1.数据分布及其指标评估.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import make_classification\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "不平衡数据的不同分布会对分类器产生不一样的影响，比如下方的分布，第一张图片正负样本比例很低(一般少数类称为正样本，多数类称为负样本)，但我们可以很容易的找到一条线将数据分割开，而第二张图片就很难了，虽然和第一张图的正负比例一样，但正负样本的重叠部分太多，很难分割，第三张图片中正负比例和第一张图也一样，但两类数据中都混入了不少噪声数据，给分类建模带来困难，第四张图片的正负比例也一样，但每个类内部又有2个簇（类内不平衡），增加建模难度"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "X1, y1 = make_classification(n_samples=500, n_features=2,\n",
    "                           n_informative=2,n_redundant=0,\n",
    "                           n_repeated=0, n_classes=2,\n",
    "                           n_clusters_per_class=1,weights=[0.05, 0.95],\n",
    "                           class_sep=3, random_state=0)\n",
    "X2, y2 = make_classification(n_samples=500, n_features=2,\n",
    "                           n_informative=2,n_redundant=0,\n",
    "                           n_repeated=0, n_classes=2,\n",
    "                           n_clusters_per_class=1,weights=[0.05, 0.95],\n",
    "                           class_sep=0.9, random_state=0)\n",
    "X3, y3 = make_classification(n_samples=500, n_features=2,\n",
    "                           n_informative=2,n_redundant=0,\n",
    "                           n_repeated=0, n_classes=2,\n",
    "                           n_clusters_per_class=1,weights=[0.05, 0.95],\n",
    "                           class_sep=3,flip_y=0.1, random_state=0)\n",
    "X4, y4 = make_classification(n_samples=500, n_features=2,\n",
    "                           n_informative=2,n_redundant=0,\n",
    "                           n_repeated=0, n_classes=2,\n",
    "                           n_clusters_per_class=2,weights=[0.05, 0.95],\n",
    "                           class_sep=3.0,random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x2316cf064e0>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABBQAAAD8CAYAAADHR/dgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzsnXeYJFX1v99THaa7J8/sLjnjDxRUVBQVQQUBUVSSoKAgoCB+UcEASFBAgiAISBIEJAkIkgQJkgUkKyBJcmbZ3ZnZmekc6vz+uDUzHap6etL27O59n2efnamuunWqp7vq3HPP+RxRVSwWi8VisVgsFovFYrFYJoLTbAMsFovFYrFYLBaLxWKxLH3YgILFYrFYLBaLxWKxWCyWCWMDChaLxWKxWCwWi8VisVgmjA0oWCwWi8VisVgsFovFYpkwNqBgsVgsFovFYrFYLBaLZcLYgILFYrFYLBaLxWKxWCyWCWMDChaLxWKxWCwWi8VisVgmjA0oWCyWGUNEQiLyHxG5qdm2WCwWi8VisVgslunFBhQsFstM8mPguWYbYbFYLBaLxWKxWKafcDNOOmfOHF1zzTWbcWqLZbnh8ccfX6Sqc5t1fhFZFfgycBzwk/H2t/cFi2XmafZ9YaLY+4LFMvPY+4LFYqlmIveFpgQU1lxzTR577LFmnNpiWW4QkdebbMJpwMFAe9AOIrIvsC/A6quvbu8LFssMMwvuCxPC+gsWy8xj7wsWi6WaidwXbMmDxWKZdkRkO2CBqj5ebz9VPU9VN1bVjefOXWoWRywWi8VisVgsFgvTFFAQkS4R+auIPC8iz4nIp6ZjXIvFstSyKfBVEXkNuBLYQkQua65JFoul2Vh/wWKxWCyWZYvpylA4HbhVVdcHPowVYbNYlmtU9Requqqqrgl8A7hLVb/VZLMsFkvzsf6CxWKxWCzLEFPWUBCRDmBz4DsAqpoH8lMd12KxWCwWy7KD9RcsFovFYln2mI4MhbWBhcCfvH7z54tIa/VOIrKviDwmIo8tXLhwGk5rme2ouqhqs82wNBlVvUdVt2u2HRaLpelYf8Hii/UXLBaLZellOgIKYeCjwDmq+hEgBRxavZMVX1t+0NxDuAu/jL63PrrgY7jDp6JabLZZFouliWjxVdyBH+K+twnuwi/hpq+zE4jlD+svWCrQ/GO4i76Kvvd+dMFHcYdOQrXQbLMsFl+0tAh34Me48zc0/wZ+jJYWNdssi6XpTEfbyLeAt1T1Ye/3v+LjIFiWD7TwDDqwL5D1NiQh9SfUXYx0Ht1U2ywWS3PQ4hto306gacCF0gAMHYW6byNtBzTbPMuSw/oLllG08ALavzdj/kIK0pehbh/SdWJTbbNYqlEtoP27QGk+4C2S5W5H+/8Lc25DJIJqHrK3oPlHILQaEt8JCdmgqGXZZ8oZCqo6H3hTRNbzNm0JPDvVcS1LJ5o8C8hVbc1C5lrUHWqGSRaLpclo6lzQDOCWbc1A8jzUTTbLLMsSxvoLlnI0dS61EhpZyP4ddfubYZLFEkzuLnAHGA0mgPnZHYDcXag7jC76Cjr0K8hcDcmz0EVbofknmmWxxbLEmI4MBYAfAn8WkSjwCrDXNI1rWdoovgj4pDFLBErvgNOxxE2yWCxNJv9voFS7XcJQeh2cDZa4SZamYf0Fi6H4PyqDjB7SAsU3IdqzxE2yWAIpvmSyaKrRFBRfQgtPQeltxoJkOVDQwZ/CnDsQEd9hjTYtmFuixbJ0Mi0BBVV9Ath4OsayLOWE3w+lN6lxErQAoVWbYpLFYmkyodWh9HLtds2Ds8KSt8fSNKy/YBklsoGZpNX4C3kIr94UkyyWIFQzAa9EILwODJ2Ib9Oa0gJw34XQypXjld5GBw+D/CPm9+inkM7jkdCK02u4xbIEmA5RRotljNbvA5GqjXFIfBNx2pphkcViaTLSth8Qq9raAi2fR0JzmmGSxWJpNonvAdWrsjGI74A43c2wyGIJJndvwAtFNPo5k1nji1L9OVfNon27QP5hTPZeCfL/Qvt2Gc1YsFiWJmxAwTJtaO5eWLwv5uYp5p90Q9sBSLvV3bJYllck+lHoPBmcuUALEIXYtkjXb5ttmsViaQKa+xcM7IPJThhJBe+Etu8jHb9qomUWSwCltwNeCCNkIbErtYFzByLvR0JzTGvUwjNo4Wk0c4tXPlGeneOCDkPuzhkx32KZSaZLQ8GynKOF/6EDP2RUrRkwaWBr4bR9r1lmWSyWWYIT3xqNfQHcReC0IxJvtkkWi6UJaPFVdOD7VPoLYQivitP2g2aZZbHUJ7wOFJ6s3S4JkA4k8W00/yjk7ve2h8z2rtPQ/H/QxQd4QQQBVSBdO5ZmoPj6TF6FxTIj2IDCDKGqkLsNTf8FtIDEvwbx7RGpLgdYNtD0RdTWjhWg8CxafBkJr9MEqywWy2xCxIHQvGabYbHMKoy/cAeavtJMKGJfQRI7LbMibZq6DChUbS1C6RW08CwS+UAzzLJY6iJtP61siw5AHNoONM82HKT7bLTwHBSegtCKEP0MaAod2M5f0LHmJDEIrzf+fhbLLMMGFGYIHToCMjcBRsRFC/+F7E3Q/SfvxjMN5yi+CpqE8PrND1QUX8dfrTlievbagILFslxi+nLfimZvB6cHSexqJwwWSxk6fJxpMzci+lZ4Bs3eAD2XITI9bpoWXzPp1OH1mh+oKL2Kb9cXQlB6F+z9YblDRLqA84ENMXWze6vqg821CrT4CuTuA4lDbGuk+xx0+CQovgyhFaD1hziJ7SuOkcj7IfL+sTEyt+DrH4+UBo++FoHQKtCy+QxdjcUyc9iAwgygxZcgcyOVUcyMSZXK/xNaPjfF8d806YKlN01KFQ7acTxOfJspjTslopuYiGx1loLmbLTVYllOUc2j/btD4QVMcNVBM9ehHb/ESezcbPMslqajxTcg/RcgV7Y1A8XnTS11bGrPdS29iw7sD8VXTJtWQDuOxol/ZUrjTonoJpB/lMprxnR3sMGE5ZXTgVtVdWevpWyi2Qa5QydC+jJAja89dCzSfQbOnBsmOFC/+Wz7EX4/lN4wP8e2Q9p/ikhoSnZPBHWHAQdxWpfYOS3LJlaUcSbIP4wJsFahaXSktmqSqLrowJ5eC7asSaHSYRj8OVp4cUpjT4n4dkDVTVDikPiGVXG3WJZXMjdAcSSYAGYlJgtDx6BuA+mfFsuyTv5Rb2GgCk2juXumNLSqov17meAEWZPRqEkYPBwtPDOlsadE7ItecEPKNsZNWWhopWZZZWkSItIBbA5cAKCqeVVd3EybNP8IpC/HBL3yXvZQFl38Y9T10T6oR3QT8MsKkjjScRjOCo+bf51HI07HNFg/Plp4AXfRjuiCTdAFH8ft3xMtzV8i57Ysm9iAwkwgXaMrAZVEweltaAgtvYM7eBTuwi/h9u+L5h8zLxQeB3eA2vSpApq+YipWm/O6Q7jJs3EX7Yzbv59RYga08ByauR7NP2nqPcuPKb4CfbtSmcLoQGI/pP3wxs5beBZ3+Pe4yXNMama9fYuvGZVc21rHYpnVaPaWsTTuciQMhX8veYMsltmG00XlxHqEsNcVZXy0NB936Neev7APmnvIvFB8Gtz51PoLeTR96RSM9s7rJnGT53r+wvfQ3H1me+F/aOYGNP9vH3/hTej7OmiRsYUXgcReSMcxjZ238ALu8Bm4w2ehxZfr71t8Ay381/oLs5u1gYXAn0TkPyJyvojULJmLyL4i8piIPLZw4cIZNUgz11OZZTxqBeQfmNhgkY9A9NNAuRBxHKKfgMjGkzdykqg7iPbvBsVngKL5l38E7d8NVb9SJItlfGzJw0wQ2xKGfunzgoPEd/A9RLVgarKcLtAC2re954gXofQS2v8Q2nm8p5Xg53yUwH13SmarO2zOW1rISCqi5h9EnRXAfQ/EARRC60DPRaORVB062mRJVGRlKBQeQWR8xWZ36LeQvhRTLuGgybPR9kNwWr9VaV/pbVPqUXy9rNTjWJz4tlX7LYTis+CshET+32TfDotl2lHNo6kLIXMVaMG0Tmz7P8TpbLZpM4O0B7ygUOsvWizLHy2bAX4aSGEk7l8WNOYvdAAOuuirJvNgxF/IP4p2/AoJ9eK/buQabaMpoG4K7dvRaB6M+guPoKEVzbYRrajQatBzCeJ0m32GjwMdojLIoVB4CJEDxz2vmzwDkn9kRNRRU39A2w7Eadun0r7SfK/U42VvgUfR9qNwEl+r2q/PTKyceUhk/Um8E5ZpIAx8FPihqj4sIqcDhwJHlu+kqucB5wFsvPHGPmnA04iW8M00Bvz1EIIREeg6EzLXo5m/Amq+2/HtzWv1zHAHjE8eXm3aOiNp5nqvBKP8+kpQegtdsCna/lMkvvO4tlks5diAwhTRwjNo8jwovQaRjyFt30VCK0PPxejAfqBpRoRXpOsUJLRizRhu+gYYPgbTg7ZoHO2a/rRZGPo12ntdQC1WHKKfndq1pC+rCCaMntf1WtiM3HuK/0OHjka6TjG/5x+h9sarXunHOOcsPO0FE0YiwS5QhOET0dgXRt8vk7r5HaMbgTt2usFD0PDaSGQ9s8/wsaYeVVpAi2j4fUjPH0edGYulmejAD7zvi/d5T1+O5u6GOTch0tJU22YCSXwTzd9Tm6UgbRDZqCk2WSzNQgvPo6nzzCQ3shHS+l0kbCbcOrCvN9E2E3HpPBEJr14zhpu5FYaOAEqev9DmBfTLVxazMHw8OufmAH8hNnUtp8zVXlCiSvuh9Kq3g7ep+DI6eDjSfbb5PfcAvhOywhOoFuuKUGrhRS+YUL5yXILkaWhsa/Ne4vkLA/sY3QhKY7YM/RKNrI1EPuj5C95ihkQ9f2FNpPsCW6a55HkLeEtVR5zGv2ICCk1D4tuh2VsZK9fz0BJEN534eBKCxE5IYqeG9lfNoYOHQfY2I26uLtq2P9K639Qn+sVX8c++ALQfho9FNYm07jW181iWK2zJwxTQ3L1o3zchdysUn4PMleiir6DF15DIhsjc+5Dui5Du85B5DyI+D3DNPwlDRxqHQFNAznyh/R64moXh31GrjhyB0EpIVeR9wuTupkYkyZeCUW0fSY0KVIweX0na3LD9HB6B8vrRwhPgLsS/1OPPZqzMNZD+qxlPhzHCVs+ii382dj4toJkbcRcfiDv4a7Twwrg2WizTgRaergwmAJA3n+vsLdNzjtIiNHWJyfLJ+/TLXsJIyyehdT+gxUx8pBWcXqT7/GnrdmOxLA1o7kG0bxfI3uz5C1ejfV9FCy8ikfWRufcgPZci3eci8x5CYlvVjlF4FgYPNoGHUX+hD9+OCVqs4y/MDcx+aJjc3QROSiooQu4eVD3fIjBwarIO66HZ26ltNwmgRsBy9JTPQektaq89h6YuNj9m/w7pP2PewxF/4UV08Y/HRtUimrkZd/FBuINHm/ffMu2o6nzgTREZUfDeEmjumx39DMS3xZQpCCaLKAadxyFO24yfXoeOhuw/MP5sCshA6hzTLW6KSHQjkDqal5qB5JmoFqd8Lsvyg81QmCSqig7+ksoHahE0hQ6fgnSfYRzm6Ifrj5P+E41N4gEKkPsHNQ9JiUHP1VNPhyr1TWRn718IYttD5lpqr6OIFp5BIhvUGSeEbwmHCBXOhbsIf2ej5KVcAumLqIkmU4T8w6i7GCThozh/Ndrx65o0SItl2ik87b9d02j+cSS+vf/rDaLZuz1nWIECyLloyzZmpbOJqYtO2w/Q+K5QeNSUQEQ3mbZWeBbL0oIO+fkLJXT4RKTHC7BFPlh/jNHSwEYoeIHKan8hAj3XTH1SVFowsf21YIIJ8Z3HJvLVFJ6CaHDmkkgI9S35rPYX+vB3b8dKPTTIXyg8iZbeA6fHZEUWn/YyrBw0cw3acThOYtfg67RMlh8Cf/Y6PLwCNHV5XESg4wSIf8NkEUoCiW+HhFaZ8XOrZr1OcdVdUDJo6g/IVDu0xLaF5O+hlMdoKPgZkQd3MdhsHUuD2CWiyaID3kOrGrehVP9RSu8SXKdVThSkA/8VgRLivtH4OX3Q4kueeFMjCEQ+PNrLWtoPgfD7fPYroH27mXrPoJFiX8b3wa8utGw59ntko+BSjxav1MMdCjiLA24STV9XFkyAUcX54V+ifsJxFst0ElrJX82dGITWmNLQqll08CDM/SGHKQvKmABk7u4pjT0tOF1I7ItIy6Y2mGBZ7lA35a2Y17xihJYbpfQOjdVvR0E6CRKVk1J9IcPx0OIbY63uGiH8vtEAhrQfBL6LDAW0/9u4bp2sh9g2BLqtsa3Hfo58sE6px+fNj+6g/zgSNhkL2ZtNEFirO9Qci7rJYBstk0JVn1DVjVX1Q6q6vaoONNsmEUGiG+G0H4TTtt8SCSYA4A4Hv1ZaNOXhRVqQ3msg/nVqurON7hSGZVXbyTIj2IDCZKmXLuR0NT5O9DPAeLXTEaMQGzgxd0AbzXIIIPcA/mKPYOyLjf0srUjHsaOvipOAyIcCjs2gw6fWOXHRZ5IVgo7jQFpMWUn+UXB6IPEdKlVyWyC0AhLf0fv1c/gGJ5wOCK1sUhxrViS88+X/U8dGy0QRkZiIPCIiT4rIMyJydLNtajrRz3hOftVtV0JIYsepjZ1/pHZcMNkPmeumNvYUcNPX4y7YDH3v/bgLNsVNX9U0WyyWpiEtBCaEygTaxEU3Z+xZHETEqMcHBh5k6v5C/l8ETkRoMS2jARPYaEU6Txg7u7QY1XtfcjD8mzonLvqUWIag40iQDjT3TzT3sCmvavt+mR2eLaE5yEh2QcsW+IthtkBoLTR7M77+gkRMq0+LZaZwesHxEzQWiH5sWk4hTrdpU9l9HjX3FIlD6z6eCLzF0hh2qWiSiMTQ2LaQvZWKtCSJQ2Lvxsdp/Taa+YvXCjIgYOCsZhyE/P3BA0U2DHxJS/ONmGFobU/12WcfBP/Upyi0fs/c3ApPQng9JL5L7ThunZbBmeuh4+Dac2oR7d/bE64sJwylV9EFnx5VZ0bi0PVHJPohNHWJWUGIbYMk9jABDUDaDkBzd3jR3RzG4Ykgncch4qCBKZ5WcX4GyAFbqGpSzFPpfhG5RVUfarZhzUIkBL1XoIt/YlJ7EQitinSdjDg9DY2h7jCavgryD0JoNaT1W0h4HYKDgYzzWmOoFiF3O5q9y2ggJHZBwmt7rxWAcE1ZhZu+0et24604ugth6DhcwEns0uB58+a+g2NE7HwzPCyW2Y1IGI3vAJnrqExjjkPrdxofJ7GrKXtwFxFY+iBzIfo5yAfdal2IfjTwHFpaAKXXIbQmEvJvW6mE8PdXwpD4NoRWhMJ/ILwOEt+1dhwNyibE1Ih3HlV7TnXR/r28bhblRKD0LrrgU2WLExEzUQq/35Q2uIuh5QtI655jmRJt+5mggbsY8zdxgCh0/NqUVkgb5t7pIzhdb0HJYpkiIg7afiQMHsJYlpEDEkfafzK952rZDO08GYZPAPcdE+Bs3Rdp/e60nsey7GMDClNAOo5Gddis7kvURP3ju41FwBsZw+mCOX9DU+dD6kJ8VxXcVyB5sv9rYL78PsKIqnl04MeQvw8jkJhH419DOo6pcMxV1RMz9B9fEl9HQitVjl140dhcfMFkJ0Q+DLmb/e3TgNTC/MP460fkIfUHKtSZNQUDe8O8+3HKUxvL7QzNgzl/R9OXQ+4hCK+OJL6DRN7nXcduaO4BalYdpH3c2lXLxFDTfHzE84t4/2a2zdNSgIRWQnqvMJoeWgh02P1Qtx9dtL3nAGeBkBEi7T4Topvg+/ZKPLBVbcPn1Tza/20o/s8L/oXR9OVofGfI3eG1lO0yCtSJPccCC6lTqU25zkDydGggoKC5+9DFB3rXpUAUus9BqiZDqgq5W43YmjsIsa2Q1r3NvdVimSVIx+GmBVzunjF/IbEzktiz8TGcNphzPZq6AFLn4y/G+A4kjyfQX4jvgUhtloNqgcy7BzM8/256VggRCuXQ2LbIaLvqMtJX+58bB0l8w+tOscfY2MVX0NQfofC8KXcIbwhc7W+fBqR7F/7jE0wAyBmxuvLuT4AO7I3MewAntoXvcOL0eP7CFZC7H8KrmPtX5P3m9cQ30Owd1PoLMYhu7G+jxTJNOPFt0dAcNHmOKS+KfARp+wESXmsGzrU1xLcOXBywWBrBBhSmgDgJpPsPXgbAuxBee1L95MXpQdoPxs3caJxzX4LSF+NI9OM1W1VzZvIxWivprWZkbkRDqyFt3x/bufjcWGvIaqKfAomhpYVIaC5afBXN3OA5MwVAzURDopgIv4+d4XX8x/Z1DjBj+joreZNqOaKZ4IM43Ujb/0Hb/9W+1vIZtHUvY/uIgySxpirOqyrk70WztwEtSGJHJLB8ZOlCTNTqcWBd4KyyllDl++wL7Auw+uq1LdKWVRqd7KoWTYsnp91MItw+xlYGjTCqDv7CdJTp+j06MPK5LwJhiH156u3h0teZicCoY100/zKXle00AMOnorhIq5ehNSKWWo27EFW37ndOSwu8aykPSKRMK7i591UIyunwycaWkVrn1Jto9kbovbFyv9LbpibamedlO1inybLkEGlBus8wgn+ldyC85qTaGYvThbT/FDd7Sx0dgyB/IYG0bFKztZAvcNb+B3L75fMRWYdozOV7R77DNt+8DQ2tbHQPPLT4MhSf9x8+8hFwOkyWgzMXSm+imRu9CX/R2FV8HuQmjPvpkxUZpCejSfyzrYLi1C7k7vV0F/wRpwNp2w/a9qt9LfpxtG1/SJ7lZUqCyXy4oGmZUsZfeMArx4gg8R2MYr9lmUSiH0d6av37GTufLXGwTAEbUJgGJLSiSfGbKtHNIRsQtQ88ecy3n7suPqwsmFBOFtKXmPrCEUpvE1gPWfgPuuAzgKAjqyoUqXyIF02LKt+HfQxpry13ACD68QBdiBD+Ky/4itWouqZkxGn3zdQox2k/EE3sZmrOnS6IfrJpInGqalT58//0Vn4dNHMt2nYATtu+TbFpOlHTV3QjEekCrhORDVX16ap9zgPOA9h4442XmQwG1Zxp+VR6E8Lvh5bNJ+yEuplbvJKBAmiJ0e4NNTsmofQW0rIZzLvH9K12h6HlM0jkA1O/mOxN+GuPVJOB5Dlo4jsmWBBa1aROV+OsOG4ATzM34t86VyF3O3hZF1paBOmLqUz/zkOpD03/BWnbx6RKD/3SlF5J1IzrrAQ9FyGhFRq4Lotl+pDQCjAdn7vo5yBzyQRPHjbP3SrO/OGF3HnFfPI5873MZR3OOmJVuua+xiZb/RnKAgqU3jEBefURTyw+gy7Y1PulBZOBWKLyu1zynndB/sIh/rZHPjoxfwHX3BurUHVBF4O0je8vtH3fZGLlHzZln9FPNW3SZTqLHeqV2Y50nbgObf0uTvuPmmKTpRYT9HkYis9AaDVo+XzDnxnVPGT/gRZfRsLrmmy7cT6jSwq31A/DJ5uuJ+ENkbZ9kfCazTbLMouwAYVZgrqp4JIBX2IgYaT7D6MTFTdzKyRP8QIEdfrHVk/KIx8IUETGSz/05nlB+4ztXPlraC2k45dIy6b+u0uHWcVw367a3uo5DtWTmIInNgXqDkL2NjT3IOTuw6xkOmjim0j7z+sGCSQ0D+LbjXMtS4D8v8qCCTCqIp08w5SmLCOTHVVdLCL3AF8EAnonLjto8U20f1fzd9WMqbcNrQI9Vzbcqk0Lz1TVT9ajhLqL0P5jofCwKeFJfAsJrzf+oY3gTKBeWJPeNbdC289g8OdUXkMM2n7awDgD+NeIFyv1Wgr/9dLHq/fNepoz+6CZv3otuPJj+5VeQxcfiPRe0fi1WSyzBBOwvH4CR8RAHKT77NEJimbvRId/i5beYJe9Q6TeW4F7/zaWMZHLOFx+2gps8oVXK4cKrx8s6liRdRjc3cnbufLX0OpIx5FIUAaitIKzMrhV9kjcC7hW+QvqQssnvR+Hjb+QfwhyI89cQeM7Ix2/qDtpk9AciH95nGtZAhQeLwsmwKi/kPojGt8BCa/WROMsAOqm0YE9ofii8WElap7HvVcioZXrH1t6D+3bxeiLaAqVVhj+LfRePaHSyJnAzfwTBss0FYrPo9lroft8pOUzzTPMMquwXR5mC7k7J7BzFDp+ZdKco0Yt2c3cDIMHeyuCdYIJANFKhWUJreI9MMsVkUc+GpNdNBaIbhwcTABzzX6dgTQL4XWr7IlD2/5IaB6a+xe6cHN06BjI/R0Ywkw+spC+Aq1SiVYtoqX5xgmbIqp5b6xGe4HXGSv7Dx9BSoywVK5SgFPdNJq+CnfwSNzURSagMosRkbleZgIiEge+AATkyS5b6OCh4PZ7f1s1+h/FV9HkaY2PkbqIxvrNh01AcOC7kL/XTObdBZA8Gx06YpJXUEV8V/zV0H2QzlHBMie+DdJ1MoTWAkLgrACtByDxr44/THRTwC+Q4ZgyrBFCc83EwW+/EQcufRm1wckSFP6LlhYCJqDrJi/A7dsVd+D7ntaKxTJLyd2D/6q8HxHoOByZez8yEpDP3oUuPghKryAUWWmNHD/53ZtsuVN/xZEL3orWdHCS0FyI74y/vzAFIh8MDiaA0YJSn5JQzUNk/Up7JA6JPZHQKmj+UXThZujQ0ZD9m8lOIA/kIHON2V4+3Kz1F+7EP1NMzMJExXmzaPoa4y8kLzTaHZYZR1NnmfJATWMyC1PgLkAXB2Tplh87dLR5dmtqZDBz7NBxM2v0eHZpAQb9MmZddOD7uO54gUPL8oINKDQR1QKauQ63/3to6kL/FMIaHGg/FCexE+KUdSZI/pbGVjMdpP3wmq3ScTy0/9TULzpzIPz/Gr2MABRKdXrpApr7Z8CEOgyJXaH9YAi9D8Lvg7aDkNYfmAf04gO8emm/h3QW0pej3nvppi5BF2yCLtwafe8TuEMnYjLxJ3g1qrjJM9EFnzBjLdgEN3m2SW+bLJLAv9REKtpdaWkBumgbdOh4yPwFhn+HLtzS1LLOXlYC7haRp4BHgdtV9aYm2zTjqJsy4mE16foFyEzg8gP7zYfNP2kDEkafJLyet2JY/lnMGr2UUpAmywRIX+tji6eIXkEM2g+q0CaQ2NYmI0ESxkFKn40u+gJafKn+OaOfhJZPUBNUjH8Jiaw/tim8gRc4qP4eRZHEt83oKlFuAAAgAElEQVSPIw5aDSHQtAnW9e1kxCIL/4HcXejAD3CT59W30WJZgqgW0cyNuP37osk/BGcJVCDQdiBOYtcq3ZFafyGWUPY+bEz3RERZ/2NZpOPI2lE7fgnth5hgodNrshamSilIU8mzOXe/v7+AA7HtoONw47eE3mcCl+0/Ne/ZwP95x/m9X1nIXId6pRFu6kp0wSc9f+HjuEPHGh2bCWL8hfPGfI8FH8cdPt2UW0wWieOfVOxU+gtuP7pwW3To18ZfSJ6GLvwCWvjf5M9tqUELz+D27YH73ia4fbuh+SdMWV3N58yFwqO4g8egxTf9x1INCBIWjfBxE9HcgwRrsuRh8Y+XpDmWWcysL3lQtx8d/p2pCyYE8e2Rth+NtgpcWjEtE/cyKbsN1Sd7xL+N0/qtqrHUm4A0QtiURESq06GF0ckKEXAnYFMQMo7D4/TiL8wkJq05dY6XypiD1Glo/j5I7MH4WRNFNHkBGloNhk+h4v1NX45KZMKtdzR9IaT+OCb8pkDyXFTakdZvT2isESS+A5r+M7UPEa0Q0tPhk7w2YSP7ZUFz6ODhSO+Vkzr3TKOqTwFBzcaXUyYQfGrZ1GstWf0dCkHvX5HSm0a3JbwB2v8NfNOLpQWKL0+pXlsLz3qrX9Wf0RZo3Qdyt0HxNWNL649wEl+rPL74Kgz+jIrJSymN9u8Bc/8ZWJokItB1DmRvQjPXAWEksTO0bFO7X8+fTJCx8LwnnhaGjmPHAg8tW0H6UmreI6cdQquZ72DpnUobyZjSo8QutlvEUoS6g+jwqZC9GRCIfwVpO7DhUqPZiqqLDuwH+ceYkL8Q2wmn7Xu12wPEHHtWKBIKK24JWuIue/58IVp8E6lpS+35CxIGjY49F6fCuP5CDyPdqiqPC5nSzOSpGH2GDKTOMgsWbf/H+KUXRTR5psncHD6BSn/hKhRBOmoXYeqh6SuNmCNl/kLqQlTiyCT1kST+NdNZq8ZfUmj5wthvw6d64t4j+2VBs+jgIciciZTJWIJwsw/A4r0ZfaYXHjNljgS1H1fIXIlmr4HuC5AJdQlpsnhwvZbwAPk7cBftCB1H4USXDUFxy+SY1RkKqnm0b2fIXAs6CNoP6T+jA9+Z2srwbCB3hxE3mYhzAOCj6CsiJp24IfLo4C9qVul16Bcw/Bsj5Oi+G9z1YSLk7scdPgfNP+b795L4TvjHtCKQushbWcxi0sbTxplKX1FnxbGM9HmQPIPa9zcD6UsmnqWQPM/Hacp47S0nh0T+n1nlIWrqQ6UVJIF0nVMZMMvdiW/QofDktKRlWqYPcVo9kdTqW2tkQrodktjdiIZWlBpETTZC/lGIfgKJbGi+++H/53M+TCpwaIp1tfnH8V+dyIAO48z5O86Kz+DMvbMmmACg6avwF5IcMA5/HURC0LIl0nkK0v1HJPZF384MEloBp/dqZO6tSM8VyLwHceJjgQdp289otYxmO4RNO83O3xhxyNxd+GZ3SQQKT9a10TJ7UC2YGuTM1SatXQcgfSXav/vUVoZnA7l7TQ39RP2FyMf8t1e1gR4hORiivavIJ7ca4rQbX2LN9Ydh6AivndwYOnQUDB9rasXdd6H0qu94EyL/OO7wGWj+kQB/YXv8XVaB1JVj+i1g/IXCk5C6uEF/4XJ0+PfUvr9ZSP9l4iULqXN8xsqYRYlJIuE1oONXQEuZvxBHus9AnI4yk/+Bb9lr8QWjJWGZOoM/onaBQIF6n7UiaAYdPKzm8y0iXlCo2h8OQ0Cb9CVGvTKkEYpPQ//OuOnrZt4ey6xldmcoZG81tcgVN8ccFF+AwmO+asWzFXWTkL3F1EhFNjL1cL7pe+NQetuo/WaugdTZUFposg1iX/VW4cofYoL/qmjOOADhdY1txTcg83caq9muRkyJhLvQ57UCpM5E0xEIrYN2nY0UnzBiNS2bIeE10M7fwtChZhz1ukc4PUYdv4YM5O9q0C4nuAWn5s17L+1o4VkT9S++DtGPm/71oXmVu6v6az1AwHU3jtO6OxrfFnL/MgI+LZshEq/aK6h+XZjlMcHlEun8jcka0Iz3OTOijNJ20PgHj4zhdMKcG9DkeSb4WHoPUCg+BcMvoslToedyE0xwh6id9EeMIvlUhbpCczxF9+p7Q0tjQcyKzJpySpD5KwT0iFd3AF18qCesKCYDovME3xa5I0hoFd8KInG6Yc5NaOZayD8IodWQxG5jCtXOHPzvlS6IzU5Yasjd5d3zyye/eaMrlH8AWjZrlmUTRt205/+8A5EPo7l7JukvmEwEN30DpM4w95HwOqZEIHUhlYE0oaO7xF+eerZqENe0hvayFLQ033x3x135D8Dp9drfVlOA1Dlo+gIIrY52/xEpPGnuPS2fRkIrQdfp6KAn6qpeG0pnXkDGRRbyd6IK43aIlXBwm1sU3EEIzUULz5v2vcVXIPoxz1/w6fDlLgoYahDV0qTbTjqJndDYlkZjSSIQ3aw2W1cidZLhmtPucllCtWgyYvxfJdjv9ii9jpbeMAGiMqTjl2j/M57+UtZ0cHPmTjg7ZrpxQp240a9A/sbxdx46HBI7zLxRllnJrA4oaOG//g9RLULhuaUmoKCF59D+bzGqRCxxoI3gdkd1CK+Fpv9k6n1HovGFp6DwAiT2hOy1ZpLrzAOi4L7lY1DJq8H2KDzNpNOqpNVEMDPX4y8GWTABhOJzsOhzqMQ8G4pox1HeA/JzpqYz/ReMA/jK5GwpR4tmNbf4VO1rTg9IG5q927RtJM9If2zNXAO91yHhVccuUQR11gT3NZ8TCVp4vrKue4KI01N/9Tq+PaQvpzLgE55QOyLLkkPCq8Pcu7y2kW9Num2kOD1Ix6G4Q27V3z9jknYGf460/Rjy9/gP0HXyVC7D0LIFJs24auVFQt6K4ThEN4PsDf6v5R/x3ayqaP8+ZhIzMmkpvYEOfBd6bzQrkbl7TElH7EtmojEO4rQhrXtA6x61ryW+hWZvo3Jy5ZhAQ8SmcC4taOG5AH8hb8phlpKAghZfQvu+yWinI4ljUqn9ygPHIbwGbupyGD6R0cWG4rNmQpz4DmSvM4scTg/Q5p+ZqCWjVD9C4VkmHciWVlO2lLka/4CE14K6+CIs/DwqLZ4NRbT9UJzW3aHlIXT4dK9dbGFcf8EtQcjzdEcWhmsCDFqAyAfNQlWNzXFwetDcA+jA/oz5C895/sI1te3zQutCyU+DWEw7wSncV8TpGsdf2BlSF1BZLhcyLbKX8lLh2cF4PnsUf72OERQWHwJzKstVJTQH5twGubvN9zO8LrR8tmltzctxek7BHZwHmQvG2bOIO3Qa0v5j32xCy7LNtC1vikhIRP4jItMnvBZas0JsZuxkkamn8i4h3Ny/vVYww8CI6nvapGT6TuI7QNYMGC0E0U9D8kyf9PssFJ/EmfcAssKzyNz7RlcUasYIrwnZW9D0X1F3yEt/nLhQIc5qEN0Usn4p+dUUzT6a8lIQczD0S7T4unmYp68y2ybdVaKcqMk26DicSlE3MMJxhwKKDh2JmUSMrO4WQIfR5Om1Q7buGnAu8XQQZg5pP8j8LSWOaf/Valpsdf56Rs9rmTwiMST+VaTtB0js85NekQIgG5A9VHwZTV/uX78sLUjhucmfcwR3wNM6KR+7w5QghOaMf3y9dM2guuvic6b0qnrCoQV04AC07xto8vfo8Knowq1xMw2snNRBoh+GjiOAuAm0ShxCayLdF1qnaIaYCX9BQqvj2xlEWqAsQDyb0cJ/0UU7mRLPGn/BjwQ4QQLKjpm8J0/DN5W/8CDOvPuQFZ5B5v6rpvvT6BihlYxQafoq1F1ssoUmGtgAkJUg+lmTqTludkPJnKPcXxg+AS28CLiQuQxzTxzfXwiVzcf8v84R02HC11+Ie2WJjo+/YOzTYZ/AbWK3QHs0dcm4Nk8FadsfIh/BXMuIv7AK0vmb8Q61NIKmqL8A10Cmb/FZ4/tWIRJGYlshbfshsS1nRTBhBKfzEKT7UvM9rkf6bHThVkt/Wbplwkznp/XHwHNAx3g7NorEv+q1WvPq6AHTfqx7qVht0MIzMLAn/tHKPMgcs9omYUBBYkj3+Wb1fNH2VK4KRqDtF4gOB6sOeyq+ImHc1KWeamw1YSi+hg6fAoRg+NfQdgiTWv1wFxlRtklTQJPnI/Eve+JOUxhqFMdE4rtORZx26L3UiHoWnofwqkbQs2Vzk7bpKzbjQv5fNVsltDpKnFrHzIXS/OkwPBCROPRcAYUnzKptaHVzjWLLHZYPgoIRpXFq/IODfJp/HE3+3og2htdH2n+EVK2amUyBvaD0WtXBOeOkBo2trlfq0YrjxHDDHzCrotVEAjQ7S2/hf81FKL3AmEPvOW6Dh6Etn62sI54gTmIXNP4Vk60l7RBezwYTZpZp9xeIfRGGTzLpwqOfEccEiVq2nLbTzBRaeAHt2x3/bk0FoNu8NhqcjCDd55pU/EVfNb7E6EM0YjojSQEN0hDwugSJhHHTV5sSixrCUHrH+AsSgqHjoO1nQAsm4NE4rjsEmZtxJv3YyqOpP3jdWyYfoBWBUglyGYdYQnBiGyNdp5uV/94rPH/haQitjLQdgMS2MIEU3+e8C/mHa88RXgWVhE/GjHrlazOHSAv0XGwyV4vPQWhViH7a+gvThA7sN94e3v8+AqIjSNgrz1rD//VZirRsgs69GxaMk5HrvoG+9yGY98CUnsuWpYtpCSiIyKrAl4HjgInJ59cb12mH3r+Yvu6Fp83G6CeRzhNmVeSuGnUHTZ1d6iLqpj5JBzL3dpNmJwmIfHRsJXPO9cbhzz8GoRWR1u8jsc8bcSAR/8l3ePWxn1N/wN8xqbJHgeGjJnB15UyDsnP+HmjdjUllSNQg0HooTvt3xrZEPgSdJxvNiNBqSGhFI/Y5+AuCb/adtdsiHwmwMQ4tn58G2+sjImYFyXcVybJME9/ZE/Oqvpco6FDAQQpRf0E2zf3TtFIbGS+/AO17BHouqNQoKD5jBNdqPvcFNH0p0nlC5biqphtK8pwxjZK2g5COo9GBPTwdhhJGFDGKdBzhb3r4A166dzUBJWIShtw/JyR66YdIfKkpo1uamTl/IQG9V3n+whNmY3RjT3yzusXp7EHdpGkbXaNpUIUkkLl3e2n5LaaGf8QP6r3edCrIPwSheUjrvkhsK7P4IFH/71OoLGsjdU7AucuekSM+R3JimXGqsPDdMN1zUkSm+mfI3Q9tB3j6CZND1awv5zIO8XkHIe3fHQ0eSuQD0HWS1x1nVSS0kunINXgYgQsufhOmyAe9EtdqYhUdnGYK4y982PyzTBtafNlbuGtg9cuZZ3zy/CPUfHa0MA2t2ZuD4zi4oY9B6fFx9syhA/sjvTObwWuZPUzXrPw04GCgPWgHEdkX2Bdg9dVXD9qt9rjw2kjvVaa/u4SQkRr8JYxqCXJ3o7m7welG4jsh4bVq93NT6KKveY74ODcd9xV0wWbQfihO6y4VL0l4DaTrlJpDRKJoYk+jXlwxoQ9BqR/3vS0gsm6wKNBsw+1DCZtVpIZbTwVlUyikTsV1Yjit3zCOwNCRkLkREy1Oo0S9bIigc8Whde+arRKag7buXfW+t5hgjxWhscwg0rYfmn/YTPC1iPnsu/jfXyJACOn8nVmpqkK1gA78hNrgRBYdOh6ZU6bS7C7CvyrO9RUw0/TFUK6UrgMwfDza8Wuk9wZPzOx5CH8Qad0rUDBSwquisS96auUj37WQd21BgcdlL5tANQf5/5gJYeTDUyubmV3MoL+wOtJ7uRE1FPERuF0yqLqQuxfN3QlOBxLfEfFEkCv3y6KLdgD3TYJ7vY/s/Da6YFNo/2lN62gJr4Z0nVhziEgYbf0epM6teuaFwE3jLtgSwmtNoO30xBGBtvbS1IMJADpgUqmdbnDn01haY4TyEgsRk2zRPbcImTPRUARp/Y5pzTl0NGSuwfgLGZQIxt+o4y8k9qrZKk4X2rZ/1fveAqG5SCKofNIy6ykt8kQv6wT+RnDnI51XoYu+7JU8j/iscUjssVS3I5bu49FF2zFu6VLhUdzMP0wZh834W+aZckBBRLYDFqjq4yLyuaD9VPU84DyAjTfeeMLJ7eIEp9jONKoFtH9vKI6IRIbR1CVo5wk48S9X7pu5ZgIPOoAkDB+BW3rJu+lEkPgOiM9KtKqXju/0QvQTkL+37NWS55QAeR8hxllLCPp2YkIZCqG1vTZVfjezDCRPQxO7oqlzy7pXjKy0ZOv/aeK7eO0sa5G2g4zidvoSo6wf+yKS2L1pTqtleSEC8V0hfb5x/AOzEiIQ/xbStpe/8jigQ78GAo4v/q9quA9R290BIAbhD5nuKJpHQ2uZ1r75f1L75crC0KGodBiBqa4zAm0rRzp/g0Y2MJ1rNG1W9Vq2gMGf1QYDtdRYa6ulCM3egQ7+nFHFcIlB97k1ZSlLG0vOX2ie+JxqyaRFFx7z/IUQmroM7fgVTqLq2ZK5sbFgwigpGD4Gt/gS5pkWMlotPpk1xl94GGiFyKchf2fZqyVwvc4Ieb+OStNHsQh9C8LEWgs15Q7qwsQy8cPQ9zXMfabBj4Wzjlm8IV8bdtQMJM9EE99GUxdD5joq/YVx/JL4V5AAvQSn7Qdo5ANmXHcAYlshiT0Qp813f8tSQGT9gGeiH2FPaPEGNHmWyaJzupHWvSH2lRk1c6aR8Fro3Htg4abj7zx4AJpcD+b8zQYVlnGmI0NhU+CrIvIlIAZ0iMhlqvqtcY6bFrT4hhEmK71uyiHiO417w9bCi+YY912Ibo4kdqg/KczeaOrRRqPU3irh0GFobIvKY3N30rhzUEb6Iu8HQTM3oG3fxWn74ZjN7mK0bzdjsxaYTMuml5+JcdsVPSSHQmy67SCf3HqIUNMXvTzxpQkd8gpmBSHgPdDFxpFLXUrdFNIaQkj7DwNveiICsS2R2Oyvx7UsO+jgLyB3awMZPE79YIKb9BzmoMN7Kn4Vpwdt3ccr3Ro5d9SIFqb/hOJivr/j3e9KJlsheyOavw/m3DL+6owmofS20WsgbsqVWj6LxneH9GXeeUOAQudvlyknXYtvoYt/QsW9S1NGz2LeA03L0psmmuwvvIVmrjAq6pGPI4mdx63x1eLLaOpycN+G6KYm26DeAkf2trJgApjPagmGjkJj21R8VjV3F5PyFzKXez8Imv0bGt8dp+PgsXHdYbR/d9NOUYtMusXjNBAKwaprFygVqQgo5HNCIQ+t7ROJFyn1FfR9cF8EIsE5TJoxIpjpi5iYvxA1mkx1IiLS8jlkCZQ4WKaOuknI3QsUTDvOUG/Za/2Qf9RoB7XuA6k/MW7Jb8wsNkpoxYYEtFUVCv822X+RD9a0lZxtOKG5uO1HwPCx4+9c+h/avz/S+4eZN8zSNKYcUFDVXwC/APBWHH62xJyD3EOeQIr3wMz9y9QizrnOtOLzwc3cBoM/N/tTgtyDJlW395pAp1QzN+F781AX8v+GFhOlUy1B8e2pXpU5V/I8NLbDaPtCHTrGBE0m6RjcdHEP5x69CsW84LrCAzd3ssEnUhx72atTEEkaD8/hr+swTaYWciTtO4iYmfRocmLDhte1AjKWWYUWXoTszTTkREc29A0maOlds0qWf5y6JQOt+9ZubTsQIhuY43UQop/xJvSNrtKUUwI3iab/grQFC1up5tC+nb00bO9+l/wDmn8cp+cCNLGj5/i1mCwhn24TWnjWCMkVnvJqyn9gxF+XAjRzLf5/J9e0FIttu6RNmjaa6i/kH0cH9h6bYOceQNMXmDbBoXn+x+TuQQdGWguP+At/Msc4Plo7gGZv9hHjA1BTTx3bwvymLvgovU/wqsyEOH0pmtgJCa9jtg7/xhNdbF4gYYSR+HwoDOmkgwiEwsp/7mtj488Nl+05srpRLytgMlpLpXGOE5AOL0N0AoRWBmfuJOyxzDZGv+fiGJENSqNtSt3UhTB8qieeDhCFth9C7g5w+0BWgOJjVPi5zsrQcUzj5y8tRPv38BYMBSia4GPnibO61M1p3QPXzUKqgRbVhbtw80/gRDeaecMsTWH2KhuOg6qigwdTOdHPgFtEk2chHUf6HFOAocOojEJnjIpx+hKk7QcBJwuKRObQgR+hUoLop0wrS3e6FP8F8vdB+Ju4ruspME9OiCiTEi45ZUXy2bHIQTYdYmBBmBnp7CJzzIokJUz94iRWYKZCZCNEHDSyERT8e91XEgWJ2LZKltlH4VEa0weII11n1GzVwgto/65emmadyUVoLU89vRKTlbM14rV/1My1RvNkUgEFgBwUxhFzyt4C7sIqe7OQfwwtPI1ENjQ9uqtQLZmAcvpPlRoyxUF08DDUXYTTuuck7V6CuAP4/q20BO7gEjdnWcD4C4dUPcuz4BbQ5OlI53E+x5TQxYdS42OUFqCp85H2nwacLMhfyKODP0cHSxD5uBFlK70xySuqOalJqQ6vY1Y6MzcyG4IJ1bz1cpSTDliDdMrhhye8bVo6yjzQPsb8hekQaJ4AkQ2M1kRk44CyrWqMDpN0nmRTuJcB1B3ygoaZyj/98G9wnU4YPg3IedlyAClInY/Mu39UFFULz6GpS03nhvhXoOVLiJTM8zf7D0xp0pdHA36AEVjP3o7mHzcBcvcdKj772dvRyIchthWavgKKL0LkI0hil8BgZjNw2vfFTXwd+r/pZQ/XoX9vWPHfS8YwyxJnWgMKqnoPcM90jhmI+05A278CZO9E24+ovdkXy1uOlZMzE/aggILUS6cd9rLw7mZCdX3j4oDEcZOXQPIEpvKQjUSVix98jqP2Wosn7h/Twdpkq6HpM3cUB7Sfsfd5STs1AvGvmZ86jkD7v+E9COqszLZsA9EPoYNHmBZbsa2R1u81/aatbhLNXG9a7oXXN9oaTqCOmWVZxOkxCmLjfU/b/s9/pX7oGAhqGzdK3LRNa8A5VhUmnHJcQdhooNQ7R/7x4BXewn8hsqH/cYOHeT3u/dKWM5A8HU18c1oV//veHeDWC+/krRfeZcNN12fLb21OLFErhjkRpGVzNHudz3ugEP1k5RYtQO4utPAsEloD4tsuNZouS9Zf6Ato+1cy759q7ee/9CqBnQ+y/4CggILUuUePrILn/9ng5LVRQsZfSF0Fw0dT77n7zmtR/nrOXF78b5x1Nsjy9f0XsMrakw0QNo4IzFu1wKrrZul/L8J5R6/MHdf00DO3hMgKbLLVEB/dPMkSn6PHdjT2dRyC9j2OEdwLWrwJGS2X6CfQ4V+j7hC0bIm07RuYFbukMP7CjVB8GsLv8/yF2TPxnK1o9mb8/95FSJ6L//MuB/lH0PB6ZvEsvA5O1/FGZHXoeBg8Eq0IuofQ1B/R9p/htO5p/lb934DiWwS3X81A6kI0eYqXVZWH3P1G3HjOdUhopald+DTihLph7q246Wtg6Bd19kzivrcp9F6JEyDKbFl6WWozFJA4gSvf7nz0vfXQ0LpIxxFIy6e9Y9rwb+VDfScg1Egt03SvwhfQ/LOQuWjKI4UjEI4ov7rgNXb98AajmQqRqDtBQaRGWMLZCDVEkbhJCZbI+tD7NyMel7kZfzE6NUKX+TvGVpZSF5mHTO+NkxYD1cLT6PCJpt2pMwda9zcP+Aa9JS2+ZdK+yXh2xY2wT+9fA9XxLcsW6vabVbOge9YoIcRLo66hUG81IAGRD0D7wUhkvcaMkhYaD27GvH3LJjcSRRK71z8stIZ3bNVkTkIQWsX3EC29O35piBahtADCq6LF19HUHyD/BITXQFr3RybYYu35R17k4C8cQ6lYJJ8tcv8193LFcRdw5kM/p2sl/5adDdHyWYhsZDo8jKyOSwJiOyLhNccuxx1E+3Yxq2KaRknA8EnQexUSbrwzwnKBxAicvLsDnr+wNtJxGNKyuXdMPX+hziJDeA3ISfD5YJzXJk6xWOTRm/v52Ma/JxoLfga/8mycg762DoWcUCo6vPzfBG++2MJvr315Bksfx3jw1k7+/c92chmTxv3Oa2PBvduv7uaaZ59hyWZ4R8Zq3cPrwpwb0eELIHsT4LdgVTJlK7l7Gf1upi81/sKcv0864K+F59Dhk0y7U6cHEt9DErs27i+U5qN9O4Gb9OyKef7CVb7dyCwGLbwIQ8fjH4BzPd/L57uqig7+rCwTLoy2/gSK/4bcfdQ+h7wy3eGT0dg2XsbB6z77VZvwHpXBjixoHh0+2bcLXLOR+I7o0LFAnUUMXQiLtsRNHIS0f99m+SxDLIFHyMwgTg9EPsJY3V053gO19BI68H00/6Q5JryGedjXXLZAaG2jgeB3rsQOGAd3qkRAeoHyFawYhD9qtkkrYzEehczFDYzZ+JdRFT66+TCtHSUOPet1dj1g4ZJfDZg2wqbusQaB7JiatYRX91Zo64gtaV9VmmoeSgvRegJ2ddDC82jf7kZhW1NG+2LoaDR1XuNjDB2DEZccscsIR+nQUZOyybL0oIX/4i7cFl2wOSxspHtBNHjVP3C1Omr6pxeehP5v4w4eYdoUjkfu3vH3GdsZwlWTdGclxrtnSWJHamcVIZAuiG6KZm/D7dsZd8HncAcPQ0vvmA4VEhnXHtUCWnwJ7dseMtdD6WXTDrj/22j2rglcG5y055lkklnyWePwZdNC3/wClxx+KJp/bEJjlSPiIN1/RDp/BdFPQ/TzSOcpNWV8Ovw7KL1VlsmQBl1sRDwtFYjTZt5L3zWUEX/hFXTgADT/qDkmtKJRda/xMcQrLfDPApD4V6h8xk+WCMhcav2FjRjxF3LZKNm0w4XHr8T9V1/LwV9fixefCs5QWWWtPFvsMEBnj/F1SiXhM9stnvbSx+rxVKGQF/54zMqjwQSDjP7LpkIz6I+EgW6f7Y4RvB2xJrQKSI7gVWO8DMxyf6EAbj+avnJSlmnxZbT/m5B/wPMX3oThE9Dk7xsfY+gEcMvtyoIOoYNHTMqm5QUd/CnBvmEM4tsBft+ndFVr9iKkTn6URxwAACAASURBVPIyles9R/Po4FGQ+ds4+0Fwe3QXcveMc2xzEBGYd19jO6dPRRd81pR0W5YJltqAAoB0/c70UJYEENQmKosmzyg75hxvlav80hWyN6CLf+R/nsgG0HYAprPAFNEc5kYSBmc1pOMopPdyZN59ENu+zK4ija1iNO4JiECirciJV73MZ760mHCEpTigUMK/fV7WROY9tPA8ZG9n4jXfOdMKbxIYR6D6IZWB1DmNTdoA8vdTm+1h2obqjAhfLF2ouxh38Cjc9zbBXfAp3KHfmN7zSzmj4kyllzGf2QINKY/nH/TfHt+V2mCoV6fszvfGz0HmBnTw0Fp7VNHMjbh9X8dduC0Unp/A1URNuU45pVfR/t1QDdaDEacH6bkMQuuYMYiY2tHey9HUBejig43YovsOZK5DF30VJealhY7D4EHo0G+9SfhIAFmBLDp0dMPfrYEFg8x/bUHN9mLB4f6bW732nJNHJILEd8TpuQin51wktmXtSk72ZmpX1lwo/BsdtyPI8od0nQjh943vLwyfXnbMGRBanVp/4RazWOHzeZHwul45xFT9hZEVUs9fkFWg/Qik9wpk3gM8/9x3OOvwVXnlmRb2+Nmb/ODYt/jNX14hPewQ5KNHYyUOOP5tLn7oOX580puIKJHo9PsBIiaIoApuyfw//41ooF0AXXOLuDP2aCsBAz7bc5A6Z/Q3Lb7qTfYm6i/k63fQqYMmz/bKLMrJQOoC1B2vXG3k9PdQmzmmUHg8MPC1vKNuvydcGkBsCyNCHF6PsaCCg3l+BjFe9p6aUqeKYITvyU1ma1AS+Swua3OcNoj5t1KtQefDws3QwjMza5RlibB0BxRCc5HevyPdF0H7gcFfsuKLY8eEV4W2w6i9KWRNfVLhad8hnLZ9IfFN/DMiGqUAjHQeKJp0JqfdrEg5XVB8nsmLnY1PvM3loJPfYp0NM0SmYwGlqdTxPMqFMfOP1t+3HsXn0MmIoBWeDj6nbx2vH0HVSLNX8XdJoZpH+3aFzNWmftHtg/RlaP8eS32wRTNXNzYxriDj3TtqkfYDoWVzzIpmu/d/B7WOT86IQLn9lfYMH29WuQpPmiBH6X8N2tQC4ff7nMc1deT5++seLZENcObegsy9B5n3AE7v5SCdkDyTytXBkgkOpK8rU+EOQk27wKB7grvIZAU1QCQaLGgbbXGh+PwS+CzWe3wvtZHiGUOcHqT3eqT7Ymj/OYFZBGXCYhJaEdqPpNZfyJnWkAElRU7rnpDYm6lVlZao8Bd0IUgMkRDidHDJb4bZ4OMDrPvBLLGE0truEkso638sxVC//2fDdFiAaEzZYscBvrbPIu67qZNcZvo/LyNBCnFMUKGzp0ixEHyeOSsWyM1YHKzOd7FUFhjMP8ak3eLSK2ipb+LHFZ7Ct1RUwiYDqSGCJrkOS7mbP4PU+8zHIHsHLN7P07BaD6JbQ/zrEPvaFM9bxPy9q/9mYjKYW7aB9p8jc26Gli199mvxFgpmL07XUZDYp7GddSHa/63JfXcss4ql/k4jIkh0IyS+S22e3QjOymj+cdziG7jJsyDpqbbWUDJtIPEmLbl/obn70JHoceldpleBOI8OlbVbaXT1epKIQDTGEqmVbCqRDcZ+dnobmGgEEZ1callQ/2AtGXsaIb4dtStcEYj/f/bOO0ySqmrjv1Odw+TZXTJLFBHJEiTnnBZQRKIECSIZAflEogIiIAi4REkCSs7JBclZJAclh92ZnTydu873x60JPV3dXT3TMxuY93nm2Z3pqlu3u6vuPffc97zvDpM1Z6nHIT+bwt3ZDOQ/MmUmCzJK1lWWf2hVGl3/LhLEaroMaX0Qafwj0voAlBLqkmBBwkvzsyFxK4ULeI/0xOD6jhODy3vRDNp/PZq4s+JOuvhaTbIVzELP9VnOQfo+D+KTmFKKEvbARgjXm2ZKvDHGKhushM9XOOeEIjY77DMXJD7+z6nrGOGD4DqI1KJEb+GDiRdWQ6J7upTVOLAWdeKFz7H7roS+P+J+H2cdK1YjjqnpF9D0v4aYUvY3jNaZyR0Z6P3j4G/ZdJrNdusiGB5xD4ahoaXycxqOKrsd3Ma7r0T47MPxuV9EzI8/APXNedbdsodAyL1vn38UJDI6yaKxIbDS0P+tFkYvLBWA9BOVDxsJ33T3v2sWfNO8tRHZBdd4IbTlfG07OC8hVpPRECqaWwdKDdKOiGrGJOx9rVgNZ0FoqxpcPeswpcMMljtbrUjLbVhNl2LF9kWsONJw9hBDQmLm2NBGSPywGvRhfGHV/wqa78FTclv70c6TF/gNoW87FpqlpVhRiO6La71T7m2040Bo3xL6/lx6l00CYLWimZfR2eujnT83P7PXNOrJgVWpTW3kMNifD/0/vH1t255nCGK+h0o1zeOBMMSPR9PPoH1/RrUXtNyEGgZKLTDyaOJG7J5z0NxHnnsgsV9QTDMPQ2SGqeX10kbdqcZWTKLmXIka5ea6BaMmUkSWFJFZIvKuiLwtIkfXqm3Nvolrjatmiyn2CwBUk2h+ttFwCayFe81mACilIi6DSTTNvIzdcSB22xbYXcejObPbKv4lkdAmRlMkWEJ7RrMOvdtB9g334yoiitU8Ewmt59y/I5E1pTs9Z6JtW6EVWDtqJ9DkvWhqlgs1ePAovDGRBGKHUfwZhyCyy6ADhGoKTd6H9l1hFolavAj61Y1HMW3pKJFYnlAkTyhis/qGvex+WBsE1vbQl7FB4seCfznnM/YPBaWT9rcVIRKE2MEU3wcCuQ/Qjp9B+xbQ9yejmu+KIPha0cy/0dk/RDsPMWUQc9bG7r8BAt93aX+M0K8G/7vF3hvg97vf815zWbH6POtv08uyK09MicwJl3zODzYzSYVQxGzO+EN+QtEgK6w2wXaRgIkXTjKbR32Xo/Yc0CClF0EhoFTWw0YTt2J3n4lmvTK5cOzKXeKF8A5DydSKbRxv5gCJMBQvTEcazvDcj28jpOFCJ4nk6JcNzlcjE4EpSN5hSvXyX9bgyhEkdrApd64/GWm4AFofB6uuQMtNrAak5Q6k5Sak/hyk9W6spsuRinpB8wes4Heh5T48JRVyT6Ftm5cth5zE/I0F1+XBBVJ3PGo1Qf81hgqNYBgFw4PQMjerJk3tq3ZRxEToPQ3qzzUDjmZx36lzMo3a4fJaqU4P7RZKbN8S9fcTBT/mfY1FJGVJpOFoozpu90DfHyqfAkZ8yreEsyAcLVMjCk1XQe85aO7DQXcEs+MwFejF2FpaEFjBnBLaCvwrQfcRFPuHZwwdMfs2mrgNbfg9VqRy0kdC66INF0Dv2YaOTwCieyF1J3p+J2LFoeUOyL4KuY/MwiGw9oLETsgBx