "# abline(v=batches*((61+196)/365), col='blue')\n",
    "# abline(v=batches*((61+197+54)/365), col='lightblue')\n",
    "# segments(batches/5, -.22, batches/5, .49, col = 'black', lty = 2)\n",
    "# #legend('bottomright', legend=c('Move 1', 'Move 2', 'Move 3'), col = c('darkblue','blue','lightblue'), lty=1)\n",
    "# #lines(gcors_s, col='blue')\n",
    "# #lines(gcors_o, col='green')\n",
    "# #lines(mantels, col='blue')\n",
    "\n",
    "# Plotting using ggplot2\n",
    "gcors_df = data.frame(time = c(1:52), cors = gcors)\n",
    "#gcors_rev_df = data.frame(time = c(1:52), cors_rev = gcors_rev)\n",
    "gcors_select_df = data.frame(time=c(c(11:19),c(25:33),c(38:46)), cors_select = gcors_select)\n",
    "#gcors_df[1, 2] = 0\n",
    "\n",
    "cols=c('Segment 1'='white', 'Segment 2'='grey', 'Segment 3'='white', 'Segment 4'='grey')\n",
    "ggplot() +\n",
    "    geom_rect(aes(xmin=0, ymin=-.25, xmax=batches*(61/365)+1, ymax=.55, fill='Segment 1'), alpha=0.5) + \n",
    "    geom_text(aes(x=(batches*61/365)/2, y=-.21, label='Segment 1'), family='Arial', size=2) +\n",
    "    \n",
    "    geom_rect(aes(xmin=batches*(61/365)+1, ymin=-.25, xmax=batches*((61+197)/365)+1, ymax=.55, fill='Segment 2'), alpha=0.5) + \n",
    "    geom_text(aes(x=batches*61/365+1 + (batches*197/365)/2, y=-.21, label='Segment 2'), family='Arial', size=2) +\n",
    "    \n",
    "    geom_rect(aes(xmin=batches*((61+197)/365)+1, ymin=-.25, xmax=batches*((61+197+54)/365)+1, ymax=.55, fill='Segment 3'), alpha=0.5) +\n",
    "    geom_text(aes(x=batches*(61+197)/365+1 + (batches*54/365)/2, y=-.21, label='Segment 3'), family='Arial', size=1.9) +\n",
    "    \n",
    "    geom_rect(aes(xmin=batches*((61+197+54)/365)+1, ymin=-.25, xmax=batches, ymax=.55, fill='Segment 4'), alpha=0.5) +\n",
    "    geom_text(aes(x=batches*(61+197+54)/365+0.5 + (batches*53/365)/2, y=-.21, label='Segment 4'), family='Arial', size=1.9) +\n",
    "    \n",
    "    #scale_fill_brewer(palette='Blues') +\n",
    "    scale_fill_manual(name='Data segments', values=cols) +\n",
    "    geom_line(data=gcors_df, aes(x=time, y=cors), size=0.75) +\n",
    "    #geom_line(data=gcors_rev_df, aes(x=time, y=cors_rev), size=0.75, color='gray') +\n",
    "    \n",
    "    geom_line(data=gcors_select_df[c(1:9),], aes(x=time, y=cors_select), colour='gray') +\n",
    "    geom_line(data=gcors_select_df[c(10:18),], aes(x=time, y=cors_select), colour='gray') +\n",
    "    geom_line(data=gcors_select_df[c(19:27),], aes(x=time, y=cors_select), colour='gray') +\n",
    "    \n",
    "    geom_segment(color='black', linetype='dashed', aes(x=10, y=-0.25, xend=10, yend=0.475)) +\n",
    "    scale_y_continuous(expand=c(0,0), breaks=seq(-.2,.5,.1)) + \n",
    "    scale_x_continuous(expand=c(0,0), breaks=seq(0,53,4), labels=seq(0,53,4)) + \n",
    "    expand_limits(x=52.1, y=.55) +\n",
    "    xlab('Weeks of data') + ylab('Correlation with survey net.') +\n",
    "    \n",
    "    theme_classic() + theme(panel.border = element_rect(colour = \"black\", fill=NA, size=0.5), legend.position='none', text=element_text(size=9, family='Arial'), axis.text = element_text(color='black',size=8, family='Arial'), axis.line = element_line(colour = 'black', size = 0.25), axis.ticks = element_line(colour = 'black', size = 0.25))\n",
