user-study-1-results.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MTurk Similarity Comparison User Study 1: Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import h5py\n",
    "import math\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from scipy import stats\n",
    "from statsmodels.stats.multitest import multipletests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('~/Downloads/Batch_3580188_batch_results.csv') # Initial test run with 5 repetitions\n",
    "df = pd.read_csv('~/Downloads/Batch_3580708_batch_results.csv') # 25 repetitions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average worker time was 5.51 minutes\n"
     ]
    }
   ],
   "source": [
    "print('Average worker time was {:.2f} minutes'.format(np.mean(df['WorkTimeInSeconds']) / 60))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "pattern_sizes = 3\n",
    "pattern_sizes_names = ['3 kb', '12 kb', '120 kb']\n",
    "pattern_sizes_shortnames = ['3kb', '12kb', '120kb']\n",
    "comparisons = 9\n",
    "techniques = 7\n",
    "technique_options = ['A', 'B', 'C', 'D', 'E', 'F', 'G']\n",
    "technique_names = ['CAE*', 'ED', 'SAX', 'DTW', 'UMAP', 'TSFRESH', 'XCORR']\n",
    "options = 2\n",
    "option_names = ['most', 'second most']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Aggregate results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "results_3kb = []\n",
    "results_12kb = []\n",
    "results_120kb = []\n",
    "\n",
    "results = [results_3kb, results_12kb, results_120kb]\n",
    "\n",
    "rows_3kb = np.where(df['Input.T0A'].str.contains('3KB', regex=False).values)[0]\n",
    "rows_12kb = np.where(df['Input.T0A'].str.contains('12KB', regex=False).values)[0]\n",
    "rows_120kb = np.where(df['Input.T0A'].str.contains('120KB', regex=False).values)[0]\n",
    "\n",
    "for c in np.arange(comparisons):\n",
    "    c_res_3kb = []\n",
    "    c_res_12kb = []\n",
    "    c_res_120kb = []\n",
    "    for o in np.arange(options):\n",
    "        o_res_3kb = []\n",
    "        o_res_12kb = []\n",
    "        o_res_120kb = []\n",
    "        for t in technique_options:\n",
    "            counts = df['Answer.T{}{}{}.{}'.format(str(c), str(o + 1), t, t)].value_counts()\n",
    "            try:\n",
    "                o_res_3kb.append(df['Answer.T{}{}{}.{}'.format(str(c), str(o + 1), t, t)][rows_3kb].value_counts().loc[True])\n",
    "            except KeyError:\n",
    "                o_res_3kb.append(0)\n",
    "            try:\n",
    "                o_res_12kb.append(df['Answer.T{}{}{}.{}'.format(str(c), str(o + 1), t, t)][rows_12kb].value_counts().loc[True])\n",
    "            except KeyError:\n",
    "                o_res_12kb.append(0)\n",
    "            try:\n",
    "                o_res_120kb.append(df['Answer.T{}{}{}.{}'.format(str(c), str(o + 1), t, t)][rows_120kb].value_counts().loc[True])\n",
    "            except KeyError:\n",
    "                o_res_120kb.append(0)\n",
    "        c_res_3kb.append(o_res_3kb)\n",
    "        c_res_12kb.append(o_res_12kb)\n",
    "        c_res_120kb.append(o_res_120kb)\n",
    "    results_3kb.append(c_res_3kb)\n",
    "    results_12kb.append(c_res_12kb)\n",
    "    results_120kb.append(c_res_120kb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1728x1296 with 78 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import math\n",
    "\n",
    "fig, ax = plt.subplots(\n",
    "    nrows=comparisons * (options + 1) - 1,\n",
    "    ncols=pattern_sizes,\n",
    "    figsize=(8 * pattern_sizes, comparisons * options),\n",
    "    sharex=True,\n",
    "    gridspec_kw=dict(height_ratios=([1,1,0.5] * (comparisons - 1) + [1,1]), hspace=0.1)\n",
    ")\n",
    "\n",
    "for c in range(comparisons):\n",
    "    for o in range(options):\n",
    "        for s in range(pattern_sizes):\n",
    "            res = results[s][c][o]\n",
    "            ax[c * 3 + o, s].bar(technique_names, res)\n",
    "            ax[c * 3 + o, s].set_yticks(range(min(res), max(res) + 1, max(1, math.floor(max(res) / 2))))\n",
    "            if c == 0 and o == 0:\n",
    "                ax[c * 3 + o, s].set_title(pattern_sizes_names[s])\n",
    "            if c < (comparisons - 1) or o != 1:\n",
    "                ax[c * 3 + o, s].tick_params(axis='x', length=0)\n",
    "            ax[c * 3 + o, s].spines['top'].set_visible(False)\n",
    "            ax[c * 3 + o, s].spines['left'].set_visible(False)\n",
    "            ax[c * 3 + o, s].spines['right'].set_visible(False)\n",
    "        \n",
    "    if c < comparisons - 1:\n",
    "        for s in range(pattern_sizes):\n",
    "            ax[c * 3 + 2, s].set_xticks([], [])\n",
    "            ax[c * 3 + 2, s].set_yticks([], [])\n",
    "            ax[c * 3 + 2, s].tick_params(axis='x', length=0)\n",
    "            ax[c * 3 + 2, s].spines['top'].set_visible(False)\n",
    "            ax[c * 3 + 2, s].spines['right'].set_visible(False)\n",
    "            ax[c * 3 + 2, s].spines['bottom'].set_visible(False)\n",
    "            ax[c * 3 + 2, s].spines['left'].set_visible(False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Most and second most combined"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABx8AAAIdCAYAAADswrOYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3X24VXWZ+P/3DQiGoomKomDqSEhiOpqhkzEooZWN9lVnkpxRBxObnB6/Wsx8r2qsq35N5ajlXCWOqU2mqT2gpqSiaFampliY4RMEAqlE8RDK4+f3xz4HN4dzzn7ea+2936/rOhfs9Xjvtfda696fe63PipQSkiRJkiRJkiRJklSrAVkHIEmSJEmSJEmSJKk9WHyUJEmSJEmSJEmSVBcWHyVJkiRJkiRJkiTVhcVHSZIkSZIkSZIkSXVh8VGSJEmSJEmSJElSXVh8lCRJkiRJkiRJklQXFh/VcSLiPyIidf+VOc+k4nkiYlKDw5QkSVKZIuLaojxtUZnznNMjv9u/oUFKkiSpLBExtyhHm1vmPBW390mSGsfio3IvIi6OiNsi4tmI+FNEbIqI1RHxm4j4RkQcmnWMkiRJnSgipkTEpRFxf1d+VvJirYjYJyLOj4jrI+KJiPhDRGzsyvN+HhEXRcTQJr8VSZKktpFFjhYR4yLimohYFBGvds33YER8KCJ2aNiblSTl0qCsA5DK8Jlehg0Dxnf9nRsRf59SmtXcsCRJkjreBcApFc7zfuArvQx/PXBM19+0iJiUUnqxxvgkSZI6UVNztIg4A/g2UFxkHAK8revv/RHx7pTS6gpjkiS1KO98VCt4CfgB8FXg34EvAo8Wjd8B+P8yiEtSTvS4irOvv0VZx9mbiBjU1T3MxKxjkaQqJGAZcDvw3QrnfQm4Dvg08N/AqqJxBwNfrkeAkvLJ/E2SGqppOVpEvAm4ltcKjwuAzwLfArZ0DXsbcEWFcUjKGfM3VcI7H5V7KaW9eg6LiE8DTwFv7Bp0QD3WFREz2LaQeRnwiT6mPR24EDgUWA/cC/y/lNKCesQiqSLH9Hj9Q+AJ4D+Khq1vWjSVGUThh9km4IGMY5GkSp2ZUloHhWdkU7hivpQlwNnADSmljd0DI+JrwK8pXCUPcFKtwUVEUGg0+5eiwZ9IKV3ax+TnAx+ikGOuAX5MIb9bVmsskrZj/iZJjdPMHO3fi8b9BXh7Sunlrnn/CFzUNe4fI+LzKaVnKn87W2MZDNwMnNw1aCPwjymlm3qZdgfg/wL/DLwBWAHcAnw2pbSq5/SSymL+prJZfFRLiYgBwG7AicB+RaN+U4dl9yw8fj6l9JmucT0n/wTwd0WvhwKnAZMj4m9TSr+uNR5J5UspPVT8OiLWAyt6Dq9VRAxJKeU1iZKkputu1Kpwnu/1MfzpiJgPHNk1aEhv05Wrl8LjFuD8lNL/9DHL14D3FL3eETiHQn53TEppaS3xSNqW+ZskNU6zcrSudrqTiya/r7vw2OUmXis+BvBeeu/ataReCo+vAKenlO7oY5YfsG1uty/wUeC4iHhbSmltNXFIncz8TZWw21W1hIjYPyISsJnClUrXU2gQAvgj8JEal9+z8Hhhd+GxD39H4QqJz1HowqLb6yl0NSEppyLimIj4YUS8EBGvRMTvIuLiiBjSY7qHIuKeiDg1Ip7oSqimdY3bOyJuioi1EbEyImZGxOld3Usc3WM574uIhyNiXUT8KSJujIh9u8btSOEHE8Dni7qomNE1/m8i4t6udayLiOci4rKGbyRJykDXMXH/okG/rGFZPQuPG4H391N4hELj1O0U8rufFg0fDXy92lgk1c78TZKyUyJH+ytgWNHr53rM3vP1YVXG0LPwuAZ4Vz+FRyjcofld4PPAvKLhb6aQ70lqIPM3eeejWt1TwPtSSlXf+dij8LgF+GBK6aoSs90DnJBSSl3LuA44q2vcX0fEW1NKD1cbk6SG2h94BLgaWEuh6+TPUOiG5Zwe046ncFXm54DFwMtdDdq3AgdR6Hp5EfA+4JKeK4qIjwH/BVxFoWuH13ct676IOJxC4vO3wP3Albx28cLiiNgNuJPChQ5nUei+Zn/gqBreuyTlUlGxcPeuQZuBi6tdHNsWHl+lcFX8j0vM962U0rld8VwMzAEmdY17b0TsnVL6Q5UxSarN/pi/SVLTlZGj7d5jltUlXu9RRRg9C48rgXemlB4pMd9nU0qfB4iIL1DoNW1M17gPRMQnU0qbqohHUnn2x/yto1l8VKtYSaGbhkHA3hSuTP8rYBzwcERMSyndUOWyuwuPm4B/SindWMY8/9tdeOxSXHwEeAtg8VHKoeJjRVci8yCFJOSbEfHhlNKaosn3BI5LKT1VNM/JFBKQU1JKt3YNnh0Rd1HUHXREvB74AvDNlNKHiob/CvgtcFZK6ZsR0X2seKG4m4qIOBbYBfi/KaWni2K6poa3L0m503UV6rUUfkgCJArdo/60z5n6tx+vFR7XAienlO4rY77ruv+TUtoSEf/La8XHAI4A+ru6XlKDmL9JUvOVmaP1fE5RqdeJyhU/Y+4PwJSU0vwy5ivO7dZHxI3Ap7sGDaPwjO/fVhGPpDKYv8luV9USUkqrU0pfTSl9KaX0MeBgClejQ6H71f+JiL1rXM16Cg/XLseLJV7vVmMskhokInaLiEsi4nkK+/1GCldGDaRwUUOxBcWJT5eju+a7rcfwW3q8fjuF58FeHxGDuv+A57v+JpYI9XcUupK5OiLe391VhCS1k6787X5ea9TaAJyZUrq6TqtYBywrc1rzOymnzN8kqbkqyNFW9Hi9S4nXf6wxtNXAyyWnKjC3kzJk/iaLj2pJXd0iFB94hgITqlxc94FtJ+COiCjnluq9Srz+c5WxSGq87wD/DFwKvIPCVVSf6Bq3Y49pl/cy/0jg5R53P8P2P2RGdP37IIUEq/hvDNt3T7ONlNIK4HgKP86uBF7o6vv+5P7mk6RWERGHUegp4q1dg1YA76ihN4tuf6ZwVTwUjsVzIuLAMuYzv5Pyy/xNkpqkwhztOQqN9t0O6jG+Z4HhiSpCWkKhNwso3K14T0T0ezzuYm4nZcv8rcNZfFSuRcRxvV1tEBEDgHf1GFxN1w0AU4Dfd/1/F+AnXYlWf/6p63bxbmf3GF+q33lJGYiIYRSOHV9MKX09pfRASulRCs8D601vx5XlwJ49jgGw/Q+Z7is6308hwer59+FS8aaUHk0pvRcYDrwNWAp8PyLG9D+nJOVbRPwdhR+Ho7sGPQm8tYauVoutAk4A/tT1el/g3ojYr+9ZgKJ8rivX/KeicQn4VR1ik1Qh8zdJap5Kc7SU0ha2vTlgUkSMKHr9D8WTAz+qIqzngfdSuAMKCs+Gu6urq8X+FOd2Q4AzisatARZUEYukMpi/CXzmo/LvbOAfI+I+Cg0+f6bwcOp3U3jeY7fVFLqDqFhKaWlETKGQXI2g0O3C3RExKaXUV9/v7wDmdsV1JIVnUHabl1LyeY9SPg2l8MyJjd0DupKYnhcQ9OchYAjwdxQefN3t73tM9wCFvuwPLHEXzwYKSdbr+pogpbQR+HlE/AeF5O1g4JkKYpakhoiI91H4QQevNVJ1+5eI6M6RHkkpfa9rntOB7/HahZCvAN8HTtv+dyUzU0qrK40rpfSbiDgJuIfCsf8NFO6AnJhS6u2qWoBpXY1lvwKOY9vueW5NKf2h99kkNZj5myRVqMk52heB0ygcZ4cCP42I6ynkX+cUzXN9Sqmq42BKaU5EvB+4iUKXjUdQePbblB7PjSt2cUQcTOHuzL+jcAdUt2919aomqTHM32TxUS1hIIVi3zv6GL8GOCOltKraFaSUnomIE4G5wK4UHnLb3UDV2wHmPrZvlIJCEfSfq41DUmOllF6MiHnAjIhYQeGChukULmoo120U7m6+JiL+HVhE4QrKsV3jt3Sta2VEzAAuiYh9gJ9QOF7tS+H4cWdK6ZaU0paIWACcEhH3Urhj5wUKV1r9IzCLwt3Zw4CPd8XsBQ6S8uJd9P0DsvhK9+soNGZB4Wr14h5YXgd8po9l3EIhv6pYSukXEXEqhR+qgyl0AzYnIv42pdTbs4Luo3BB2Xt6DF9KGVfLSmoM8zdJqkrTcrSU0pMR8c/Atym0Nb8RuLjH9D8HLig3+N6klH4QEdOB7udOTgB+HBHvTCmt62WWuRTuhOppPn2/L0l1YP4msNtV5d/VwBUUDjTLKVyhsIFC384PUEgWxqSU7qx1RSmleRQam17pGrQ3hQaqA3qZ/HPAmV1xvULhYPRDYELXciTl198Dv6HQj/u3gIXAReXO3NXX/MnAvcAlwI0Urpz6fNckq4qm/RpwOoUfcdcDPwY+2zX9b4oW+y/AJuAOCseVcyg88HoT8B/AncD/AH8BJqeUevZvL0nqRUrpJxS6T93SNWgchecEDe9l8g8AH6HQILWewvONrqOQ3y1pQriS+mb+Jkk51nW30eHAtcBiCm13qygUHf8VmFRNbxa9rOdbwIVFg94OzIqIns+Pg0IB9rPAs13xLAO+DkysRyySSjJ/63Cx/fM6JUlSpSLiagpdzexh9y2SJEn5Z/4mSZLUWszfWofdrkqSVKGI+ACwI4Wro4ZQeA7tPwOfM/GRJEnKH/M3SZKk1mL+1tosPkqSVLl1FLrmO4DCc8Sep9D1y6VZBiVJkqQ+mb9JkiS1FvO3Fma3q5IkSZIkSZIkSZLqYkDWAUiSJEmSJEmSJElqDxYfJUmSJEmSJEmSJNVFs5/5aB+vktS3qGBaj6dtKqL018Au03vl/lO5crdZZtsrpVTWPiFJGeqo80/3MdlcJL+Kz5st8Dl11P5TJ7nP3+C17555nKSc8vyjlpaDnLysfcg7HyVJktrMkiVLOO644xg3bhyHHHIIl19+OQAXXXQRBx98MG9+85uJiB9GxOszDrVfEcHiNVtYs8Hfe5Ikqb31lb/dfPPNHHLIIQwYMICIeEvGYZYlIiw8SpLU4aLJ1VFbjsrw2GOPcdFFF/Hoo49y9NFH89WvfpVDDz0067AkNZ5XXsk7H6vn/lNk+fLlLF++nCOOOII1a9Zw5JFH8qMf/YgXXniB448/nkGDBhERXwZIKX2qxOIy2V5P/2kz/3Lvqzy1cgs7DoQPjB/Mv791iHdDSsqbjjr/5OAqa5XgnY+tq6/8LSIYMGAA559/Pvfff/9RKaVHy1hcJttr05bEFx9ez3d+t5EtCaaO3YFPTxjC4IHmbpJyxfOPWloOcvKy9qFmd7uqEtauXcsJJ5zAH//4RwDuuusuTjzxRJ5//nmGDBliY5sktTGP8aqXkSNHMnLkSACGDRvGuHHjWLp0KSeccELxZA8Bp2cRX3+6i4sfnVsoPAK8uhmueGIDb9lrICe8wfRVkiS1n77ytylTpmQcWfm++esNfPM3G7e+vvrJjew6JLjwyCEZRiVJkrJgt6s5c+edd24tPHZbvnw5c+bMsVFakiRVbNGiRTz++ONMmDCh56hpwJ0ZhNSviOCPr2zhiRVbtht3z+JNGUQkSVL7sp0hn/rJ33Lttue3z9Vufc78TZKaxfO68sRLxxuk0h196tSpfPe732Xo0KG9jt9pp50AeO9738usWbMqWnYLdLUiSZIaYO3atZx22mlcdtll7LLLLluHf+ELXwDYBFzf23wRMR2YDnDllVcyffr0qta/z1VrKpp+5x3g6XOGMWxw8Poh8Of1247fb1ghv7rqNxv47EPre1lC35adN6yi6SVJkrLQV/5WSlb5G8BP/s9QDt1jIK8fsn1b2G47FoY9++fNTLx5XcXLNoeTJKk1WXzMiblz57Jp0yZOPPFExo4dy4IFC7aOO+yww5g4cSKvvvoqP/3pTzOMUpIktYqNGzdy2mmnceaZZ3LqqaduHX7ddddx++23A5yZ+rhCKaU0E5jZ/bLhwXZZuxHmvrCJSaMG8bG/HsJ/FBUY9905eP/BgwH48UKvoJcktbccPMtHGegrfytHVvkbwJwlmzh0j4Gc/+bB/Gz5K2zpWnsAH3zzDgDcs3hzM0OSJEkZs/iYE8uXL+f666/n7LPP5v777+crX/kKjzzyCEcffTSf/OQnGTBgANdccw0rV67MOlRJUsZsjFIpKSXOPfdcxo0bxyc+8Ymtw2fPns1//ud/cv/997PnnntWful5E1z2+AaO3Wcg0w8dzOF7DuAnv9/EyJ0G8PdjdmDXIcFPl27i4RdtvJIkSe2lr/ytFXz7txuZdshgjh89iB++ZyjX/24DW4Az3rgDf7PPINZtTFz72w1ZhylJkpoomtxwWfXKqun2oVHK6fKhmv6Vd9ppJ2bNmsXkyZO3G3fHHXdw2mmn8eqrr1a8XBunpZZRyYHDHbsNVXru8Pi+DfefIg8++CBvf/vbOfTQQxkwoPCI7y9+8Yt85CMfYf369ey+++488cQTTwAPpZQ+WGJxTc/fTjpgEJdM3JFdBm/7sd63ZBP/cu8rrK6i7couuyQ1SEedf7wAqnmq3dbF+WQ580ZElp9nR+0/pfSVv61fv54Pf/jDvPzyy2zYsOElYF5K6cQSi2t6/vbWvQdyzZTXbe1mtdvqDYnz7n6Fny6r7uIxczhJDdC255+Mz+tqkhzk5GXtQxYfq9Co4iPAgAEDOOmkkzjrrLPYe++9WbZsGddeey2zZ8+u+svkAUdqGW2b/Kg8Fh9r4v5TuXK3WSb52047wCkH7sDY3QawblNi9qJNPLFiS9XLs+FKUoN01PknBw0dHaOZxcdKpq+zjtp/6iTX+duwHeAf3rgDR48cSAAPv7iZm57euN1zvCthDiepAdr2/GPxsTPkICe3+NgojSw+NoIHHKlltG3yo/JYfKyJ+0/lct14VW82XElqkI46/+SgoaNjWHzcjl+6go7K38AcTlJDtO35x+JjZ8hBTl7WPjSg0VFIkiRJkqT2kaeLbdud21qSJEmtaFDWAUiS1OlsVJIkSZIkSZLULrzzUZIkSZIk9cuLpSRJkiSVy+KjJEktykZASZIkSZIkSXljt6uSJGXIAqIkSZJqZU4pSZKkPPHOR0mSJEmSJEmSJEl1UVPxMSI+HhFPRsT8iLghInasV2CSJEmSJEmqXkR4V6QkSZKaruriY0TsC3wEeEtKaTwwEDijXoFJ7aT7B18e/iTlg/ukGm3atGmMGDGC8ePHbx128803c8ghhzBgwAAi4i0ZhidJkqQezN8kSVK7qLXb1UHA6yJiEDAUWFZ7SJIkSarVOeecw+zZs7cZNn78eH7wgx8wceLEjKKSJElSX8zfJElSu6i6+JhSWgp8FVgMLAdWpZTu6jldREyPiEcj4tGZM2dWH6lUJOu7B72TUJKUdxMnTmT48OHbDBs3bhxjx47NKCJJkiT1x/xNkiS1i0HVzhgRuwGnAAcAfwZujoh/TCl9p3i6lNJMoLvqmKpdnyRJkiRJkiRJkqR8q6Xb1XcAC1NKL6eUNgI/AP6mPmFJktS+vGtaedepPVdk3ZuCPStIkqRqdWr+psbKOic2R5ak1lX1nY8Uuls9OiKGAq8Ak4FH6xKVJEkqS/cPsJTsXED1Y88VkiTlh/meymH+JkmS8qSWZz7+ErgFeAz4TdeyvLRKkiRJkiRJkiRJ6lC13PlISumzwGfrFIskqQV45XU+RYSfibYxdepU5s6dy4oVKxg1ahQXX3wxw4cP58Mf/jAvv/wywI8jYl5K6cSsY5XyKE9de3l8lzpLI44/5oqtwfxNkiS1i2hy8ln1yva5ak0946jJsvOGlZzGxorGarXt22rxKjOVfFEy+SCLv8t+l6rXqGNC8WfS2zra/DPL/f6TQ+VuM/O3DLT5/toS/D6oTB1z/um5T/i9bJy+jj/9bfO+cr9S+XvG+X3H7D911FH5G5SXw6lx8pQPgece1U3bnn+82Kgz5ODGkLL2oaq7XZUkdY7eHu6etx8h6v1zkiRJkiRJUmewbUh5YfFRkqQ2Y5IpSZLqpdUasFopVkmSJKldWXyUJFWt1RqjOp2flSRJkiRJkqRGG5R1AJKkfLNgVZ5y+lvPw7bMQb/wkjpUHo6B3TwGSpLUHvKUX4A5htqb+5ukSnjnoyRJNcpbAi5JkqT2Vk7+aY4qSZKkrHjnoyRJVeqtQScivPpOkiRJuWZhUpIkSY1k8VFAvn542GgvKWu1FhDzdEzti0VSSZLUl0q6kzefaK5WyDMlSZIki4+SpF5V0rDRLoWs4vfc2/svfo82/EiSpHbXSvlOK8VaLxaAJUmSOlfe22NreuZjRHwrIl6KiPn1CkiSpCyU+9yc7r92UPx+2ul9SZIktYp65GHlzmu+17qmTZvGiBEjsP1Nai+9/SbP8k+S6qmm4iNwLfDOOsQhSVXLOjlrx0Stnd6LKlPq7k+1vtmzZzN27Fgi4tmImJF1PKpe1ue8djz/SVKe1et463G79ZxzzjnMnj076zAkSVLGWimPq6n4mFJ6AFhZp1gkSS3MBuj2UfxZ+rm2l82bN3PBBRdw5513ArwJmBoRb8o4LEmSOlLec6y+8kDzw+abOHEiw4cPzzoMKfeyviDPC/SUFb9z6ikP34la73yUJLWRPJyYmq0T37M618MPP8xBBx3EgQceSEppA3AjcErWcUmSstVuPR+0an6Xddzt9j2QJEmqRNa5WLsZ1OgVRMR0YDrAlVdeyfTp0xu9Skk1ytNBNs8PzVXvIrZ/2HH3dyqrz7Ov77Tfr/L19rmq9SxdupTRo0cXD3oBmNBzunrlb8vOG1bVfFlpte+48TZWq8UrVaMR3Xi677S2vu507GtaP+/8sP2tNeSpvQU8ZktSvfU8znfycbbhxceU0kxgZvfLRq9PklSdRvwI6rnMchooyp2mWPH0lbyPvP3wy7usi8iqXR+f3XYDzd8kqfVVet6uNi8yn+psvX3+5orZqFf+5sVjjWW8jWW8jdVq8ao52qWtKA85bX8X9LVie2fDi4/1YvLTWMbbWMbbWK0Wb5406mTU33LLWWelceXlpNopvLuhdY0aNYolS5ZsMwhYllE4kqQeqjnH9jZPX91nVnvBViX6iqe399PMO+fycpeeeaskSZ2tVYp1leZO/U3f13sulafWaxtV8156i6ueKnmP5bazlpNvN7NNr6ZnPkbEDcAvgLER8UJEnFufsCRJ9eTDz9Uo5Xy3+prG72XzHXXUUTzzzDMsXLiQiBgMnAHcmnVckqT6KHUubfb5tnhdfZ3zS01TTsz9dQtazvLLmVflc5vV39SpUznmmGPA9jdJyp1K8q++2kt6Dutt3mri6iuOcmLtGV9fOVWp5fT1fvqbt5zpK2mPKvc9VvJeylXrfJXM29f0WeW10eQKe77L+ZKUrUrOACWPp/7oV7soM1ep6/7Tzu644w4+9rGP8cwzzzwPfCul9IUSs3T09pKkEhqav1XaFX27q6X7qXqtu3idKaWtr2u9ar3dmL81TLnbzO0lSX1rWP7WjF4mpEaqZw7X1OJjRMwG9mjaCiWptaxIKb2znAk9nkrbKXv/UWU83khSv8zfpOqZvzWIxxtJ6pf5m1SbsvahZt/5KEmSJEmSJEmSJKlN1fTMR0mSJEmSJEmSJEnqZvFRkiRJkiRJkiRJUl1YfJQkSZIkSZIkSZJUFxYfJUmSJEmSJEmSJNWFxUdJkiRJkiRJkiRJdWHxUZIkSZIkSZIkSVJdWHyUJEmSJEmSJEmSVBcWHyVJkiRJkiRJkiTVhcVHSZIkSZIkSZIkSXUxqMnrS01enyS1kqhgWo+nORWx7ceYkh9Vk7j/VK7cbeb2kqS+ef6pQXHeZM7Ukdx/Kmf+Jkm167jzT0S0bK7VnS+2avxtqqx9yDsfc6jnjuSOJUmSOk2r5T+tFq8kSZIkSe0sIra7SUDNY/Exh7Zs2cKNN97IhRdeyC233GJjliRJqti0adMYMWIE48eP327cV7/6VSIiRcQeGYRWUkqJiODRFzfzxYfX860nN7B6Q37zoe54f/aznzFjxgyuuOIKVq9enXVYkiRJTdPddvXCmi1c+th6/utX61m0ess24yRJncPCn6LJCYDZRj+6G65OOukk7rjjjq3DTz/9dG6++eat4yW1rY7r9qEd2e1qZtx/enjggQfYeeedOeuss5g/f/7W4UuWLOEDH/gAd91112LgyJTSihKLymR7/c/8DXzmF+u3vh49LJj93p3Ybcd85kKXXHIJF1544dbXBx54IL/61a94/etfn2FUkprA808N7Ha147n/VC7X3a4+/tJmTv/xOl7ZVHg9ZCDc+O7XMWHvZj/1SZL61XHnn2Z3u1rPHK+WblftsrVh6tPtakR8KyJeioj5vYy7MM9Xzbea7ivmiwuPALfccgvz5s2z8ChJOeYVXcqbiRMnMnz48O2Gf/zjH+fLX/4y5PhH1MYtif96bP02w5asSXx3wYaMIurf+vXr+fznP7/NsOeff56rr746o4gkSVKrWbJkCccddxzjxo3jkEMO4fLLLwdg5cqVTJkyhTFjxhARd0fEbhmH2qevzduwtfAIsH4zXPZYPvM3SZLUWOVcenQtcAXw7eKBETEamAIsrn9Yra/SBuhzzjmHa665hoULF/Y6/vnnn+fwww/nfe97HzfddFNFy7ayL0nZ8Sor5cmtt97Kvvvuy2GHHdbvdBExHZgOcOWVVzJ9+vSq1rfPVWsqmn6vocHjZ+7M6vWJP6/ffvzvVxf2oxsXbOQTD7xa0bKXnTes5DSV5m+jR49m8eLFrFy5klWrVm03/vnnnwdg5syZnH/++RUt22OGJEmdZdCgQVxyySUcccQRrFmzhiOPPJIpU6Zw7bXXMnnyZGbMmEFEzAFmAJ9qVByV5m8A958+lDG7DWTxmi3bjft917AX1mzhrTf+peJll5PDSZKk/ClZfEwpPRAR+/cy6lLgk8CsOsfUkRYsWADA5MmTGTx4MBs2vHZl2NChQ5k0adI200mSJFVi3bp1fOELX+Cuu+4qOW1KaSYws/tlQwMr8sdXEn9en9j9dQN48x4D+PWKbRuwJo8eCMBzq7Zv2MrCiy++yKpVqxg5ciSHHXYYTzzxxDbj3/3udwPmb5IkqbSRI0cycuRIAIYNG8a4ceNYunQps2bNYu7cud2TXQfMpYHFx2pzwQtuAAAgAElEQVQ8t2oLY3YbyPGjB/HUym3vdJw8etDWaSRJUuco2e1qbyLiZGBpSumJkhOrLL/4xS+YP38+I0eO5JprrmGPPQo92Y4YMYL//d//Zfjw4Tz88MPbNWpJkiSV47nnnmPhwoUcdthh7L///gCjgMciYu9sI3vNpgTfe3ojAJdP2pGDdyukqkMGwgWHDebE/Xdgw+bETV3TZG3Dhg1cd911AHz729/mTW96EwA77rgj//Zv/8ZJJ53E+vXr+fa3v93fYiRJkraxaNEiHn/8cSZMmMCLL764tSiZUloOjOhtnoiYHhGPRsSjM2fO7G2Shrn+d4Xc7KOHD+adbxi09UFQk0cP5MIjhwDwnd/lI3+TJCkLnfiopoqf+BwRQ4H/B5xQ5vR16barE5x33nncc889vP/97+f0009n0aJFHHDAAeywww6sXr2aD37wg1mHKEnqRScmEGo9hx56KC+99NLW1xHxAvCWlNKK7KLa3mWPrWfSqIGM3W0g956+E4tWb2H4jsEugwv72X88tJ6XX8lPl6QXX3wxJ5xwAm9+85t58sknee6559hzzz3ZZZddAPjIRz7CihW52sSSJCnH1q5dy2mnncZll122NZ8oR1Y9VwDMWbKZW57ZyOljduBbJ7yOl9ZtYUuCvXcqXEh22/MbuWPhphJLkSRJ7aSaOx//CjgAeCIiFlHiqvmU0syU0ltSSm+x8Ni/hx56iLe97W3cdtttDBo0iDe+8Y0MGDCAH/7whxxzzDE8/vjjWYcoSZJaxNSpUznmmGNYsGABo0aN4uqrr846pLKs2gCn3raOa5/cwNoNif13GcAug4Nfv7yZaXe/wrW/zddV8ytXruTYY4/lm9/8JmvXruWv/uqv2GWXXXjkkUc45ZRTaPadB5IkqXVt3LiR0047jTPPPJNTTz0VgL322ovly5cDEBEjgZf6WURmPnb/q3zuoVdZunYLI4YOYO+dBrD8L1v44sPr+dC9rza3GipJkjIXKZU+/Xc98/H2lNL4XsYtovyr5jsm16j1Lpjdd9+dPffckxdffJE//elPNS2rnM9YUi5UcuBwx86Rco75Hosbzv2ncuVus6q31z5Xral2VgBeNxD23XkA6zYllv2lto9t2XnDSk5Ta/42dOhQRo8ezdq1a1m6dGlNy/KYIbUMzz81KD7uetzrSO4/RVJKnH322QwfPpzLLrts6/CLLrqI3XffnRkzZhAR/wYMTyl9stTiqo2j1vxtQMB+w4IAFq9JbK7xkysnh5OkCnXc+Scimppr1TPH615WNcupNI5a1lXOstso3y1rHypZfIyIG4BJwB7Ai8BnU0pXF41fhMXH7eSpC742+lJL7a7jkp92Ue4x3+NxQ7n/VC73xcd6akbxsZ48Xkgtw/NPDSw+djz3nyIPPvggb3/72zn00EMZMKDQUdkXv/hFJkyYwD/8wz+wePFinn322XuBv08prSyxuLbI38Dio6SG6LjzT5bFR6it8GfxMZfK2odKPvMxpTS1xPj9ywxIkqS2kqdChSRJkqTWdeyxx/bZKDlnzpzu/05uWkCSJEk1KFl8lCRJkiRJkiR1lrxdcNtGdw1JUtuz+Cg1QZ6SNRM1SZIkSZIkSZLUKBYfJUlqkkb2HS9Jqi8vHpMkSZIkqToDsg5AkiRJkiRJkiRJUnvwzkdJLc87E5SFPH3vJEmSJEmSJCkvvPNRkiRJkiRJkiRJUl20zJ2P+1y1JusQtlp23rCsQ5AkSarKggULeN/73scTTzwxr2vQgcBnUkqXZRmXJEnKj4iwVxdJkqQmarf8q+o7HyNibETMK/pbHREfq2dwkiRJqq+xY8cyb948UkqHA0cC64AfZhyWJKmDRIRd2Eu9mDZtGiNGjGD8+PFbh61cuZIpU6YwZswYIuLuiNgtwxAlSZLKUnXxMaW0IKV0uA1XkiRVxsY25chk4LmU0u+zDkSSpFbUzoXUdn5veXXOOecwe/bsbYZ96UtfYvLkyTzzzDMAc4AZWcQmSZJUiXo989GGK0mSpNZzBnBDz4ERMT0iHo2IR2fOnJlBWJIkSZ1n4sSJDB8+fJths2bN4uyzz+5+eR3w3mbHJUmSVKl6FR97bbiSJKndeAW42kVEDAZOBm7uOS6lNDOl9JaU0lumT5/e/OAkSZIEwIsvvsjIkSMBSCktB0b0Np0Xj0mSpDypufjYX8NV13iTH0mSerCIqRx4F/BYSunFrAORJElSbbx4TJKk+rPtrnqD6rCMfhuuUkozge6qY6rD+iRJklS7qdhzRUvL04+glEzzJdUmIjyW5JSfTbb22msvli9fzsiRI4mIkcBLWcckSZJUSj26XbXhSpLUEfLU0C/VYt26dQBTgB9kHIokKSPmNVJrOPnkk7nuuuu6X54NzMowHEmSpLLUVHyMiKHYcCVJktRShg4dSkpp95TSqqxjkSRJUsHUqVM55phjWLBgAaNGjeLqq69mxowZ3H333YwZMwYKbXBfyjhMSZKkkmrqdjWltA7YvU6xSJIkSZIkbb0zs1W6+7RrUtXDDTf03rHYnDlzuv87uWnBqCHydte5xy1JUqPUo9tVSZJUpbz9+JQkSVJ5OimPi4iOer+SJEmqTU13Pqp95OlHhFddScqbPB0jJUmSJEmSJDVXrT1dtFrPHrXyzkdJkjLmleSSJEmSJEmqJ9ub8qm/z6T4M2v1z887HyVJ6kcrn+QlKU/ydDztlCtN88zvg8rlsxSbK0/7piRJai2dkEd02t2LtbD4KEkS2yZIWSUQNq5JkiSVpxm5m41LkiRJjdcJRctOZLerkiT10N2tgcmPJElqN62Q49QaXyPfYx63XR5jkiRJtfMcr1bmnY+SpI6Wt0TOK+wlSVKztEveUa/eI/paTt7yxUYp533aU4fyJm/7p/tHtvw+SI2XVS7Q6TlIK+btNd35GBHvjIgFEfFsRMyoV1CSJDVD3n6YSM0we/Zsxo4di/mbJClvyrljsa9parnbsdJ5a4mzHbTze8uz2bNnYxucJCkrWZz/m7m+LN5bf9u0HXKtqouPETEQ+G/gXcCbgKkR8aZ6BSb1p3jnzPpPkhqh0uOLxyaVY/PmzVxwwQXceeedYP4mSVJd9ZaHtUJeZv6Yf905HLbBSZKUa9XkVJXkYsXT5T1/q+XOx7cCz6aUnk8pbQBuBE6pT1iSJDVOJzSwVJq49NVY5gUY7eXhhx/moIMO4sADD8T8TZKUhVoaZKq5OKta9Zy3UXmT+Vjn6M7hbIOTpPbViJ4Yek7fyerRTlbNOhuhVE8gPQuUWV0gV8szH/cFlhS9fgGY0HOiiJgOTAe48sormT59elUrW3besKrmy0or9b0LxttorRZvq3H7qhKtlGx1x9rfd7ycLr8q2UeqadSrdB8sXof7b3MtXbqU0aNHFw8yfyvSat9H422sVou31bh9O0uW+VepvKOVcsP+FOeN5eSQPedTvpWTw5m/tQbjbaxWi1fqS7Xn5+I2mlY4x7dCjFD+c7Ermb6R669lmnqqpfjYW6TbHeFTSjOBmX2NlyTlQ38nIBP4bFVT4Ct3ufVcTl+NXf2tx0Jkc/Wxjc3fJKnDNeJ83Kj8pZz11mOaRmjGNumv4atZn0dWn307KyeHM3+TpPbW1/m1mrsf+2qzKbctp79ll7veatTjAqtG5Cnltns1Qr17+qjntqml+PgCUHzZ1ShgWW3hSJIapV4nozw2JLTKlVK1qMfnV0sSWcl6ql1+X/Pk8TvXqkaNGsWSJUu2GYT5mySpQfKWo+W5MNls9XifPX8jdMq2y4I5nCR1jlq7xyw336mlraXaYmK9L4LvGUOtvXNVOp/tVf2r5ZmPjwBjIuKAiBgMnAHcWp+wJEm1atTt/uU8B7CcvtErXU4169C23GY66qijeOaZZ1i4cCHmb5LUvmo555snNFaluXKj1l1vfm8aqzuHsw1OkvKnnHN6tfM2Um/rrqbQ2YrtTPWKO8/vPQ+xVV18TCltAv4V+AnwFHBTSunJegUmSapdHk40Fgyl/Bg0aBBXXHEFJ554Ipi/SVKu1XIxV89pevt/LXF4cVj99LV9+5rGbdyZunM4bIOTpNzq7xxd6lxeTdGvUbLOM2rJV3v+X9mKJt8a6n2oktS3Ss6MJY+nnmjVLsrMVeq6/3SIcreZ20uS+tb0/M3uLdUKzN8axvxNkmpn+5vUh3rmcE0tPkbEbGCPpq1QklrLipTSO8uZ0OOptJ2y9x9VxuONJPXL/E2qnvlbg3i8kaR+mb9JtSlrH2r2nY+SJEmSJEmSJEmS2lTVz3yUJEmSJEmSJEmSpGIWHyVJkiRJkiRJkiTVhcVHSZIkSZIkSZIkSXVh8VGSJEmSJEmSJElSXVh8lCRJkiRJkiRJklQXFh8lSZIkSZIkSZIk1YXFR0mSJEmSJEmSJEl1YfFRkiRJkiRJkiRJUl1YfJQkSZIkSZIkSZJUF4OavL7U5PVJUiuJCqb1eCpty/2ncuVuM7eXJPXN849UPfefypm/SVLtPP8o9yJe+5qmlLuvYVn7kHc+SjmWwwOLJElSLpgnSZIkSZKUTxYfpRyLCH72s5/xtre9jZ133pl3vetdLFiwIOuwJEmSmuoPf/gDU6dOZdddd+Wwww7jRz/60TZXgkqSpOxtSYlLH1vPkd9dy19fv5YvP7qezVu8WEhS54gIf6dIXaLUFcMRMRr4NrA3sAWYmVK6PCKGA98D9gcWAf+QUvpTifWZcUgVWLFiBQcccABr167dOuyAAw7g6aefZuDAgZ7M2o/dPkjVc/+pnN12KfdSSkQEkyZN4v777986fODAgcybN4/x48dnGJ0EeP6RauH+08O0adO4/fbbGTFiBPPnzwdg3rx5fPCDH+TVV1/liSee+BXwoZTSwyUWlcn2uuo3G/jsQ+u3GfZvRw3mw4cPySIcSepLw84/3W219tKiWnVKt6ubgP+bUhoHHA1cEBFvAmYAc1JKY4A5Xa8l1dGsWbO2KTwCLFy4kJ///OcWHiVJ/Zo2bRojRozYrjjz9a9/nbFjxxIRT0bElzMKTypLRPDCCy9sU3gE2Lx5MzfccENGUUmS1BjnnHMOs2fP3mbYJz/5ST772c8yb948gM8Auc3fvv/sxl6GbcogEkmSlLVBpSZIKS0Hlnf9f01EPAXsC5wCTOqa7DpgLvCphkQptbhKC4XTp0/nyiuvZOedd+51fPfwyZMnc++991a07BxeKSFJaoBzzjmHf/3Xf+Wss87aOuy+++5j1qxZ/PrXv2bIkCGHRMSIDENUh6k0H3rDG97AokWL2HHHHRk4cCCbN2/eZnx3PvTlL3+ZT32qsp8h5kOSpDyaOHEiixYt2mZYRLB69erul7sCyxoZwz5Xral4nrtPHcohuw9kpx22P9fv1NXy+NyqLbz9pr9UvOxl5w2reB5JkpS9ip75GBH7A38N/BLYq6sw2V2gtPFKqpN77rkHgFNOOYX9999/m3HHHnssRxxxBKtWreKXv/xlBtFJklrBxIkTGT58+DbDvvGNbzBjxgyGDCl0fZVSeimL2KRyLF68mAULFrDHHntw5plnbjNu11135eyzzwbgrrvuyiI8SZKa4rLLLuOiiy5i9OjRAF8F/i3jkLZz/wuFC4TOPWSH7cadO34wAA+84B2QkiR1krKLjxGxM/B94GMppdWlpi+ab3pEPBoRj86cObOaGKWO8/zzz/P973+fHXfckQcffJCPfvSjTJo0iU9/+tP8+Mc/BgoNyH/5S+VXDUqSOtfTTz/NT3/6UyZMmEBE3B8RR/U2nfmb8iClxCWXXALAVVddxSWXXMLxxx/Pueeey0MPPcQ+++zDY489xpw5czKOVJKkxvnGN77BpZdeypIlSwA+Dlzd23RZ5m/X/nYDr2xKvPuAHfjOO1/HO98wiBP2G8g1U3bk1IN2YP3mxLee3NDUmCRJUrainC6HImIH4HbgJyml/+oatgCYlFJaHhEjgbkppbElFmX/RupI1Tyfcdddd+X222/n2GOP3W7cTTfdxJlnnsmmTZVfOWg3Y7nWsAdeSx3A/acXixYt4j3veQ/z588HYPz48Rx//PFcfvnlDBgwYALwPeDA1P/JoWO2lxqr2udVf+UrX+HCCy/cbvjvfvc7TjjhhO7G2IqYD6mOPP9I1XP/6UXP/G3XXXflz3/+MxFBRAwAVqWUdimxmKq3VzXdrgJMGjWQq97xuu26Xn1lU+KCe19l9u+ru/PRblclNUDDzj/dv3n8vaFaFf9+zuH3qax9qOQzH6PwLq8GnuouPHa5FTgb+FLXv7OqCFJSH1atWsWkSZM4+eSTOfPMM9lzzz1ZsmQJ11xzjVf4S5KqMmrUKE499VQigpTSwxGxBdgDeDnr2PKg2uJYI+Twx0VmLrroIm688UamT5/OwQcfzOrVq7nlllv43ve+x6uvvpp1eJIkNdQ+++zD/fffz6RJkwCOB57JNqLezX1hM2/73l94/8E7cMzIgUTAw8s3c/2CjSz/i3mNJEmdpuSdjxFxLPBT4DfAlq7B/07huY83AfsBi4G/TymtLLE+sw11JBszVSav/JWq5/7Ti55Xzn/zm99k2bJlfO5znyMixgJzgP2887HA83VjuX3Vpjz/SNVz/+lh6tSpzJ07lxUrVrDXXntx8cUXM3bsWD760Y+yadMmfv3rXz8MfCil9KsSi2r6nY+N4p2PkhrAOx+Ve+1w52NZ3a7WUe62ktQMNrapTP74lqrn/tNDb41X//RP/8S0adOYN28eTz755OPAhSmle0ssqiO2F3i+bjS3r9qU5x+peu4/lSt3m1l8lKS+WXxU7ll8rFzutpLUDDa2qUz++Jaq5/5TuYY3XrUaz9eN5fZVm/L8I1XP/adyFh8lqXYWH5V77VB8LPnMR0mSJKkaeWq8suFKkiRJkiSpOQZkHYAkSZIkSZIkSZKk9mDxUZIkSZIkSZIkSZmLiFw9tkPVsdtVSZIkSZIkSS0nb43TOXwulyRJmfDOR0mSJEmSJEmSJEl1YfFRktRwdpcgSZIkSZIktT/bAAVlFB8j4lsR8VJEzC8aNjwi7o6IZ7r+3a2xYUqSJEmSJElS6+q+MDcvf5IkNUo5dz5eC7yzx7AZwJyU0hhgTtdrSZL6VcuPG38YSfVx6aWXcsghhxAR8yPihojYMeuYJEmS1L/LL7+8O397MiI+lnU8kiRJ/SlZfEwpPQCs7DH4FOC6rv9fB7y3znG1vKyvXKr0KqasY/SqK6l91bpfe2yQ6mfp0qV87Wtf49FHHyWlNB4YCJyRdVySJEnq2/z587nqqqsA3gocBrwnIsZkG5UkSVLfqn3m414ppeUAXf+OqF9IkiT1zkKkVLtNmzbxyiuvEBGDgKHAsqxjkiSpE5jHqlpPPfUURx99NCmldSmlTcD9wP/JOi5JkqS+VFt8LFtETI+IRyPi0ZkzZzZ6dZKkNmIDjVRf++67LxdeeCH77bcfwHJgVUrprp7Tmb9JkiTlx/jx43nggQeIiN0jYijwbmB08TTmb5IkKU8GVTnfixExMqW0PCJGAi/1NWFKaSbQnfWkKtcnSWoT3QXFlDwlSM32pz/9iVmzZrFw4UJGjBixD3BzRPxjSuk7xdOZv0mS1BgRYR6sio0bN45PfepTfOADH7gbWAs8AWwqnsb8TZIk5Um1dz7eCpzd9f+zgVn1CUeS1E68c1HKl3vuuYcDDjiAPffck5TSRuAHwN9kHZckSZL6d+6555JSOiKlNBFYCTyTdUySJEl9KVl8jIgbgF8AYyPihYg4F/gSMCUingGmdL2WJKlstRQmLWpK1dlvv/146KGHWLduHVHYkSYDT2UdlyRJjWb+qFb30kuFTsciYj/gVOCGTAOSJHW8avKriDAv6xAlu11NKU3tY9TkOsciSRJg45DUKBMmTOD000/niCOOAPgN8Divdc8lSZKknDrttNN48MEHfwtsBC5IKf0p65gkScqC3di3hmjyh9Qx34g8NZyX8xm3Wrytxu2rMlXyRWmJD7Lc737P72U58/ldVg9tt/80QbnbrOrttc9Va6qdte6WnTes5DSerxvL7as25fmnReShkcpnn2/H/adyHZW/QekcLk/5BZTev1stXqlNNez802nn+r7yq/7yrnK2Ualp8pDXNVrx+SKH77WsfajknY95kafkp5zGK0nqZHn7QSVJUqXydC7L4Y9NqWXkaV+WJEmSOkXJZz5KkpRH9hG/PbeJJEmSJEmSpKy1zJ2PkqTOYPGsOsXbrec29I4ZSZIkSZIkSc3inY+SpLqyeNhc3u0oSZKyZB4iSZKkZqu2PczctXm881GSlDlP/I3VaQ88lyRJtTF3qC9zXUmS1E56y23Md9STxUdJkqpQademEVG3BjwTOkmSJElSo+Xtt6cXxUhS67D4KEnaqhWvcq9nUa/c9fU3vDiWntPWY/vm7cefJElqfX3lU83OsxqtVC7Wbu9XkjpN3n4vl3ORcp54DpRUTzU98zEi3hkRCyLi2YiYUa+gJEnZKk6Au/tQL9WXeic8e7Dc91fOtspCo9bb8zvS7t+DVjd79mzGjh2L+ZskqTfVPjvH87/UWLNnz8Y2OEmS1CqqLj5GxEDgv4F3AW8CpkbEm+oVmNSf3hq6s/qT2kVvd+n11Yd7p+0HlbzPRhf4Gr2cSj/XUneCKl82b97MBRdcwJ133gnmb5LUcTr9QrJ6qXVbua1Vqe4cDtvgJEkZM4dpvlbd5rXc+fhW4NmU0vMppQ3AjcAp9QlLktRKsj4JNrIBJw/FxEbEUE6xudYCsw1r+fPwww9z0EEHceCBB2L+Jkmdo1PPyXl4z3mIQa2vO4ezDU6S1CqqzT/NndpHLc983BdYUvT6BWBCz4kiYjowHeDKK69k+vTpVa1s2XnDqpovK63WR7bxNlarxavO4Qm9b52wbSp5jz2fgVTLvMrO0qVLGT16dPEg87cifk8bq9W2b6vFK5Wr+xxe63e8lXOl/nKTSraPOY6apZwczvytNRhvYxmv8qKZOUJxTpZS6jWXqSSeSucvlRPWsi2K5zXvaj21FB97+1Zt9+mnlGYCM/saL0lqvlZuLOpPvRKRZm6fWhK4ZstbPKpcH9818zdJajP17F2hXsXKVteonigq2a42unWucnI48zdJUjmFQNj+vFJqeKPjqmXecoqv9YrBPKwytRQfXwCKL7saBSyrLRxJkmpTS0KQVXGtr6vK2omJWj6MGjWKJUuWbDMI8zdJyo2srpJv5DzN0rPhqdzpixup6hlDPRUvt9JGw2rXZc6WL+ZwkpR/ec6TelNJUa+c6WuNpZL8rdLx9cz1Ks2RKslR63lTRR5yuVqe+fgIMCYiDoiIwcAZwK31CUuS1CitlgxVq9L3mYftUuszFltB8Xss90/1c9RRR/HMM8+wcOFCzN8kqXX0d47M07mz3Dsp+4q10vjrtZxy1lOPafqar9Q2KbW+Wt9rqe9OHr5bna47h7MNTpLyp9pzZM9zeanlVHru7y1nrCSmZugtp63nskutt9rPrbf5ylleOe1e1S4/y1yyL1UXH1NKm4B/BX4CPAXclFJ6sl6BSZJqYyGnshO/1AkGDRrEFVdcwYknngjmb5KUS5Xkb83MYcpdV6UNJr29z2a+93oV7uqp3Ny9vwarSv4qicHfGNnozuGwDU6SWkY5xaVKz7X1Ktbl4dzdjBiqLQj2HF5qvkrG1aqS3wmlCpGNLP4CRDNvv4yI2cAeTVuhJLWWFSmld5YzocdTaTtl7z+qjMcbSeqX+ZtUPfO3BvF4I0n9Mn+TalPWPtTU4qMkSZIkSZIkSZKk9lXLMx8lSZIkSZIkSZIkaSuLj5IkSZIkSZIkSZLqwuKjJEmSJEmSJEmSpLqw+ChJkiRJkiRJkiSpLiw+SpIkSZIkSZIkSaoLi4+SJEmSJEmSJEmS6sLioyRJkiRJkiRJkqS6sPgoSZIkSZIkSZIkqS4sPkqSJEmSJEmSJEmqi0FNXl9q8vokqZVEBdM2/HgaUX44KXl4V+Zytf+0iHK3mdtLkvrm+afFFee85rRN5/5TOfM3Saqd55820J3Dmb9loqx9yDsfJUmSpDrzB5AkSVL+Feds5m+SJNVPyeJjRIyOiPsi4qmIeDIiPto1fHhE3B0Rz3T9u1vjw63ec6u28P1nNvLcqi1ZhyJJktRw06ZNY8SIEYwfP37rsHnz5nH00Udz+OGHExGPRsRbMwyxbXQ3VK1YsYIbb7yRn/3sZxXdPS5JkqRsbNiwgdtuu43bbruNDRs2ZB2OJEltI0pd1RMRI4GRKaXHImIY8CvgvcA5wMqU0pciYgawW0rpUyXWl8klRP/56Houf/y1BOJjfz2YT75lSBahSFJ/ctXtg92uqsXkav/JgwceeICdd96Zs846i/nz5wNwwgkn8PGPf5x3vetdRMRJwCdTSpNKLKojtletbr31Vs444wxeeeUVAKZMmcLtt9/O4MGDM45MUoN5/mlxdruaKfefytntah0tXLiQ4447jt///vcA7Lffftx3330ceOCBGUcmqcE8/7QBu13NVH26XU0pLU8pPdb1/zXAU8C+wCnAdV2TXUehIJk7z/15yzaFR4DLH9/AQu+AlCRJbWzixIkMHz58m2ERwerVq7tf7gosa3Zc7SalxObNm/nQhz60tfAIcPfdd/Od73wnw8gkSVIrWbJkCccddxzjxo3jkEMO4fLLLwdg5cqVTJkyhTFjxtAKPY+1kk9/+tNbC48Aixcv5jOf+UyGEUmS1D4GVTJxROwP/DXwS2CvlNJyKBQoI2JEH/NMB6YDXHnllUyfPr2qQPe5ak1F05920CC+ftzreGLF5u3GJeDXKzZzwK4DuODeV/jhc5sqWvay84ZVNL0kSVIeXHbZZZx44olceOGFAF8F/qa36eqVv7WaSrtKHT16NIsXL2b58uUsXbp0u/GPPPII06ZNY+bMmSOs6FkAACAASURBVJx//vkVLdurNyVJ6iyDBg3ikksu4YgjjmDNmjUceeSRTJkyhWuvvZbJkyczY8YMImIOMAMo1fNYx6imq/unn36aMWPG8Mgjj2w3rnvYs88+y5gxYypetjmcJEkFZRcfI2Jn4PvAx1JKq8s9uaeUZgIzu19WHGGVXlpXWNXhew4keqw4uoYDvLjOpECS1DnslqKzfeMb3+DSSy/ltNNOIyI+DlwNvKPndFnlb63mj3/8Ixs2bGCfffZh9OjRLFmyZJvxEyZMAGDZMm8wlSRJ/Rs5ciQjR44EYNiwYYwbN46lS5cya9Ys5s6d2z3ZdcBcLD7WZPny5YwZM4YJEybw9NNPbzPO/E2SpPoo2e0qQETsQKHweH1K6Qddg1/seh5k93MhX2pMiNX5+fLN/z97dx8eVXXu//9zhyiCYhU1NBJQUYxIFEQr9bSlKuWpcqoFa0vtERqPOW3le2r9lVNar/784vnqpbbHgrVHiT96StXS77FWsZ6aqtSo1VJLxWIsRVBSAkQsghrkKQ/r90cycRIymT0ze2Y/zPt1XV6SPXv2vmfP7L3vWfdaa7R9T4dGfahE35hwePcktCbp+gmH66SjS9TU0qHVbx46MhIAkJlsepsCKLzly5dr1qxZiT8flHR+gOFE3t69e/XQQw+ppKRE//mf/6kjjzyy+7Hp06fryiuvlCSmXwUAABlpbGzU2rVrNXHiRO3YsaO7KNk1A1nKmcfMbI2Zramtre1rFXT56U9/Kkm66aabdMopp3QvP+WUU3TTTTf1WAcAAGQn7chH62xRXiZpvXPujqSHHpU0V9KtXf9fmZcIs9TupP+9+oBqPzVI3zx3oC4ffZjW/b1d404YoJOO7qy5Llp9QB305QcAFInkIrGZMfqxCJ144ol65plndOGFF0rSxZI2BhtR9N14442aPn26Zs6cqaamJq1atUojRozo7jW/ePFivf766wFHCQAAomLPnj2aPXu2Fi9erKOPPtrz85i5wrv7779fNTU1Ov/887VhwwY99dRTkqRPfepTOuyww/THP/6RzmMAAOTI0jU8mtnHJT0n6RVJHV2Lv6PO3338b0kjJW2R9Dnn3K40+8s6+cn0Nx8TLjmlVDdOHKiKIR8M8mxq6dCi1Qf068bMfusxgd98BJAnmQwfzPuXyUxHM1LICr/e72nM3rNQnT9hMGfOHNXX12vnzp0aNmyYFi1apMrKSn39619XW1ub1q1b96Kkrznn/pRmU0VxvKTsR3GfddZZuvfee7sLjpL07rvv6o477tC///u/Z3Wuxez8BOKM+0/EJV/7ufYWHOdPL62trZo5c6amTZum66+/XpJUWVmp+vp6lZeXy8xOlFTvnKtMs6miOF5S9vnbhz70Id1999363Oc+p9LSzrEZbW1tevDBB/XVr35V7777blbb5ToCRAL3nxjgZ4UC5ekcSlt89FnBi4+SVGLSRz88QMMGm3bsdVr9ZntOIx4pPgLIk1AlPxQf44fiY7dYvfAceD1mRXO8cp1Cety4cRozZoz27Nmjp59+Wu+//37W24rZ+QnEGfefiKP4GCjOnyTOOc2dO1dDhw7V4sWLu5cvWLBAxx13nBYuXCgz+7akoc65f0u3ubwGGyK55m/Dhw/XBRdcIEn6/e9/r23btuW0Pa4jQCRw/4kBio+BoviYLxQfgxem35fjAgcfhSr5ofgYPxQfu8XqheeA4mMv5BcAssD9J+IoPgaK8yfJ7373O33iE5/QWWedpZKSztm7brnlFk2cOFFXXHGFtmzZok2bNv1WeZ55LGrClL9JXEeAiOD+EwMUHwNF8TFfKD4GL0zJJRc4+ChUyQ/Fx/hJ9Z7G5L0L1fkTERQfeyG/AJAF7j8RR/ExUJw/mSN/6yVM+ZvEdQSICO4/MUDxMVCezqHSfEcBAACCFbYv5AAAAAAAAADiqyToAAAA8WBmFLkAAAAAAAAAoMgx8hEAcAiKiAD8wLT5AFIJU67BVE0AAAAA4C9GPgIAUOTC1AAMAAAAAAAAINooPgIAAAAAAAAAAADwBcVHAAAAAAAAAAAAAL5IW3w0sx+b2Vtm1pC0bKiZPWlmG7v+f2x+wwQARAVTeALhtmTJElVVVcnMXjWz64KOBwAAAAAAAPHiZeTjTyRN77VsoaRVzrnRklZ1/Q0AAIAQa2ho0L333qsXX3xRksZJmmlmowMOCwCAgjMzOs0hdKqrq1VWVqaqqqruZbt27dKUKVM0evRoMQAAAABERdrio3PuWUm7ei2+VNLyrn8vl3SZz3EBAIAA0AgXb+vXr9dHP/pRDR48WM65NknPSPps0HEBAABAmjdvnurq6nosu/XWWzV58mRt3LhRYgAAAACIiNIsnzfMOdcsSc65ZjMr8zEmAADgE6/FRIqOxaGqqko33HCD3n77bR1//PGDJX1a0pqg4wIAAIA0adIkNTY29li2cuVK1dfXJ/5cLqle0rfyFcOJ97bka9NZ2X7NkKBDAAAAWfAy7WpOzKzGzNaY2Zra2tp87w4AAAApjBkzRt/61rc0ZcoUSaqT9GdJbb3XI38DABQjOmMhjHbs2KHy8nJJnQMAJPU5AID8DYiexBTgYfkPAPyU7cjHHWZW3jXqsVzSW6lWdM7VSkpkPS7L/QEAgAJJfOlwjtt2HF199dW6+uqrJWmSmd0iaWvvdcjfAAAAooX8DQAAhEm2Ix8flTS3699zJa30JxwAABAW9HyMp7fe6uwzZmYjJc2StCLQgAAAAJDSsGHD1NzcLElKNwAAAAAgLNIWH81shaTfS6o0s61mdrWkWyVNMbONkqZ0/Q0AAGKG6VfiZ/bs2TrzzDMl6VeSrnXO7Q44JAAAAKTwmc98RsuXL0/8yQAAAAAQCWmnXXXOzUnx0GSfYwEABIwiExB/zz33XOKf44KMAwAAAD3NmTNH9fX12rlzpyoqKrRo0SItXLhQV1xxhZYtWyZ1DgD4XMBhAgAApJXtbz4CAAAAAAAA8MmKFX3Phr9q1arEPxkIAAAAIoHiIySFa7STc/wuOhB1iWsK53N8mBnvJwAAiIUwff8FAAAA4ijtbz4CAAAAAAAAAAAAgBeMfAQQeWHquczIMAAAACCcvHxvYAYPAAAAIHeMfAQASApXERe5MzPeUwAAUNTIhQAAAIBgMPIRAAAAAADEEiMZAaB4hK3TCfceAMWM4iMAAPCExjsgXMLUuMJ1AQAAAAAAJFB8BIAiF6bGawAAAAAA4ips37/pQAYAyJecfvPRzKab2QYz22RmC/0KCgAAAPlRV1enyspKkb8B8ZL4rd8w/AeEEZ9NRF1dXZ1ogwMAAFGRdfHRzAZI+pGkGZLOlDTHzM