6vqayJSB7wqIo+p6phX/OKfjhKhcOccs8PuW2KszU8YVDNGfDPpZPElAvETwNfsMDAGxqowhNY37geJmygew+oQ//LYyYeh+yQGx4/8l2j6CWi+HQmsOHi0xA5DUw87OgIDC5IIxA4sTHhZU6ne9jUEsf2dbm8LfZc7tF23Uq4E2CkjiKh9ptQjvDVSd/ygBZtm/4N2HIAplRigio5U0A/gbcwKQ/0ZSHhH1P4G+q8zSWTNQHiLQdFDzX1qSmpImzFBwuBbBppvKnB9aVm0iR0Om8HNZ91Aut+idbEs2+7dgT8AZJ5Hs/9BAqtW9elVAzNG3GXE3LLvgX8pCG1WU1vMhRkSOxKVOuif6YzTFmbeHz7/lAtuU2jP+aDdFMcLZ0Pd6WDFwM4Uvw5A0JQhaRVU32FCxFvtvwWfPd/C0ivOHZUGgm1DIKSccsVn1Z8MjCZeiMRsTr/2U7rm+uhq99M0rYXXXz+bRE+SDbb7GpGzPbbUAv7pTrJntPFCyMQLfRehuXecZGXYoVVMA7vbiR3EJPdRYxkZWAM6D6Y4Tsua/uTeQZP/QOvPxIruWrEXElwDGi9Be84Aew7gh+ieSN2vPL8TsaLQfKtJAufeM+NVcJ0FKV6YJxD/UjBllmGW5L+CwPfRjn1LHJ0xQt39l9fiwhDeHrHiqH9ltP/P0H2CWUxLBI0fixUzOgQiYpKTge+P/brzAFZgReyWe2HuTpUPtr9E5+6HtN4y/h2bRM2x0DAUAEQsrPjBWNOeR5quNIFg5bOG/d92JvcSmfKec8wiL7BGibbSzoRfxSBu1WP3Xorm5yISgvDu3s+tOQYC9rHgawisjET3MjuUXnY4rUWwpj0LjZeYiXC0ECB5hwmuBxdLCUOF9jUhzbdD3WnGxz7/OeBHgqsaW9HInphdAreaQycp1fNrVL1pXFiRbZAp/0KmvohMexWr/hSk2tKL/P/Q1GNo5iWn3GbBEVdS1a9V9TXn/73Au4C751+1CO+AsS8c/pz5THAe2qwml5gIaM/pw9SeUyYJ2ns2xH9laiStVrAWh/gRSONlSOznYDUxxJISIAwNZwMW9J5FYZBrRN20tzCpJ/6lkZa/m89K6g3ltv40JD6CRBJYrXrxJ/8qSPwocx0JmuvE9jfvwxW2cUGx55qFWfIudO7uhiGgNtp5BGif8zxnzPFYmHElCMQhcjgVS9GsRZGWv2FFdkJEsOqOQ6Y+jzTfhEx5CqvxYjP+grED0y6nhMI21859iA4TbwO45dw7ueG3t5Po9QFC+1dBfn/40rz6VBzIoP03V/fZjQIiFhLaCIkfgoS3mUwmVAERwYrtjzX1WaPDJCEqs1yEofnBdjYPStx7vb+H5r9D8Acl2so4O6JVxAvSiN17CZpvIxAMsPTqo48XLAvCkbFQjEcfLzS25Jn+nTQNjXPYbI9mdvz5VjQu6jFekGasRZ6Hxj+Bb8XKx5dsx2+Sudm3nPHFZjBekJD57upPNxau+c8BCwmsihVax2HDlooXbEy8cLpnoWAJb4ZMmTUsXvi/6p/l/Gdo+lE08yLYXzKeelzjBRHxicjrInL/xF0ziIS3Q2IHObo/paDGYlIrCSpWgG8ZM+9YcTPPdZ8CfX8ZmuO0G3rPwJ673yDDcEGHFfgONF7l7eDcK9jDhHEnseBgoUooFMDuxBMFVharotF+0IzZaXattRW8uzM4yH8M/TPR9m2wsx9B+sEq+jM/IoemHME1//fNoqQsDFXX/mZVaN8Y8tWoyLvA1bZHIfexUXDuOQOyz4PdBtmX0I6DIPMvrHqjnk3sCLBK1S0KZN/03BURQay66hMJgJ18BG3fDRI3Qup+tPv/0LbNsOdsjd0+w6k/XzDsdkRkOrAGUBOBA7HiSMttzr3lNz/BHyAtty0wVEA797WjCu6i7p24AavhdKypz2FNnYUVP8yo/vtajAZC/HBDqQ/vjLTcihXZ1iyAXQVEFbKvF/4l/STadQyknzGe5/FjsKJ7Fu1miQjUn0l1An+9gM/Uk/f9GVIPIbHDsabOwpvqfQ7yHcbBIPd2CV2EPCZo95v/Z+6j/EJEoOla49Yz/K9WHAmsjPiGdE3U7obsOxQvlDLGZnKgl9kct59/D+lEYZIvnbL463mLmPPttorvdhLzCdTNAcAFlvPdekIa7C6k7uQSpT9U0dbA4Z9B/1Vo+7bY2feR9N3zpVPTgMNDJeTzeVJd/wIwLKrguhXO8IHVgP3NatC+AeS9z8fFncw7GkkujND8bCNC2/0bIyBrtxnHhM6fYycfxao/EWm5HWJHlk6Wiq9o7C2HoXih+jlMU7PQ9p2g/3pIPYB2n4HO2RR7zrbY7bti999e0hZ9PsPRmM2HcYeqjSbuxJ77I+z23bD7r0dtM3+VhkfXjZIII/GjkcB3sZMPo3PWg9RdFCd/FLIvoO27oZl/j/Ga8wes8CYQO9Tbwf1/xu65enw7NImaY6EqeShA4AdmwigHiRiadDVJ+vyHhiYvdSMowwA+49xQNdJGkHHu7hTRuIsQgcCakH0Zkw+aV+URZdB3LrZaWHX7QfNf0d4/QfJWh1I4QCFVTIY/C/lPqInYpdqU9aJ3LSdJoZ1HolYr+L8H9ScZ2mDma5cG7DKBYe2gmoGek0f0NQl2EmhzhPLPgOwrSMO5496fsUBE4sAdwDGqxT6fInIocCjAUkstNfLl0u36l0FabkftPsAylM8FBGp3QMfulFxMlKnRFKvR1NzGjxjxQoySC3+rdejaqVlo19EMlUV8At2nYOdnQ+5NSP/LjIvRvZHYIUh4OzT1iCM25mHHK9eOdh4Amdedawj0/BaNHmjYEJ52d5Jo+lXEt2yZYxz2ETi7h6UGcQtCO2MFlvNw3YF2K6O3s59s1p0O/+XHIUyZyuYerzmJeY7Amk4pYzlEKK23UwK59yAywzyDRfep36G4V8sSyJgynbl7Unn+jxjNlMyrmHhhYrQSclkjxFgJlqX8+5Gr+KZjKrv9cgbSNNPYKCZuckoQh8cLQSDvlFDVgq0nholQ6uPv/wtu8QLdx2L3TjG72fUnmu84/blLA1pi46m2UM2h3b+iKF7QpGHc2pgy0OwLxmp9PoWILAHsAJwDHDfe19PuEyD1BIPPRO9/TSLbWhzsT8fpqik0PcvEAL3nUblcJ4l2/RL1LQ4SQqI/gdDWC2wpi1V3Anbmv5D1IF6aOB879wrSdDkyaqHUSUwkFtpvSfxLQHQvF8quYCamsKnxlRKK5+6tIpowNNPma8GaYiYMiTttKmOb6MpN9n5TQ994MVbLdciUpyCyE/NtTqj/HFNDLCGs+hOxpr2KtcjbSMv9EPmRUYEPrsvgLuOYEYbgOhBcy/1l35JO4sINGeNpn3kM2rd2aG8u943VavQWxhvZt6i4K6xJSN6H5kZb/zr+ELPVcgdws6re6XaMqs5U1bVVde0pU6ZUfw0rvkAlEwC07y8l2ARgBBYrsXpczpIQRHajSNxLIgW7Atp7Aa6siL7fm2BKe0xStO8KI0jb9weQJojujTemQqdjvTZwDQVykLiqOqpo/kNTmuRJ2HWAej1yOhOI/ARp/J3ny4rV6NRLj3yvQYjsPPhbfXOcYMi9b0utYBS8JTovy9cmUQ3ENw2iB+I67g/EC6HNCpJzHloFzZqd56arDbuhpvFCuWSCDwJrII0XYDVfj0x9yiQ2PLGEaoOsh/yjCKy1aS+N4fN48+l3jStN3TFY014x8ULrQxDZy4kXNsDEC7VIJoSMyHbgB7iOa1arsZh1RdaJF56A9u2Mto2b+KbEHSHvcUbuAyp/JklIPY5mP6xw3DzFxcBJjL3utiI0+65xiyqIuVPms4zsQXWsvCqRehh6f4dn7Q/7G6OllXkO7T7JaG0swLBaroC6C70dnPknmrhpfDs0iZphoU0oAEjdqUjDHyCwLibQ9mPoTDYQgfTzoF/jffAIDi4oxb88MuUppPFKo6pd/xuPmg2jQQBpuR2r5RYkbGrExdfi+MvPr1C0/4aiv0pgBSSys9E5yDzP6AWVRiB2iKmtT//T/fX8Zx7ZBWoWP77lMYuZmAkCrSlI018mJDOsmi2jZj8MVVIqJxJiPqhrgHdVdf7dFplgqN0HqfspHQCGkLpfjqptqT/NJEkJOjtjEYgdhkSGCYPlSyWglMI4LgWZp43AbeoWSNyGt3FyrIskB9n3EPEjjRc5SeFKtdV+41AhEWfBFoLQdoCiiVtMKUOlnqua2s3cBxRsW0oU/MsisSFWiM/v46en7U4oWuj6E4rAAWdsYpS5q9WfmMQ8hVV/PNJ4MQR/iFkgDo8XQpB9Dez/4T1eCCGBgXhhuqmRb5rpxAtnV69P4hl+pPkWrJbbkPDW5vpWM4Q3YwLWagD4/HDNOYuS9kCI8Pthw+07efT6u4teE/+ySGQ380xmnqV28cKBEP0JpB/ClaJgz64iXrga/CsAASdhFAOrBWm6ekJ2VlXz3uIFxNzD8yFEZEdgjqq+WuG4Q0XkFRF5pa1tDCVlmVdx/d41AYmb3V+rGdKMeo7UpHGbyH1Syw5NOKzYThDzKFLeezZ2puxtMYn5BPPp9nZtICIQ3gpN3ovZxRpOUe0cNreK8zN8EAlhKL469HtgZWMllrwTCW8LgTWRkFPzl362xBA0st1RwPc9tPNoY4EYXAepO94IqfVeRnkV6nmMfHH5h9rdaOchTrlIrRDCqjsKe84mlA44sk5RZ5jKNFGF/LuGDhk7Egl8F4LrTYifs+bnQPexePtexbBk5k9sAOwLvCkiA0WAp6rqgi4SMmrYidug52zKlg74l0OJg2rVySuRINJ4Pmr/2tT8+hYvXtT6FimTVHDttfPvxFClh2BYSxLaAG28CjoPKH+4b0mk9WHIvYvmP4Oe30HmSWeciaB9l0DLrYh/ecAkD8i8CLn3wb80BDdCk/dA/7UUBns+8K+KNF9X9PzvcdxOROJhbj7nDjq/6Wap7y7OoRfsxxpbV88wmcT8AQlvhqYepDhe6B7GKnJ7LkOY+2bgeQmSyq7ALae/RTrxGhvOWJdVN14ZGRBozLyCusYFtYgXVkK7TzQuS8G1nHhhOei9FCU3nnuvgJlmRWDGIW08ensz623VTeui+bJaD5YPLC0uM1S7D+38mRFmrRn8WHXHYbdtRfl4oRtv8QJG74UgxA5HAis78cL4h9dqd0DXEXhKFInlOPfMl9gA2FlEtsd86PUicpOq7jP8IFWdCcwEWHvttUf/oPhajCinunz/Wt7G2CAMwc0g+4Jzn0yknpUFmZeNy8kCDKvuEOz+mUDlZD8dP8Gu/zNWdKtx79ckRo8FIqGgqVlo36WOrcv3kLrjigS2Sp6rWaf+t9wCTQG/2dGyGiG2NyIxQw9OzzILS98KZjc4+wZm1+t2iOxibMhEDH1foi66CgNWVGNAfpgoS/oxNPMchHaGvHev43mC4JDivqpC6l6092KK7RnHCN/S2MnHwXbTPSjoEIQ2cVgMlXY6ckaDI/MvJH6wp25o5g1Dz7LnQmgLJDqj6l1K7bsM7C4PR1ogjRBcr6r2Jwqq+gzjyhucf6C5zw3bRuIQ3sz1O9fsO04yocJ9l3sb2jdGiaJ1J2DF9il/vBskbsaMzEvY1uJGfFE7DWU7sq9j5Vqjnb5xgQWhjY3oqN3mMDoqBGy2EdWTwMrOM9jG0JifBE2h3ScjLf8wi5SOfY0g7mDd/MCYPXKeyDu7ehlG0ppFhB1/vjU7/nzr0b7RSYwDNP2MSSDlPofASkj8WCToLclj5qkHKb+DqEAAgtuArx6iP0WsJideeBwI8NF7q3PyjC76e+5DbZuHr/0nG+2+Hided6RjA7emKbccl3jhraH/p/+JZl6A8J6Qe2NCBuSBxMG0JXPsuF8HuexQkqEcvr/pmoP/H/getO+iEsKsY4BvCezU48ZdpiyCENrUEXqulFTIA0lIP47EvQnPafYtNHEj5OdAaFMksmfV5XumfK6DykkoMfFpaKOq2p8oqOopwCkAIrIpcMLIZEJNEdocb+V0brAgfqRh2lqNhombuNthL40ns8GBWGadsjBg6pMwZ20qjXmqQPeR2KkDsJpPnYieTWIUmO8TCnbiLug5ncEBPfM0OvcVaLkJ8eTLOpLKWwISRqK7IMPKCKTR1Plo/mu0bWsKg/AkpO6ByG6oJtDec8Fux6yh/CYJMWA5VlOoaTO1APi0Jq5Bozsivqloz1nG0rHSTqd/HfAvY44b3CmqgPwH0H1E5eNEkMi2aP1p0LYplWln6tSDV4bdfyv0nou5RxQyr6LJW6D579UFCekncX/PDp2SjEl0+JdFGv88KVYzj2H3nG+cOLDMRN8j0HSN8RYfBk3cSnWL+AT0noktfqzoXp7P0vw3aMfeZoGtaQrupdS9mMCyzllIz6cuIVYzBDdF2zZykmse6KHahaYeQiLbQ+oxip8hhezbqJ1Aey/0WHc8AAG7D3yTJQzzO+zkg9A9TNA28xza8Ro0X48E1yx77hA8LOglhES3HSwpAJDG8wDonN3FsdsdQSY11E6qP83Td7zAVvttwuobJtGec5wE+LB4QfN42g2vCk68kLy+xu16gwgEghQ4PpRKLmy29ZVofmvEt5hJziRupny8IBBYy7AvSEHqAbzFC584u/oVe4+EN0Xrz4C2DfEkSpt9A9V8RTajnbgben7DoBVu5lU0cQu03IFYVQh/pp/AfRwLOqVfSSNW7VsKafrzhLAmFgSIhKD5JnTublRffmBD35/M5k9gFaTxUjS4HnTsOZqeOP9Wk4gImI2xhQCWFcNu/ZdxbCmDgfFC09djdyaxms6agN5NolrM16sRVdt4ObuIiGmvN1EPkSAE1qCyyF3esYRyQfop9/M1hfbfiHYeDrkPGUpeWBDaGlru8dTH6jGfLgRGQueiPeeg+a8heTuVadMB0DYk9hOsxj8YBshIkTl8jDoPpnkIboTlmwLxX+JJoMqDLoba/U4yIcXQxJCE3Odo8h/V9dGqK/1a68NIyz+QKQ9jtd5jhEcnMc+g6WcheTMmUZA0O2naZ8QMRyrG5yuxZ0qg5yzsORtiz9nY+M+71Mqq3YvdfwN298no3H2MS4T24x5cqxFeHNUY4mf0om5xIOwkxeIgLe6HhXaFhguh9xyHZVBFsNd3ibHeLZnEFZP0Sd1XXbtWQ5VifJOYF1BVR+zMxcmn93xPbRi24fpUDo1sUz7kglceeQPLX7ygTPWnee/Zu9GOgyH3LgXxQnATmPIw47PDWbpNtfGkczBWDE8giFBkKSkCQifa/Vs03w6JG6gcL/jAnoNE98BqvAAiP8Y9XhhtqWIeQpti+Rqh7gRMWUslBKl076imofe3mPt0YBxOQf4rNHFzdV2U+tKvtdxttFxaH8Sa8gCygFDkVfVJVd1xvK8jgRUdUc7RIAukTQKp80CjxzYqKOWfeb9JDEncMEysRZHmG8y6ZiGB5Z8CLY9Q/OwWQwQ0dRt24t7x79gkqsZ8nVAwu2wl6G7Zt9z/7gJpONvsypW8Yf3gX94MMBhGgt19BnbbDtidh6G5r3D/qCyHDjsygEkbsR+rmaoEHxc62JB+CJ37Yzxlfys37wAAIABJREFU98lC/mO0Yx/U7kbqz4DI7pjvLQDWNMfzuVrdCMfGreH8wey/Ff850nQ5BDY0bUozxYFHCCIess7Z/5h6vCKkIPVIdV2N7u8i1uWH4HpYvmYjBuor4Xs9iUGoKpp5FU3e71nASDWPpp9Fk/d4cs/QxO0lyndyjujTMHjeHR2JrLGWs7+B/qvRjp+ZhdNAH3Kfo21bQu+FkLzT+NRXuSixbXjrxRgvPVFHf0+ZKSG0G6OmY/sWR1rvRpr/ikx7AWn5O/iWofCZC5hxs+csRlWSkf8Y5u5EyWRBYA1Ewnh/DwKEoe70Bdam61sF7XdYgi7Iere2l/oznIVaKUaKH3xLgN8wJDU/G7vnbOz2HbE7DsFnfeq6A29ZwnqbP0fxvZ02pRLlrF+LMFq69hDsPPR1+9hvnZU5YfflaP964nav3R8nhcyT6Nw98BYv5CD/Gdp5AJqfi9T/2ggtDsYLUxwHhmrHrJD5qT8bsZoAsGIHIE1XQmBjsJYAaaU4JjROOxXHiuw7LucCpCH1aFU9lZibM4kfAqth+Rcx8YJ/yara/DZB4j+j9HPuBTmjw9N/Ua26NAx+Mw75FofQltB4BTLlyUGh14UJVmAZrEX+A/HfA8UJxyL0nICdfmn8OzaJqjB/85+sOkrmPEbsDqjdD7l3wGoaFN4agPiXgymPo8m7HSZBENIPg90P2EbosPEPANjZT2HubpjseB7yHwHPUXq3r8RumIJoLxrYFLKzPLzZWlgnzqewvYjcDIPm0b4rnM82DY0XGQFMqxHt2MdZNHlEZC9zP4S3RXyFgkQS2niwxEU1hXb+wgi1ScBQwkM/NIJWlWDVU3LH1wlIvEIiexpLo+TfQYJOecMKg/fnJCpD821ox35DlGLNoeGtkIYLSlJRNfeZqavXHsxznUcjuyD1Z5UJEEvRk22KFg3hnaDvYsa2A5k2Y1z2FXDE3bTnzDGJQn3yfohTf7IsiT4fIpDLCof+5kt2OqDD5fJ3MGpZjPwHiH/Zod/9S6DNtztU4oGxz0kE5D8a3TXMyaVfCqxu/g1t6egylDrWZxKMwVWR2M+R4Opj6M8kJgwSMYwyt02IEWO/2gmjVyINZnwd9oyLf6lh8cL7mHjhUbB7AduxY7wIEcHOfQntuwD9mPvpA9ZZ7yXs/HeLuhAIwWJL9+I6BogfsdvR8PYOdb8SRs9SVB1KIl5y0pJ0zQ3Q85KfE2Ysz7XPvoc1r7eZ7K+qO15zaN/lGN2jFDT+wVhsW41o56GQL2X/6ILwDLNgC2+LjIgxJbQBEjLUbNUM2nUMpJ924oUcBNdG6k+pfA2rzilvcXutytr48I6QfRsSN5l4gTz4piONl1TXzrcUEtoYDe4Imb+PsaVaM4sEUzLdAbkOyH1sdJpa7zPaWQsprPgM7P5rQEvbmw4O1Z37YDfdiRVaZWI6N4mKmK8TCiJBNPoTSPyNwuA9jMR/Mfib3X8j9F7gqLbmUP90Y880bEIQqxGJHTD4u+r/Qf4LsOLGUgmwk49C93EUZsfVubbbQsRnHm51EdETjHVQs/FzJ/NUhXe7ECcUqkbSoTzmATW7/KHNoeFCJLo32v0W3lTnLaT+VGdXsjxEwkjz1WjuYzN4+5dD/Et7665/ZUOJzn9OwcQiESRana6QiCANp6Pxww0t1loECXynqja+7dDu4x3BreH6AY+jgRsLxoCCc7qONFZhwwP11H0QXBuG2y4OR3AjR7BrZGMJNLDaUHVk7jOYuwc1EVzTnGFnDarFP8voFhcW+bzNKT9ejo45foYnCq4/b1G23buDQBFpaiylVhaqGUg9jGaeBWtRp5xoAsu3xOzqSt2v0MzLjpCj2zjih5a7sPzzrSL6JFwg4kOj+0P/dRR+rxGIHTn4m3FbORfEZxZ2vsWh+aoC5pdY9Uhsv8HfVX/jxAuxwXhBU7Og62hGJhaj8ST/d9X/OOvgJbEsNQv4vLDP8XMJhOJGhG8kNAfWNKThj6idhcxjlF+kFI4j2YygqgQ9sPJvu6yVv12yCKnEUExj54Wudj+vPx1jrU1qLII47khD8hYGS0jSj0BwA2i8FIn+xDzrnuIFQepP86RhIBJEmi43Y3vuI/BPL0yYljvXvzzqWwLy/6Vg/JMIEtvXUxtD/RCk/mQ0drBJkFlTjSvVJDzBTj4GmWK70nkKaXCYj8PXIRmwO9H+G0ZtK73AYMqd9L6/PnWNfRWFXOmcgV1/KVZ0mwnp2iTKY75OKABI3YkotuODjsnCxo81to2Apl+E3j8AqaH5N/ch2nko0lq6zkbEAv9Sg79r5g3oPoHSVDu3hYBTR+VmLSSNaOZ1JLAK0jTTeAVnX4Cuo2qvWrxQYthiUBPGmSHzEoS3h8wLZge/0mIksg+auBPNvQv+7yCRXSsGC+JfxhGF9A4RI8SnnQc6isuWYTjEjkJCo3NhEN/Uol21SVSG2l1OucFIRlEKEreAS0LBBIWfUnQ/aRJN3IyUSijYneBq8xaArl9iE0bCW6KpR8agWzACEjALoOHXGpV1rM3bL7WQTFiMZB2su1UPatfAvm44ghugc2eYRZkmMFPPaPo90s7XKywkvAUA4muFKY9A6iGjg5F6lMIFhw1dB6ItdxrxrkksMJD4USZxNeAlL36IH4UVNc+wZv5tkgkkh26h/P/QjoOg9aGSbCQRgWHUcc2+g7okEwbwg826+Nu/u3nxsQbSSYsfbN5D66I5sFtxjResBjTzGhJcA2m61IkXXoWuw/FimViRIjwMX30cLkgmDGDLH3WwxkYLamwyLD7ThEm0Zp42GxGRH0HyRiqOv+E90eQDaO4tUwIb2Q2xymgU4LBZhsWRXiHNM9GOAxydGAs0A7FDkdCmVbcFzpjmWziE+iYKmnnd2UCsVpRxvCBAyJS9Jq4190QBMka0eyFPKFhWiPrvvMLDl+7ENj/6sHJSoecobOuvWOH1J6R/kyiN+T+hIH6k/tdo3fFG8dtqRcSP5r4wKvrJuyjOPuch9wma+6io/KEUtP9qRmelljGZ6q7jQTuH/mx/DZ37oQjIFKg7ESu6C3b8GOg9j9EF099iaBJNP44VWhclh7l1hw+4w9VyIxA7DJI3QDIBJEEixnq05e8mCBjZfOYNNHGdsSYNboDE9h3cifIC8S8NrU8YPQXtNLRYqwHN/deIekoEwttU1eYkRgFNU7JMqpRdqSaNWJ9bUG6XcWmx23E/KWtKZ1A08yKFYp0j4F/NUKntL6hcN2yZOuuQsWO1s++Y0q/8yGRIAPzfNfWX2WdKtrbiap3senCA2y6dip0fmrUXXyZNMFxjCqf/u5B4kaExdjTjnw/iR0NfOYG9EqrZ4R0KXIFEQoZ5ElgLTT0MwDefB3jn5RjN07Ks+sMv8CUfgOiMUfRzEvMKIj6k/iS07miT8LNaEAkYp6bEzZC8m+J4wTZlebl3wKsddf91VHpeY3XK5jNGshfTSNNf0O7jjL3wYBfmQNdBTrzQCvFjsGJ7YtedDD1nVryWz6f4PEZz31unn6fubSxIKqyybh9Hnv1V5eB9QYEm0NSjWKFNnHghQHF8N5A0jUD0YEje6pRCJYCIUfFvuc2VefDha//jH3+8n6//9w2rbfo9Zhy9A03TvFPRxbc4tD4KuTfNJkRgNWM9mvvEsSoPQmgbkyiYxLhA+65gvrFP9q0A/iWR2KFgNaH9M0scN21i+zWPYFkW2x/9AN3vr0e8vqPyuNR1IPbUN7GssWvLTGL0mO8TCgMQCQ/qJmjmdbMbrBlKBqbiH/Qm94R89WJmRp15QzT3gXvZg+kt6BzoORG7fybEf+HUxk8mFKqDDySG5tsgeR/FAdbw7y4HiaucnVBnoaVJII32/BZpvrbgTDv5AHSfwqDlY/YdNHkbtNyD+KZ47qFRCB/yOx+yFFTAgp7fQePFSHhzz21OokpYDrMj//mIF/wQ2sr9HP/ymF3DkcmDEES2K3kpCf0QTd1TQkdlmNtHOeTeNzto2Vcg9z8Gdy4lYmi7+S8g91/zt8D3kYY/AD7sjoMh868RjflAQuBbBmm+Ds19Dh3PUmpcC0dt9vrFbJb7XpKzD5k++Pcv/hcilwV/LefmxF8Zc/AmESeBXAoB8K0I0d2g/xqT8LFaIf5LrOju7qdkX8dWPxcdN4Un727C5zefVUNLjgvue4Kpy7Sb3W5NQGhjpO4ExLfo2N7HGKGagtSDaPY9Rx9mB8SKzdM+zW8QCQ3FC9m3jPaOZim9G2mNIl6olnFkQXB9NPc/h8nmBgVtg95fYyeuhvjxeHIpEMjlwO8hott0ly5uuXga7V8LuaxJvh55zpee38WCAZN8VbvTYTO6JWQGKa2QGCiTGWA6JEFTaPdpSEuhRfdz97zMuXtfTCadRW3lo9c/5qGrn+CK1y5gyhIl3GtcICIQWHXwd7vvUuibyWC8wO/RhvOxysxBkxgDcu/P6x4YSBPWlELdFPWvaBKcBeubCBL72YR2bV6jboXnYM4qVN6AsGHOGtits4xrxCTmCea1/M6ooN2nOEF8mZtM86a23SuC61BdfsUyO4Dxw43CupdkRP5Do/VgLUktVJrdEQLfkixAuSKP8CGRXcyiqyINOevQREcGfLbxJR+ukq856PkthbvIGbC7S2eJPUAzrwyzFMw47afQrmNRuzKFdRKjg9GgOM9xyhh4xiJG60TA7vwFdv+NRsR18Bwf0ngBJqngPDcSAd+SSPSA0hcLbQH+FfBid1QaKUjeBg3nm2Sj/7uG3VJ/BtJ4GVbrvcjUp5Gpz2G13Ir4l0D7b3BJJgDkIXoINN8BuQ+gYy8qjUvhqLLO5j0stULS+T3PN58F8Y3Waa2kaGMpAcsqoCnj5FAK4W2xptyFFdsPabzY2P9JADLPoDl3oUeVFh69Nc5T9zaSSVsk+30k+33M+SLAmft9CX2XGbaZdkPqAbR9V7NIGe1bUBtNP4P2XYYmbq96LND8bLRtayPGmbje2PK2b4nmvhh1nxZ2aPdpTrxQhtqs2YLFXUUE16E6ZyYx1m/xYx0rbC/xwsfQe5bZvRxxrWzaPGf5PKQSwv1/bUE95jdCEeXSBz9kh33n0jxNWGT6FJZYrpTzwoKKIBKZAbnPHLHCcsgCfRSXtSpkXzMlNA5s2+ain/+FdDKD2uY7zKZz9HUluPHM0Qv7afZN6LuKwnghDd0noXb3qNudRBnMF2zREMQOKvqrcRRZBWO17FhG1p2CBEdrc7lgwrIsmPoy3sSgM9C+AXbGuwPgJGqLBW7VqXaXy+7jSFhQd8Lgro3mPkL7rzU7fcG1kOj+yAjqkMQOMqrO6ig5V4JMQVrvQtNPUlUNlnZC/amQetCxFMxRHTOiXG1zBFruwAosj92+UxUZWGHoga21SJoPk7caTZ2a3yQPNAf1pyP+ZVAJudSWVdnmcOQ/KdG3rKlX49ejuoom7zELoJEQCzLPgKMBMonaQ4JrQ+tDaOJvRhvBmgrJ2yFxK5CD9NNo/zXQeudgCYqENoLW+w0zJf8VEtwIIjuUraEX8UPzTWjiVkjeaxYl+fepnukkSPYVJH4oxA8tftVqRu0+7L6Zhg5bzjK3/xKz0BSXOu0S8AeVXQ5q54VHG9hoh242m9GBVJ1qFmAJ4CvGV2C2VBI5iISNMJOmZhXWt+e/QNOzoPkWJDCUZNbsO9D9K+67voF0sjCDYtsWn74Hc77MM3VQssI2VOrE35D4EVX3XDWNduwPufccxlTYlL8131jQr7Jt9Jzj1F0P20m102jP6UjzNVX3aWGHasrDPCgQP3qwXl5z/zMlDbkPDBU9dmARK0WiBzi2sXk83e9WM9JyN5p9naqYOtoHdaeauSh1PwNzVdvXft5+KUay38fjf28mVp9nu592UDz2uMcL9c15jjh7LkdccSVWYCXs9h9B7t/e+2VUp6n9s245P6MsiZKwGYfrTkICK6H59jHGCwP9MZj9aRvJvmLWWT6X55WHq/n8CqGurEsAnxnzS+n4TGL0kLp53QPjPORbzmxs2V1o4gbIvGxEPuvPceLfLvCv6ElcfGGEZcWwp74Oczy6LXXsBYtMJhXmBRa4hELl3WkARaJ7m/+lXzAuC2SAPGTfNjtD9WebgdruRMJbQ2RnpPVutPdPkHnOoSSWWgQLhNZHfK2D9bfeoYj2Io1/RFXRzsMg48VWcuj80khD5/7Y1tKOyFw1bdba9saBhCF6IPQPZN+rgG8pJP5LY984YKdk94PVYGpOy2Lg1h4emAQgvF2h8JY0lC4/qdLysRCOQ4UbSllGTaJmEN9iSN3x5hlr25TCxXUS7Bzad4XxLh84x7+U2UFM3osm74TUPRDZE8LbGxFXt+tICIntjwa+h3Yew6ieI/GVDW7U7kXn7gL5Njw9Q9ptfjzC5/Ox434d7LhfKRq2FyjwDSaBOF73d/lFhvacZQLzzOsUft9OIqD3PKT5r+bYgcW9dpPsd3/OLZ+STo783tNO+9VD+693POgH+maEAbXrGGh9pLJ/PTiuIiM/Xxsyz6Jql7xPv70YSGiXuycFiRk3Hs28inb8jKF44S00+Q+0/hzDCsq3Q2gLJLor0nqP0eVJ/8splyi1cBUI/ADxTUN7Hqyu+5pHtBtpPA/V36Ndx0L6QRabnuW5h/1cf96i+HzKqVd+SjDkNvaUG48y0LE/tv87o6B/K+PxnCtR465R5NThAdYiSN0JJjYb2H3WHlP2VNGK0i1e8ENoc5M4dhBriGLn3Ddd6loqu0OUhNq4flcCE+qE821C2TlytKLBXuFs4mVfge6X0W7/oEsdZCD7Opp8AGm6YtCq9NsMy4piN94IXV5cUDLY7cchLRd6m1MnUTOMOfoQkSVFZJaIvCsib4vI0bXoWOnrRSC0CeVLBvyo5rD7/258iAtq4zKGhdB9DKTuhsyTaO9Z6Ny9wGrFajwPa+rTlKfYGCtDzf6nqsDdnJofpFaam72WojC22b3KvUJNKMa1gKYgsDKjCj7ybWhwo8FkgmbfRTv2dHboysEP1jTwf8dQxQibf/3LIfWnFRwpvikQXJPi3NrY6tUkvKNDux8BzUFoo1G3O4kqYX9VojY6C6nHCv4ykODTnjMg+xJknkd7fo12n1T2Epr/Cu08yGiljAaq5tp9V6KpJ8xuxfCXEzd6TyZUDYva+lrPK8XsjLH9TD9W+nvIvjH0//QsBgLGjXfsJhAsDtqjcZvFl3X5zEebaEzeheu4nP/GaGV4gZSqRSl265gfMfHxQsBhg5WjvVvm+Uve5SQThscLTvlc97Hm+8s8Bb2/Q+fuCVYDVsM5Jl6QcvGIQnqWKYOzS2ktlTm3RLywx2Ht/PWFdznynC/5zuplxGPLta2dxn2q2sX7OCGdSIPvO1QWqHWB3YEGNxyy9cx9ZJxl7K8rnOgDaTaCnBJhMF7wTUcaziw4sr65jjW3WhV/sDBeCMdC7HHcTtX32YFEtse1dE7zMErnh0lUQGAV3JdAQWi8DPyrjNeFMWO1jRlnbMy6JMHQfW8DKbT71IIS3W8zrPC60HCVt4Nz96Ozv2sY7ZOYMNRiOyMHHK+q3wXWA44UkSrEC6qHNJwL/pVwFysKQGhb6PyZqT90XVg7fsUDGWFNmnKI5DCbSf/SFXqRRLtPdwb7UsGKW//SaOeh2MlHzXsJ7wi4LDznKWoZmOYZ/fvrhbZ1sNu2N0KcvRc6g26FATa4IdJyO9JyJ9J0FVJ/KtI0E2m5B7Eaig6XxouH1avVMVjXFhqDt21wfQjvjHnvFmYSCUHDuRWtqCZRQ0iEkjs8IxM+mefNjsHw4FqTkHrU0ONLYIj6PBoEwIpBz2/QvovQ7hPQ9u0L6/RTTzCuatTaW6OGchhbx3mJMrt5w2tm7a7B72zPI+YwdfEM4aj53R+wCUXynPinz7DcZsj0rIK6au8oNa5qmddGILwDxcl0P4S2XFB2YyY+Xqj/rePe4EbI9ENwM+g6wsznJRfWNkP3VhJyn6KJYTXzvmIngEKk0e7fOPFCqWfErX9ptPNIIxwMSHiHgnGrZVqOrX/cSWPrvGS91TZeeP8/o3U2SELbetht26CZl9Hei53SokrxwnpI6z+Q5tuRputMvNB4BdJ6v6sr08k3/pKV11uRUCRIrCFKIBRg5yO3Zct9Nh5lv0GCa0L0R5ikwrB4of70SWeocYLEDjUM2gKEjctK12GQG0/avEfWiT3XcZOaBIAV2QRa7vR4tI3O2chY8E5iQjDmhIKqfq2qrzn/7wXeBRYvf9bYIFYj0vIPaL4RfMtgBuGoswu9gtkBzr1Ndbv0STT10NA16k6iYnCcexciPwarhcKkQhjqTkamvgCxgw2tfvikm/8cuk9AU7MgsrOzmB3vpILXRaxAaLcaXjeA2J87AZ1b4FEpGMlD/iPj6uGJamwhTX9CfFOMQF/wB0h0LyS4TmmPcasZq+V2pPVupPFyZOozWHW/HFOALiJYDWchLTdB7HCk7lhkyiNYkdHvYkyieojVDIHVKE7uhSG6T8FfNPM87o4Neci8UPoi+U8Y1Y7aAOw2jMOEgvZD/nO053dDr1uVlMPHsoh3dkcqQoAY5RXnFcNQGPncDPweYDBQxrsaenl4FcaLQHSYPkVwncH/xuptrnj8Aw474ys22bmTGYe2MXPWB6y5cRr38SnnMByqRGR3XHchfYsj/iU8NSF1J4N/WYd5FQSJGfHQht9W3595gHkTL9RhtdwGzX9zBA5DDMULy0Bke8iOLJOphBQMK3eUuhOo+BzmPzEJId+0EceGIX4MMvUliP0caKTgvrO/hO5TsJMPGbZFYE3GP16ogr4f2mXMV1MFOw8XHjudd1+cA8G1cQ9PvcQLH6OdB5ta9IolaH6k8SLEt4gTL6xp4oXQ+iXLh+KNMS588gyu/PcfOOOuk7j1i79wyO/3GXNCz6r/NdJyK8SOROJHI60PYkX3GFObk3CHatZopET2At9KmMR+C4S3Bi1VHlOrxFk1LD51xvpJDMAKrAL1l3k8Oo3O3mxc+zOJIdRUQ0FEpgNrAC/Wst0S10KCa6OtD0P2P5D70AQHgTXRntNKLAzKtgjWEPVXQpug9adAzxmUnpT8ZsHSeq9RXk8/AVarEXEaqHuKn2B2MIvaSKF9fzRicP7pTkBTTnDR43soeX4CE3x7WDxIqgZ9GYAPzb6PCaSDTpvD++DxGprFkzOGb7FRi9eIf1kTrNcQEvg+Evh+TducRHWQxovQjv2M17wKkIPwNkj0J4XHWc0oIYrZAIGCsaEIgfUgNQv33c0gxH4J/X8ocbJbcJGH1EPA+ebX6N6QebL09Qd91ser3EBMUiZ+DHQeXOHYkbsBAtZyRngw8zhoPxrcCObuWcX1S+kyWJQcz6TRzAESMEyE2M+Q6I+HXvYvi0Z2cVhpSUIRZbuf9rDdT3vN9YLrgdRD+v7itjVvdo6qhMT2MyK+ubccwdYwiB9pvMR7G1YdtNxjElwDc15wA6RkKcT8i4mMFwCs4Gow5QE0+xZk3zMsxMDaaO+5o4gXMFo+DiS0Plp/JvScTOk5TRBfE7TcZcqYUo+C1YTEDkBCm5gj6o7HTv4D7OJ4gb4/Qngb8E0HXmLsc7RF6Z3SFLiOhcZZ4rmHGuiYHeB76yRYfu0Bhs3o+yICmazw1acxfrTWWwzt0o+MFzxCMy67zy6wmp0Nn+qxxAqLssQKtbWQlcDKngVaJzE6aP5LdO5PDDNPs0a7ILCqsVru2K/cmRPWx0H4pk8mFFxgRbfGtm6ArnLf1wC+wf5mA2TaMwsKi2+BRc0SCiISB+4AjlHVHpfXDwUOBVhqqaVqdVlzgwRXMz8O1GqitKiKHzNRjQz+w4NCjoNtR3ZHe88vEWxYENnNZLGlAak7CuqOKjpKs28bYSA35D5D+y6H5H0l+uoFlmFlWNMguDr0/anEcV7b9zmLmVoNnllI3oEJUMYSdGTBvwjkMpSmpYaRuv8bZfuTWFghvqnQ+pBJ2uW/hsD3Ef9SaP5rVJOonYLkTY6QqcvCVSwIbQ2Aqg3ZNwFjMycSRCI7o4mZkJ9N4aI+CJHdkejuaPq+KoXPhvohoU1QopikYKljx4vWJ9DyCELG0aOpdpxSsD9B7E+HFvSZV9GS72U46qH1Pkj/E/rOd+jLw1FqMRRBGs6G4LqQn2MYAFZxUCb1Z0JoQ+PSoWkI74xEZyCOzZymHkUz/3Qf/wPV23eJBA2rLvMiZP8NvqkQ2ta1b+XbsSD0Q/OzgGJexQsAEljFYQUamHihVEJuwDlgZKldBBnBcJLIjmjv2SVKiMQRBA6CBI1LiItTiJ39uHSyKv8l2n81JO8s0VcvsMC3vLn3AmtD/8Uljsvj9qyrDfdc08oNFyxCPieIz8/am/yHX8/UMVjNGuSy8OsrP2LR6R9BJol7vOA1fsibZIGdchk3BhCGutMmFxnfMmjXiY6otzN/aMaIr/ZdCfZYHEGGIwi+xRz24hiQ/xxN/h2J/qgmvVqYYIXXw266Fzp39nB0Gzp7G5h662QJ0TiiJpLQIhLABAc3q6prgYuqzlTVtVV17SlTppRtTzWLph5D+29AM69VLUoikT1wz5XEoPVxpPUOswCXGEgcCEHd0cZuruB9hSB+HK401cBaSP2plTvTdXbp13yLQeIGKlMtB2jCIxGGyK5YrfdhNV+NBH7A2L/S0dhYlnvNx9D789JuAHdadcjszsR/YTK2EsWILy4G1qJml675OiQ8OnqTZt9BE7egqcdHWR89ifkZg3TWyA4gPuz2GWjb1mj7TtCxq0l6ZV9mSNwuZsYHqwVpuhax4mj2TbRtI7Rjf7TjYHT2emhqFmJFkZY7DJNA4gw9ExlI/gOduzPUnYTrOFIK/hUL+k79/zFv9AlCkH4Y7fixB6X0UsiZxOrgr594vPQPsfyLmkRv/DhnJ9GPoWOXKnXwQ2BFtPs0tH0HNPWA2YFygYhYnzshAAAgAElEQVQg4W2wmq/DarkFK7bXYDLBXH9zk6wd/r1JBMJbAFk0X/3nISJIaD0kfhgSmVF1MmFhQO3jhRya+qcTL7w8inhhV9znnDC0Poy03GXmmeHxQvzwIvV1kUDp5zywqtFyqITuM0u/5lsEEtdTWUDR7/Rh5NwcNgms1vuwmq9FQutTOl5w/wzTKeGuq1pJ9vvIpC3SCZtXnozx0M1eg/TS8UIwBIssbSOD78/L9+inpLBeeAuIH+98b1EM02xRkMUguD7SdBVWZHT2zZp934kXHpmMFxYgqN3riPOOTEanIXEH+GpVfZWBUcwPxUhB3+U1aGfhhBVaCerO93j0J+icHxor2UmMC8bMUBCT3r0GeFdV/zjW9jT3Odqxt1FW1iyID/yrQfPVhcFeuT75p6MN50HPKZjJxtQhSdNViH8xc9CUpyD7Kth9EFzTVawPwIrth/qWQPtnQv5L8C8PsZ9jhdar2A873wN2mbr//OdU3mmIQN1xSHAdU1aRfcuc41vM7Pg5O6cA2ncR1VsMjYU1EMD45Pa5vGaBLA76eZXdCRhv3sxrDCUiLPP9Rfc2tPTYfman2WpFrDFYNeEEo12/hPQzmPvEbxYNzbcg/uljansS8x9UbbRjX2eyH/6sDDwDGUwwui0S3QsCqyDiM0yGjgOKdh+163C09UEs/7IQP9zsdhc8T1mwe4z+h0QcqnslCMQKdy+t6O7Yua8hUYqBNF5IGzcMHaMoZOqfEHMsnwIr4smWK2vGThEx1pzRfUETaOpR6D0Td395P2TfNm0r0H8Nmv030nw9mn4R7T3H1M5KI8QOQmIHlbEE9UPzTWjib5C8ByQIvhUhdb8pXdAcGlgZafwz4hutkNy3CzWPF/LfGIcm7R6iL/tXgubrjCOUlz75FoPGi9Hu4xmcDyWMNF6JDIgzT3nC3I92NwRXL7nLZUV/jPoWNbud+S8MZTl+KFZow8rvRZOQe7b0AfnZVJyrJQKxI5DQZsbTPvsfs+vqW8Tscoa3HdyR174/UU28kMvC6Qcsy5wvCxOb6aSPB29q8WA7a+bxkfGCyf9YBCJLQ/5jz/0BTPwRWNPRSxgeL4SR6H6Ibyoa3cuM91aLKRkaA1TzaPeJkHqcwXiBILTcjPiXH1Pbk5gIlGHyaQdk/lnDa9Uo0TQpylgWVmxX7OwXkPISG9lo28Yw7e1JZtI4oBYlDxsA+wJvisi/nb+dqqpVGi4baPfxjkjZAB0J48nafxUSP9JzO1ZkOzS8mRPIhyGwWkHgKGJB0BttVcKbI+HNq3gXDvr/XOGADKUz9j7zI37jQBHZCavxdyWOdeB1528QARMka3+V5w10cVkjGOWG4HqAQKaahELQtNl4FSSugcTNhmoa3AipO2EwiBMJGd2JGkATtzrJBCcY0bRZtHQdhbTe536OqrNbG5pcSJSBiFwL7AjMUdXx8mCqDpmXHBvJcoF0BnJvIsHfYSfvRfsuMdR51wDBhs4jYcpDkH3XeZ5GLr7TThmRlwDDD4GVkfBWBX/V7IeQ9GiZVBEC1AF9VF5QiEMPHaMnd/YlND/HlJ/4VzGaDNnXy1/fN62wJ2KBxNHg2iUSHAGnveF9TUPmdezEXdBzOkPPeQf0XYba3Uj9CSW7IBJCYgdA7AA084ZJRpEaWtdl30Q7D0Fa76rwAUzCQY3jhZOMXejAQkEzkH0b7bscqTveczsS3hxCLzrxgh8CqxfoUph4YS1vbYU2RkLVK/5r38wKR7gJng7AwoRzPsh/BlYTVkMZdiSYuKIKfPVJhPdeb8FVVyHrITj3TXfGkkKIAP7VwddcZUIhYFiejVcapmfiBpOsCK6P1J1kxhqcUqNabQ4k73Jcd4bFCyTQziOg9RHXRYqJF74G/IN9msS8gViNqH95yL1HYXLO5/w+HzoCWIvN6x7M97Aaf4Gd3Rnmbunh6BzacTjScuW49+vbhlq4PDyjqqKqq6rq6s7P6IIDu9PZXSpBR6oSImGj1htco+Qu1Lgi5SLoVYSBQW1gIrKcHx/Gm7bX0Kbbd0PtCnXH/uVKvODmUW5BYONR7jz6zE5Dw5m4LggkYoLwsJfapoH3G4HILkjzDVhWACt+GNbUp7GmvYrVdLFnFfSqkbyN4pIThdwnrpRmzbyCtm2Otm2Htm2KPffHaL6Sz/W3FtcDo+OUjhfsNm/HST1259HQfYLDJEpTcncw/z80PxesqaBuC2/LqcOukFCwphi70qZrwW43StQOtO+yCs+qYEoiBnZlS413YWj6KzLtZaTpeiM8WLaUwnZdBAzBq8tC0NlFdBgHzddCdF+HQu42PkWQ2OGDv2n2fbRvpmFp9V2O+3cRoWTSpv8aip/zJCRuQO0Eqmro8omb0fSzRitjBDRxPcWLqRzk/ovmPnK/7jhBNY96YrvMX6htvNAPmVcoXgSkHZ2B6iASRELrIsG15o3IZfJuDwe5xQvO5gMZs6BO3oXO3dXQu8thWFlVIdzjhSW/uybxhuLnKxi22WKPzqK/F7YXgYZzSrweQeIHQsSLW8RAvBCGyI5I89+ceOEgrKlPOfHCZYi/trobA9DkrRSXnCjkv3FNhmjmDbR9K7RtW7Rtc+z23dFclazNSdQU0vAHjD34wFwZxZTRjTFpPmaU2N+1v1kgx/qJhhVYCpo8TiXZf2J3V0i4TqJqzINVdhlontIZ+PF52FUV1VTVdZe1h4BvaQhuhBlYhk/cObC70OS95ZuI7EGxE0IEovs79ccDtZ0BIOiwC6p439JkBJ0iOyMtd2IF10AaLnTaDTE40Ye3heAmpizBtTbV+Y6l0dhlTXsHa5E3sBrOGXMJQ9XQUiUnVtFrmv8G7TzI+dxSQAayb6AdP3VdgHzboar/AirxYCcWgdVKLPqHIwKBdSH9UIXjBmCB3YYEVjAlUUWBQRCiPzXlPKUgcWi8wiQe2jYxSas562D3zTRjU/ZNSu7mS8RowrTcjTSeb6xsA+tSvNj3Q93pSHANyL0D/qWRqc9Dw/lUtoMsBY+0TqHQRUciWPW/xpr2GjLt38ZOj6CjjxKDuhOR8BbYdh67+yx07u5o34WmZCF1J66fhW8pXOvXRUrTRsVCcx+jc/dAOw9Be85Du36Btm+P2iNu3fzXuH4W4oe8x0TVGKGawe4+E529Bjp7dey2bdH0hJgkzIcoM+aOk/f4+MYL1VBwBaylnHjBorB0Mgd2L5r8R/kmIjNwjRciPzFz/eCCyw/4Ee3i1Ms/JhzNEwybzz4SyzN9pSS7/Mzl+ZIGEy+Ed0Ba78AKruU4moSdHyfRENoMQluViRcG2quH+JHItLexFvkPVsN5iOXVDrtGKBUviFu8MBftPMAwRgbihdzbaMfeBcniSUwsJLACMmUWUv8riB6INJwL8UOoseldFQiYOS+wtvvL4of08xPbpQUUVmh5iJzi7eDkDdhzfzW+HfqWYV49Qa4QXyvqWxryH454JegEnLWDqqKJW4wrgnaD1YTGj8EaZi02ZkR2gf6/4k2R2W9qtqUJzb7sspuZNJoP7FV0pm2noOswyLzK0O6CgrSYCTj6U4gfifacA//P3nmHSVJVffg9VZ17cthAEgSUjOQkSaJkBAQEQURQWBGJKvBJkCA5q0SRJLJEEZC4BMlZEFBy3p3dyTOdu873x60JPV093bM7cbfe55lnd7urq27Pdt8699xzfr/UPe54spB/t8RYvOykbHPTzH8EmQwaeg8JfB2JbAPNj5qSbu2B8OZgTTOlqKkH8S4hi0DNyViTQbk2sjP0Xk3RzqNV7y5QBtDE7R6BqmNK6DPPj6rqujpdJoGU/xgJfgsi21esIeJTGgksh0Z3geQDFO40ibmpaxZiB7laJZWSR7EQQBquQduPNir+YgMhqDkD7GUxuyAlqoy0B7pOcduWBpXT9lxpHB5KJiNsqL0ICW9ldlWDK0JwbXT+dhQv9nOQuAntPgPzHU+b8ZFk4XVUKkQTqMQ9l0wiEaTuItTpNAr39jLgdOO0zzIOD5WWoebfpnjeCphki7UMZJ/2GJcDib+6Dhzu70uB/Kdo56lI/eUDx4a3cCvohswVmoVxsnrTjhMgPYf+z0j+Q7T9ENRe3lgQxvY3C7gloD9UrGo0sKqx4Sz4/AZNUnuUcRKzoedicNpAatGqnyOxA0fvdx3dE3qvorIknRiXqcDyaPZlj4VuylRvxA8peqXjZKFjFmSepTBeaICqnyKxH0H1cWjXuZC6nf7NnPx/WH1D+MvzvTx6Rx0Lvgyx1qY9bLRtD3Zg6PwRMHNW/iMghWbfRQIrGZvs5jmQut9UX4Y2Q62ZaNfJkPwH3t/1MFSfiNWnwTKRRHaDHvOeCpCYK+A6gCbv8kheO2auT/8LFlJA2gt1etDkvZD/yNhUR3Y0raE+nohVDbEfmE++0wOtezExFQphqD0dieyMdv56mKWCv2FVKVbtITgShMQwArd9ZO/GaXkfa9rIK+B9iplUCQUAqbsAbTvQvUGmgBjYM43N0iiiyduh+zz6FxVOK3SdjUMQK/a9UbmGxGeh6WfdRUKC8gKIgWFUZkPGc3wITvJ+6DyJ4jK8AARXwYobayvNfgCpe8tcv2/gESP2mLrfHXMAIy7n2lnlP4XOE1ARJLK96QuMH2yu43SgC74LTgelFwJ5JLhW+XGMAxI/FE0/Yt6TJoAwiI3UXVQcKOY/xzvYU3DmjtqYNPs/V5g0CyRRucMkvhpnlxQPnaqMpT1cyWvWnAXB9dDEzaBJCO8I4U0RTRg1drsRZ0Fx4q40QST7JgS/gVgNSONNaL4FtAvs5dHE36DzRMouFHJeCb4k9Py+xAtM2a8V2abw4fSTlCw+y7899KLDj2nUUGjbH6fxbqzgKp5HiFULVq0p52/bz4jgjqin1SMJGtoMqT3HWO61DRZuA1O9dYDb9uSRfEk/iuNksSyTzJHYAUZzxWllIPKLmkXYOHwvNd8C6ce8x5p/H/KgnW9B5kWktoJgajFA6s51RRnNXInEwGpGqn85qtdxEvdC95n0WxBqO/RcYBKJ8QNG5RpSdRiaeQqy/8Pcz4eLF8SIhtrLlDgmCIGvFz2qqUeg8wSKrVBtCKyA5SYgNPcJpO7wPHddU469fza4IiEIkb0hdTcDLQm5gaqg/OfQeRKO5rFiuyN2I8SNf7w6PbBgR5OkGWYumjzxwgFo+kHIve/+DkMmXqi9uLitNv8FXnoTaH5044Xch2jrvpiWlyQqMei5FBrv8C3yKkB7r3P/r8YZa2kTZ4bWMf+O7Iam5xR/NzXvapL5VIpVcyBO/l1I317+YOdNnPn7YDXPHvuBLeZMvoRCcFVoftzdnf0UCa4DkW1Hf3e25zKKF+FJ8/hoJRSsODTeBZmnIfs22MuYyb7jKIpvngKR7d3dtOlu3/agYySIRPcpeIVmXoHO3+BtO5mDzIuo0212cnoup7IspwWhTbHqzkP1dHA60da9PfqoU2j3BUhk+4JHNfE3cHoZfiGQw9ure/wRKwaNd0L6UTT9onHPiO7pKbYooQ3R1EMUfW7UgVEMeLTzRLMY7X8gYRZE3Zchtf83ateZDKjq1cDVAOuvv/649B2JWBDbG4ntXfqg6I7Q/QYVLWglZCpaBj9kTwOmmV337t9THFiOxF2lRCIitCFSc7r3eEZUPj1e5KF1b7T5MWSI4GIB6afcxcgiJjtkGlaDK2RpN0H9VWj3mWYxILUQ/zESPxxN3lJ6vC2r40gc4rOMI0TTPWjv9ca1wqpH4ocgkUqEoEaB/BdG1X5YLY6k6aGPHzrgULAYI4GVoHmOGy98YhaeY1HN1XvpQDKhD01C7xUwWgkFiUDD3yDznHFnsGei0mCqD4u+C7apwrCXM9VPuQ8oihdi+xcON/sW2nEc3vFC3giM5lsRuxHtuZLKknkCoQ2x6s5E9WQ3XvgBOJ8POS4FPRdBrFAnwezid3u8v8FkTBxTwWjGGvN/dBukH0PTz4M93Y0XiuczCa2PJu/BsyptVOOF37jxgns/0QTkMyY+qz171K6z2JL6B6PmyOBJn316noJ7vnYaK9g+wluZ1p/Uw5jvaBCwoPZcs5bwGRFW/Zk4HbWQqkDMOv8GTsdJWHX+92VRmFwaCi5i1WLFf4hVczIS3WnUgwOjului53UUM8dgFi8S3hKpOgKJ7mp2E6tPxmgO9PUShqHmFMSeaY5vuBlCG2EmlCDYKyENNyJ2oR+39l6Dd3DQf/WB4LMiRecQSBVS/St37FEjFFdKlC0/NGjAFckqJyDjoJ2/wWk7BKd1P5zem8fFy1lV0ex7aPadAs0DkaDx5w5vYZIwHUfj9N5aPKboLmBPo7A3PQrhrUbNMkqdTsgNbfkByEJ6obTLfBYCiX4frOUpzrmW8DwvZQuXedH0QBahJc41AuxlzWd3KOFtGPMWhoUmg/beNPwh+Y+H0TYZATpv4K+ZV9Heq4ztX3BTqD0bie5jkkvh7zBs77b2Qs95aNdpiNWAVX08VvMDWI23jF8yAYxSfSXzpNhu+9uSgVjVWPEDsGpOQqK7jE1rWL5EXOC0oqOo12Dihc3ceGEPrMgWUHM6A/FC2PxUn4AEvuYKnP4FQpvRr41kr4DUX2/sMAehvdfhuWM+cHX6792e96ChmN5vqTnZHXsUsWe4DkgeOF8Va09kXi5O1BSh0HUyTtuhOK374vTeMC4idaqK5t5Hs28X/B+LBJDIDojrIqYdx+D03lg8psj2xoWiIF6IQHgTJLj6KI0xaZJPRfN9zl2Y+pRlrFtJI/vQb18/GE2jvX8ZGIYIUnse0nADxI9Aqo9Bmh/Gik4uXeuphFV3AtReVdnBqTvGZR2yODPpKhTGAxFBraW9LQ/tsS+7tuIHoJHvuF7GYiowBmUqxW5GGv5sygHJlC5b81rQD8ZeBsTC6TrHLSksgcyAwNeMv3bshwXWRiI2ajV7J2CGBCyAcZrIPEPZ3cX8p65YEZB9G03dCw23ei+QRgHNvoO2zwJtpb9fvu4SxLUO1e7zIHEr/dUH2TfR1N0FYxKJQOMdZmGSfNC0hsT2NxoVo8Zw4nhT6+sqIn8FtgKaRORz4FRVvW5iR1WIk7jTlIc680y7Ufx4JLwBSBRpvgtN3AWpR8CqgdiPEG03FSQo4IDUIfVXlV7ESKzElQVvrZJKCZdsjxKrBq29EDqPpjL9lnEm+5Lnw6o5NDHbaBqMSiuGSdho+hm0/Qj6F0tOC3Q8gxJAg2tC9SlmYeN0UVy1Nojk7WjVrAmzfhOrHo3u7ToYDLcIE2PB5zN62MtB/sPix62ZY+4IYcX2QcNbQPpRUw0X2RaxZ/Y/L1aD0W4pFy/kPmPYRKPVhFKFdp/v2uSWQJpNO0VoLTdemFH4vD3Tu4TcmlHcRhhYEdJBys5TzleQcd2Usu+gybtNC+AYLQY1976ZM/ItRjtKwq5WjdFJcrovhcT1A8mQ7Fum2qLx9v4xiYSg8Xa092qjDyFBiO6LjKoWhJcjRx8T4FQyFYnuB93nU34jbGGohdD6RuC5aLGadZNBA4gIhNYZaIPwWWSs6NY4gaegtbyVr7ZsiTZcizVKCb8ljam1QhlNqk+Azl9TOIlEkOoTxuXyYs+EMjeWso4HwXUh9yHewXcUak5HW7/nBgel1IlrkKY7hg+Sq46GrrMoDGIjUHWsx8GBEuMZjpTZEUk9AtGdRvja8qgm0baDTIlZ/4MJtP0waHoMY0t6M4W7NynIvmey/NEBQVCxapHqE6H6xFEfpzl/FRpcB7IvU7jYDEN0rzG55lihqvuXP2ricBK3F36u859D1y9RAoBA+NtI7bnFPdLh54w4n4QhsOrwomyhDSlWUgcIg1Xr9uMP/r7EjNtL/j2G/R6JjUT39HxKM29A18n0a5+MKe7viiwVJ0jslYseUlW0Yxakn2f4xXKIistTw0b0TLvOpGRbWPYNk3hpetBYS/b+kdLvIY+2HwI1pyOhEorcY4hm34P8lwzo2pgxFS4SBSQKoRIVMz4LhVT/Cu04mqHxAtXHj8/17elG72O4Y8rFC6H1IPcO3nNCxNg6tu3jJgNKzRtxN16YWeJ5IH6MEZgd/LuSKFR56VoszByVgvwnplQ9OjrtqYNRzaCtBxqdDNR8vbTXJBiaHwJs6L2GwnkoZQQok/8oaJkVqwqpPhaqvWKlRUckjIY2dTdxBlfKhCq04fSR2P5GT8nD9nPRSUDXqXi3DwUgsOoYXNNnKFZwBk5gS8g9OfyB2gqte+LUXIYV8ytDRsqkbHkYD6zoTkjdhWCviNntWwmpuwSJbFf2tZpvMVZjg8rm1Ung9N6G03EMTvelaKkSyVFAc5/htB4IyUEKzP3YENwYaX7YLNLzbXjfsGPGXq75+YJkgmoap/tSnJbNceZtjNN5qinrrTkFLPc4aymoORNryOJfM69D4saFfFMJI0g1FqQew3NxpnnTe5t5yVXkH0oCTT8xNmMaBqk73+hoSJx+K73gmkjVEeM+lsUVVYWeS/BevOYwLSb/Qtt+UvSsSAgJrYMEV+tPJqjTjuaLrdNEgkjDtaZnX+Lu/2kYqn4B1WdQaHMoQMJ1XxmujDoO1SehTi9O9wU4HcehybtNIKxJs+jV9hLvbbTJgUyDwBoQ2Q8Cq2DKsocbvkciNftGyWTCu6/VcNIBG3DSAd/imrM34D8v1RW/fijShNSeY/6fvXaW+zEuLZJ9AwltjKft5GBy76FtP0bTHo4RY4jm3kfb9oHME5ie7BwQhOg+A58tomAvhzTcZAT7fEYNiWxtLA8D38DEC19H6s7Hiu5a9rWan2+E8waVzaum0MRsnI5jcbovRsdQFE7zX+G0Hezem4fGAjYE1kWaH0LyX7ntjV7xQhQie8G0FwqSCaoZnJ4rcVq2xJm3EU7nyWYXv+Z3YLnHWTOg+tQisWvNvu26WizMm0qgqScW7rXlSD+J2VwYWs2RN1UI2VfxdN3RJJp+bGzGNAxSe7apFi2IF1ZFPBM4PkXk3itf7VuO6C+g7hpj5VpAn1hsX7vSICSEuGLmPmOP1XQN8I3KDu462rjn+YyIJTrqkMh2FSUQ+tD8PLNLkX0LsMGKQ+3vIbimWwnQhgmIQ2jiz1B/PRJad1THrJpC2/Z1WxgG76SJWfAH10XiPzZ/z7yA96IijtSeiQzaeTfnVrT9cMi8Sv9ufXI2mnkaaXoAie2DqpbckdXkbLx7NN1dTAlibJSU4qAlAFaxEOKo4Cwo0ZOdNgFUcAW8ywYDYDWOzZiGQeyZ0PyYEafLfw7BNSC4zhJhBTdSTE9ufiEWULkB15KSZM0CMvuOEYv1un7+SyN0lv03IKi9LFJ3PhJco/8YCa4F054xNm1OL4Q2RjPPQudPh55tyJ+lcKDrdAZXBWj6Mei9FqKHjI7+wEjQLyA3z1gn1l1u3GF6/ojnXCD1SMDjpp59Ga+FzJsvxDn5ByuSTprdwFfm9HDfdctyytV5NtxmiLCrTIPwJhDa2O2lNwGcSp2bYCk1fgfyX0FkJyrTnkihXWcjzQ9WcOzooN2XgaYoHF/S/K6bn0HyHwARCKzkzxNjhES2Nn3zFaL5BWjHL10L2QD9VQChDdDWvdzKwSQQRHtvgPqrkPDoqrmb3fZ93ZbFIUlKaYbQOhD7EVgz0MyLHu4PgMSQ6t8iHmLV2nEUpJ9lIF64G00/jTQ9iMR2LxMv3IV3tZG7WJfgoLnMIxFiNTMmOK0U20MDZCA/z1SGemKNXQwzDGJPg6aHjfh37lMIrgrB9f15oEI09TCLXMmXugGpeQEtpcGmXRD9ASTvBDIQWB2pPQ0JjI+zlY/BmvEPnLmbAOViP4WWddCm+xEPtxwfb5bYCoWRoqpo24/MThoZIAnOArT952jX2YOCA8zzmkA7TywWIVpUUg+7fXtDy3LV3AjTD6FtB6GdJ7iWUh4LLVHwUlnPvQmZ1ylcCORM8iJlBAGHvUk5PXgH5DZUn4Y03gPTngWp9jgmUORiMWqE1sezn1BiSHgjV9DKa1c1gMS+PzZjKoMRfvoOEj8ICa3rBwdDUFWc3uvRlo3ReavjtGyNk/znCM5QYbJIbLfM3GsMOaNonn0NE5BkIP8B2vZDdIhmiUgICW9lknhWA3SePIKxxij8HicZCIDceUATkPsEei5geOG1sSJnKqYWbGcEYOuvAqIUJupsiOzC4EWEag6n5yq05yq8qoiuOm1p0snCuS6dtPjj/w3Vj4hCzclIZBvIvIx2nGAqrdIvQ2gbhs+dq7ELtWJQcxJlKyzA/D9XOLdr5kWc9llGUK7nKrfXfYRk38C7FUMRZx4SXBMJruzPE5MEk5w/1Oxmu/EA2oZ2HIt2nWMSWP3xgmsP3HnC6McL6Tmui4LHAlnbIf0IdBxqEgPWUAHBPsQzXtDsfyH9HMXxQieavNe8crjPo/bi/ZkWqP61iRean3cddIaeJ4jERmLtOwJKtTNJzFRfhDYAqfIYU2jsxlQGEdvcX+IHIaEN/HlgJGT/u+jn0LSxSpdSFW5BpOZUZPq/kelvYzXdNWmsUJc0ZNojYK9ZwZF5dMGuoyq6u7jjJxQqJfuGEQYqujHnTO+/V4YzP8+IvY0m+U+9dxH6xoIDJI2Ak70cxf3btimnD65X/PLsf/BMCGjCtDOUQaI7lhChSxsPb6cFy6pBGm40wYvEzI1Zqowf7xjZnElwDQhvjlng9BEx/WuhLdyy9BtNiWb/mOLGrsfPTk5KtPca6L50oM/V+QI6T0RTcyp6vYgYbZCCz4TXhbJm592LzL9cXY4hQbHm0MTdpiWh9xac9qNwui9Ac25ZpdNGxQJQEoeqw6jMwi0z/E78eJG6D9qPgJrfur38fd70eUjdic7fHs0bkVftPBF6rizUNxnER297B2hffRImlw27gX0U4kdC4s9ox4mQuhPS/4TeK6H9B5C+g9J6FBEIb4EETdWEFdvXuOxEdofgJpRsgVhA2iUAACAASURBVJD6ioJ2p/cmtO0ws3DLvgY9V6Ctu488qVBCgBPNTciuqE8Zcu8at5Kiz13GfDa9duadrgGh4tEi/5lZ7HjiVgtq0uxu200erX8WWHUQ8qicyP3HCBYWkYRseZcRCW9fIl7IGJG8/BdYdrW5N9vLDrk3n4MEi7VYRgMJrASRHYzuQz8R0x4b3sYs3hv+4ope940pBjW/K1nJ5jM5cdIvQ2Y02lTECDdH98U7Ie249uttYy7i6jM8YlVhNd8J9bdVcHQWnbe+7/5QIUt0y8OIcObhnX/JuaV5ni8ywm2jSWBVc/PS3uGP0wRknkHq/4R2/spkT3FML37dJd7BsL2MCSiK3ksE7OXLjy28HQRvd60jhwYxSSNqNO05E7w3zzECUZqC4Bpjptbch9RdCsm70OTtppwxujsS29/YxoE7picg95YJsIJr95dM+0wuVPNu7+3Qdp4U2nNJxWXJVmxfHAKuy8NcBr7ffQmCCER3KS1Alv+qRGlsGnL/Q1t3GWiDSgeNXWLDNUZvoFI05+6gVLpzOVZ2kWGz0595sMJrJE21hN33XgdVUmga7ToLqo9zk7El2qSkgbrmKAu+LE6+xGpjBGY+iugCsFdwe5vfpbJETdTsekrYdWk5sOBZCa2NhNY2o+69AXouptDaLgrxw8peRZ1eD/XwNORb0MQtSNXQlpfSSNUstP1IioRxozshllfFl8+E4rTgrbKfh5JVCPlhdjgXksAqICG31XAYNAnpx5H669COE9x2MMf04tdd6r0IspctcbKQcYEoR3hLk6hIP0PxHJAyIq3TnjNJ/aZHTJJGEyaGGet4ofZcSG2KJm4zCZnIrkj8gP7WOjOmRyH3tqnMDK1tXKB8FgoRWRa4EZiBuVlcraqXjuU11emA9kMZnXtmCm3ZHCK7QmgTs9lQkEzMQ+4dtOMopPHWUbiez6JihdfFqT4Dun9b5shedN630WnPY1n+Hvxw+AmFSgmuVaI3OQLBzdwJZHDgaJu+d6ve83Tq9EDqfjT3CRJaE8LbVmaZGN7S7KTnP6F835cg4U2g+UnTiy9RxB5mNyu0qSnHzqco2BGVIBLzVpQvuJoEoP5atO3AEjsUCpkXjXq+SOmd376jNQvaA1Lbv/BfWERsiO2DxEq3VZgxVVIK5TOhaI/bT+5B/rMRncqK7QUx456h+bloz6WQfsLsOsUOGt4SNLgGntobEjO6HQXuKlkgi3aciDTPAfvrZcQCMeeOHbjwQqejhvsew+tC5nEqrq7QFGSfprisOQ/px4x7igS9d1Dtr4Mzl32P+oxrz2gknRz4/odjYfY+dleswHRgunupf1Q+LpLQcB9WBf2rEjsY1W6jT9FH7BAkfmj5y+TeNr3zRfFq2rz/kSQUwpuhNadD9zkDFWrR3ZCacsGQz4QQWJ1imzgwibk+Vf7BySELAqsYNwcP1OmF1ANo7iMkuBpEtq9sUR3a1GwG5N6nvDuKGPeS5seN04OEEXsYnYLg+oNikUGLJwki0fKtgiIW1P3BtIZknvEcD+lnjE2miNEGGIbRjRcsiO5Z0knHHCPg28uNFjngOFV9VUSqgVdE5BFVfXusLqiJv1GxY1BFJ2yD5F8wemENGIvyweSMHXl+brHVqs+EYMX3w5E4dB1X5sgOaFkTp+kprMD466pNFfx0S4WIPROiew4pgwuC3Qi1Z0N4KyDilsDFwF7WuEh4oLkP0fnfMdoLiWvRjt+gC3ZBHe+y34JxSABpvA2i3wepA2rxtKWTKOJaKokIElh2+GQCbh9ew62u1V3A/ARWQRpuKZkY8TrHsL3pFQjGqeZwus4xpUYt30bnfxsn+Y+Kru+zBCDVJUplAXuFhT+tPQOr9hysac9hNT+CFf9hUWCqmkFzn6BOFxJc0+23HbwzFTSCqLn38Uz4OW3gfAkNt7rf3+EGVO22ZYxnD59XGb9i2pbOYkRe3SV7pAGyZpHkuegKmmSLdrHrQV/y/VnziMTyRGIO4ViQ3WftwA9OGiIQN9LdwQpVvUUEq+ooZNqLRph22otY1b+srEdZ6kpUsLBQbQpWbA9k2rNI86PI9Bexas8c851an4VD7CbX5nFIvGDVGWHGyPZA2I0X4mAvhdRf5nkuzX2Kzt/GVPUkrkW7TkEXfLdIp8VzHGKZFp7YfiD1QA3eOglRxE2smnhhmeGTCX3HNdxsdmQJmPdnr2xcRsq8dvD4hv8uVBIv5HG6LkRb1kdbNkdbNsNJ3FXR9X0mB6r6laq+6v69G3gHKNHnNUpkX2ds7q3qkUxwkYBbLewzWbBiu+Jt7T2ULCzYDM3+b6yHNGXxKxRGgNScDsFvoYkbTSY8sj0SPxyx6qD+MjT3gXGAsGdCsLQwjnac4PYM921dJSD/Odp9KVJbfsdJrBqk9lSoPdWcL/0vtH2W+2wWCEBkZzfJMcL3aE9HGv7i9vjmEat25OeI7oqm/4WxNxuE5iG0UdnXa/fZkLiD/sWLswA6T0KtOiTs+6sv6YhYaNVR0H0hQ0vApbpcpnnhcXpvdUUPHaOTENkB6i6G3lsgcTOQhvDWSM0paGspcS4HJIplNcD0F3GSc6DrRA8NAWMJKPn3UGkEnV/ifGHMPBKg6Ps2YsJmfCWD+NFsp3Bcr3qvUmzBlIvnEIEDj23h+0fOp60lSv0KhxJtLradlNj+aPa1Ia0JpRBvp4nhXiHh0joGpV4TXBkNLGtEKgsC1ygSWzi7MBHbW1DXZ9Ih1b+G4OrGwUE7Tf991c9Mcr7ufDR3pHGHsaZDaMOSu+raeRKo27IIpkIln0G7LzCWgeXGYVUhNacY62eMSKi2uy07mgWCENkWwjuM/D3ajUjDdSY5SNbEQiM9R3QXNP1IsTaU5lzB5OHR7osgcRP98YK2QtdpqFVrRFp9phQisjywDvCCx3OHA4cDLLfcIjokBFaG9NOMapVCWazK2oF8xpfGf0BrJfOfg7bugTbOwQr69+GhyKirClfA+uuvry+//PK4X3cyoE432rIRnoG0NGBNfx7VNKQeRnMfmt7+8DZld6LUaYfUQ+B0m5aCCRQHUnXQjp+7dlIJTPbPhtpzsIZYVRa91km4vx+PMujguliNlQip+ACIyCuqWkKyevIx0nnBSdwJPZcbS7TACkj1r5Dw5mMyNk3NMRZwBQmMsEna5d53BVsBdYx1YXAT03tfcLztfoZvKTx39t9o28Fun3Oa/ioBe3lXDX6oJ3rYnCuwoqmGqDoakrdBcjaLbH814UQGlYR7zAHRg5DIt01lQ2hjxKoBXFX9rtMr+x0ENyj6PxgrNP+VWbzlPnP1aXJQfTxW/KBxub4Xi/u8sDihmkbnfQvPnVSpxpr+ihEMSz2K5t5DAitW1A6hTiek/glOJ4Q3LbC5HW9UFe08zrQBaQqTTAxAzammJW3Y12bQeRvgaY8dWA2r6Z6xGPJiyWSYF0SkCngSOEtVhy0zWdR5QfNz0QU7DklkCaUT58aieWQEMLG+AGGo+R1WbPeRD9ZnzHEyH0Dbdys8OgjTXsKySlTKLkaMZF7wKxTGnWFKZcU2k1zr90G7jLsCMbAboGE2YpduJRCr3pQ1umh+AZAGa6lxtxAyvZFXQuY5NP0ESA0S3Q0kjtN9qbHSslcyFkdDnR2cNqMc7TWnV1im7LNkMFj/wAt1OszuYPpxsOqR2I9G5CPffx7Nod2/pzhoTUP6YfodDPoffg4Ca0JkO5PkkwCgYE1H6i4qOr8E14KmR40FpfPpwLnyH3mMxgJpNrtwuTeBALQ9A+HvUtrJYCqRcpMJXu8lDMnZaMqNMzWL1pyGFdvLlF/XnobGf4ymnzJCllYQ8r2QfgAzoTgmCVHnXVo+Fog9Exrvg9x7Zpc5sBpiVY3b9X2mOsPduy00vwBt28eUUWsvKjHoPg8aZ5fUYwBM5WFs3/5/q9MGTgLspScgXhCovRCyL6Gpx0BiSHR3sGpweq6A9IsmYRw/uNh1yemi5CKvhN2vz+REjIjYncAt5ZIJlaCZl9DELeB0IZGdjObMoESb2DOg4Wa082TI/Q+wwF7KdVrxCkBHmkwAQlsYbQVrJhL/cb/Yr8/kwwqtiNPwKLRtW8HRWWjdH22a7bcdDsJPKIwzYlWhwXUh+zKFE1QYonugXaeaHdf+BUpfeePZJTUZBqP5r8xOavY/GCubJqg734gtjSMiYnY+wpuaceU+Qxd8180GZ4CX0NQdUH9d4djsaXhLe4gRuvLxqQB1utAFu7tq5aakUbOvo7kjsUYghqeaQdsOci3gPI+gePcwBcnZWNOeRHNHQfZNsGdAcL3SwbrTUsKW1uN6Ovi4nPlJ313R+5kaeLk+WJjKA6cg1ssu+D/e/zjGapvv6PZ+L4cECl0bVM+A3EdgNVbc2z2aGPG2kbVY+PgAiITQ0GZukm3w3BCEyK5o15nGnrovAdfvonIGUn9l2fNrvgXtOMbYYvdZRNaeh4Q9bCLHEBExbR+hDd1xzUXn7+TqsKRNsiF5N9T/qT+mAFy3loi3sGtwlfEZvM8iI+bGeB3wjqoWZ91HiNNztbEjJgUomn0Fkn+DhlsKkwrBNZCme1EnARIw8Xfyk0W9/ACB5bFq/jR65/MZU6zQcjiNT0LrtpStdMy/g87bEZ32CJblW4GCL8o4IUjdeaZUWeJA0AgzBVeB2BGQforiRUXOtVcbHlUHbTvADQ4yQBqcL9D2Q9H83NF/IyNAu883VRf9/Wo50CTaeUrBcSIhqPoFhWJWYPrjjx6HkfosDpidiTYK+iM1CT1XoE7XCM5zB2TfZsT6Aa4LhQS+BuEt0OxbaOeJOD3XGbuqoWTfxtvTvejEjK9I42RB8NqtFcnxn8d/zx9++WcANPMyTvtROK374vT8EXW6EQkjwVUKkgmqaZPkLOUW4uMzSZDas0xCcnC8EFgZqT4W0o9SXM2Th/QcyrWzqqpptcq+ipknU+DMRdt/iuY+HZs3UyHac7GrK9OXKMgDKbTz5IL3JWJD1bF4xgtVY6en4zPqbAb8EPiOiLzu/uy0MCdSpw16LsNUFLqfFU1C9n9o8l6c7stxWrbCadkcp/O35j7Re63ZhAtvX1rwuSTDVPT4bg5TDis4E5n+BtiVJCQ/h9bSznFLGn6FwgQg9lLQ/Bik55gy/sBqrlihosOWOJYh8zw47RSVZmkeTdyOVP9iUYa9aGSewbNkLP+pUcx3e6EBrPghqNWM9lxhqjWCayDVJxi7LB+fSkg/iedOtwTdwGGTys6T+julnQ1CJsjX9iGPB4zIGW5lTuveboIhCTyE9l6JVh2DRLY1JfHgCv6NdalxFKO/MBYq0yFKiluFdgRLTPvHwpSNAqWSKJYN4UiGB699jH1/kaYh/if6W1Oyb6PJ2WjdVYhVi9jTzCKq90rovcaNNRWNHYhUH7/IVnM+PmOB2NOh6WEzp+U/gcAqENrEiNMuypyRfa1EVVQOTdyK1Px6UYa9aHhurGAEmp0FMCg5aMX3R+16tOdyyM+FwKomXvDLy6cMqvovRusGmHnJtSMeej9KQvf5rnCvGxskBzS5tPc6CH/H6B9lRiLWGMB8Vofe24JIZPuFeQc+E4xIAJruRVsPg9xTwx+cfwun9Uisxj+Mz+AmMX5CYYIQCbrWUQWPouGtTaKhYNchCJEdy5/UmYv3TmoG8p8t9FhHBakC7fZ6AiRc/Gh0FyS6y9iPy2exQnOfGN2EXImyRc2PzLKvpB2hBVLj8ZmOglWLVB1jKobaDx+ScEiZr2j32Wj3eWh0T+MeE9oQrGbIpxkTLYTAGkjNSWjXBZB7dfTPP1x5YOYR9zuuGIHW0ROOTCctnn2olmAwR03oSgqDwDTkv4DW3VBsNLCiCRh7r6dADyNxM2rFkapZDEXTz6C9V5t+7NBGSPxnSGCZRR635udD5l8YYc8tESu+yOf0WXwx8YJHb29ke0g9SOGcEXCFnMusz/Jz8V7DZSdHvICX9Z4zxLrbPTyyI1JJjOSz+CM1pZ8rcFcbStLE3rXnQWhj6DmrsutZjRDZDxJXYhILFmBD9a/M5qHPlEREoPEadN46lHXQyj6K03kRVu2x4zK2yYq/JTPJkJrTjC2YxAHb9aheBqn5TfkXB9cyKvNFRJEK7BrHlNiBwNDFWQgi2xlLNh+fRUQzr6ALdjO9kp4+0DYElkeCK1d8TontR3E5LYBlxJYKqiBsY/XY9CBiN5k2n/wHJc6cN69N3oa2HwnkkIZbTCBDwJyr7IZN0J0nYpjKg1JvoglpuNloleT/W+acQ4lAYG3K556V0uPNu9opbiAX2gzslVmY208+LzjuFJfstXj1qSpee6qKldZMoZ7n62sRyUDuXej9A8Ximinovb6oRNxJ3IG2HwGZ58zOcPJOtHV3NLdoiy2n98/o/K3RrtPRrpPR+Zuh6WcW6Zw+SyZSc7KpbiqIF2YgNaeWf3FwTddZZihRdx6aQGIHUzzvBiG8hS9q6jM8oQ09k06Gcq2LCej6FfScj7kHlyNsfhKXmXMHVoGqnyNND2DFB7R8NPc+TtuhOPO+hdOyBU7P9ahnrO4zmRARaPo75eIfVdDEn+j66LCyrWaLM35CYZIhdjPS9DBSey5SdTRSewHSdL9xcSj32sBKZgeu4EYcMuWBE7zbL/Efu1UWYZBqIALBbyE1Z07ouHwWH4weRxLvHf4ABFZH6q8Z2UnDO0B0d0zgEDUBu1TjbSGVB+dzxIqjTg8kbq7sGpnH0fYjELsZq+F6ZNqL0PQIwycJYmCvBsGNoPE2pOluwCvQjiC1pyJ99kZWaaeYQkJABGIHQO6/VFY1UcmNNGsEEuuvNoJqnsma0lhWkFefrOWJe+o496jl+N1PlkdV6OoIEAhWMr4SY9QeBpdYq2ah+xwK213yRkm/94oRjbngMtm3oftiIOMK6PUaN5+OWUYYzMdnBIjVYBKYtee78cJ5SNNDiF2+CksCy0J0Zwq/g0HjiBP93piNuRIk9gOI7oppK6vCxAurI7W/n9Bx+Ux+RGyk/oZBOmUjrP7SXkylWx6TJA8A00GmUXhPtjA6ZZ9gYoGcsZBOP22+W32ny32Gtu5jKtI0YSqJey5Bu363CO/SZ7ywAsvBtJfAKl2ZKGJ+4pEn6f70Z+M4usmF3/IwCfFuh6jwtXUXGkG6xF9N33ZkR6TqZ0jJ0u3xQcRG6s5D88eYBYq9jEmATBGeuedFbj7zDhZ83so3N1yZQ8/anxXW/Fr5F/qMC+r0mF3kktgQ3RWxp43ovMaO8Aw0fghkXkKlzthQpkq4WuVNZYTmv6Ry8UQ1586+jQRdW0GxUavObWMqGJF5L5qH/BuQt6H1GbT2IqNW3X6wq6Mipoc0doARmuojfiR0nVx+bPEjkfjBkPrHSOUoy+N8Cb1XIU2PoMl7IHGdaU2oAJEMy662IYdtlsYO2ESrQR3lJxcchwROde2/FkK00l7O9E32kf+8xHkcY2O3kGjybrx7cwUyT0KkUh9sHx+DSJ9mSyV2Z0NeW3M2GlgbEjeZxU5kexMvTHALjoiF1J6JVv0csu+YeGEElWUTzQv3v8KNp8+m5dP5rLzu1/nxWT9gpXVWmOhhLTFI8BvQ/BRkX0MzL0Hv1W7SeKQoYINdZZLx2dcHPedVYZCB7H9wOk6B9COuELmNafMbfCdNQfIOtPooxGpYiHH5jCeWFUeb/o62bAGU/hyJQDw0B6fzCqzan4/fACcJfkJhMUPERuIHQfygiR6KJ2LPhD4huinCfX96iKuOv4l0wpS3v3j/K7wx5y0ue/YsP6kwWZAQw7cIpKHnD2jsoILeYtWcCTSkZlhRPgmsAIEV0N6/QeqBEkeFIfId89fcW4xMC0GMWKQrPKqJu4y3vNdx5AedO29+Oo9Dpz0LDXcacdbEjSZxl7oPteIQ/ykiQSS6J5r7GBJXMWxFQe+f0NBGIHUg9ohNLsqSvAuqjjTjSdwyopdOX66K2XOv5LXH3sKyLdbZZg3C0TCavwptPwxyfT7iXmKafVoOgxf1EVM6PhirvkQ5OK617UKiKUoKU3pZ3/n4jCEiFhLfH+L7T/RQPBF7xpRTyn/oL3O4fNZ1/fHCS/98nX8/9Q4XPXk631hvxQke3ZKDiAWh9UCiaO8fF+FMadO6WLJ90eP41O2D/l0iwS0hyH0IIT+hMBUQqwptfh7mr4GqSR54HidA8jIcUayao8Z1jBPNqLQ8iMiOIvJfEXlfRCZQGtjHZ3TJZXNc95tb+4MDMP1S6USGG377twkc2eRnPOcFkZDbUhMqfZB20reQVHVwui9DWzZAWzZDWzbBSVTw/5n4EyVdH6wmJOb2TSbvH8nwjWWkPaikLv1oies4eK/uU9CyIczfDDqPMVZw2gNOC/RcjXb80lxGBKvmWIjszvAJmBS07w+dR4/NQldCkH3T+H6PVAAusArRqiib7r4BG++yHuGoKUMVewZW031I42yovQSsr1H4eYhAcG1ouAVCm5uS2NDGSMN1SHirwuFZdRDemuLPUxSJHz7CNzvovJEdvG3JNAehby/0eacSfrzgs7jiOA7XnHBTQbwAkE6kuf6kWydoVEs4gVXBmknxcicCwbGYcyvMvmum8J7vM+mx7BDUXgWYdcCwJC7H6Z1t2ieXEBa5QkFEbOBKYDvgc+AlEfm7qr69qOf28ZloWr9sJ5ctzjCrKu++8N4EjGhqMBHzgtScjuZbIPsinjd1q5G+BaKxDryOfoE+bYeus3EIm50NpxPCmxS35eTbSg+g6ohBjiWlyu5D9FcV9GMbh4fBwqlWI2bBP5LSgFIVESlIP4nTdgTk/wf2UhDeFVL/pLQlZh99CQwvzYhFwUGtRlehfoQ33FzxTpE6bWjvLaYdQbvMbpIEMb/nqHH2iO2NxA81yaeG68peRmp/j3aeaOz6JAgoVB2DRLYZ2XgHE9rMJCpSj2M+exYQgurjKup7n+r48YLP4kzngm4S3d5z6n9frnSH22c0ERFouA5t+4lpt8M2CdzqXyESQrOvUCzUu7BUep8MQ3grU4HjM6WwolvjWI+ibRW0mHWfgib+CA2zEbtS/aqpy2i0PGwIvK+qHwKIyG3A7oAfIPhMeWqaqlHH+wYxbbnFfwGwCIz7vCBWFdJ4E07iNuj6HYUL1QhUHYuIoJovtg4E8++uX6ESMwFHN2j0e0jNaQNtEsHVIPuK9wC6zkJ7roDG25HoHmjmteJrSATqrzc6BrkPzWOhDY2Y2qCWC4kdiKYepnDBP9IEw2AykHncvD7/GWTegOi+kLrPdasYDgUi5vjkDQt5/aEEyWSWxc7nsSsR0x5M+hFUU/26MJqfiy7YfZCYVt+w+/6uEN4Sq+qIEV1GrDhSfyWabwVngXEIWURHGhGB2osg+iyaeggkikT3QIKrLtJ5pxB+vOCz2BKvjWHZ3pVfzcss/guKyYrYS0PTA6YN0OmE4BpGPFmT0HMFOFkGEvKuTpHEXA2E4YiAVWfaEyViqt7y/xvmeAsIQnTP4jY7nymDFV6OTOS32KkzSrY+GBTyX6DdZyJ1F4/X8CaM0Wh5WBoYXLP6ufuYj8+UJxqPsO62a3o+t9uRO4zzaKYUEzYvWLH9kPorIfANjMvJ16D2HKzYXmh+gdl11t4Sr1b3ubT5Sd0D6cf6n5XqX1Nsf9pHEpx5aOfJENkVwhsNKm83rglSdzFWaC1Tmj/tGWTaC1gNf0bs5oIzSehbUP0bcy2pMueR6UBZK4NhGCIKlbrHCFeFd6a8RWUS8ztZlOsPHkqKU3Y9n68+KT6fAkgtfbcnVchlzc/8LwPm8XzLwPHdF7uBn5fYIfQLYC1k6aHYjUjwm6NmbysiSHgzrNozsGp+syQlE8CPF3wWY0LhIJvsur7nc7v8dLtxHo3PYEQECa6ChDfqFx0ViSKNd0FkNwbubW7VmA7juiM1ENoSqTsfaX4Ca8a/saa/iNSeRcm2y+B6yLRXkemvYdWe4dulT3FC9Qcikd2MZeSw+zwKqfvR7JvjNbQJYzQSCl6RaNGvV0QOF5GXReTl+fPnj8JlfXzGHlXlo7eKe7zFEv79lL+pNgwTOi9IeCuspn9gzXgLq/kRrOjOqCbR1r2GEVX0GnESTcweOG9obWi4ldJaDQqZZ4E8UncVUvdHI4hYfSzS/BgS3nzgXFYd5D7AaTsMZ/42OO1Hodn/9j9vxfdHpj2H1F0BVb90f6MjEXos9956IfcZZF+r4GAbgqsv2uXUJATOPWpZLjl+Ou+8+AmXHL8MqYSQc9f6mbSQzUSRxjshuBEP3tLA7Vc0c91ZMzlks2/ywC2NJHtTaOtuOO2z0NynkHmK8s4OOV/0cHLgxws+izUfvPGx5+NvPv3O+A7EpyLEbjTaNv0F23kgwbD3Wu2G7L/BXqGwsjC0NtT8FiP+O2h5Fd4aqb8GsWKFbkI+Uxqr/gKk5pwyVQoGbf0+TubdsR/UBDIan+zPgWUH/XsZ4MuhB6nq1cDVAOuvv/6oO5H5+IwFHS2dtH1VrLavjvL8/a9OwIimDJNvXkje7zonjNRWsHDnW6wqlPJ1+iJidBjCm3g+r+mn0fZZ9Lc15L9A009Bw40mMMGU3atVBd0XUV7voGgEQAAkWqJ0Mwfd/+eKVZb71YdMG0fqIcg8R2GwFXSvVapCwB2NQF1Tjh32a+Ok/Vckn3N48/kqZu3wDb532HyW+0aat1+OMX/BThz1x5l8+cXGPHTbfN55JQYIJ17+CZt9t5NITM3uUfoxNPOC2S0qhz3D9ST3mWAm37zg4zNK9Hb28uUH8zyfe/mhN8Z5ND6VoOqgvdcwMh0FBW1H238KzY8XOEdZse+jkV2M6LAEILAGluVXIyyuWPG9cOxm6PhJmSPz0H4IOu3Zgs/L4sRoVCi8BKwsIiuISAjYD/j7KJzXx2fCCcfCaIl6plh1dJxHM6WYdPOCZt+gZNBgLYt3K4PpcS9AogwrscsaiwAAIABJREFUvBTawAj/lRtP1xkUJgkcIIl2/77wuO6rMO0GIyECNWcg019Fas+GUgmQ7OtUJCJVdzkgEFzH/IlgEgkhiOwE9bdS/Psrvr0EQ7DqegnqmwfaDz7/IMJlv16W47+3EtefvRTfXC+PtmxCU92VnPPXD7nppXdYb+suNt/ZTSb04xgbRnt5oNR3UTC2kKcttjfxKcakmxd8fEaLYDhYcp6JVpVqlfMZb1QdNP00Tsfv0JaNIfvywp3IaUNT96FOT8HDYsWQ8EZYofX8ZMISgBXZAuwKKji1FW3ZGs17Jx2nOotcoaCqORH5OfAQJmq9XlX/s8gj8/GZBMSqo6y//dq8/NDrBW4P4ViYPY767gSObHIzKeeFwIqYRe+