    "ggsave('time-cor.pdf', width=3.5, height=2, units='in')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#### Simpler python plot\n",
    "cors_over_time = %Rget gcors\n",
    "import seaborn as sns\n",
    "sns.set()\n",
    "sns.set_style('ticks')\n",
    "sns.set_context('paper')\n",
    "\n",
    "f, ax = plt.subplots(dpi=150)\n",
    "plt.plot(range(1,53), cors_over_time, color='k', linewidth=2)\n",
    "sns.despine()\n",
    "ax.spines['bottom'].set_color('gray')\n",
    "ax.spines['left'].set_color('gray')\n",
    "ax.tick_params(colors='gray')\n",
    "ax.axvline(x=10, ymin=0, ymax=.955, color='gray', linestyle='dashed')\n",
    "plt.xlabel('Weeks of data')\n",
    "plt.ylabel('Correlation with ground truth')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Spatial correlations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "spatial = nx.read_edgelist('ground/space-syntax/ss-distances.csv',delimiter=',',nodetype=int,data=(('angular',float),('metric',float),('topo',float)))\n",
    "topo_adj = nx.to_numpy_matrix(spatial, weight='topo')\n",
    "angular_adj = nx.to_numpy_matrix(spatial, weight='angular')\n",
    "np.savetxt('ground/space-syntax/topo-adj.csv',topo_adj,delimiter=',')\n",
    "np.savetxt('ground/space-syntax/angular-adj.csv',angular_adj,delimiter=',')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "topo_adj <- array(as.matrix(read.csv('ground/space-syntax/topo-adj.csv',header=FALSE)), c(21,21))\n",
    "topo_adj <- topo_adj[c(1,2,3,4,5,6,7,8,9,10,11,12,13,16,17,18,20,21),c(1,2,3,4,5,6,7,8,9,10,11,12,13,16,17,18,20,21)]\n",
    "angular_adj <- array(as.matrix(read.csv('ground/space-syntax/angular-adj.csv',header=FALSE)), c(21,21))\n",
    "angular_adj <- angular_adj[c(1,2,3,4,5,6,7,8,9,10,11,12,13,16,17,18,20,21),c(1,2,3,4,5,6,7,8,9,10,11,12,13,16,17,18,20,21)]\n",
    "gcor(angular_adj, ground_adj)\n",
    "write.csv(angular_adj,'ground/space-syntax/angular-adj-trimmed.csv')\n",
    "write.csv(topo_adj,'ground/space-syntax/topo-adj-trimmed.csv')\n",
    "graph.Cor = qaptest(list(-angular_adj, ground_adj), gcor, g1=1, g2=2, reps=5000)\n",
    "plot(graph.Cor)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%R\n",
    "graphs = list(topo_adj, angular_adj, interaction_adj, ground_adj)\n",
    "graphs[is.na(graphs)] <- 0\n",
    "gcors <- gcor(graphs)\n",
    "rownames(gcors) <- c('topological','angular','interaction\\n model','overall\\n ground')\n",
    "colnames(gcors) <- c('topological','angular','interaction\\n model','overall\\n ground')\n",
    "library(corrplot)\n",
    "pdf(width=3.5, height=3.2, file='space-cor.pdf')\n",
    "#corrplot.mixed(gcors, lower='number', lower.col='black', upper='shade', tl.col='black', tl.cex=0.6, cl.cex=0.5, number.cex=0.6, family='MyriadPro')\n",
    "\n",
    "gcors_upper = gcors\n",
    "UT = upper.tri(gcors_upper)\n",
    "gcors_upper[UT] = gcors[UT]\n",
    "diag(gcors_upper) = NA\n",
    "corrplot(gcors_upper, type='upper', method='shade', tl.col='black', tl.pos='d', tl.srt=90, family='sans', addCoef.col = 'black', tl.cex=0.6, cl.cex=0.4, number.cex=0.6, family='Arial')"
   ]
  },
  {
   "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}