QnX6JI7VngKZde0UHyBjdguD6ENm58HB7OhpcFbJvUlztEEdqzi47FNVMabvE7FsAOMn7oPtCV5naCxuzYPYozZQIEt3N9GlGtkftpSH/qcc5woPEC0sQ3AQrsgVO+88g/SwDYpcKxE2LR+pe46TAdHC+AqseyBsxwyHkskJtY47WuUGmLZ2lt9silxW2/34bG23XxbpbvgnqEAoBIQjHHE656hMyaQgVxeQZ8/6ju0LyHmNHqRlTjWDVQWBFpOpnSHCt4d+jz7gwKecFH59RIhQJsdkeG/DsvS+RTQ/My+FYyNdcmiSoZtH2n0DmdRbd3SEJnSejKBrdA6QRxEaiOxU7RPks1kjTXWj3hZC4FegpfaB+ibYfhTTdPm5jGy9GpZlHVR8ARtCY7OMzdTjhhlmc9N2z+OTtz7Fsi2wmx1b7bsrus3ac6KFNaibbvCDRPYwLg6YZKPEPuJaNm5heyOaH0OS94LQj4S0GHh96rrrL0LYfgjMfNA/kILguuH2S5VC1Mespj+oAqx4neT90nszwbQ42hLeB9COYxIaa92OvgNRfisigXfvQlpC8leIESNa8f+critsegsZi0mnB6TgO0s9QWCmRAzoG3oKzAOiB2vORyA5o2488EwoiMPNrGc68+SNiVcblIZ8HUCKxYoEjywIRJRwprhRSAkhgZaza/0Orfg7Zd8BeBgmuPMzvzWcimWzzgo/PaHLMVT9lwRftvP/ah9gBm1wmx6a7b8g+x+820UNb4lF10M7TIPMCo2eD7N4Tk30LxADaey1a/Uus+I9H6Ro+kx0RQWqOh5rjceatZ3Q2SpF7HWfuahD9AVJzkmeMORXx1UF8fMpQ01DNFS/8ng/e+Jh5n8xnpW8tz7Tlmsu/0GdSIVYdNN6Gdv4asv8BBELfRmrP7p/QxZ6JVP2s/LnsGdD0sCmVzLdAcC0ksGzZ1/WT/CveAU0Q4j+BngsZPpkQhdiBWDUnoE4vaBeqWUQCiL1U8eHWdIqTCQEgB85cBpIJApG9ILAc9PwR8p+Yh/Mfe7zeixT0XIRm/uW2UwzgOJBJCXdf08QJl35a0L4QUPqFjbwqhkXgq8/CzFgmQ2hwYkEDSPxgc4w9w2gl+Pj4+EwQ8do4lzz9Oz5661PmftTCCmsux4zlp030sHzA3PtTf2f0kgle5MxP94U42f+CBJHIdyG0qd92t6RQfSZ0HV3moBwkb0SdFqT+snEZ1ljjJxR8fCpkxbWXZ8W1l5/oYfgsAhJYCWm8wyzCJbBI1k0iAqENFu7FvX/Ae4EeRGI/RLvPKf1aa3mo+gkS3ceMw4oD8ZLGj5qaA71Xep2I4nYJhdz7kH6AwnLQyoUs85nPuOQXsNr6MbbeM0Mk5uA48L83Ytx4wVLsuF87wXBhtUG5OEsEzp21LHv8ZAFb7taJZSuffxBlmXWvxg4sV/HYfHx8fMaDFdZYjhXW8OemyYJm34XUPxnbZMJgssaaGUWT90FkO1O95ycVFnus2HdxUndA5unyB6cfRvNzzYbIFGfxqLPw8RkBcz9u4cnZz/HWM++WFFz0mXxo7nOc7vNw2n+G03Md6ng5F1SGWPEJ84FWVc9WAEPC6BHYS3s/bc3AmvYwVuz7FQcm2vsHvHtFS2gn5N5kUW4NX30c5OG/1fGnU5fisK2+SUerjWXBKusk+OmZ0/nWlrXY5U0y+snnwzx5XyPvvxnngqO/xh4rr8kP1l2PR+8/nlD1Rgs9znKoptDEbJyOX+J0n2csKn18fJYoWj6dz5Ozn+PNp9/BccZrMeoz6oxqm0Ol9MWXSdOamHlhnK/vM1FI/dUQP5qSgtj9qHGrWgzwKxR8lhgcx+GSn17NY7c8RSAYQFVpXLqB8x87laalGiZ6eD7DoJlX0PYfg2aBHKSfRRPXQeM9iD3FyklLiTG6aOdvIfRt00pR0PYQgapjF+J6pUQdSxErL9RYglRCuPasme7fbXJZ4S/nzeDoc78AAnxtta+DtTr03sDwNpNiRDEDaxKo3Zf4Mo1MW+4vLPiijXA0zK5H7MyBv917ocZYCer0oK17QX4uJhkTRHtvhvorkfDmY3ZdHx+fyYGqcsUvruOf1z1OIBRAHaVuWi0XPH6q3/I4FbHqQAILfW9bZDSFph5BwhtPzPV9xhURG6mehcYPQXtvhN6LShyp0HkSTvY9pPq4KV3B4lco+Cwx/PP6Ocy57V9kUlkS3UmSPSm+fH8uZ+5b6ovuMxlQVdP7qEkGSvRT4HSgPZdO5NAWCu25pMwRKcg8BdX/B/aygAXW0lDzO6zYHmjuIzT7Nqoe7g5eBNcCz4aIEDC0SiMC8YMgtC7GFrISbBSLLz4Kc95Ry/HcP2v7n8llLZ55oO/fASR+IBL7ERCj8PZjmR+pMokEe3mk6UGsxhuR6M5s/r2NufmjP3Bv543c3X4DB5++L/ZIyhxGiPb+2U3E9FV2ZIEU2nkCqv4upY/P4s6cv/6Lh294wsQLXSZemPfJfE773vkTPTSfhSG8HaWXPCO9lyzMos+CCgSbfRYvxIphVf8MQtsNc1QWkjdB5slxG9dY4Fco+Cwx3HvFg6R60wWPOXmH/738IW1z22mYUT9BI/MZFqcN8l95PJGD1GNQe9a4D2mRyDxXwUEOaAfS9Gh/xlpzn+LM39lUOIgNBKH2XCSy9bBnkqpfopnn3IRMXwlmBKqPh+wbkHoIJGzcLyI7IVWzQHvR9sONhaUE3cqQrPcFrGbS0Yc4fKtDCqxV+wiFAWsmUnsOEljBtHyE14f0nEFH2RA7CAmtDdY0CK5TlKkXESKxcWpTST1AoaOFi6Yg9wGUcJFQTQEWIqExHZ6Pj8/Ycvfl3vHCJ+98wdyPW3yhxamG0wKB1SD7kvtAAIhA7amQ+6/RDgqsbqyHE3eB84lJcFtLQeAbEFgGsm+bhHd0T+g6A5wvKHZHKkUQie4+Nu/NZ9Ij9ZehnadC6g48W280ifbeioS3Gu+hjRp+QsFniSHR7e05bNkWyZ7hFPV9JhQJU/KmPRUz/lIHtJY5KAc9l6P5j5Das1HNo20HmqAIp//XoR1HQ9N9SOBrpS8X/CY03IZ2Xwy5N8CagVQdCeHt+Nfd3+Sf11fjZBNs+8Ot2Gr/HRAyaMfxJngiZBINwe0g+xTQO+TsNoS3JVYdZe2tVuf1OW+Rzw3cLEPRIN89fCek+UcDCYLMC5B5hkKhxywkbob4wZNDnEhKfK40D4PtOPsezr2Pdp4E2TcBQcNbI7W/Qyy/lcrHZyqS6Ep4Pm7bFskSsYTP5ETzX5kWNu1lIJawILo7VtTDzrMCpycNb4Im/gqJ2yD/EYUxStD8W8KuD3Ieqn+FBFZa5PfiMzURsZG6M3HSu0L7zyiOpQCn1bQ+2EtXZD8+2fBbHnyWGL79vY0IhopzaFV1cWZ+ffoEjMinEsSqgtCmFOc/IxA9YCKGtGjEfwwUL0qLSUHyH2j2LbMI126KM9s5NHFb2TNJcBWshquwpj2P1XQPEtmeCw/9A+cdfDkvPvAfXn7kIy454lZO3+sCnM6zIPM8Zoe+F8hBdo5RqSbCFx9F+c1+K7DTsmux24qrcemJdSR7kpz4l58zY4XpRKsjhGNhwrEQa22xGvv9+sCCagNNPWJ2+ouwID05Sv4kfqBH4sCCwNeRwDIFj6rTgbbuZ6o9yGM0PuagbQf6oq8+PlOULfbZhGC4uO0rGAmy3GrLeLzCZ7Kivde6FXqD759pSM5GnbaFO6nTYe5joU0h/guw3WSBVEH8MJj2gqnKqz0DmfYkVnwKxio+o46EvlW6Yyb3JrTujLash9N59pRrr/QrFHyWGPb/zZ48fcfzdMzvIp1IEwja2MEAJ/z5SCzLz61NZqTuXLTtEMh/AlimBD+yLRL/0UQPbcRIdG809zEkbjTllZrCLES9bh4ZSD8NdqkANgeOVzvIAOq0QfYdsJdCAisA8MEbH/PE7c+SThiBqmDY4ds7fcUaG33EV+92s9TyQ8v9U5B5mm65jaN3OYOejjyqkE8Kj9z4PJ++M5+Ln/od179zCW888R/mfTyfldZZgZXWWaF4QFaffsKQ9gixSlcGjDeRPSDzCiTvNUJeAFKH1BXbb2riTlfoa3DyIGc0GDIvQnjsnCh8fHzGhr2P2YU5f32GBV+0kU6ksQM2gVCAE/88a0z1W3zGgMxrFFskY+6/uQ8hNLJKMk0/i3YcYSrWyLgCwt+ExjewrEGJ6MiOizRsn8UPkTBacyZ0/gbTRuplyZ2H5E2oXWtaUKcIfkLBZ4mhpqGaa968kIf/8gSvPfYWM74+jd2O2IGlVpwEJdY+wyJWAzTeA7m3IP8FBFZDAmPj8a25j8yORvZtCK6CxA9DAl8ftfOLCFJzAlp1mOnbtGeiqceh+zyK+/aDIHEIrgNeIowSRUJbeL8PVbT7bEj81S29zKLBNZD6P/HaY2/2tyYstXyKyx54j1i1g21DroRUAk4vD930HumUjerATTCbzvL+q+/y36fO5hubHc4631mz5HjI/hucbrxT9A6Eh9eDGC9EBKk9E40fbioPrGYIbYiIR+Ix/z6FbhwuqpD/GPATCj4+U414bZw/vXY+j9z4JK88/P/s3Xd0VNUWwOHfmZ5e6B0EASkiHUFAQEBQQJ+KvWBBLFixYAEVG6hYwAKKDZFmAem9iIACKigqIEiRYgiEhCTT57w/JoQMMwkJJJmU/a3Fei+3zZ6YTM7d95y9N1O5TkX6DelFrUa5tPQVJZepHnj+IChpr11grF6gS2ntRR97JGvGw4mNmf6kfeY0tKmmv86OqT5Yu6GU3GaJQIaIy9DmRujMGWD/Omv26am8kD4JJKEgRMkUER3BgPv6MOC+PuEORRSQUgrMzf3/ioh2/+avVaBdgBc8f6Ht8yHxc3/BwEKkDPFgaeP/wtYXfTxU9XDlL5RorIiOuArs33Ky84DVP3Mh4rLQ7yXjY38yAdfJVlnuzejUJ4hOGIjJbCQ61s77S7djjdCcWJVgCtncQYGlHTt+3oXLHtx2Sykve7bM5dxGX0OFGShT/cBYtEanPQmOhVkzMk7cmFuzZgBoVPy7/uUtJYgy1YbTJa5MLYD5nPzvknNf40KLRWuNzpwOGe+CL9nfCSP2SZS1a6G9hhDiJFuklX5DetFvSK9whyLOgoq607/ULiDxawXrhagCJhTwbCNkwV4ckP4GWpn9yQYVAYZESJyBMlY48+BFmaRMDVCxT+Gzf5PHUelo7UWp0jEjSuZ5CyFEFp02KuvJw4kn8F7Ajk57oUhfVxkroBLe8U+dVNHZ7RNV/FsoY0X/MbEjUXEvgbmV/0Y1+j5U4gyUCux8oLXGl/YGpI8BTr35d4NzNZ2vPA9lUFz3QFJAMiE0/ywJFfskDVrWwxIR3MFAa6h9bgbodP/38FSuVVnJhBOdJk58f134axM0C1nssDRQEf3AEENg6zErmJtltewsHDrzM0h/BXz/AV7w7kSnDEU7fyi01xBCiLJGmc9DJbyXtXTQDFjA1gcVd7oWzn7atRnfsSfwHb0T7ViatdQhFFdW4Uef/3+9B0P/PRTiBHPT3PcZ6pSaZALIDAUhhDjJvSX0ds/vaK2DWhkWJmW9GCqvB2dWW0lrB1SOm2ylFERcjoq4PO8LOZf56zPk2s7KQGQMvDT3KSpYB+aRTFBgugAsLVFRt6GMVelzRxWmj5mN2+HOLjhotvg4p6mdhi2yntC7fgy6krZ/FzhF9OQe/3Q/90/oo7dBwgcoa8e8318JowxRUOEb9PEx/u89Zoj4HyrmoUL7edHaC+njQ3wPHej0N1HWToXyOkIIURYp60VQcRnoVP9SQZW/FsS+jC/h+Kv4ZyVocG0g9Lr3UDzgXHKGEYvyQMUMQx+5nuCHPyZU3IhwhHTGJKEghBAnqGjQaSG2RxVpMiH7ZZQNbGdXR0BnfkHI6fcnGOLAUIXmnaviPdwQvL+FPs7cBkOFKQGbYivE8M66lxl330f8umIzZoum+/9SuPu5A9nH+HyWEFPf8jMZzoE+/jLKOjcfx5YsylgZFf960b2ATsulMwb+omJCCCHypJTKatucP9qXnpVMyPnZa8c/yyFruV72bIWshIMQBaDMzaHCNHTa81mtujWYzvXPSLW0Cnd4BSIJBSGEOCHyRsj4hMABhA0ibghXRAXnC5EQyWZFxb2UnRwxRA9Cpz5FcFFBKyp+bMgr1Dy3GqMXP4svdSTa/jUqR2bd6VAsmhZDRNUV9L7tZGJERVyJdi7JZZZCDp6/8R1/G3QmytYDzG2LJZGTF+35F+2YBzrDP4vE3LL4Y1KxWYU1g+tXYKxbvLEIIUQhU0pdCryNf+3YR1rrV8McErh/zUoaBO0AFJjbgLU7ytIenT4ua4ZazuLJRrB2L7ZwRemkzM1QFWaGO4yzJjUURImStPcwGxdv5tDupHCHIsohFX0/2PoAFlAxgBVsvVAxD4Y7tPyzXQqEms5pgsSp2UX8tNZoUwuIuBr/E5cI/zGGulBpOcpYJc+XUbFPciSpLg67gYw0Aw674vf1UXz4QhXG3T8Je0aOJIWlY9br2LJeKzc+yJgAmZ+iU+5Cpz6evbSiKGit0fa5+I7cjO/IdfgypqFz3LT77HPQyX0g/R3ImIA+Ogid+mSRxhSKUkaIugf/f6OcbKiYh4s1FiFKCu09hHauQXv2hjsUcRaUf6H4u0AfoAlwvVKqSXijImsMEKqdM/iXP2wC32GUuSEqdgQYKvu7MoH/fw2VUbHPFle0oozTvjS0fRY6cybaW/LukWSGgigR3C43o28Zx7rvNmK2mnE73bS9tCVPTX0IizWvGxAhCo9SZlT8aLR3GHj3gLEOylgp3GEViIq8EW2f5W+viR1/3tgCca9gsDQDQLs2oI89Br6jgM9f5DHqNpSpMcp8bv5eR0Uw5uE2HDsAtRo42bvdxt4dNgAiYw38uX4HrXo0zzpWQdRdaJUI3l3gywDXWkK2WzzxhEfbwbEYIvqDtfPZfEtypdOGg30B2UtE3H+gHXMh8TN/K7DUpwms6G33F5eMuLzIYsqNirrDX0E84wPwHQFjbYh+AlXMcQgRblp70KnD/b+LygLahbZ0QCW8E1B3RpQa7YC/tda7AJRS04ABwB9hjcrc3L9EIruY8KkckDkZHf2wf5xQaTE4lqI9f/s7Hdl6olRwEWMhCko7lqOPPYA/waUBAzr6MQzRt4U3sBwkoSBKhM9GzmDdnE24HG5cDjcAGxb9wkdPfsG9bw4Kc3SivFHGSlDKEgknKEMUVPwGnTkLXCvBUAUVeQPK3AgA7T2ATrkzcPmB5w//U/iKiwr0WtHxUWxeEcGebYGDeO3TRMbYsr/2pb8H6e+BMgIG0D6w9fYnDBRZ61A1/qmkOdnRjrlFctOs3dvBPp/ApIYDPL+DczXg8scbNI60o+3fFfuNvFIKFXUrRN1a5AVChSjJdMYEcCwCnKCzEn6u9ei0l1BxL4Y1NnFGagD7cnz9L9A+TLFkU8oAiR+jjw4C38HQB+lM/EUaTf7kQURf5JNZFCbtS0Mfu5/A5TReSH8Vn6UjBkvDcIUWQJY8iBJh7oTFQf3tXXY3Cz5aHqaIxJlSSl2jlNqqlPIppdqEO57yRPvS0Y6F4FyBirgMQ8IHGOKez04mAOjM6aA9p5zpBV9yVgXr01zfdyz76/739MYWFbi8QimISYymUdsG/nNcv0L6BPwttexZbbXs4FwKlVejKsyDuFf9NQKCKEIv3ygE7g2EfOqkM9GuteT+51H519WGkSQTRLmW+QXBs5ucYJ/l74giSptQH2hBH85KqcFKqY1KqY2HDx/O14W1Zy++Y4/hS+qML/lK/9/HggRmOgdVaQUYc5m5ZzwHFea/B6Js0/Y5BCYTTvDBsTvRuRVsLmaSUBAlgiPdGXK7M9NZ7OuVxVn7HfgfsDrcgYSb9h3Hl/4xvpTB+NJGoT27i+y1fPYl6KRO6NTh6NSn0Emd8NkXBB/o3UvwTAA4lmwg5dBuPG4P00bP4ub693FdzcGMGzqJ1KRd+I4OQie18183+XK0eyutLjmfax8fgNlmJjImgogYGwlVE3h5wdPZN73a/jWhlzYolGsjylQTZe1O6LWqVlTElWfxXcmDis8lMWABQwWwXkToaa42VMT/iiYmIcTp+TJy2eEh/y39RAnyL1Arx9c1gQOnHqS1nqi1bqO1blOp0ulnEGrvfvSRK8ExB3z/gWcr+tgT+NI/LFBwShlQcS/jr2Fz4rZJATZ/7QQhipLnr9z3+Q6hU0tGnQ5Jq4kSoWmnRmxZFbxcrlG7BvI0rpTRWv8J5ecpqtYuf4XnzOn+J/CWDqjYp0DF+gczvmP4b6hN6MyZkPA+ytqpcGPwJkPqI/inAOfYkfo42tISZax6cpu5PTiWc6JuwL87Lbxybx12b7Oh1LeYrfNwOz24nf6kw/wPl/LjnHlMXP4XtsisLLlnO/rozVBxCTc9ew2XDe7J72v+IiYxmuZdzsNoNOYIzkHoG3OdPV1ZGSIhfjw65T78A7WsdYLRd6EsLQvlexTE1gPSRobYYURFXOFfix3/DjplaI5lGQoib0RZ2hZNTEKI07O0A9f3BH2umM6VNeul0wbgXKVUPWA/cB1w1q2VdPqErCUJOZPVdkgfj4680f93J5+UpQVUmIFOfxc8f/p/1qLv8bf9E6IomZuDfXru+x0L0L6RKEN08cUUgiQURIlw39u381DnZ3A7PHjcHkxmI2armQfevTPcoYkipJQaDAwGqF27dpijOTP62EPgXEP2U3jXavSRX8Day188L3s2gAfwoFOfhEqrCzfh4lhI6FmjbnRyX/+w29IRFfM4KqI/OvND8B7CaffwyBXnknbUiNYK8OJ2BrZ29Lg8pCYrVs2Opff1R3O8cTfa/hUq+m4SqsTT+aoOIUNTtr5ZLSMzA3doD1g7njzOehFUXuNfCqHtYOmCMtU8k+9Gvihlg8TP0Sl3gz6O//tnQsWPzU7AKGsXqLwKHFktL62dUaZ6RRaTEOL0VOxw9JGfsxKSbsAEyoyKfSHcoYkzoLX2KKXuBxbhbxv5sdZ661lf2LWBkDNWlBG8u8FQsEYSytwIlfDOWYclREGoiH7otOcJNbPUzwg6FZCEghCcc34dPtwylm/emsf2TTtp0LIe/e7tzcaFv/Lm4A8wGI30vesSeg+6OPDppwgLpdRSoGqIXU9rrWfn9zpa64nARIA2bdqUurUt2rMHnN8T2AlA+5/KOxcS8g+AL9XfgaEwb5a1nVzX2Ol0//91LkW7fkRVXICq8DU6/QN++HYlLseJZELuHJkGflsfFZhQwOkflJ2OtStYOvufKOpM/ONFI0TdDSou4FBliIGiWuIQgjI3gUqr/E+ctBvMzYLWwypDPEReU2wxCSHypkz1oeJ8dMan4N4MpkZ4LTcx/8MdLPr0CZRSXDqoG33u7IHJLMPc0kBrPR+YX6gXNVYH784QL+YCQ+ksuizKH6UiIGEiOuUOQi4NVRYw5N3muzjIJ60oMarUqcQ9b94GgNfr5dGLR/L3L//gzPQXa9y9dR8bFv7CyK+GhTFKAaC1viTcMZQInh2gzCcrjWdzgc7t49UHBZhqmS/WrpA+jtBJhRyvq+3ozC8wxDyEin2Sw8ca4XRMI/de234Wq4/q55z6HiNR5lanDU0pA8S/A6616IyPwbUeMEHGRLRjHiRMQJnCNztFKQOYm4bt9YUQBaeMVVGxTwKgtebpXqP4Y912nJn+z6k9f+xj3ZxNvDRveLlZficCqei70Uc3EFjDx+KfaZaPLk5au/zJf0MFlCG2yOIU4nSUtRM68Ss4ej3g4uRyrwiIGV4iCoNKUUZRIm1ctJldm/dkJxPAX6Bxw8Jf2b4pRMZZiHAw1Q3RMQHADJZW+Is4BZwAllYoQ2KhhqHMDSHy2qzXy2vw7AL3L9lfNWpbH2vE6dccG80GLr0+ZyE0ExjiIeLy/MWnFBirZE1BdQOZgAO8/6CP3orWeSc0hBAiN5tXbuXPH3dkJxMAnJkufvv+D7au3RbGyEQ4KUs7iB3lL8BLBP5kQndU3OunPdeXMQWd1B595Ep0Ukd8xx4tMdX08+OX5b/x/NWvM6zHc8x+byFOe+jC56L0MFiaoSrOAVt/MNYES3tUwnsYIktGkeizSmkopV4D+uFPl+wEBmmtj+V9lhCnt3nlVuzpwR/ePq+X37//i4at64chKpEfSqkrgXFAJWCeUupXrXXvMIdVJJSpAdpyAbh+xv8xeGKH2T+QyXgf7LP9X+MDY01U3NiiiSXmKbD1RNtn+5dVOFcGxgSACUwNsr+6oHsz6jarxc5fd+Ny+JdnmK0mTGYTLqcbg0FRtW5lHvt0CInnLIPMmf5rWnuiYh7yT8XLJ50xleAlID7Qx8C90V9oTZRZMl4QReW31X/iyAgeL7gcbn5b/SfNOjUOQ1SiJDBEDkBHXAbeA2CIz9dMA+1YBsfHcKJwMQCOxWgMqPjXii7YQjJt9Ld8Merr7ATbXz/uYMFHy3hn7UtYbFK0tDRTprol9mfwbGcoLAGaaa3PB7YDw8+8/8hHAAAgAElEQVQ+JCGgQvUELDZz0HaTxURC1fgwRCTyS2v9rda6ptbaqrWuUlaTCSeo+A8goj9gAQxgao5KnILBVAND3IuoSotQca+iEj9HVZiDMlYsmjiUwmdszcwJbbm+mYst6624Xad8xCszKvLW7C8NBgOvLRvJwMcHUKVOJSrXrsg1j/Zn+sEPmXnoI6bsfp+P/3yb89qfhyH6fgyVV2GovA5D3AsFn2XhO0SuLd18Rwp2LVEayXhBFImEqvFYI6xB2y02s4wXBEqZUKba+V62oDPeJyCZAIAzq5p+eqHHV5hSk9P4/LmZQbN19u84yLIp34cxMlHWnVVCQWu9WOvs+b7r8feOFeKsdb+hM4YQxRdNJiMdB7QJQ0RChKYMkRjiXkZV2YKq8huGil+jcqzJV8bqKFsvlPn8067l1Z69+NJexHd0EL7j49Degt1oj79/EpNfmMnRgymMuLkOq+fE4nYqfD4D2tgAlfBJUL0Ca4SVW5+7li/+eY8pu99n0IvXExFlIyYhmoQqhTcYV9aLCV4CQlYxxCJqDSlKDBkviKJy8bUdMRqDh7NGo5EuV4fuPiNErrz/5bLDCL6juewrGbau3YbZGjz53JHh5IdZP4UhIlFeFGYNhduBBYV4PVGGbN+0k/EPTOKNO97jx/k/4/PlvWY6oXIcL89/isRqCURE27BFWqlarzKvLX8u5JMIIcJNKQNKBc+qyS/t+hl9pD9kTgXXD/6ihcl90J59+Tr/2OFUFn22MrvuiD3DyJihdbiiYTOuadqcu7o2JCOz0RnHd9Yi+oGxBpDj91dFQORN2W0a86K1xpfxJb6kzvgONcGXPADtXFd08YqiJOMFkaudm3fz3kMf8/rt77L2uw2nHS9Ex0fx6uJnqVgzEVuUDVuUlcq1KzJ66QgiY/K/LEsIAMytCXl7pMxgrFbs4RRETEI0Wgc3zFIGRULluBBnCFE4TltDIT/t4ZRST+MvLz4lj+uU+n7zIv9cDhdHDx0joUoc3723iM9GTMftdOPzaVbOWEvbS1vy7IxHsp/Yer1eVs1Yx9IvVmO2mLj09u50uLw1U/d9wJ6t+zCYjNRuXEOqNYsyS6c+k9VW8QQnaDf6+Gv56n397/aDWKxm3I7AOgUet4H0VHDY/+Pz56Zz71u3F3Lk+aOUDSrMRGdOAcd8UDGoqJvA2itf5+uMDyH9XbKnonr+RKfcDYmfoCytiy5wkW8yXhBnwuV0c/RgCvGV41j86QomPjY5e7yw6qv1tOjahOdnPZ7dMtrn8/H91z+y5POVKIOi923d6HRFO77c8wG7t+5DKUWdJjVlvCDOiIp5AO1aldWO+UQyKwJiHj+rhwbFoWmnRkTHReFId5Azr2Cxmbn8njK9+lSEmQqVySrQBZS6FRgC9NA6YDScqzZt2uiNGzee1euKkklrzRejvmLGa7MB8Pk0bqcb7Qv8ObNFWRnx1TDa9r4ArTXPDhjN5hW/48jwr/uyRlq45OauPPT+4GJ/D2WFUmqT1rrUrA8pz58L2peOTmpHyLaPKgZDlU2nvUbygaPcUv9+3M5TCx+elFAljhkHP/K/ptYs+GgZM8fOIT0lnZY9mjPoxeupVi/8/YxPpbULndQedEbwTkt7DImTiz+oUiqcnwsyXhA5aa2ZPmYWU176BrTG5/XhcXvxeQNnJNiirDw5+QE6XdEOrTUvXvsmPy34OWC80PmqDjzx2dBwvI0yQcYLgbTnH3T6eHBtAmM1VPQ9KGuXInu9wrRv236G93mJtOTjKIPC6/Zy79uD6HundPsWBVOQz4Wz7fJwKfAE0DW/gwNRts0av4DpY2YHFIQJxZHhZPVX62jb+wJ+Wf47m1duzR4cgL+IzLwJSzCajNz39iAMBulwKsowZSHXdo8qMl+XqFg9kQv7tWbd3I24HaFaWYIhxzrjiY9PZu4Hi7N/71ZNX8uGhb/y4ZY3qFijQoHCL3K+o6BzKejo3lG8sYgzIuMFcaoFk5YFVKPPjSPDycoZP9Dpinb8sW57QDIB/OOFpZNXY7KYeOiDwdkzGYQ4E9p7GJQFFfd6qZzlUqtRDSbvfJftG3eSkWbnvPYNiIiWpT+iaJ3tXdp4IAZYopT6VSn1QSHEJEqxaa/OOu3g4ISMY/4x5YaFv+AI0SISYP6HS/n02WmFFp8QJZFSFrD1wt8pIicbRN4IgHaswJd8Ob5DLfAl90M7VwZd54nJD9D3jktQhuBBkNlmpuctXQFIO3Kc2eMXBgzKfT6NI8PBV2/OLay3VXgMiZDbwM5Ut1hDEWdMxgsiwNRXvi3weGHTks3ZdWJOteTzVXz4mMxWEmdGe/fjO3IN+nA39OFL0Ye7o10/hzusM6KUolHbBrTq0VySCaJYnG2XhwZa61pa6wuy/g0prMBEybdv2342LdlMyn8nW4mnHk7N9/m1GtcAICYxOmRVWgCPy8O378zHlcc0biHKAhU7CswtABuoaMAKtktQUXfisy9GH3sQPNsBO3i2oVMe8PfLzsFiNXP/uDv4bMc44irFYsiRWIhNjGbg41cAsHvrvpBtWT0uL7+t/rMI32Uw7UvDlz4JX8q9+I6/jvYeCDpGKQtE3k5wlwgbKvrBYolTnB0ZL5Rv+/8+yKYlmzlyMCV7W8qhY3mcEahWI/94ITo+KtfxgtftZe6EJdgzQj+gECI3WnvRR24E92+AC3CCbz865XZ0rl0fhCh9tLajfUdDFu88GzKPXBRYRmoGj148kntaPc6ogWO5se69jBv6ET6fj7rN8ldAy2Qx0bZ3CwAuualLnksatE+TduR4ocQuREmlDNEYKkxBVfgaFfcGquICDPFjUcoE6aOBUwfJDvTxV0Ney2Aw4HK48OWoXXI8JY23Bt2K73BfKibOC1lrQRkUNc4tvirW2vsfOvlSSH8bnEsh4xN0cl+065fg2KKHQvR9oLIqVRtro+LfRlkvLLZ4hRAFY0+383jPFxjcYhijBo7llvr3MXbwB3i9XupfUDdf1zBZjLTr628t2+26Tqg8xgvKaChQokIIAFxrQadysghjFu1B278OS0hCFCbtO44vZSj6vzbopM7o5J5oV+G1EpWEgiiw1wa9x5/rt+O0u8hIzcTtdLPok5XM/3Ap94y9DWvEqdO2CVgebrIYOef82jTt1BiAyrUq8tTUhzAYQ09pNllMxFeKLYq3IkSJo8znomzdUKaagL9wGd5cWkfmsv3rN+cGdXtw2b2sW6hJ2rOHqhU/p3kHe9CTPovNzMDH+p/9m8gnnf4m+FI4mSxxg85Epz4VdKxSBgzRgzFU2YCq8ieGSktRtm7FFqsQouDeGjKR39f8iStrvOByuFn+5ffMemc+d79+C9bIU8YLioDxgtFspGbD6rTs0RyAhCrxjPzqUQym3IevFWskFsE7EWWa9z/QodqTunL/+ytEKaJTBoNzOeD2//PuRR+9C+3ZXSjXl4RCOfLfnsMc2p10VtNcMo/b+Wn+z7hdgUXfnJlOvn5rHi0ubsqYZSNp1fN8EqslcEG3Zjw74xE6DmiL2WoiItpGr1u7MWbpyIBiNx37t2XM0pFBNzjWSCs3j7wGk/ms6ocKUWoppcBQKfROQ+WQm3f8vAuPO7iIocWi2bfTCjh55sN/6HhZNcxWE2abmYo1K/DsjEdpcEG9Qoz+NBwrgBDFFr170b6U4O1ZlJKia0IUpaR9yRz857+zGi+4nG6+/2o9buep4wUX37yzgCYXNuKNlS/Qtk9LEqsl0LzLeTwz7RE6X9UBi82MLcrKJTd1YeyqFwJmMba9tCVvrh6F2Rq4bMsaaeX6J6/AYgvxUEOIvJjPB0L8rKtIlKVdsYcjRGHSnr/BvRV/MiEnFzrjs0J5DblLKwf++X0vL147lkP/JKGUolKtCjw97eHT3ji4HC4WfbqSVTPXEhUbyeVDenHO+XUgRME3gPQUf0u3Jh0aMnrRswH7ulx9+mnJLbo2ZfTiEUx8fDL/bNlDYrV4bnr2GnrdenH+3qgQZVXU/ZD+alZf7BMiIPr+kIfXv6Auf/64A+8pSQW3S1HjHH8RtMioTJ76yIfT9DH2dAcJVeKLv6K1igCdW+LAWqyhCCH8tZFGDRzL/h0HUUqRWC2Bp6c+RKO2DfI8T2sX2Geh7XPBEImKuBaXvW2uCYmMVP94oVGb+rw8L3BGUtdrTj9eaNKhIa8vH8nExybz96//EF85juuHXymt8cQZUeaGaGsXcK7m5Iw5CxiqgK1vOEMT4ux59oEygz516awXvLsK5SUkoVDG2TMcPHrxSI4fTc/e9u/2gwzr9hxT9rxPVGzolnRul5uHu4xg7x//4siqwvzz0i387+HLiK8Uy+F9RwKONxgNtMmqiXA2mnc+j3HrXj7r6whRlqjI69B4IH0c6OOgYiD6QQyRA0Mef9VDl7Po05UBCQWLzUebi9OoWutEhtoMxupEREeErwp05A2QPp7A+hAmsHZGGfLXLlMIUThcTjePdBlBavLx7ETAwV3/8fglLzB517vEVogJeZ7WHvTRW8HzR3bSUzvXERl5PVXqVmb/joMBxyuDolWP88863iYXNuKtNS+e9XWEAFDxb6Ezp0DmNNBOiOiLirobpSS5LUo583mgQ3XHsYK5TaG8hCx5KOPWfPMjHldwT3qvx8uqGetyPW/VjHXs/fNkMgH8vaBnvjaHO0ffhDXSmt3T3mw1Ex0fyaBR1xX+GxBCoJTCEHUzqvJ6VOWNqMo/Yoi6MXt/8oGjrP1uA9s2/I3WmmrnVOGNFc/RuP25KKWwRWr63nSU4e/vzXFVIyoivL+zKup2sHYDrKCiQEWCqSEq7pWwxiVEebR+zkacDlfQrAKPx8uyL7/P/UTnMvD8ecoMKjtkTuHxj688ZbxgIioukjtfvTH0tYQIE6VMGKJuxVBpAYbKyzHEDEMZQifRhChNlLEqRFxOYKcsg39JT1ThfBbLDIUy7sj+ozjtwVkpR4aTI/uP5nreujkbA3rUn2C2mDAaDIz/8RW+fnMu/24/QIuuTRhwfx8SqsQXauxCiEBKGfw33lm01rz/8KfMm7gEk8WEz+ujSt1KjF48goat6zNu3cv4fD6Ubz/62H3gsYAyAjZU/BiUKX9dWYqKUiZUwtv+okCeP8FYA0zNi3/phRCC5P1HQz6AcNldJO1NzvU87VwBOjN4hzJyXqsU3ts4mq/fnMvev/bTtGNDrnzgMipUSyjM0IUQQuRBxb6INjWEzMngS/fPBI15FGUonCK2klAo487r0BBLhAVHeuC6mYhoG407nJvrefGV4zAYDfi8gVVv3W4PM8d+R72mtel3b28atq5fJHELIU5v2ZTvWTBpGS6HG1dWV4d9fx3ghWve4O2sqcAGgwEMtVAVv0N79vmfIprql6jChspUF0x1wx2GEOXaeR0aYjQZg4ooRkTbaJbVlSkkQwX8w8mT57mcihXfxvHD4h+IrXKA/vf0pnG73MccQgghio5SRlTUIIgaVCTXlyUPZdz5XZvQsPU5Aa0cLREW6jWvTeueua9hvHzwJZgtwfkmt8PNtp92suizlTzSdQQLP1leJHELIU7v23fmBc0k8nl9/P3zLpL3Hwk6XplqocwNS1QyQQhRMjRu14Dmnc8LaOVosZmp0bAaHS5vnet5KuJqcj6fcjkUjwxowLvPVOTHhXtZ+vlqhnV/ju/eX1SU4QshhAgTSSiUcUopXln4DLc8N5BajWtQq1F1bn72asYsHRHQhulU9ZrX4aEJg7FFWYmMjQhq56h9Gmemi3cf+DigzoIQovhkpNpDbjcYDWQeP7WarxBC5E4pxQuzn2DQqOuo06QmNc6txvXDr2TsqhcwmnJPQipTPYh71V8DRUWz9Ouq7P3bhjPTP8bQ2j9emDjsczLSQiyNEEIIUarJkodywGI1M/CxAQx8bECBzrvkpq5c9L8O/PXjDt5/+FN2bdkTdIzBaGDbhr9p0bVpYYUrhMinjgPaMuud+bhPWfdsjbRS49yqYYpKCFFamcwmrnq4H1c93K9A5xki+qJt3cG1mTWLvsGZGdyKzGg28cfabbS9tGVhhStEmfHH+u0s/nQlbqebrgM70vbSC6SekCg1JKFQzuz9az9rvvkRpRStezZnzvuLWTHtBzweL216tWDo+DupUqdS9vG2SCsXdGtG5ToVQyYUfF4fUXHS3k2cpJR6DegHuICdwCCt9bHwRlU2XffEFayasZbUw2k47S4MRgNmq4lhk+7FaJRlDUKIM/fvjoOs+Xo9Xq+PNr1bsGDScpZ9sRq300PL7s0Y+u6dVK9/MnGplA2s7YmpsBaldnFKswi0lvGCEKF8MWom00bPxmX3d1lZ/dU6LuzfhuFfPChJBVEqSEKhHJk2+lsmv/BVdm/6T56ZilLg8/n/6m9Y8Av3t3uST3eMIyo28I/+gPv68Ouy3wOWNyilqFA9gfot6hbbexClwhJguNbao5QaDQwHnghzTGVSbIUYJm55gwUfLePnpVuoWq8yA+67lDpNaoU7NCFEKfbNO/OYNPxLvB4v+DSfjZyOUiq7UPOmpVu4v/1wPt3+DrGJga31+t97KevmbMIZMF6A6PgoGreXwoxC5JS09zBTX/k2u7Ay+DuxrftuI1tW/UGLi2UGcGmxbcPfTBj2Ods37SK+UizXPXkFlw3uWS6SQlJDoZz4d/sBJj8/E5fdhdfjxevxorXOTiaAP7Fgz3Cy9IvVQee36dWC6568AovNTGRsBBExNirXrshL854qF78oIv+01ou11ifm4K8HaoYznrIuKjaSqx/px8vzn+aBd++SZIIQ4qwk7T3MpCen+McLbi9erw/t0wFdn7RP47K7WPjxiqDzm3c+j1ufH4g5e7wQQWL1RF5Z+EyetZuEKI82LPwVFeL3wpnpZM2sn8IQkTgTu7bs4dFuz/Hb93/izHTy357DTHj0cz5/fma4QysWMkOhnPhh1oagFpChODOd7Px1d8h9Nz5zNZcP6cXWtduIrRBDkwsbyuBAnM7twPTcdiqlBgODAWrXrl1cMQkhhMjF2tkb83Wc0+7i75+DayUAXPNof3oP6sbWH7YRHR9F006NZLwgRAi2KBsGQ/CDOYPRSGS0LQwRlT9Jew+zfdMuKteuyLmtzjmjB6UnHtrm5Mh0MvP177j28QHYIq2FFW6JJAmFUuzQ7iQWf7aSlP9SadOrBR36tc513bTBoPxzDk/DGmml/gV1c90fVzGWjv3bnmnIooxQSi0FQlX9e1prPTvrmKfxNyafktt1tNYTgYkAbdq00bkdJ4QQ4swl7Utm0acrOHoghZY9mtPpina5dm5Q+RwvWCIsNGhZL9f9sYkxXNivzRnHLER50KFfa94aEjz8MZmNXHJzlzBEVH74fD7eHDyBZV9+j9liwuf1UePcary66BniK8UV6Fo7ftmFPrVwDGAwKpL2JlO7cY3CCrtEkoRCKbV+7iZevHYsXq8Pj8vDgo+WYYu2ctVDlzHgvj7EVghc03jRVe35dMS0PK9pMChskVYuuUk+wETetNaX5LVfKXUrcDnQQ4f6hBVCCFEsNi3ZzMgrX8Pn9eJ2eljw8XKsERaufKAvVwztEzRw7nRFWyY+9nme11RKYbGZufT27kUZuhBlXlRsJM998xjPX/U6yqDQWuN1exn82s3s33GIfdsO0LJHcyKiZLZCYZvzgb8wvdvhxp1Vw2LP1n2Mvnkcryx8pkDXqtmwOv/tPhy03ev2UqF6QqHEW5LJ/LNSyO1y8+ot7+C0u/BktYvzerxkHMtkykvfcEfTh0nefyTgnGr1qnDXmJuw2MyYrWb//9rMNG7fAIvNjNFkoHXvCxj348tBBRmFKAil1KX4izD211pL03EhhAgTr9fLyze+jTPTiduZNV5we8lMszPt1Vnc0eQhDu1OCjinYo0K3D/uDiw2c8CYoVG7BlgjLBiMBlr2aM649a8EPbwQQhRc654tmHHoIx775D4emTiEYZ/cx0dPTuGVm95m9C3jGFjlTtZ8+2O4wyxzZo9fEFA8FsDj9rJ55VaOp6QX6Fo3PXs11ghLwDZrpIVet3UrF/dVMkOhFNq2YSfaF/qhr9ft5fjR43w2YjqPTroX8CcgPhs5g7kfLMbt8lCldiU6X92BK+6/lMq1/S0itdZSXFEUlvGAFViS9TO1Xms9JLwhCSFE+bNr8x7cTnfIfV6Pl/SUDCYNn8LTUx/O3jb5hZnMHr8Qt8tDpVoV6XxlOwYM7UO1elUAGS+IwiNtpk+yRVq56Mr2pCancWPde3BmBq7Hf/Wmd/h0xzgqVk8MU4RlT+ZxR8jtyqBwZDiJSYjO97WadWrMszMeYfzQj0nal4zFZqb/vb25/aUbCivcEk1mKJRCZqs514QCgNfjY/38n7O/HnPreL59Zz4ZqZlon+bQ7iTmfrAYb44ijTI4EIVFa91Aa11La31B1j9JJgghRBicbrzg82k2Lt6c/fWbd0/gqzfmkH4sA+3TJO05zLwPl2bPhgQZL4hCtQRoprU+H9iOv810ubb6q/UQ4ldWa82q6WuLP6AyrMNlrTGag2vJxFeJo2KNgidu2l/Wms93jmfWsc+Ydewz7hp9c661asoaSSiUQue2qkd0fFSex5xYa3X43yOsnb0hqPKoy+nmq7FziyxGIYQQQoRXnSY1Sagan+cxJ6bppiSlsvzLNThPHS843EwfPavIYhTll7SZDmY/bsfj8QZt97g8ZB63hyGisuuW564hrmJs9meg0WzEGmll2KR7zzhxqpQiIsqWa5H8skoSCqWQwWDghe+eICYx2l+NOYQTvwj7th3AbDUH7fe6vbm2exJCCCFE6aeUYtTsJ4irFIvBGHrIp7LaOR74+xAWW/B4wef18fcvu4syTCHA32Z6QbiDCLc2vS8I+VTbbDPTrk/LMERUdiVWTWDS1je55flradenJf3v6c2EX1+jZffm4Q6t1JGEQinV4IJ6TNs/kcbtGoTcf+TAUfb/fZCaDavhCrF+0mgy5tnuSQghhBClX50mtZi67wOadz4v5P7jR9P55/e9VK9fBZcjeLxgMBrybCctRF6UUkuVUr+H+DcgxzGnbTOtlBqslNqolNp4+HBwNf2y4pzz69B7UDdsUdbsbbYoKxcP7EijtqHH/OLMRcdHMXBYf16a9xT3vjWIGg2qhTukUkmKMpZiFquZ3BrymSwmkv89SouLm3Jhvzasn7spYNmD2WbmqocvL6ZIhRBCCBEuZos51ym8JrORw/uOUK9ZbS6+tiOrZ64LWPZgsZoZ+PiAkOcKcTqF1WZaaz0RmAjQpk2bMt2Oeui4O+g0oC2LP1uF1ppLbupC20svCHdYJZrT7mTZlDVsXPQrlWtX5PK7e1KzYfVwh1VuSEKhFNm5eTeTnpzCnz/tILFqPDc8dRUXdGvKzl//yW4HdYLb6abe+bUBeHLyUD55ZirzJizFnuGgcbsG3D/uDqrXrxqOtyGEEEKIIrR76z4mDZ/C7z/8RXylOK59fAAtujXlj3XbgmYhuBxuGrSsC8AjHw6hQo1E5ry3iMzjds5tdQ73j7uDOueV+6XtogjkaDPdVdpMn6SUonXPFrTu2SLcoZQKGWmZDG0/nMP/HsGR4cRoNjJ3wmKenf4I7S9rHe7wyoVCSSgopYYBrwGVtNbJhXFNEWjPH/t46KJncWY60BrSUzJ46+4J/O/hy4iMiSDdm4HX4+/aYIuy0v++S4lN9PeHNlvMDB5zC4PH3CLtnoQQQoSNjBeK3v6/D/JAx6dwpJ8cL4x/4GP633cpUXGReD3peLOKvtkirfS+vRuJVRMAMJlN3PHSDdzx0g0yXhDFQdpMi7P29ZtzObT7cHaLXK/bi9ftZdS1Y3ngvbvoes2FWCOsp7mKOBtnXUNBKVUL6AnsPftwRG4+e24GTrszYImDI9PJN2/P5521L9Hr1oupVLMC57SowwPv3cWdr9wY8joyOBBCCBEOMl4oHlNe/BpnpitgvODMdPLd+AW8+f0o+tzRnUo1K1C3WW3ufXsQ9719e8jryHhBFDVpMy0Kw+qv1mcnE3JyZrp4554PubHuvfy742AYIis/CmOGwpvA48DsQriWyMW2n/4O2Uta4e8j/ciH9xR/UEKUY16vl5RDx4iKj8pu05pTyn/HmPPBYnb8vItzW53D5Xf3zH4KKEQ5JeOFYvDn+u34vL6g7UazEUeGkwffHxyGqIQof7xeL7s278FsNVOnSU1J0hWRyJjgMdgJTrsLl8PNqze/w/j1rxRjVOXLWSUUlFL9gf1a683yS1K0qp1ThaS9wbNDvR4vCVXiwhCREOXXsi+/5/2HPsWR4UBrTbfrL+KBd+/EYvP3Mt77134euPAp3E43Loebn5ds4Zu35vH22pdkLbIol2S8UHyqN6jGv9uDn8a5nR4q1kgMQ0RClD8/L93CSze8hdvpRvs0CVXieH7WE9RrVjvcoZU5A+7rwz+/7cWR4Qy5X2vNzl93k3b0ePZycFG4TptQUEotBUJV73saeArolZ8XUkoNBgYD1K4tv0wFddOzV/PXTztwZp6svGyNsND9houIiosKY2RClC+/LP+NNwd/EPC7uGLqD3hcHp6c/AAA4+7/iMy0zOwpxy6HG7fTzfihk3ht6chwhC1EkZPxQslww1P/Y/PK3wM+oywRZjpd0Z64irFhjEyI8uHwv0cYccUYnJknb3AP7kpiaIfhXHxtJ5L2HKZlj+ZcPqQXMQnRYYy0bOh+w0VsXbeNhR8v9y99yK0HSJnuDRJeKo8OLXmfqFRzYBlwoiprTeAA0E5rfSivc9u0aaM3btx4Rq9bnq2c8QPvP/Qpx1PSUQYDl97ejSFv3IrZYg53aKIEUkpt0lq3CXcc+VVaPhceu+R5fl3+e9B2s9XMtP0TiE2Mobf52pBTjg1GA4vc04sjTCFCCsfngowXit8Ps35i/NBJpCangVL0vBLibGAAAB0gSURBVKkL971ze/YsKiFykvHC2Us+cJR1320ErTm4O4lv35qHx+3N9XhLhIWYhCje//k1EirLTOPC8N+ew7z7wCR+Wvgr3hzfe6UU57Y+h3d/ejXgeHu6nU1LtqB9mlY9zycqNrK4Qy7RCvK5cMZLHrTWvwGVc7zobqCNVG0uOhcP7ESXqy8k7chxImMjsVglkSBEcTv0T1LI7SaLkZRDx4hNjMFiM4ecemexye+sKH9kvFD8Ol3Rjo4D2pKanEZkTIQkEoQoQgsmLWP80Ekog385l9vpCflQISeX3UWqx8vUl7/h3rcGFUeYZV6VOpUYPuVBHuk6kv07DmLPcGCLsmGxmXly8tCAY9fN2cjL17+FwejvT+D1eHnsk/voOrBjOEIv9QqlbaQoPgaDgfhKkskUIlyadGxE0t7koMGC1lC1nv+eqddtF7Nw0vKAfu8Wm5let15cnKEKIcoxpZSMF4QoYkn7khk/dFLA3/v88ri9rPtuoyQUClFEdATjf3qFjYs2s33jTqrWrUznqztgizzZNvLY4VReuu5NnHZXwLljBr1Lk46NqFSzQnGHXeqdddvIE7TWdeVpgxCirLv52auxRloCqjVbI63cPOKa7D7Hd42+mWYXNcYaYSEyNgJrhIWmnRpx15ibwxW2ECWGjBeEEGXFmm9+PKvzoxNkmn1hMxqNtO/biptHXEPPW7oGJBMAvv/6R3+bvFNon4+V09cWU5Rli8xQEEKIAqjZsDrjf3yVT56eyu8//EVitXhuGP6/gGlytkgroxePYPfWfez9819qNa4hlZ2FEEKIMsbr8YVs654ftigr/3vw8kKOSJyOI8OJ1xO8JMXj9uLIcIQhotJPEgpCCFFAtRvXYOTXw057XN2mtajbtFYxRCSEEEKI4tZxQBs+fXZqvo41W00YjAZMJiMup4e+d/bgkpu7FHGE4lTt+rbk0xHTgrZbbBba9W0VhohKP0koCCGEEEIIIUQB1WhQjRufuYovX/oGt8uD1jrXGQudr+rAwMcGkLz/KOe2qkdi1YRijlYA1DmvJv2G9GTexKU4M51o7Z8t0v2Gi2jUpn64wyuVJKFQSDxuD8ePphNbIQajyRjucIQIG6XUKGAA4AOSgNu01gfCG5UQQpQMXo+XtCPHiUmMxmSWYZgQpd0NT11FxwHtWDVzLW6nh9njFwR1erJFWel168XUb1GX+i3qhidQkW3IG7dxYf+2LJ28Cp9X0+PGzrTs0TzcYZVa8pfsNJIPHOWrN+bw64rfqVq3EtcMG0DTjo2y92ut+fLlb5g+ZhZetxeTxcQNT1/FwGH9A4q2CVGOvKa1fhZAKfUAMAIYEt6QhBCiaKX8d4yZb8zh56VbqFyrIlc/2o/zuzTJ3q+1ZuYbc5jy4ld4XB6MZiPXPjaAG56+SsYLQpRy/iWO1wLQsnszRl75Gkr5aywog6LPHT1odcn5YY5S5NSia1NadG0a7jDKBEko5CFpXzJDWj2G/bgdj8vLrs272bh4M8Mm3cvF13YC4Ou35jL1lW9xZvozkS6Hmy+en0lkjI1+Q3qHM3whwkJrnZbjyyjgzKoVCSFEKXH0UAp3X/AY6ccy8Lg87Px1Nz8v28LQ8XfS+7ZuAMybuITPn5uRPV7A4Wbqq7OwRlm5+uF+YYxeCFGYWvdswdR9H/DDtz+RmWanzaUXULtxjXCHJU5Da83yL9cw54NFOO0uul13Ef3v7R3UJUIEK7S2kWXRF6O+IjM1E4/LC/j7zDszXYy7/yO8Hv+2aTmSCSc4Mp18+dI3xR6vECWFUuolpdQ+4Eb8MxSEEKLMmvbqLNJT0vG4PNnbnJku3n/4U9wuf3/6KS9+HTRecGY6mfrKt8UaqxCiaC3/8nvua/sEb9/zIfM+WsrBnYfCHZLIhzcHT+CtIRPY+sM2/v75Hz4fOZ1Hujyb/RkucicJhTxsWrw5ZFsRl8PNwX+S8Pl8pCYfD3luyn/Hijo8IcJGKbVUKfV7iH8DALTWT2utawFTgPvzuM5gpdRGpdTGw4cPF1f4QghRqDYs/AWP2xu03efzse8vfwmZ3MYFacnH8fmCxxpCiNJn4SfLGTt4Agd3JeFxe9j7x7+MunYsP87/OdyhiTz8u+Mgy6asDqh94bS72Lf9IGu++SmMkZUOklDIQ3zl2JDbvR4vsYnRGAwGqjeoGvKYWjK1SZRhWutLtNbNQvybfcqhXwJX5XGdiVrrNlrrNpUqVSraoIUQoogkVIkPud3j9hJbMQaAWo1rhjymeoOqGAwyHBOitNNa8/FTX4aYieRi0vApYYqqfNBak3bkOC6H64zO/33NXxiMwZ/DjnQHmxb/erbhlXnyFywPA4cNwBYVuG7GbDHRutcFxFbwDxCGvHEr1ghLwDHWCAt3v35rscUpREmilDo3x5f9gb/CFYsQQhSHa4b1D1pna7IYadapMRWrJwIw5I1bQo4Xhrwh4wUhShOv18u6ORv5+JkvmTthCRmpGQC4HK5cZy7v33GwOEMsV35a8As31buXa2sM5oqE23j9jvdw2p2nPzGHhMqxKENwcVyTxUTFmhUKK9QySxIKeehyzYVc+8QVWGwWImMjsNgsNO/ShCc/PzmD+8J+bXhh9hM0bn8uMYnRNOnYiJfmPUWbXi3CGLkQYfVq1vKHLUAv4MFwBySEEEXpwn5tuPm5gVgjssYLERaadGjEM9Mfzj6mdc8WvDT/KZp2akRMYjSN2zXghdlPcGG/NmGMXIjyI/O4nd/X/Mn+v8/85t6e4WBo++G8cuPbTH35Wz549DNurHsvOzfvxmKzEB0fGfK8KnVkFmZR2LZxJy9c8zpJe5PxuDy4nW5WTF3DqzeNK9B1WvdqgS3SyqkNd4wmA5fe3r0QIy6bpMtDHpRS3PTM1Vz5QF/2bN1HheqJIT8QWl1yvrSCESKL1jrXJQ5lhdaaFVPXMGv8AjLS7HS+qgPXPHI5UXFR4Q5NCBEmA4f1p9+Qnvzz214SqsZTrV6VoGNadG3KW9+/GIbohCjfpr82m8nPzcBoMeF1eah/QV1emP0EcRVDL2/OzYzXZrPnj39xOfyF+pyZTpzAyze8xbsbRtOyR3O+//pHfN6TdVGskRYGvXh9Yb4dkWX6mFm47IFFE10ONz8t+JkjB1OoUC0hX9cxmU28vuJ5RgwYTfL+oxiMCrPFzJOTh1K1buWiCL1MkYRCPkTFRtLkwkbhDkMIUUK8++DHLPpkRXbxnhk7Z7Ny6hre/+U1IqJsYY5OCBEuEdERMl4QooRZP3cTk5+fidPuArt/jf32jTt5/uo3GLvy+QJda9kX32cnE3I6uOs/7mz6MKnJaQHJhNiKMdzz5m10vqrD2b2JMsbtcrP1h20opWjSsSFmi/mMrvPvtgNoHdyd3GQxcXhfcr4TCgC1G9fgk7/eZu9f+3E73NRrXhujyXhGcZU3klAQQogCSNqXzPwPl+F2nhxQuJ1ukg+ksPTzVfS7p3cYoxNCCCFETl+NnRNUKNHj9rLtpx0k7Uumcq2K+b6WwRi8zv7E9Y4cTAloHQtgsZrpcUPnggddhm1cvJkXrx2bnQgwGAyM+OpRWnZvXuBrNbmwEXv/3I/XE9hlx+3yULNh9QJfTylFnfNCF9AVuZMaCkIIUQB//bgDszU4F+vMdLJx0eYwRCSEEEKI3OTWstVkNpF2JHQRxdz0uq1bUHFVpRRGkzEomQBwPCWDAzsPFeg1yrJjh1N57n+vkZGaSWaancw0O+nHMhgxYDRpR/P+b+Hz+VgyeRUPdHyKu1sOY/qYWVzxQB+skRZUjuIHtigrVw7tQ3R86GWoyQeO5vozIc6MzFA4Da01Py34hTnvLyIzzU7XazvS5/buWGyW058shChzEqvGo33B0+uMJgOV6+T/KYcQouzZuHgz3727kOMp6XS+qgN977okqPuDEKJ4tevbigM7/wu+4VdQO5en0W6Xmw0LfuVYUirNu5xHrUb+dvBXP9KPTYs3s33jTtwuDxabGYvNQkxCFP9uDy726PP6sMlSyGwrp6+FEEsUtNasnrmey+/umeu5Y+98n1Uz12UvN/13+0FWTPuBN1eP4rOR09my+g/iKsRw9bD+XHbXJUHn//3rP7xyw9sc3J0EGuo2q8XTUx+iRoNqhfcGyylJKJzGJ89M5dt35mf/8G7ftIvFn67krTWjzni9jxCi9GraqTEJVeJw2l0B6yRNFpMsdxCiHJvy4ldMGz0re7ywY9MuFn68nHHrX8YaIUkFIcJl4GMDWD7le46nZOB2ulEKLBFW7n1rEBZr8Fh+99Z9DOv+HG6HG6/Xh9aabtd14tGP7sFiNfP68uf4fc1f/PXjDirWrECnK9qy+LNVfPDoZwFLKwxGAw1a1i3QOv6yLj0lA5czuAaFx+khPSUdgCMHU/ht9R9ExUfRqkdzjCYj+7btZ8W0HwLqV7jsLvb/fZDdW/fx/LeP5/m6x1PSefTikWSm2bO3/f3LPzzceQRT9rwn93RnSZY85CH5wFG+fnNu9uAA/NOa9/75L6tmrAtjZEKIcFFK8dqykdRvUQeLzUxEtI3YCjE8M+0RajeuEe7whBBhcOxwKl++/E3geMHu4uDO/1j2xfdhjEwIkVA5jolb3mDgsH40bH0OHQe05dWFT9P7tm5Bx2qtGTFgNKmH08g8bseZ6cRld7FqxlpWTF0D+McBzTufxzXD+tPtuk5YbBb63tWDLld38I8LYmxExNioWrcSz0x/pLjfbonWquf5QUtGAExWM616ns/kF2Zy8zn3MXbwB4waOJbragxm15Y9bP1hGwZj8G2rI93Jz0u3nPZ1l3+5JqjOgvZpnJkO1s/ZdOZvSAAyQyFPv3//J0azEU6p5urIcLJ+7iYuualLmCITQoRT5dqVeG/jGA7tTsJ+3E7tJjUxGqUSsBDl1R/rtmOymIKqvzsynaz9bgN9Q0y/FUIUn7iKsdw26npuG5V3+8bdW/eFXF/vyHAyd8ISuudSYNFgMPD4p/dz4zNXse2nv6lQI5Hmnc/DYJBntzmd1/5cOlzemvVzN2UnYG1RVjpd2Z70Y5lMHzMbt9OdXfjaftzOU31f5sH37wr5vTRZTFSqWeG0r3todxLOTFfQdrfTw+F9R87yXQlJKOQhJjEaRXA1V4PRQEKVuDBEJIQoSaQ3sRACIDYxOmTrMoNBkVA1PgwRCSHOhNvpRhlCd3Jw2oNvSE9Vo0E1WZOfB6UUw6c8yJpvfmTxZytRStHrtm50uqItL177ZlA3DoDM45lExUVii7JiT3cEfNYaTUb63NH9tK/b5MJGREQvwZ7uCNhuNBtp1K7B2b+xck4SCnm4oFszrFFW7On2gPohZouJywbnXjRECCGEEOVHk46NiEmIxpHuDBjsmm1m+kttFSFKjfot6mK2mLETeONpjbDQ40Zp/1gYDAYDXa6+kC5XXxiwPTMtM+TxSilcDjevr3iOkVeMIWnfEQwGhSXi/+3dfZTU1X3H8fd3H2b2CSTIQghPC6IYn0Bc92hJ00DQEElAjRQ9ktpW5SQnoEkkRsWQNNqeqD0hjUlrOWA5LTSaEpP0JG0MtKamTS0CgkIAY63AQiwbjVqedln49o8ZcIFhZmd3mXtn5vM6Z4/s7rC/j7P3990vd+/cm+DeFXcweGRjzmv+zoxmho4Zwq7te46vfkjUJji/5VwuuPK83v9PlTmtw8misqqSR9YsYvCoRmobaqjrX0ttQw2fX/opmi4cETqeiIiIRKCiooKHVi9i6DlDqEn3CzX1SeY9eivnThwTOp6IdFNlVSX3rLiDZF2S6kTq9641DTWMvnhk1hMIpPc+NHsSNfWnbmB7pPMoF04ax4hxw1j2y2/w2IaHWfzzB3hyzxIuu2p8t752ZVUli3/+ALPu+jhDmhp539j38skv3cCf/tN9Jxw5KT1jmZbonWnNzc2+bt26gl83my2/2M6KB1ex++Vfc17zGOZ8adbxSQN351cbXuXgvkOc3zJWuzVLUTCz9e7eHDpHd8VYF3725H+wfNGTtO36DcPPex+3fe1mLp92aehYIj2mutB729b+ihUPrGLn1t2cM76JOYtu4JzxTUCqX/jvja+x/+0DjGsZqyMjpSiErgtmtgB4BGh099/kenwh6sLenW08vfwZ3tjzWyZOvYRJ17ZQWaW9ks6kwx2HWTD5K7z64g4O7W+norKC6kQV8799W8YNNLv63x1t/N1X/4EX/uUlBgw+i9l3zzxlBYTkJ5+6oAkF4LkfrefB2V8//tooqzCStQkWP/sAYy8dHTidSM+EbhDyFVtdeHr5Mzw6b+kJm/gkaxN8+akvcPlHJgRMJtJzqgu9s371Jr583cPH64JZatntw2sWccEVWjYrxSlkXTCzEcBS4HzgslgmFCSMzsOdPLvqOX7xw+cZMLg/19w2lTGXjMr6d9pa32Du+Ls48M7B48d5J+uS3Hz/J7jpnusKEbsk5VMXyv4lD+7Oo/OXnrDRih91Du1v56+/8LcBk4lIKO7Osvv+/pQdgdsPdrDsnpWBUolIaN+64/ET6oK7036gncc+vzxcKJHithi4Gyj8bzglOlXVVUy56QPc/8TnmPfNW3NOJgA88dAPOLTv0PHJBID2A+2sfPB7HNx38EzGlbSyn1A4uO8Qb+z+bcbPbVv7SoHTiEgMOg518HbbOxk/t2v7ngKnEZEYdB7uZPfLme//V174nwKnESl+ZjYD2O3um7rx2Llmts7M1rW1tRUgnRSLF/9tC52Hj5zy8cqqCnZuU89WCL2eUDCz+Wa23cy2mNnDfRGqkJK1CaoSmV8TNaCxf4HTiEgMEjUJ6s+qy/i5IaMGFTiNSGko9n6hsqqSmoaajJ/rf3a/AqcRKQ5mtsbMNmd4mwksBBZ15+u4+xJ3b3b35sbG3Lv6S/kYMirzeDjc0cnZQ3VsbyH0akLBzCYDM4FL3P1C4M/7JFUBVVZVMn3uVSTrEid8vKY+yewvXhsolYiEZGbcfP8nSJ60oVqyLsEtX70xUCqR4lUK/YKZcd38a07pF5J1SWYtmBEolUjc3H2qu1908hvwKjAa2GRmrwHDgQ1m9t6QeaX4zL772lPqcnWyigmTL2LQsLMDpSovvV2h8Gnga+7eDuDue3sfqfBuf2gOU276AImaaur61ZKsS3LDXR9n+u1TQ0cTkUCuv3M6f/TgjfQb2EBFhTFw6Hu44y9v5/dmaddgkR4oiX7hD77y+1z9h5Pf7RdqE8ycN43r75weOppIUXH3l9x9sLs3uXsT0ApMdPfXA0eTInPx776fzy35FP0GNlBTn6Q6WU3LRyey8DufDR2tbPTqlAcz2wj8EJgGHAIWuPvzp3nsXGAuwMiRIy/bsWNHj697pux7az9v7HmTwaMaqa3PvKxRpFhoN/e+4e4cbj9MdbJaZxVL0QtVF0qtX9j/9n7aWt9kyKhB1DbUho4j0isx9AvpVQrNOuVBeupI5xFef20v/QY20H+gXobWW/nUhapufLE1QKblRwvTf/89wBXA5cB3zWyMZ5ilcPclwBJIFYLuhCu0hgH1NAyoDx1DpCTke650rMyMRE0i9wNFylw59Qv1Z9VTf5b6BZG+kl6lINJjlVWVDBs7NHSMspRzQsHdT7vu38w+DTyVbgjWmtlRYBCg7VdFylj6XOmrgJ2hs4hIYahfEBERKT+93UPhB8AUADM7D0gARfubSBHpMzpXWkS6Ur8gIiJSgnKuUMjhceBxM9sMdAC3ZFq+KCLlo+u50tpzQETS1C+IiIiUoF5NKLh7BzCnj7KISJHI8Vrp+4Cru/l1um6+1mf5RCQu6hdERERKU29XKIhIGTrda6XN7GLePVca3j1XuiXTUVDFsPmaiIiIiIhk1qtjI3t8UbM2INQ5UIOI63WbseWB+DIpT3anyzPK3RsLHaarfI6BClwXTlYs3+NQlCe32DIdyxO8LuRD/cIJYssD8WVSnuyi7Rfy0Yd1IZbvTww5YsgAceSIIQOEz9HtuhBkhULIomVm60KftdtVbHkgvkzKk11seXoqpmYmtudUebKLLQ/Elym2PN2lfuFdseWB+DIpT3ax5empvqoLsTwfMeSIIUMsOWLIEFOO7tBLHkTkjNG50iIiIiIipau3x0aKiIiIiIiISBkqxwmFJaEDnCS2PBBfJuXJLrY8pSC251R5sostD8SXKbY8xSC25yy2PBBfJuXJLrY8ocXyfMSQI4YMEEeOGDJAPDlyCrIpo4iIiIiIiIgUt3JcoSAiIiIiIiIivVQ2Ewpm9riZ7TWzzaGzAJjZCDN7xsy2mtkWM7szcJ4aM1trZpvSef4kZJ5jzKzSzF4wsx+FzgKpYxDN7CUz22hm6yLIM8DMVpnZtvRYujJ0plJgZrPS98FRMwu2w66ZTTOz7Wb2ipndEypHlzyqo9nzqI6WAI3znHk0zrtB/UJxMbMFZuZmNijAtR9Jf19eNLPvm9mAAl8/aK8RU42LoY4U471aNhMKwHJgWugQXXQCd7n7+4ErgM+Y2QUB87QDU9x9PDABmGZmVwTMc8ydwNbQIU4y2d0nRHKUy18AP3H384HxxPdcFavNwPXAs6ECmFkl8G3go8AFwE2BawSojuaiOloalqNxno3GefepXygCZjYCuArYGSjCauAid78EeBm4t1AXjqTXiKnGxVBHiu5eLZsJBXd/FngzdI5j3P3X7r4h/ef/IzVYhgXM4+6+L/1udfot6AYbZjYcmA4sDZkjVmbWH/ggsAzA3Tvc/a2wqUqDu2919+2BY7QAr7j7q+7eATwBzAwZSHU0Zx7V0RKgcZ4zj8Z5kVG/kNNi4G4CjWN3/6m7d6bffQ4YXsDLB+81YqlxMdSRYr1Xy2ZCIWZm1gRcCvxX4ByVZrYR2AusdvegeYBvkCrwRwPn6MqBn5rZejObGzjLGKAN+Jv08qylZlYfOJP0nWHAri7vtxLwHxGxUx09rRjrqPSQxvlpxTjO1S8UATObAex2902hs6T9MfDPBbxeVL1G4BoXQx0pyntVEwqBmVkD8D3gs+7+Tsgs7n7E3SeQmhltMbOLQmUxs48Be919fagMpzHJ3SeSWhr2GTP7YMAsVcBE4K/c/VJgPxD8dfbFwszWmNnmDG9BVwF0YRk+pmN5MlAdzSziOio9oHGeWcTjXP1CJHL8vF8ILAqc4dhjFpJa/r/yTOfpGi3Dx4L0GiFrXER1pCjv1arQAcqZmVWTunFWuvtTofMc4+5vmdnPSL2GNNSmVJOAGWZ2DVAD9DezFe4+J1AeANx9T/q/e83s+6SWioV6nX0r0NrlN0OrKIKiEwt3nxo6Qw6twIgu7w8H9gTKEi3V0ayirKOSP43zrKIc5+oX4nG6n/dmdjEwGthkZpD6ObvBzFrc/fVCZOiS5RbgY8CH3b2Q/6CPoteIoMbFUkeK8l7VCoVALFW5lgFb3f3rEeRpPLarrJnVAlOBbaHyuPu97j7c3ZuAG4F/Dd0cmFm9mfU79mfgasI1UKR/2O0ys3HpD30Y+GWoPNLnngfONbPRZpYgdR/8Y+BMUVEdzS7GOir50zjPLsZxrn6hOLj7S+4+2N2b0uOnFZjY15MJuZjZNOCLwAx3P1DIaxNBrxFDjYuljhTrvVo2Ewpm9h3gP4FxZtZqZrcGjjQJ+CQwxVJHCm1Mz4qFMhR4xsxeJFVcVrt7FEcvRWQI8O9mtglYC/zY3X8SONN8YGX6+zYB+LPAeUqCmV1nZq3AlcCPzezpQmdIb9A0D3ia1AZF33X3LYXO0ZXqaE6qoyVA4zwnjfPc1C9IPr4F9ANWp+/vxwp14Uh6jdhqXGhFd69aYVfViIiIiIiIiEgpKJsVCiIiIiIiIiLSdzShICIiIiIiIiJ504SCiIiIiIiIiORNEwoiIiIiIiIikjdNKIiIiIiIiIhI3jShICIiIiIiIiJ504SCiIiIiIiIiORNEwoiIiIiIiIikrf/B9O63sPo5wNRAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1296x288 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (18,4))\n",
    "plt.subplot(1,4,1)\n",
    "plt.scatter(x=X1[:,0],y=X1[:,1],c=y1)\n",
    "plt.subplot(1,4,2)\n",
    "plt.scatter(x=X2[:,0],y=X2[:,1],c=y2)\n",
    "plt.subplot(1,4,3)\n",
    "plt.scatter(x=X3[:,0],y=X3[:,1],c=y3)\n",
    "plt.subplot(1,4,4)\n",
    "plt.scatter(x=X4[:,0],y=X4[:,1],c=y4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "所以，在对不平衡数据建模时，我们不仅需要考虑正负样本的比例，还要考虑其他的影响因素，并尽可能对其进行量化；对分布的不同指标的直观理解，有利于后续的处理方法选择以及模型选择，总的来说，影响因素包括这些：类别不平衡比率、重叠区域大小、训练样本的绝对数量、类内不平衡程度、噪声样本比率、样本维度、离群点等"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.类别不平衡比率\n",
    "这个是最常用的指标，假设正样本数量为$N_+$，负样本数量为$N_-$，则不平衡比率为$IR=\\frac{N_-}{N_+}$\n",
    "### 2.重叠区域大小\n",
    "该区域可以通过$k$近邻的点来看，如果与某一样本点相距最近的$k$个样本点中，既包含同类样本点，又包含异类样本点，可将该样本点视作处于重叠区域\n",
    "### 3.训练样本的绝对数量\n",
    "训练样本的绝对数量也是一个很重要的指标，如果绝对数量过小，则所有训练样本可以视作整体样本的一个随机抽样，这里会引入偏差\n",
    "### 4.类内不平衡程度\n",
    "类内不平衡程度是指正样本可能还有多个簇（这无疑雪上加霜，本来数据就少，还被分在了多个地方...），这可通过密度聚类的方式来评估\n",
    "### 5.噪声样本比率\n",
    "正样本中的噪声样本对建模的影响很大（如果后续采用了过采样一类的算法，还会进一步扩大其影响），同样可以通过$k$近邻的点来判断，如果某一个样本点周围都是异类样本点，可以将该点视作噪声样本\n",
    "### 6.样本维度\n",
    "样本维度也会存在一些影响，本来数据量就很少的正样本还被分散在了一个很高维的空间，无疑会增加建模的难度，这通常可以通过降维算法来降低维度，但如何“无损”降维需要更多的尝试\n",
    "### 7.离群点\n",
    "离群点也会对分类算法产生影响（不同的分类算法，影响程度不同），可以通过某样本点距离同类样本中心距离和平均距离做对比来评估  \n",
    "\n",
    "这些影响因素中，我们需要重点关注**类别不平衡比例、重叠区域大小、类内不平衡程度、噪声样本比例**，下面构建一个类来对这些指标进行计算："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.cluster import DBSCAN\n",
    "import pandas as pd\n",
    "\n",
    "\n",
    "class ImbLearnUtils(object):\n",
    "    def __init__(self, k=3):\n",
    "        # 计算重叠区域和异常点所需的k\n",
    "        self.k = k\n",
    "        # 每个label包含的样本数\n",
    "        self.label_map_num = {}\n",
    "        # 正样本标签\n",
    "        self.pos_label = None\n",
    "        # 负样本标签\n",
    "        self.neg_label = None\n",
    "        # 类间不平衡率\n",
    "        self.ir = None\n",
    "        # 所有标签集合\n",
    "        self.labels = None\n",
    "        # 各样本最近k个样本的类别分布\n",
    "        self.k_neighbors_dist = []\n",
    "        # 记录该点是否在重叠区域\n",
    "        self.overlap_marks = []\n",
    "        # 记录该点是否是异常点\n",
    "        self.abnormal_marks = []\n",
    "        # 正样本密度聚类类别数\n",
    "        self.pos_db_scan_classes = None\n",
    "        # 负样本密度聚类类别数\n",
    "        self.neg_db_scan_classes = None\n",
    "        # 正样本中重叠量\n",
    "        self.pos_and_overlap_nums = None\n",
    "        # 负样本中重叠量\n",
    "        self.neg_and_overlap_nums = None\n",
    "        # 正样本中异常量\n",
    "        self.pos_and_abn_nums = None\n",
    "        # 负样本中异常量\n",
    "        self.neg_and_abn_nums = None\n",
    "\n",
    "    def fit(self, X, y):\n",
    "        # 确定正类和负类\n",
    "        tmp_labels = []\n",
    "        tmp_nums = []\n",
    "        self.labels = set(y)\n",
    "        if (len(self.labels)) != 2:\n",
    "            raise Exception(\"标签类别数:\", len(self.labels))\n",
    "        for y_label in self.labels:\n",
    "            self.label_map_num[y_label] = np.sum(y == y_label)\n",
    "            tmp_labels.append(y_label)\n",
    "            tmp_nums.append(np.sum(y == y_label))\n",
    "        if tmp_nums[0] < tmp_nums[1]:\n",
    "            self.pos_label = tmp_labels[0]\n",
    "            self.neg_label = tmp_labels[1]\n",
    "        else:\n",
    "            self.pos_label = tmp_labels[1]\n",
    "            self.neg_label = tmp_labels[0]\n",
    "\n",
    "        # 计算类间不平衡度\n",
    "        self.ir = self.label_map_num[self.neg_label] / self.label_map_num[self.pos_label]\n",
    "        # 计算每个样本的最近k个点的类别分布\n",
    "        knn = KNeighborsClassifier(n_neighbors=self.k)\n",
    "        knn.fit(X, y)\n",
    "        distances = knn.kneighbors_graph().todense()\n",
    "        for index in range(len(y)):\n",
    "            indices = np.argwhere(np.asarray(distances[index]).reshape(-1) == 1).reshape(-1)\n",
    "            self.k_neighbors_dist.append(y[indices].tolist())\n",
    "        # 计算重叠区域和噪声\n",
    "        for index in range(len(y)):\n",
    "            label_set = set(self.k_neighbors_dist[index])\n",
    "            # 该点附近既有正样本又有负样本\n",
    "            if len(label_set) == 2:\n",
    "                self.overlap_marks.append(1)\n",
    "            else:\n",
    "                self.overlap_marks.append(0)\n",
    "            if len(label_set) == 1 and label_set.pop() != y[index]:\n",
    "                self.abnormal_marks.append(1)\n",
    "            else:\n",
    "                self.abnormal_marks.append(0)\n",
    "        self.overlap_marks = np.asarray(self.overlap_marks)\n",
    "        self.abnormal_marks = np.asarray(self.abnormal_marks)\n",
    "\n",
    "        self.pos_and_overlap_nums = np.sum((y == self.pos_label) * (self.overlap_marks == 1))\n",
    "        self.neg_and_overlap_nums = np.sum((y == self.neg_label) * (self.overlap_marks == 1))\n",
    "\n",
    "        self.pos_and_abn_nums = np.sum((y == self.pos_label) * (self.abnormal_marks == 1))\n",
    "        self.neg_and_abn_nums = np.sum((y == self.neg_label) * (self.abnormal_marks == 1))\n",
    "\n",
    "        # 计算类内的不平衡度\n",
    "        self.pos_db_scan_classes = len(set(DBSCAN().fit(X[np.argwhere(y == self.pos_label).reshape(-1)]).labels_)) - 1\n",
    "        self.neg_db_scan_classes = len(set(DBSCAN().fit(X[np.argwhere(y == self.neg_label).reshape(-1)]).labels_)) - 1\n",
    "\n",
    "    def show(self):\n",
    "        \"\"\"\n",
    "        展示各统计指标\n",
    "        :return:\n",
    "        \"\"\"\n",
    "        show_list = []\n",
    "        show_list.append(['样本数量', self.label_map_num[self.pos_label], self.label_map_num[self.neg_label]])\n",
    "        show_list.append(['重叠样本数', self.pos_and_overlap_nums, self.neg_and_overlap_nums])\n",
    "        show_list.append(['异常样本数', self.pos_and_abn_nums, self.neg_and_abn_nums])\n",
    "        show_list.append(['类内不平衡度', self.pos_db_scan_classes, self.neg_db_scan_classes])\n",
    "        return pd.DataFrame(show_list, columns=['index_name', 'pos', 'neg'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index_name</th>\n",
       "      <th>pos</th>\n",
       "      <th>neg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>样本数量</td>\n",
       "      <td>26</td>\n",
       "      <td>474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>重叠样本数</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>异常样本数</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>类内不平衡度</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  index_name  pos  neg\n",
       "0       样本数量   26  474\n",
       "1      重叠样本数    2    4\n",
       "2      异常样本数    1    0\n",
       "3     类内不平衡度    1    1"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imbu=ImbLearnUtils()\n",
    "imbu.fit(X1,y1)\n",
    "imbu.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index_name</th>\n",
       "      <th>pos</th>\n",
       "      <th>neg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>样本数量</td>\n",
       "      <td>26</td>\n",
       "      <td>474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>重叠样本数</td>\n",
       "      <td>15</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>异常样本数</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>类内不平衡度</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  index_name  pos  neg\n",
       "0       样本数量   26  474\n",
       "1      重叠样本数   15   41\n",
       "2      异常样本数    7    2\n",
       "3     类内不平衡度    1    1"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imbu=ImbLearnUtils()\n",
    "imbu.fit(X2,y2)\n",
    "imbu.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index_name</th>\n",
       "      <th>pos</th>\n",
       "      <th>neg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>样本数量</td>\n",
       "      <td>40</td>\n",
       "      <td>460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>重叠样本数</td>\n",
       "      <td>9</td>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>异常样本数</td>\n",
       "      <td>15</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>类内不平衡度</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  index_name  pos  neg\n",
       "0       样本数量   40  460\n",
       "1      重叠样本数    9   38\n",
       "2      异常样本数   15    2\n",
       "3     类内不平衡度    3    1"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imbu=ImbLearnUtils()\n",
    "imbu.fit(X3,y3)\n",
    "imbu.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index_name</th>\n",
       "      <th>pos</th>\n",
       "      <th>neg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>样本数量</td>\n",
       "      <td>27</td>\n",
       "      <td>473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>重叠样本数</td>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>异常样本数</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>类内不平衡度</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  index_name  pos  neg\n",
       "0       样本数量   27  473\n",
       "1      重叠样本数    1   10\n",
       "2      异常样本数    2    0\n",
       "3     类内不平衡度    1    2"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imbu=ImbLearnUtils()\n",
    "imbu.fit(X4,y4)\n",
    "imbu.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}