70KzAAAJCdfDeu0XgXXe3t7br22mv1+OOPS+RvAAAAkZDI4UQbHAAAiIhcpl09X9Im59wbkmRmP5d0qaS/+BFY1IWpYZYpFILH5wFhVYgiFZ85IDxefPFFnXbaaRo1apSccwfJ34DUyN+A/oX5t6D9OH+T81hyWgQtkcO9/vrrtMGhoMKUD0nhvOcAAPqWy7SrwyU1Jf29tWsZACDEmBYNfmDKvWjatm2bRowYkbyI/A0AkLFivNcXMscpxuOL/pHDAQCAqMll5GNf2fAh3U/MrEZSjSQtXbpUNTU1We1s+zVDsnpeUKLWE4d48ytq8UYNxzfc6CleOGFqqOovFkYRBCfF8SZ/6xK1z2PU4o2aqB1f4kVQ4nw/D1NuheLmJYcjf4sG4s0v4gWA8Mil+LhVUnK3qwpJ23uv5JyrlVSb+DOH/QEAshTVhpPkuIs9KY/qe9iX5NeS7nUV+/vut4qKCjU1NfVYJPI3AECOwjAFaxhiAPLFSw5H/gYASCXbzmJ95VdRbKsLqrNcnDvpeZHLtKt/lDTazE4xs8MlfUHSo/6EBQDwQ1imwEw3RaeXqTvD8DoKjWlMi/N9z6ePfOQj2rhxozZv3izyNwCA3/rK3/y+lwedH+U61bzX5yRvP+jXjOAlcjja4AAgmvq6r/f1eO/HMr33p8vF+mtz629fXnKR/l6Tl+34letk8hr9/AmhTH+SKNVnIk55X9YjH51zbWY2X9JvJA2Q9GPn3Ku+RQYAyEkcblK9ZfKaevfK8runUTbHN10McXzPEC6lpaW66667NG3aNElaL/I3AAidTHqTZ5o7pNpe7+0kT5GeqVSNSZnkYqn223sbqfZVSH6+X1F5PVEc8RB1iRzukksuoQ0OAEIm21zAayGvr/VSjUTsb7te1vMz98t2X8l5Y6Y5R75eY6b7zWR/Xp7rnOvzWPSXM6c7dl7z7VxYgRNFslIASC2Tu1/a6ymFLMSFx1zF1/OnSHg9ZhwvAEitYPlbLgXBvrbjx7bQU6bHtnfDULb7ymR/fa2fyWfLa8zkb3lD/gYAuaP9DZ71l9NE7b33ksf5mcMVtPhoZnWSji/YDgEgWnY656Z7WZHrKXAIz+cPMsP1BgD6Rf4GZI/8LU+43gBAv8jfgNx4OocKPfIRAAAAAAAAAAAAQEyVBB0AAAAAAAAAAAAAgHig+AgAAAAAAAAAAADAFxQfAQAAAAAAAAAAAPiC4iMAAAAAAAAAAAAAX1B8BAAAAAAAAAAAAOALio8AAAAAAAAAAAAAfEHxEQAAAAAAAAAAAIAvKD4CAAAAAAAAAAAA8AXFRwAAAAAAAAAAAAC+KC3w/lyB9wcAUWIZrMv1tEDMPnhbnOOwhxjnT+a8HjOOFwCkxv0HyB7nT+bI3wAgd9x/gNx4OocY+QgAAAAAAACgqPTVuZQOpwAA+KMoio8793Vo0er9uuxXe7Vo9X69va8j6JAAACFnZj1GPQJAnBw8eFDf//73deGFF2ru3LlqaGgIOiQAAIpaU1OTLrroIo0ZM0Zjx47VkiVLJEnf/e53dfbZZ2v8+PEysyfM7MSAQ40F55zMTGvWrNHXvvY1fe1rX9OaNWtkZhQgAQDwgRX4hlrYnTmnDidN/uVevbb7g4Jj5bElemrWYJWYaFgGECZM+xAifd0f+BIaapw/mWPariL25S9/WT/5yU+6/x4yZIgaGho0cuTI4IICoon7D5A9zp8kzc3Nam5u1oQJE9TS0qJzzz1XjzzyiCoqKnT00UdLkszs65LOdM59Jc3mYn+8/PDkk09qxowZam9vlyQNGDBAv/71rzV16tSAIwOQZ9x/gNz4M+2qmf3YzN4ys4akZd8zs7+a2Toze9jMjskl0nwxMz23vb1H4VGSNuzu0AvN7RQeAQAZ4b6BKKmurlZZWZmqqqp6LP/hD3+oyspKmdmrZnZ7QOEhQLt379Z9993XY1lLS0uPYiQAACis8vJyTZgwQVJnp6AxY8Zo27Zt3YXHLkeKhnDf3Hrrrd2FR0lqb2/XbbfdFmBEAADER6mHdX4i6S5JP01a9qSkbzvn2szsNknflvQt/8P7wIn3tmS0/mdPLdWPLh6kfW1952R7WzuXf3XVPq18oy2jbW+/ZkhG6wMAABTavHnzNH/+fF111VXdy55++mmtXLlS69at08CBA8eaWVmAISJHmXaIOOmkk9TY2KgDBw70aGhLeP/99yVJd955p77+9a9ntG1GhgMA4J/GxkatXbtWEydOlCTdcMMN+ulPfypJV0q6qK/nmFmNpBpJWrp0qWpqagoUbbCy6SC6YcMGnX766dqxY8chj7355puSpI0bN+r000/PeNvkRAAAdEo78tE596ykXb2WPeGcS1TsVkuqyENsOXn5750NKhdWlOr4QT0TkRMGmSZVlPZYDwAAL/gtSETFpEmTNHTo0B7L7r77bi1cuFADBw6UJDnn3goiNgRj69at2r59uz784Q9rypQpPR4bMGCAvvjFL0qS/vCHPwQRHoCIIScC8mPPnj2aPXu2Fi9e3D3q8eabb1ZTU5MkPSBpfl/Pc87VOufOc86dVyyFx2ytW7dOkjRr1qxDHps9e3aPdQAAQHbSFh89qJb0uA/b8dXm95x+29SmQaWmFTMG6ZPDB+jow6VPDh+gFTMGaVCp6aktbfpbCz2SAAA90ZCGuHrttdf03HPPaeLEiTKzZ8zsI32tZ2Y1ZrbGzNbU1tYWOkzkSXt7u+6++25J0gMPPKCrrrpKQ4cO1TnnnKOHHnpI48aN05tvvqmHHnoo4EgBAChOra2tmj17tq688so+C2OSfiZpdoHDip177rlHkvSd73xH11xzjQYOHKiBAwfqmmuu0Xe+8x1J6s6ZAABAdszLdABmdrKkx5xzVb2W3yDpPEmzXIoN9Zr24dxse19lOu2qJA0bbPrFzME69UOH1lhff7dDs3+1V2/ty7z46GXa1TA1XDPlAxAZ/OB1SHi5hnNtDR3Onz40NjZq5syZamjo/OnuqqoqXXzxxVqyZIlKSkomSvq/kkalyuO6FM3xipps8s3DDjtMDz/8sC655JJDHnvvvfc0Y8YMvfDCCxlvl2siiljR3n+Sr0FcA5Cloj1/+uKc09y5czV06FAtXry4e/nGjRs1evRoSZKZ/aukTzrnLk+3ufxFGi7Ztr8tXry4e5r5AwcOSFL37CDZTEGfwPUQiISiv/+YGdcr5MLTOeTlNx/73rrZXEkzJU3ur8HKOVcrKdFlvqCf6B17nWY+8r7mjT1cl48+TMMGmXbsc3rwtVYt/8tBvXuwkNEAAOKERA1RVFFRoVmzZiU+vy+aWYek4yX9PejYwqAYOo+1trbq0ksv1bx581RTU6MzzjhDe/bs0UMPPaTFixfrjTfeyMt+AQBA/55//nndd999OuusszR+/HhJ0i233KJly5Zpw4YNKikpkaSpkr4SZJxxcd111+lPf/qTrrvuOk2YMEGS9NJLL2nx4sW67777Ao4OAIDoy2rko5lNl3SHOntbZdJYlXUrSjYjH/MljiMfoxYvEFNF3/MqLLxeE7lehQrnTx96j3y85557tH37dt10000ys0pJqySNZORjp6jlQ1GLF4ipor3/MPIRPija8ycHXo9Z0RwvP/Kho446SlLn723miushEAlFf/+hQz1y5M/IRzNbIelCSceb2VZJN0r6tqSBkp7susmvds7R8woAACAk5syZo/r6eu3cuVMVFRVatGiRqqurVV1draqqKkn6uaS5aQqPAAAAKCJh6vwveRsAkCs/io4AAKAnTyMffcTIxwDQcx6IjKLveRW0TK+FXK9ChfMnc/Sc7yVq+VDU4gViqmjvP4x8hA+K9vzJQd7ztzC1v0np2+DClA9JXA+BiCj6+w8jH5Gj/P7mIwDvwpQMc2MBAAAAgPijYREAAABBKQk6AAAAosrMQtW5AAAAAAAAAACCRvERAIAiRwEVAAAAAAAAgF8oPgIAAAAAAMQQM3UAAAAgCBQfAQAQo/8AAAAAAAAAr2hLQ38oPgIAAHrFAwAAAAAAAPBFadABAAAAIJ5OvLcl6BC6bb9mSNAhAAAA9Ku6ulqPPfaYysrK1NDQIElasGCBfvWrX+nwww/XK6+88rCkLzvn3gk2UgAAgP7lNPLRzI4xs1+Y2V/NbL2ZXeBXYAAAREVURwwy2rE4/eAHP9DYsWNlZg1mtsLMjgg6JgAAAEjz5s1TXV1dj2VTpkxRQ0OD1q1bJ0mvSfp2ELEBANAX2paQSq7Tri6RVOecO0PSOEnrcw8JAAAA+bBt2zbdeeedWrNmjZxzVZIGSPpC0HEBAABAmjRpkoYOHdpj2dSpU1Va2j1x2WpJFYWOCwAAIFNZFx/N7GhJkyQtkyTn3EGmfQAARE0+emglthnFnl9RjBmZaWtr0759+2RmpZIGS9oedEwAAADwpFrS4309YGY1ZrbGzNbU1tYWOCwAAICechn5OErS3yX9l5mtNbP/z8yO7L0SyQ8AAEA4DB8+XN/85jc1cuRISWqW9K5z7one65G/AQAAhMvNN98sSW2SHujrcedcrXPuPOfceTU1NQWNDQAAoLfS9Kv0+9wJkv6Xc+4PZrZE0kJJ301eyTlXKynRauVy2B8AAKHXe+Rg4m/nonMLjGLM8Gb37t1auXKlNm/erLKyshMlPWhmX3LO3Z+8HvkbAABAeCxfvlyPPfaYJF3pSNIBAEAE5DLycaukrc65P3T9/Qt1FiMBAAiVVFOJ+jnFaFSnWUVxeeqpp3TKKafohBNOkHOuVdIvJf1D0HEBAACgb3V1dbrtttv06KOPyjm3N+h4AAAAvMi6+Oice1NSk5lVdi2aLOkvvkQFAIDPgi4OUphEGIwcOVKrV6/W3r17ZZ0fysmS1gcdFwAAAKQ5c+boggsu0IYNG1RRUaFly5Zp/vz5amlp0ZQpU2RmL5vZPUHHCQAAkE4u065K0v+S9ICZHS7pDUlfzj0kAACQbxRDi9PEiRN1+eWXa8KECZL0iqS1+mB6VQAAAARoxYoVhyy7+uqrk/8cX7BgAAAAcpBT8dE597Kk83yKBQAAX/VVYDOzwH7LkN9SRBgsWrRIixYtkqSqoGMBAAAAAABA/OTym48AAIRWfyP7gp6CNSo4RgAAoC/kCAAAAAD6k+u0qwAAhAqNYQAAAOEV5CwUxYJ8GAAAAEGj+IhICtOXKb44AwAApBe1/C1q8QIAAAAAEBYUHwEAKCJhakwHAADx4OV3rZNzEH4HGwAAAIg3fvMRABAbUSmsRSVOAAAAxAP5JwAAAAqJ4iMAAAAAAPCFmWVd6MrluWETp9cCAACA8IhKnknxEQCAIhGFxAQAAMRPpjkIOQsAAAAQbTkVH81supltMLNNZrbQr6AAAACQH3V1daqsrBT5GwAgn7wUEIuhyJjP1xhEr/dieM/CqLq6WmVlZTKzhqBjAQAA8CLr4qOZDZD0I0kzJJ0paY6ZnelXYADgVeJLdxj+QzA4/vnDsY2X9vZ2XXvttXr88ccl8jcAQJ6RR+RPkMeV97Xw5s2bp7q6uqDDAAAA8Kw0h+eeL2mTc+4NSTKzn0u6VNJf/AgMAACES7pGJudcgSJBtl588UWddtppGjVqlJxzB8nfACD6EvfnMNyHcylI9X5upq8rTMehP8mvM1WsZtbjsXSvLXn9XIuCUTmOxWbSpElqbGwMOgwUobB1NEh3bYpavFHD8QWQiVyKj8MlNSX9vVXSxN4rmVmNpBpJWrp0qWpqarLa2fZrhmT1vKBE7eJHvPkVtXijhuMLhEPvhjKEz7Zt2zRixIjkReRvSaL2+SXe/IpavICXopaf+yjUNgvxuvyW6jWlKrB63VbQox0Rbn7lb8ivsJ1LUbmuxhWfBwSJDkfIt1yKj31dHQ/5pDrnaiXVpnocAIBchC1Z9ypqSZ7X4xy111VsUrwv5G8AUET6uqc75yJxD+89wi/daD+vHaP8KnB6LTrmsi2/1veyvUxHU0bhMxRn5G8AAMRflNpBcyk+bpWU3HW+QtL23MIBAPgp1Q2pr0abqInSzbY/qabVSpbrdGNBHKt0nz0Eo6KiQk1NTT0WifwNAEKj9/0z2+nlUm3Hz+JYUNKNBPQyurC/45ptITJKxzCVTEdm+jndK8KDmSvyi3jzi3jzK2rxxoWfs0B4yQWzbTNMtX6U2x6l8MSfLu8NQ4y9leTw3D9KGm1mp5jZ4ZK+IOlRf8ICAOSqvwYAM+t+PPnfCEbiPegvCexrnUyXh0HY4ik2H/nIR7Rx40Zt3rxZ5G8AEH7p7vOZbsfrunHnNVcKYy6VD7m8zr6eWwzHDACA3nrnF5nmbKnyPr/i6G/9vp7fX2xec6hs89Vsc5N8tYX5tc1sj4mX5+RyzPOVu2U98tE512Zm8yX9RtIAST92zr3qW2QAgIKK0ki1Ym/QiPrrzzT+MH4Go6q0tFR33XWXpk2bJknrRf4GAJFRyPt/qt7v8DZrRVglYiW3ip45c+aovr5ekirNbKukG51zy4KNCgAg+ZMLeO3Ik2qkm5d7fLbFvHxtJ1mq3Kr38zLJS/0audl7nUx4fV19PSfXz1Wmz89HnmiFTDrNrE7S8QXbIQBEy07n3HQvK3I9BQ7h+fxBZrjeAEC/yN+A7JG/5QnXGwDoF/kbkBtP51BBi48AAAAAAAAAAAAA4iuX33wEAAAAAAAAAAAAgG4UHwEAAAAAAAAAAAD4guIjAAAAAAAAAAAAAF9QfAQAAAAAAAAAAADgC4qPAAAAAAAAAAAAAHxB8REAAAAAAAAAAACALyg+AgAAAAAAAAAAAPAFxUcAAAAAAAAAAAAAvqD4CAAAAAAAAAAAAMAXFB8BAAAAAAAAAAAA+KK0wPtzBd7fBzt2TmbW/X8ACKFMLk6BXU/xAb/uJ87xdvqA8ydzXo9Z4PkbAIQY9x8ge5w/mQt9/gYAEcD9B8iNp3Mo9iMfEw26B9ulDbvbdbCj53IAAIC4aWpq0kUXXaQxY8Zo7NixWrJkiSRpwYIFOuOMM3T22WfLzB42s2MCDrVfZqbX3+nQ7v3kbQAAAEAUbdmyRdu2bQs6DABAgVmBi3CBtBw9sqlVN7ywX7sPSMcOlG752BG69NTDgggFAPpDz6uIYeRjqHD+JGlublZzc7MmTJiglpYWnXvuuXrkkUe0detWXXzxxSotLZWZ3S5JzrlvpdlcIMdr/a52/cuq/dr0TocOL5G+PPYw3fjRIxgNCSBsuP8A2eP8yRwjHxEZb7/9tj7/+c9r1apVkqTPfOYzeuCBB3TUUUcFHBnA/QfIkT8jH83sx2b2lpk1JC0bb2arzexlM1tjZufnEmk+vbW3Q9c901l4lKTdB6Trntmvnfs6gg0MAAAgT8rLyzVhwgRJ0pAhQzRmzBht27ZNU6dOVWlp96z7qyVVBBVjKoli/HX1nYVHSTrYIS19pVWPN7ZSeAQAALGUauaKXbt2acqUKRo9erTM7EkzOzbgUAFPbrjhhu7CoyQ9+uijuuWWWwKMCABQSF6mXf2JpOm9lt0uaZFzbryk/7fr71B6YfsHU60mHGiXXmhuDyYgAACAAmpsbNTatWs1ceLE3g9VS3o8DRYoQQAAIABJREFUgJD6ZWbaua9Dr7x9aEexp5vI3wAAQDyVlpbqP/7jP7R+/XqtXr1aP/rRj/SXv/xFt956qyZPnqyNGzdK0ipJCwMOFfCkrq7ukGWPPx66rx8AgDwpTbeCc+5ZMzu592JJR3f9+0OStvsb1qFOvLclo/Wnjhygn0wbrJOO7ru+evKQzuVX/WavntqSWUPW9muGZLQ+AABAEPbs2aPZs2dr8eLFOvroo7uX33zzzZLUJumBvp5nZjWSaiRp6dKlqqmpyWr/meZvRx4mbZh7lD400HTsQHXPXJFwcldeV/vKQf3v1Qf62EJq5G8AEE1mxhT1KArl5eUqLy+X1HPmipUrV6q+vj6x2nJJ9ZLSTZsP+CKbWUcaGho0duxYnXrqqfrb3/7W47FTTz1VkvT888/r4x//eMbb5n4AANHhZeRjX66T9D0za5L0fUnfTrWimdV0Tc26pra2NsvdZe657e1694DTOWUDNOPknjXWT59cqrNPGKDd+52e30YPegAAED+tra2aPXu2rrzySs2aNat7+fLly/XYY49J0pUuxbd351ytc+4859x52RYes/F+q1S/tV2HlZi+ee7AHo+NHGKaU9n5m92PvdFasJgAAAAKLXnmih07dnQXJZ1zzZLK+npOUO1vQG+/+MUvJEk33nijBg78IKc/6qijdMMNN0iSHnzwwUBiAwAUjnnpMdI18vEx51xV1993SnrGOfeQmV0hqcY59ykP+8u6e0qmPeclaf64w/Wd8weqwznVNbbp5b93aPwJJZp+cqlKzPR//nBA/7nuYMbbpec8gDzhB68jxu/fnqMXZ044f5I45zR37lwNHTpUixcv7l5eV1en66+/Xs8884xOOOEEr8esoPnbR4YN0C9mDtJhJaa1b7XrN39rU/mRptmnHaajDjfVb23TFx/fl/F2yd8A5An3nzxj5GOscf70Yc+ePfrkJz+pG264QbNmzdIxxxyjd955J/Gwmdlu51y6330smuOF/MrmO29ZWZnWrVunYcOGafPmzXrggQdUWlqqL33pS6qoqFBjY6PGjRun9957L+Ntcz+AT7j/ALnxdA5lW3x8V9IxzjlnnXehd51zR/eziYSCNl5J0oJzD9f88YfrsJIPjkdrh9NdLx/U9/6UeeFRovEKQN6Q/EQMxcdQ4fxJ8rvf/U6f+MQndNZZZ6mkpHOii1tuuUX/+q//qgMHDui4447Tn//85z9LWu2c+0qazRU8f5t2UqnumHSEjj2i59v6xN/aNP/pfdqTxcBH8jcAecL9J88oPsYa508vra2tmjlzpqZNm6brr79eklRZWan6+nqVl5fLzE6UVO+cq0yzqaI4Xsi/bL/zjh07Vg8//LBGjx7dY/krr7yiyy67TG+88UZW2+V+AJ9w/wFyk9fi43pJX3XO1ZvZZEm3O+fO9bC/gjdeSdKwwaZLTy3VsEEl2rGvQ49satNb+7K/btB4BSBPSH4ihuJjqHD+ZC6UIx8TBg2QZo4qVeWxA7S3zek3jW16dVdH1tsjfwOQJ9x/8oziY6xx/iRJNXPFggULdNxxx2nhwoUys29LGuqc+7d0m8trsCgauXznLSkp0YwZM/Sxj31Mzjk9/fTTWrVqVU7XdO4H8An3HyA3/hQfzWyFpAslHS9ph6QbJW2QtERSqaT9kr7mnPuTh/0F0njlNxqvAOQJyU+E+F14lIL9IpV4PRH+Msf5k7lQFx/9Rv4GIE+4/+QZxcdY4/xJkmrmiokTJ+qKK67Qli1btGnTpt9K+pxzbleazcX+eKEw8vG9NxfcD+AT7j9AbjydQ6XpVnDOzUnxkJeRjgAAIEJo4AMAAAAK7+Mf/3jKPHzVqlWJf04uWEAAAAA5KAk6AAAA8iFsPTTDjGMFAAAAAAAAwC9pRz4CAAAAAAAUEp2jAAAAgOii+AgAiBUaqgAAAAAAAAAgOEy7CgCILTOLRTGy0K8hLscNAAAAAAAAQOEx8hEAgAwEVZRL7Nc5F8j+AQAAAAAAAMALio+QFK5pCmlYB4BOfV2bzYzrJAAAgAfkTQAAAEAwmHYVABAbYepIAQSturpaZWVlqqqq6l724IMPauzYsSopKZGZnRdgeAAAIAtMjw8AAIAoyGnko5l9XdI1kkzSvc65xb5EBQCAj+j1jmI0b948zZ8/X1dddVX3sqqqKv3yl7/Uv/zLv+iZZ54JMDoAAAD0Vl1drccee0xlZWVqaGiQJO3atUuf//zn1djYqE2bNj0p6Qrn3O58xXDivS352nRWtl8zJOgQAAAe0PaG3rIe+WhmVeosPJ4vaZykmWY22q/AAABAftFzPt4mTZqkoUOH9lg2ZswYVVZWBhQRAAAA+jNv3jzV1dX1WHbrrbdq8uTJ2rhxoyStkrQwiNgAAAAykcu0q2MkrXbO7XXOtUl6RtJn/QkLAADv4lhE6+/1FOr1xu2YAgAAAGHWV+exlStXau7cuYk/l0u6rNBxAQAAZCqX4mODpElmdpyZDZb0aUkj/AkLAAAAQTGzGjNbY2Zramtrgw4HAACgaO3YsUPl5eWSJOdcs6SyYCMCAABIL+vio3NuvaTbJD0pqU7SnyW19V6vWBuvEqNSwvAfAABAJpxztc6585xz59XU1AQdTsEEnbORvwEAgGwVa/sbAAAIp9JcnuycWyZpmSSZ2S2StvaxTq2kRNbDL44CAJACP84NAAAAINmwYcPU3Nys8vJymVm5pLf6Wo/2NwBAsaNdLVxymXZVZlbW9f+RkmZJWuFHUAAA+M2P0USF+p1FRj3BD3PmzNEFF1ygDRs2qKKiQsuWLdPDDz+siooK/f73v5ek/zGz3wQdJwAAyBz5YvH4zGc+o+XLlyf+nCtpZYDhAAAAeJLTyEdJD5nZcZJaJV3rnNvtQ0xA7ITpiyG9PxA3YTq/ooTjFn8rVvTdJ+yzn/1s4p/DChYMEEFhuk6SvwFAcZgzZ47q6+u1c+dOVVRUaNGiRVq4cKGuuOIKLVu2TJKmSPpcwGECAIAUGH35gVynXf2EX4EAAIDM9E5oEg3lJDkAAKCYhanzAJCJVJ3HVq1alfjn5IIFA0RQ2K7/fDcHUMxyHfkIAEDspCrqhVFfsdHLCgAAxAkdrAAAAIBoofgIACgq/RXmkgt5YS44ekEjHQAAAIC4C9v3Nr5/AQDQqSToAAAAyIaZhe6LZhjleow4zgAAIMrIZQAAAIDCY+QjAAAKX4/ZfCmW1wkAAAAAAAAgGIx8BAAg5ujxDwAAAAAAgLijDSw8GPkIAMib3jd7v37/giQCAMJ1LfRyfY9avACCE6brRZhwXAAAABAVFB8BAL5L1TBiZqFo8E3E55wrqkacYnqtAAAAABA2YftOFobv5wAQB2G7vqeS3CaabxQfEUlhOplJ1IDoCtO1BAAAoNj5kZuR3wEAAADBy+k3H81supltMLNNZrbQr6AAAP5IzHNeyEaYfO0riNcCxFFdXZ0qKytF/gYgSMn39aD/QzzwfiLu6urqRBscACBsyMGQStYjH81sgKQfSZoiaaukP5rZo865v/gVHAB4EaYbXJhGwvY+LoUcVt+fTOMI0/sLRF17e7uuvfZaPfnkkzr11FPPFPkbAABA6CVyOEkzRBscgICErX0m6PYtvwXZbheWNsNshOUnnnCoXEY+ni9pk3PuDefcQUk/l3SpP2EBAIoZoxHCh/ciHl588UWddtppGjVqlMjfAACIhlR5Mfly8UjkcLTBAQCyQb6A3grxmcjlNx+HS2pK+nurpIm9VzKzGkk1krR06VLV1NRktbPt1wzJ6nlBiVq1nXjzK2rxRg3HNzP56hGU6U2LnklA4W3btk0jRoxIXkT+liRq1yTiza+oxRs1HN/i4iXvy6UBxMtzs8k9wzAKgMZCSN5yOPK3aCDe/IpavFHD8c2fXHKZ/nKcuOURYcjNoib581HoWfJyKT729ck9JErnXK2k2lSPAwCiz49kJm4JEfxBkdpfKY4l+RsAIGPZTqWfy309OV/MZju5Pj/dtv3cJrkxknnJ4cjfACD8vOYLyXlTqpwgebnX4mVC8nZzyamCaq/xK+/KpqAbRek+H/nKkXMpPm6VlNztqkLS9tzCAQAUgl9JQq6NIjSqAIVVUVGhpqamHotE/gYAyBMvjWVhlU2MXhpu8tG4U8gGsrg1xkUFORwAhFem9/98z/bg5zbz2XEr0/2kGrXndfvJo//ynctk8nrinFdZti/OzEolvSZpsqRtkv4o6YvOuVf7eVp8jyQA5C6TDCLt9dTrTTiH+0BWz0P0FSJJyyIp9PX8iau2tjadfvrpWrVqlUaNGjVQ5G8AkCvf87d8jMjrEUQG05/219u+L723HYV8MV/TlKXqRV+Ifafbfu99ZZJ3eYktg88w+ZtHiRxu8+bNo+StDa6ojxcApJG39rdimf40IVWu2Ht5X9N+9pdnJh/HfB+7VFOS9hVD7xGjmeRFXl57NiNS/T4+frbBZT3y0TnXZmbzJf1G0gBJP07TcAUACKFMbmpxTZaQmUJ8Dvis5UdpaanuuusuTZs2TZLWi/wNAEIn03uglyJXttvP5Te9o3Ivz/fvoaebLi2fx8vrNv0ajZG8jTj34g9CIoe75JJLaIMDgAiIYk6UDa+zXPS1nl/5aq4ymakjm5wpk9fe3/YLWZD1S9YjH7PamVmdpOMLtkMAiJadzrnpXlbkegocwvP5g8xwvQGAfpG/Adkjf8sTrjcA0C/yNyA3ns6hghYfAQAAAAAAAAAAAMRXSdABAAAAAAAAAAAAAIgHio8AAAAAAAAAAAAAfEHxEQAAAAAAAAAAAIAvKD4CAAAAAAAAAAAA8AXFRwAAAAAAAAAAAAC+oPgIAAAAAAAAAAAAwBcUHwEAAAAAAAAAAAD4guIjAAAAAAAAAAAAAF9QfAQAAAAAAAAAAADgi9IC788VeH8AECWWwbpcT4GeOH8y5/WYcbwAIDXuP0D2OH8yR/4GALnj/lMkzDrfaud4G33m6RwqipGPvT9cfNgAAADgJ/JLAAAAAACATmmLj2Y2wsyeNrP1ZvaqmX29a/n3zOyvZrbOzB42s2PyH2522p300MZWffeF/XpoY6vaaRsCAAAx1tTUpIsuukhjxozR2LFjtWTJkh6Pf//735eZOTM7PqAQY8U5JzPT008/reuvv1533HGHdu/eHXRYAAAABZPoiLX53Q59b80B3bbmgF5/p6PHYwAAoHhYugTAzMollTvnXjKzIZL+JOkySRWSfuucazOz2yTJOfetNPsraLaRaAj6p7q9WtXU3r38UyMH6KfTBnc/DgAhwbQPQPY4f5I0NzerublZEyZMUEtLi84991w98sgjOvPMM9XU1KR//ud/1hNPPLFF0rnOuZ1pNhf74+WH2267TQsXLuz++6STTtJLL72koUOHBhgVgALg/gNkj/Mnc6GednXNjnZd8T97tb+rCe7wEulnMwbpH04s9K8+AUC/uP8UCaZdzRt/pl11zjU7517q+neLpPWShjvnnnDOtXWttlqdxchQMTOt2dHeo/AoSU9tadfat9opPAIAgFgqLy/XhAkTJElDhgzRmDFjtG3bNknSN77xDd1+++0SX6J8s3//ft1yyy09lv3tb3/TsmXLAooIAOLFzPj+jqJQXV2tsrIyVVVV9Vj+wx/+UJWVleqakez2gMJL64cvH+guPErSwQ5pydqDwQUEAAACk1HXIzM7WdI5kv7Q66FqSf83xXNqJNVI0tKlS1VTU5NxkJJ04r0tGa0/+7RS/fCiQdq2p6PPx5v2dOicsgG69rf79PDrbX2uk8r2a4ZktD4AIP7oTYWwamxs1Nq1azVx4kQ9+uijGj58uMaNG9fvc/zK36Im04btESNGaMuWLdq9e7fee++9Qx5vbGyU1HkMv/KVr2S0ba4lAAAUn3nz5mn+/Pm66qqrupc9/fTTWrlypdatW6eBAweONbOyfMaQafubJD1z+WCNPnaAtu45NH9p6mqXa2rp0MSfv5/xtmmDAwAgmjwXH83sKEkPSbrOOfde0vIbJLVJeqCv5znnaiXVJv7MPtTMbOqaV/7jJw7QEQPUo+fVoAHSx7umfEisBwAAEDd79uzR7NmztXjxYpWWlurmm2/WE088kfZ5QeVvUfPmm2/qnXfeUXl5uc455xytXbu2x+MzZ86UJP31r38NIjwAABAxkyZN6u68lHD33Xdr4cKFGjhwoCTJOfdWAKH1a9O7HRp97ABNHlGq9bt6jnT81Eja3wAAKEZpp12VJDM7TJ2Fxwecc79MWj5X0kxJV7qQdc/+884OvbKzXccNKtFdFx2hskGdPdmHDTbddfERGnqE6c9/b9crb5P8AAD8w5RgCIvW1lbNnj1bV155pWbNmqXXX39dmzdv1rhx43TyySdLnVPmv2RmHw420uhqbW3Vf/3Xf0mS7r///u4RpYMHD9aNN96oGTNmaN++fbrvvvuCDBMAAETYa6+9pueee04TJ06UmT1jZh/paz0zqzGzNWa2pra2tq9V8ua+9a2SpOvOOVz/OKpUJdb5Y1DTTyrVgnM7i6b3/7W1oDEBAIBgpR35aJ2tqMskrXfO3ZG0fLqkb0n6pHNub/5CzN43n92vX8wcrE+fcpimnlSq7e87nXikqbTE9N5BpwXP7c/bvsPU+ByyujAAxE6YrvmA1Hnvv/rqqzVmzBhdf/31kqSzzjpLb731QUd5M9sq6Tzn3M6AwoyFm266SVOmTFFVVZVefvllbdmyRccdd5yOPPJISdK1116rt99+O+AoAQBAVLW1tWn37t1avXq1SkpKFkj6bzMb1XsQQJAzV9RvbdfPN7TqC5WHaenkQdq1v0MdTjp+UOeYh0c2taquMbOfPAIAANHmZeTjxyT9k6SLzezlrv8+LekuSUMkPdm17J58BpqNV97u0D8+ulePvdHZu2rkkM6X+6s3WvWPK/eqgVGPAAAghp5//nndd999+u1vf6vx48dr/Pjx+vWvfx10WLH0zjvvaNKkSbrzzjv17rvvauTIkTryyCP1wgsv6NOf/nT3yEgAAIBsVFRUaNasWTIzOedelNQh6fig4+rt/3l2v777wn41vtehoUeU6PhBJdra0qFFq/drfv1+5vEHgAigcz38ZAUeFZf1zrL5wetkQw6TjhtkenufU0uOMz14+bHrMJ2ojHwEIiOTCwcndoj0vuZz3Q0E50/mvB6zojleueZvAwcO1PDhw7Vnz54eo0yzwXUEiAzuPwWQuD5zbYwdzp8+NDY2aubMmWpoaJAk3XPPPdq+fbtuuukmmVmlpFWSRqb5+aPA2t9M0olHds672vy+U0eO75yXNjgAyFBk7j9dHU9iu798I4fMG0/nUNppV+OipVVqaeVDBgDFohAJU6pCBckN0CnXxis/FaLh6sCBA3rjjTfyvh8AABBPc+bMUX19vXbu3KmKigotWrRI1dXVqq6uVlVVlST9XNLcNIXHQDlJ294PbXgAAKBAiqb4CAAAAAAAAITVihUr+lx+//33J/45oWDBAAAA5IDiIwAgthiBCAAAAADxFaafPZL47gkAQALFRwBArITtyycAAAAAAAAAFJOSoAMAAAAAAAAAAAAAEA8UHwEAAAAAAAAAAICAxG02N4qPAIDYy8fN28s245Y0AAAAAAAAAEA6FB8BAAAAAEAo0ZkLAAAAiJ60xUcz+7GZvWVmDUnLPmdmr5pZh5mdl98QAQDwhsYp4APV1dUqKytTVVVV97IFCxbojDPO0Nlnny0ze9jMjgkwRAAAAAAAAMSQl5GPP5E0vdeyBkmzJD3rd0AAAOSDmVGcRFGZN2+e6urqeiybMmWKGhoatG7dOkl6TdK3g4gNAIBCIPdDnPzgBz9Q10CABjNbYWZHBB0TAABAKmmLj865ZyXt6rVsvXNuQ96iAgAAQE4mTZqkoUOH9lg2depUlZaWJv5cLami0HEBAJAtiokoVtu2bdOdd94pSec556okDZD0hWCjQjFIdOINy38AgOjI+28+mlmNma0xszW1tbX53h0AAAC8qZb0eF8PkL8BAACES1tbmyQNMrNSSYMlbQ82IgAAgNTyXnx0ztU6585zzp1XU1OT790BAJBX9LhEHNx8882S1Cbpgb4eJ38DAIQd+RiKyfDhw/XNb35TkrZIapb0rnPuieR16DwGAADCpDT9KgAAhBuNT4B3y5cv12OPPSZJVzrnXNDxAACQreQckFsa4mz37t1auXKlJJ0i6R1JD5rZl5xz9yfWcc7VSkpUHTkhAACe0a6GfMj7yEcAAIoZIyURJnV1dbrtttv06KOPyjm3N+h4AADwCzkX4uypp57SKaecIufc351zrZJ+Kekfgo4LAIAoIEcMRtrio5mtkPR7SZVmttXMrjazz5rZVkkXSPofM/tNvgMFAMAPJBwoFnPmzNEFF1ygDRs2qKKiQsuWLdP8+fPV0tKiKVOmyMxeNrN7go4TAAC/UYhE3IwcOVKrV6+WmQ22zg/3ZEnrg44LAAAglbTTrjrn5qR46GGfYwEAIJRovEIUrVix4pBlV199dfKf4wsWDAAAALI2ceJEXX755brpppteUufvdq/VB1OsIkLC9t2SKasBIHhhuzf4hWlXAQBFJ1Vv+MTy3v8BAAAgmsjnEBeLFi2Sc+4M51yVc+6fnHMHgo4JAAAglbQjHwEACDO/GpPy3ShlZvQqBQAAyBGFRAAAEHfkO4gDio8AgKJGQgcAABAf2eR2iefQUQwAAIQNeQqiiuIjAAAAEEFh6jzBF2EAAAAg3sL0/UPiOwiQq3wXtvnNRwAAAAAAEGupGkzD1pAKAACA3JHjBY+Rj5AUrpORXisACiVM1z4AAAAAAACgL2ZGuzkiheIjAAAAAAAoGjTeAQAAFJe+BiCQD+YXxUcAQCRFcdQiPxIOAAAQnCjmjwAAAOnQ3pSdMHRIK0R+mvz5KORnJafffDSz6Wa2wcw2mdlCv4IC0jGz0PwHAECU1NXVqbKyUuRvAICw4TsWkFpdXZ1ogwOA4paPNmlyr/jx8p4W4n3PuvhoZgMk/UjSDElnSppjZmf6FRgAAHFFYoegtLe369prr9Xjjz8ukb8BAEKMQiTwgUQOJ9rggH4FPUCBAQvFLfk9L9Rotkz3w+czc16PlV/HNZPt9F4v3XML/b7nMvLxfEmbnHNvOOcOSvq5pEv9CQuIl6CTnbjfWII+pnE/vsgPPjMIwosvvqjTTjtNo0aNEvkbACBMCp0X9ZWLkZshrBI5HG1wABA+QRSd/NgO7VKdMj0OyesH1SacSQyFLp4my+U3H4dLakr6e6ukib1XMrMaSTWStHTpUtXU1GS1s+3XDMnqeUEJeq7gTBFvfkUt3qjh+EZL8o0s+b3r6waX6r2NU3JkFvz88ige27Zt04gRI5IXkb8lidq5GLV4AcCrsOR6qfJWoNC85HDkb9FAvPkVtXijhuObXl+jH8Nw3DJpe/Ir7qBev195ZLbb6S9/zLTI2ZdM30e/181ELsXHviI65JU752ol1aZ6HAAQPpnedNLd+LLtfZOcHIWlESofMim8ArlI8bkifwMABCZsOV66vDXKOVq610CxNby85HDkbwAQTr3btvq6pueSD/WXu6Tr9A9vwnTswhRLOrkUH7dKSu52VSFpe27hAACCkmui43fy5Oc2osjL6/azUShVY5SXRqg4NMYVi4qKCjU1NfVYJPI3AIgsikXIhpfRD8zOES7kcAAQbVFs2+pvQECqNsBU6/eVs+aSa+SriJvJAAiv6xRzPmU5vMGlkl6TNFnSNkl/lPRF59yr/TyteI80AKSXyR3Sl+tpFJMfpJYuKcxkylu/Yimggp8/UdTW1qbTTz9dq1at0qhRowaK/A0AcpW3+4/XAlGPHWRw//WrMaR3fuEl34jLzBb5yncybeDLZpteFKDjGfmbR4kcbvPmzaPkrQ2uqI8XAKTh6/0nl9nD8p0LRSXn8jI9abq8MtXzkBmP+Z2nA12SQxBtkuZL+o2k9ZL+O03DFQCggHr/4HDy3/yodDylel/7Wp7vz0Ffnzcv0+zyOc2v0tJS3XXXXZo2bZpE/gYAoZfJvT15eX/30Uyem+l9uNju215fr5dj3t966babr9ypr23292/yt/xJ5HCiDQ4AIo975aG85BHZ5LcIVtYjH7NEzysASC3QnldAWPnZ6yqxySxDiRuvx4zjBQCpFSR/C1uP7t4jHNFTvt6vTI97kJ8b8re8IX8DgNzR/gak4GcOV9Dio5nVSTq+YDsEgGjZ6Zyb7mVFrqfAITyfP8gM1xsA6Bf5G5A98rc84XoDAP0ifwNy4+kcKvTIRwAAAAAAAAAAAAAxlfVvPgIAAAAAAAAAAABAMoqPAAAAAAAAAAAAAHxB8REAAAAAAAAAAACALyg+AgAAAAAAAAAAAPAFxUcAAAAAAAAAAAAAvqD4CAAAAAAAAAAAAMAXFB8BAAAAAAAAAAAA+ILiIwAAAAAAAAAAAABfUHwEAAAAAAAAAAAA4IvSAu/PFXh/ABAllsG6XE89MOv/kDrHYYwRzp/MeT1mHC8ASI37T8wk8kfnXI9cshjyxt65cwFeM+dP5sjfACB33H+A3Hg6hxj5CAAoWumKkwCQi2JoqAYAAAAAAOgtbfHRzEaY2dNmtt7MXjWzr3ctH2pmT5rZxq7/H5v/cLPz8t/bddmje3XSshZd9qu9+vPf24MOCQAAIK+qq6tVVlamqqqq7mUvv/yyPvrRj2r8+PEyszVmdn6AIcbKjh07dMUVV2jQoEE6/fTT9bOf/UxmRgESAEKOzmgIi6amJl100UUaM2aMxo4dqyVLlkiSFixYoDPOOENnn322zOxhMzsm4FBTau9wun3NAVXdt0djf9qim188oLYOciEAAIqRpWsQMbNySeXOuZfMbIikP0m6TNI8Sbucc7ea2UJJxzrnvpVmfwXPON5vdZq44n3tOvA6CacIAAAgAElEQVTBroceYfrjnCM1qJQvGQBChWkffOalMYnCQGxw/vTy7LPP6qijjtJVV12lhoYGSdLUqVP1jW98QzNmzJCZXSLp35xzF6bZVFEcr1xNnTpVTz75ZPffZqbVq1fr/POp7wIxx/0n4np3FGHa1Q8w7WphNTc3q7m5WRMmTFBLS4vOPfdcPfLII9q6dasuvvhilZaWysxul6Qwtr9J0j3rDuqmPxzosezfzjtc150zMIhwACAV7j9AbvyZdtU51+yce6nr3y2S1ksaLulSScu7VluuzoJk6NRvbetReJSkXfudntnaFlBEAIBCoBc7it2kSZM0dOjQHsvMTO+9917izw9J2l7ouOJo586dPQqPUmeD7YoVKwKKCACQjpmRLyJUysvLNWHC/9/e3UfLVdaHHv/+Tg4poKElQJAkICppDAmivIcraayNoOKlErwacRUWyJFVENDrC633LkCLV9BbQVFMuCrpqpcUrkFTWg+gNSAWTJGX8GZIIkgSIuGlrYlAyMtz/5g5YTI5c+bl7JnZM/P9rDXrnNmzZ+/ffnv2b55n72cfAcC4ceOYNm0a69at413vehf9/f1Do90DTG5XjNXcvHrLrsNWWf8mSVIv6q8+yqsi4mDgbcAvgP1TSuuh0EAZERMqfGcAGACYP38+AwMDDQU68bqNdY3/vjf2M/+dezBut+F/TLx2bGH4R29/iX9+sr5E6OlzxtU1viRJUh5cddVVnHjiiXzqU58C+Apw/HDjZZW/dZp6K6EnTJjAM888w+67787YsWN55ZVXdvp8r732AuBrX/saF154YV3T7oU7bCRJ0vCefPJJ7r//fo499tjyj84C/mG477Sr/g3g9lP3ZPo+Y4atgxs3tvB39X9s54Sbfl/3tK2DkySpM9Xc+BgRrwW+D1yUUvpdrZUzKaUFwIKht3VH2KCfP72NzdsSb580hunj+3jkhe07PpuxTx/HHzCGl7cm7lnv8x8lSVJvuPbaa/nqV7/K3LlziYhPAN8G/qx8vHblb51mw4YN3H///bztbW/jnHPO4Rvf+MaOz/baay/OPvtsAG699dZ2hShJkjrMpk2bmDt3LlddddWOC5kALr/8coCtwPeG+14787efrtnK9H3GMHDYWO5e/9JOMx+YUWh9XGoPZJIk9ZSq3a4CRMRuFBoev5dSWlwc/EzxeZBDz4Xc0JwQG/PCy4l/eHwLfRHc+N49uehtY/mTyWO46G1jufG9e9IXwaLHt+zSJWtWhrpwycNLkiQJYOHChZx66qlDb28CfCDhKF155ZUAXH311Vx77bW8+93vZmBggGXLlnHQQQfx0EMPMTg42OYoJUm18Pez2m3Lli3MnTuX008/vTRnY+HChdxyyy0Ap6ccdo+w8NEt/H5L4l2v72fRe/bgfW/s571v6OfvT9qD//qm3Xhpa+I7j7xSfUKSJKlrVL3zMQrZ97eBx1JKf1vy0RLgDOBLxb8/bEqEo3DJ3Zs5YM8+5ry+n88ctfPDrW/7zVYuu2dzhW9KkiR1n4kTJ3LHHXcwe/ZsgD8FVrY3os63aNEipkyZwuc//3nOPfdczj333B2frVixgpNPPpnt27ePMAVJkqRCl+tnn30206ZN45Of/OSO4YODg1xxxRXccccd7Lfffi+2McSK1v0+cdZtL/F/5uzBCZP6OWHSq9WNL25JfOwnL/HE73LXZipJkpooql0wFRFvB34GPAQM1Zz8NYXnPt4IHAQ8BXwgpfRClfk1nGk00uf8kFmTxnDalN3Yf8/gmRcTNz2+hZ893Xh3q7X0N5+nKyZzeFGcpOHVU3B4YFdRR/fgTY5ELeLxU2bevHksXbqU5557jv3335/LLruMqVOncuGFF7J161aWL1++DPjLlNIvq0yqJ9YXjC5/mz59OgMDA0ybNo1NmzZx8803c+ONN7J5c2MXu1k2SR0jN+efiLDsqENpmT+03sqHDTdONys/D7ZgmXNz/OTBXXfdxQknnMBhhx1GX1+ho7IvfvGLXHDBBWzevJl99tmHBx988EHgnpTSuSNPrT31b+N3D+ZN3Y2ZB4whgF/8dhs3rNjCsy81vvl85qOkJvD8I41OTcdQ1cbHjLUl+cmajY+SmsTkJ0M2PvYcj5/61brOemZ9mb9JakBuzj82PtbHxsdd2fjYEZqev+Wp/g2q18HlKX+D3igrpC7g+UcanZqOoardrkqSJEmSpHzKW8W7JEmSJNn4KEnqKlbASZIkSZIkSVL79LU7AEmSJEmSpDzxgjZJkiSpcTY+SpIkSZIkSZIkScqEjY+SJEmSJEmSJEmSMuEzHyVJkqQOlKcuAVNK7Q5BktSgPJ1PJEmS1B2881GSJEmSJEmSJElSJrzzUZLUNRq5anvoO961I2Vv4nUb2x3CDk+fM67dIUhS05nXSJIkqR0iwhxUO6l652NEfCciNkTEwyXDxkfE7RGxsvh37+aGKUmSpCxcffXVzJgxg4h4JCIuanc8kiS1W0TY9ahy4ayzzmLChAnMmDFjx7CbbrqJ6dOn09fXR0Qc1cbwJEmSalZLt6vXAyeVDbsY+ElKaQrwk+J7SZIk5djDDz/Mddddx7JlywAOB06OiCltDkuSJEnAmWeeyeDg4E7DZsyYweLFi5k1a1abopIkqTIv4lIlVRsfU0p3Ai+UDT4FWFj8fyHw5xnHJUlSTUxwpNo99thjHHfccey5556klLYCdwDvb3dckiRJglmzZjF+/Pidhk2bNo2pU6e2KSJJkqTG1HLn43D2TymtByj+nVBpxIgYiIh7I+LeBQsWNDg7SZIq8yorqTYzZszgzjvv5Pnnnyci9gTeAxxYPp75myRJUmcxf5MkSXnS3+wZpJQWAENZj08clSRlwsZGqX7Tpk3js5/9LHPmzAEYBB4EtpaPZ/4mSZLUWczfJEndbqguMCVPc52g0Tsfn4mIAwCKfzdkF5IkSZKa5eyzz+a+++4jpTSLQtf6K9sdkyRJkiRJkrpHo42PS4Aziv+fAfwwm3AkSarOux6lxm3YULhmLCIOAk4FbmhrQJIkSZIkSTWyXrAzVO12NSJuAGYD+0bEWuAS4EvAjRFxNvAU8IFmBilJkqRszJ07l+effx7gH4HzUkr/3uaQJEmSBMybN4+lS5fy3HPPMXnyZC677DLGjx/Pxz/+cZ599lmAf4qIB1JKJ7Y7VkmSpJFEi/vHbXhmE6/bmGUco/L0OeOqjpOn1nf7QJY6Rj0FR08e2M0sWy0rO57HT/1qXWfmb21QS5nUafFKXart55/hygKPyerqLUO7eZ1WWhctWOa2Hz8dqKfyN6iew+UpH4LuLiukLuL5J0Ol5XCzy8BWzksjqukYqnrnoyRJedDsH5URUTFxaVVy08h8Sh+2bRImSZLUHKU5lyRJkqSR2fgoqePl6UpHKyOy1+7tWz7/ZlQ8VVrGao2JpZ+3Ik5JktQZRrqoSpIkSZKazcZHSZKGkUWjZ6XGw1oqBG1MlKT2avfFL6Us+6Vs5en4liRJkrqRjY+SJBU1syIqy8bEeuL0zgdJeZGnyn7LRUmqzovfJEmS1Ki+dgcgSVInyqoSvd7pNDLfiMhVpb8kSZIkSZKk7uWdj5Kk3OqUBrPy7lU7Ie5Kd0R6p6QkSZ2jE3IOdT7zQ0mdIm/nRctOSb3MxkdJUu7k7QdDJcPFmefYR+r6tfQzu9iSJEmSpOzl7feiv/kkSc1it6uSpFzJ24+xXuA6lyRJvSCLrui7LW/qtuWRJEmtYx6hkYyq8TEiToqIFRGxKiIuziooqdsM/cjNw6sb41XzDbc9st4+bvP88bjsToODg0ydOhXzN0nqbp63pe4yODiIdXCSpF41XL2kuW6+Ndz4GBFjgG8A7wYOBeZFxKFZBSaNpN0NYlbCqxtV2p9Lh2e973ssSa21bds2zjvvPH70ox+B+ZskqcuV57FZTtccVq00lMNhHZwk9YS85Bl5iUOdaTTPfDwGWJVS+jVARCwCTgEezSIwSVJrjKZSZrjxy58ZUTpO6WcmL50rIup+Nki1faWRaap+y5Yt45BDDuGNb3wjKaVXzN8kSeosQznVaPMmc/HOMpTDrV692jo4tVTeygp/M7aX+0M2sjqXS3k3msbHScCakvdrgWPLR4qIAWAAYP78+QwMDDQ0s6fPGdfQ99rFwqO5Om39Gm9zdVq8edKMxHGkaeYtUVXjakmWq23vWrvy9RjPzrp16zjwwANLB5m/lei0fc14m6vT4pX0qmbd8djprOzsXLXkcOZvncF4m6vT4u00rt/seW4ePS9mz6/RND4Ol8HvspVTSguABZU+lyS1npUwyor7UmepkJCbv0lSTjSj8qTVFTLDVaK1MoZW5yaVevkYzbQqTWekRtVq8x5tg2w9laNWAmavlhzO/E2S1CrVHt3UKXlAr+UsrV7e0TQ+rgVKL7uaDDw9unAkSVmxUUhSucmTJ7NmzZqdBmH+JkkdZTQNR/BqZVClxqvRNKbV2gNG6bwanUdKadj/26mRCrfh4h7tNlb3MYeTpO4wUqNdo41CWV4IVWm6ndJI1+w4mzX90TTaVvpuHhqC+0bx3X8DpkTEGyJiLPAhYEk2YUmSJClrRx99NCtXruSJJ57A/E2S8ikimtqQVGn6ww0fGlb+qmc+1eY13P+Vvlvp7r28NbzVuq7yFvdoNXvf7WVDOZx1cJKUX/XmSrV8v1IuNtp5jCbuPJ7ra827hluHtaz3StMpHTaa2KtNp1JsIy1fPbE1a5s23PiYUtoKnA/cCjwG3JhSeiSrwCRJkpSt/v5+rrnmGk488UQwf5OkjpKnhp1mxDFSZUuelr0RlSq0pFoN5XBYBydJuVNPQ1D5OFnPt3zeleZXT8NatflViq3afGtpLGtkHY2mcbDeZa92Ed1Iy9rI9htpm9a6XPXGMxrR4mc/DAL7tmyGktRZnkspnVTLiJan0i5qPn5UH8sbSRqR+ZvUOPO3JrG8kaQRmb9Jo1PTMdTSxkdJkiRJkiRJkiRJ3Ws0z3yUJEmSJEmSJEmSpB1sfJQkSZIkSZIkSZKUCRsfJUmSJEmSJEmSJGXCxkdJkiRJkiRJkiRJmbDxUZIkSZIkSZIkSVImbHyUJEmSJEmSJEmSlAkbHyVJkiRJkiRJkiRlwsZHSZIkSZIkSZIkSZmw8VGSJEmSJEmSJElSJvpbPL/U4vlJUieJOsa1PK0gYuTVmFJnr7ry5ev05cmQx0/9al1nri9Jqszzj9Q4j5/6mb9J0uh5/pFGp6ZjyDsfNWpWfEvqJNUaJyXlhzmGJEmSJElS56na+BgRB0bETyPisYh4JCIuLA4fHxG3R8TK4t+9mx+u8igiWL58Od/97ndZsWJFu8ORJKnnrVmzhne84x1MmzaN6dOnc/XVVwNw0003MX36dPr6+oiIo9ocZkVDjY5btsOy327lN7/b3uaIJEmSmqtS/jbkK1/5ChGRImLfNoXYlV5++WUWL17M4sWLefnll9sdjiRJXSOqXVEeEQcAB6SU7ouIccAvgT8HzgReSCl9KSIuBvZOKX22yvy8fL2LpJSICC644AK+/vWv7xh+ySWXcOmll7YvMKlz2e1DBmq5s7ET76aqtFyduCxN4vFTYv369axfv54jjjiCjRs3cuSRR/KDH/yAiKCvr4+Pfexj3HHHHUenlO6tYXJtWV/3PrONj97+EhteSgRw6iH9XPUnu9MX3sEsKVc8/0iN8/gpUSl/O/TQQ1mzZg0f/ehHue22254CjkwpPVdlcl2/vrKwatUqZs+ezbp16wCYNGkSS5cu5ZBDDmlzZJKazPOPNDrZdLuaUlqfUrqv+P9G4DFgEnAKsLA42kIKDZLqIRHBAw88sFPDI8AXvvAFnnrqqTZFJUmSDjjgAI444ggAxo0bx7Rp01i3bh3Tpk1j6tSpbY5uZEMN6p+882U2vFT4PwHfX7WVJb/easOjJEnqSpXyN4BPfOITXHnllWAleKYuueSSHesYYN26dVxyySVtjEiSpO5R1zMfI+Jg4G3AL4D9U0rrodBACUyo8J2BiLg3Iu5dsGDB6KJV00REXa+BgQEAfvnLX+4yre3bt3P//fcDcOqpp9Y9bUnSziwbNRpPPvkk999/P8cee2y7Q6lJRPDsi9tZ9R+7drX686e3tSEiSZJUib/jm6M0f1uyZAmTJk3i8MMPH/E7vVr/Vm+dU0SwatUqAO69d9dOQIaGrVq1qqFpS5Kkgv5aR4yI1wLfBy5KKf2u1hNqSmkBMJT1eIVWl1i7di0AM2fO3OWz/v5+jj76aICdriCTJGVv6Hxs96sazqZNm5g7dy5XXXUVe+21V83fi4gBYABg/vz5Oy46qtfE6zbWNf4e/bDijNcyfvdgvz2CZ1/aeb9+8/jCdXPXPvgKX1i2ua5pP33OuLrGlyRJaofS/K2/v5/LL7+c2267rer3rH+r3dq1aznkkEOYOXMmjz/++E6fDdVzWZ8lSdLo1HTnY0TsRqHh8XsppcXFwc8Unwc59FzIDc0JUXl0++23s379eg499FA+97nP0ddX2JXGjBnDFVdcwcSJE3nsscdYtmxZmyOVJKk3bdmyhblz53L66adz6qmn1vXdlNKClNJRKaWjGm14bMRLW+G232xlTF/wP4/9A8aUXOt26Pg+5k3dDYAfrN7SspgkSZJapTx/W716NU888QSHH344Bx98MMBk4L6IeF17I+1sCxcWniL1hS98gSlTpuwYPmXKFD7/+c8DcP3117cjNElShrwrvb2i2p0SUdg6C4EXUkoXlQz/MvB8SulLEXExMD6l9Jkq8/PKq5xq5CD84Ac/yKJFiwB44okneOCBBzjmmGOYNGkSW7du5eSTT+bWW2+te7revaMe5gOvM1BLedZp5Uw3LlMTePyUSClxxhlnMH78eK666qpdPp89ezZ33HHH0SmlXfuaGmZyjcZR752PANPH97HklD3Zoz944j+38+OntnLAa4ITD+5nt75gya+3cO5PXq57ut75KKlJPP+op42yFw6PnxLV8jeAiPgNcFRK6blqk8s8wJxqpD5r7NixLF26lJkzZ7Jt2zaWLl0KFHLkMWPGcPfddzN79mxeeeWVuqft7zKpI3j+6RH2FtY0NR1DtTQ+vh34GfAQMPTwnb+m8NzHG4GDgKeAD6SUXqgyP7dyTjV6BcBpp53GlVdeyRve8IYdw371q19x0UUXNdTwCBYG6mkmPxnoxoa6blymJvD4KXHXXXdxwgkncNhhh+3oneCLX/wimzdv5uMf/zjPPvssr7zyygbggZTSiVUm19LGR4DjXjeGq2fvzoHjXu2kY9v2xI0rt/DXP9/M5gYe/Wjjo6Qm8fyjnmbjY3Yq5W/vec97doxj4+OuGq3PGjduHF//+tf58Ic/zG67FXrX2LJlCzfccAPnn38+Gzc2lsf6u0zqCJ5/eoSNj02TTeNjxtzKOTWa24/7+vo4/vjj2X///Vm3bh333HPPqGKxMFAPM/kZpTqeR9zkSLLVrcuVMY+f+tW6zlre+AjQFzB78him7t3Hi1vgx09tZd3vG990Nj5KahLPP+ppNj62XNPzt04z2u70JkyYwHHHHQfAL37xC5555plRTa/Hf5NJncLzT4+w8bFpbHxU7fLU97GFgXqYyc8odVsjXaNlc6csX8Y8fuqX68bHrNn4KKlJPP+op9n42HI9lb9B9RwuT/VZ0LO/xaRO4/mnR9j42DQ1HUN91UeRJEmSJEmSXpW3Rh9JkiTlh42PkiRJkiRJkiRJkjJh46MkSZIkSZIkSZKkTNj4KEmSJEmSJEmSpJ5hF/LNZeOjJEmSJElqCSt5JEmSpO7X3+4AajXxuo3tDmGHp88Z1+4QJEmqKCJIKbU7DEmSJEmSJEk9yDsfJUl1iYiOv2K9G5ZBkiRJkiRJkvKoauNjRHwnIjZExMMlw8ZHxO0RsbL4d+/mhilJyps8Nd7ZmCjt6qyzzmLChAnMmDFjx7CbbrqJ6dOn09fXR0Qc1cbwJEmSVGa4/O3Tn/40b37zm3nLW95CRNwcEX/UxhAlSZJqUsudj9cDJ5UNuxj4SUppCvCT4ntJkiTlxJlnnsng4OBOw2bMmMHixYuZNWtWm6KSJElSJcPlb3PmzOHhhx9m+fLlAI8Df9WO2CRJkupR9ZmPKaU7I+LgssGnALOL/y8ElgKfzTCujpenO3B87pekZhkq63qpnCkt33tpudV5Zs2axZNPPrnTsGnTprUnGEmSJFU1XP72rne9q/TtPcBpLQxJkqSOkqd2mV7X6DMf908prQco/p2QXUiSpE7TKyf2XllOSZIkSbl0FvCjdgchSVK38FFOzdNo42PNImIgIu6NiHsXLFjQ7NlJkpoojyfjVsRUKRFpZoKSx3Wt3tGr+dvQMZ2HlyR1G8s3aXQuv/xygK3A94b7vFfzN0mSlE9Vu12t4JmIOCCltD4iDgA2VBoxpbQAGMp67J9OkqQW6MUucZUd8zc1Q54aHSwbJUmdZOHChdxyyy0Ap6cKJzHzN0mSlCeN3vm4BDij+P8ZwA+zCUeS1Km8ml2SJEmSsjU4OMgVV1zBkiVLSCm92O54JEnqdNZhtkbVxseIuAG4G5gaEWsj4mzgS8CciFgJzCm+lySpZUwUpJHNmzePmTNnsmLFCiZPnsy3v/1tbr75ZiZPnszdd98N8E8RcWu745QkSVLBcPnb+eefz8aNG5kzZw4R8UBEfKvdcUqSJFUTLe5yqOGZTbxuY5ZxjMrT54yrOk6eKsRr2cadFq/Upeo5EFt6oNRTRrTqGM6i3MqqfGzGMmdVLvdQmZrb4yfHal1n5m9t0I3HrutXXcrzT41KywCPwe6QwTb1+KlfT+VvUD2Hy1N+AZZvUofw/NPFhjsvlJfNtYyjEdV0DDXa7aokSZIkSeoS9iohSZIkKSs2PkqSOo4VY7VzXUmSpF5jQ6okSZLUXjY+SpJGlKeKm16oSOqFZZQkSdkzf1A7uf9JkiSplI2PkqRM9WLFQy8usyRJUqO82EqSJEnqbjY+SpKGlbdKoWbE0q7lqzTfPK1vSZKUTyPlC3nL3yRJkqSsmOt2lv52ByD1gjwViimldocgdYSsj9vy6ZW+97iUJEndLC8XfJlzSZIkSa1h46MkqS1sfHtVsyvkhqbfzvXs9pYkqTNEhOdqSZIkdZw83QAkGx8lSU1QqbFrpO5GK1VydUrDXL3L3Eu861OSpObJItcwX5Gk1shbeevvMUlSs4zqmY8RcVJErIiIVRFxcVZBSZK6Qz0/rIb6bS9/tUPefhBKWTrrrLOYMGECEfFwu2OR8qzSeakdL0nKE8um9hgcHMQ6OEmSmsPcJnsNNz5GxBjgG8C7gUOBeRFxaFaBSZLapxnPO8z7STyrSt7S7+d9mZutE7Z7LzrzzDMZHBxsdxiSpFFoduN0q87h5glSbbZt28Z5550H1sFJkrpIVjmn9U/5NJpuV48BVqWUfg0QEYuAU4BHswhMGkmeChO7qGg/94d8y9P2aZVeXOZyta6DPDyPstfMmjWLJ598st1hKAN5Kms8htvP/UFSXvjc0OwtW7aMQw45hNWrV1sHp5bKU34B3ZdjdNr67bR4JbXXaLpdnQSsKXm/tjhMkiTlUKvuYMjbDxJJkrpBPefYvJyL85YX5C2eTjTS+nP9Ns+6des48MADSwdZBydJUsbMZbI1mjsfh9sKu1xuEBEDwADA/PnzGRgYaGhmT58zrqHvtUunXXlhvM3VafF2GtdvfUY6ibou1Yisushw/8uPrPI3SaokTz/qPf+0Vyf3gmD+olapsJ/tNND6t85gvCrVaeu30+JV9yrNwfL0u0I7G03j41qg9LKrycDT5SOllBYAC4bejmJ+kqRRqOVk7Alb9cpyn6k0LX/gtF5W+ZuVV81lvM1lvM3VafH2qpEa17JuNMyqIa/WnLfacoH7qb8N8mPy5MmsWbNmp0GU1cFZ/yZJ6iT15hnmJfWpJVdvdt47mm5X/w2YEhFviIixwIeAJdmEJUnKkidoDcmiC4mhabhfSZKkLPOCTswxOi3eajpxG/SCo48+mpUrV2IdnCTlVyecOyud46sNL88PaplOvY8MGG78rLp7H24ZatUJ2xXyGWfDjY8ppa3A+cCtwGPAjSmlR7IKTJI0OlYcKGvuT91h3rx5zJw5E2BqRKyNiLPbHZMkqaAZDXndbDTLWEsFV70VdI3GUD69WveD4SoEK71qjUf51N/fzzXXXAPWwUlS7lRqcKvnPD3Sebva+byWc36t0yzPR6pNq97/S+c9kkYbMbNSqbG13mUZab1XW8Zat3e1OGvNDZuxnqPF3YjY7YMkVVZPCV+1PLXyQNWU5gB53l9qzFUyPX56RK3rzPUlSZU1NX/L4lk2WZ3va51O1s/fqaV71GbMd6Q4Ks2j2jjly1JrrM3Yho3EkcW8ymddzyQbDKXbmL9J0uhZ/6bMdcszKLPM4Vra+BgRg8C+LZvhrvYFnmvj/OtlvM1lvM1lvPV7LqV0Ui0j5qA8lfKm5uNH9clJeZOHMroexttcxttcxlsf8zepceZvTZKT8qbd5XO9jLe5jLe5jLc+5m/S6NR0DLX6zse2ioh7U0pHtTuOWhlvcxlvcxmvJCkrnVZGG29zGW9zGa8kKQudVj4bb3MZb3MZr6Q8aviZj5IkSZIkSZIkSZJUysZHSZIkSZIkSZIkSZnotcbHBe0OoE7G21zG21zGK0nKSqeV0cbbXMbbXMYrScpCp5XPxttcxttcxispd3rqmY+SJEmSJEmSJEmSmqfX7nyUJEmSJEmSJEmS1CQd3/gYEa+LiEURsToiHo2If46IPy5+9omIeDki/rBk/NkR8Z8R8UDJ688iIoqfX1r8Gy1chm1l8VxcHL40IlZExPKI+FVEXBMRf9SquIaJ83MR8Ugxngci4tji8P0iYktEfKxk3HHFbTKl+H63iHho6POwr2AAAAlsSURBVDtNjnNofT4SEQ9GxCcjoi8iTixZx5uK6/aBiPi7iLg/It5a/H5/RPw+Ij5SMs1fRsQRGcZ4cEQ8XDbs0oj4VERcHxEvRsS4ks+ujogUEfuWDHt/cdiby6b7UnG5Ho2Ib0XEiMd5ROxTsl5+GxHrSt5fUmGbD+2bQ+OdVhw+tO4fjoh/HNpfy+Iaev1F8bOzivvG8uL3TikOv35ouiWxbqqwDAdGxBMRMb74fu/i+9dHxB8Xy4VVEfFYRNwYEfsXx3t7RCwrHl+/ioiBsu0xtC4ejYh5JZ9dX5z+A8V97J0jrWNJ0q7CHK5VMZq/ZRej+Zv5myT1tDB/a5nogBwuzN96Ln8rfmYOJ6k2KaWOfQEB3A2cWzLsrcAJxf+XAT8Dziz5fDZwyzDT+gjwGeBLxb8faeFybKowfClwVPH/scD/Bu5o07qeWVzXf1B8vy8wsfj/XxbX89Ky7/w34Lbi/38FzG/1+gQmAD8GLqu0bovvvwH8ZfH/I4H7gG8W378G+HdgTIYxHgw8XDbsUuBTwPXA8qF9kMJFAsuBtcC+JePfWFzvlw43XaAfuBM4tY64LgU+VcM232n9VVj3C4HPVVre4vDJwGrgD4vvXwu8ofj/9cBplaY/zLQ+Aywo/j+/uM/tDqwE3lcy3juAGcDrgKeAI0qW75fAe4dZF1OA3wG7lcdWnN7KVuzbvnz58tUtL8zhWhWf+Zv5m/mb+ZsvX758ZfLC/K2V67ojcjjM33oyfyt+bg7ny5evqq9Ov/PxHcCWlNK3hgaklB5IKf0sIt5EoSD9H8C8ShMo+d7fA2soFJ5PFd/nRkrpFQqxHRQRh7chhAOA51JKm4vxPJdSerr42TzgvwOTI2LS0BdSSjcC2yPiM8C5FE5ELZVS2gAMAOdHjHgl3c+B44v/Hw98i0ISDXAMcF9KaVvTAt3VDcAHi//PLsa3dejDiHgt8F+As4EPDTeBlNJW4F+BQxqMYaRtXou7gUlVxpkAbAQ2FeexKaX0RCPBAl8FjouIi4C3U/ih8GHg7pTSPw6NlFL6aUrpYeA84PqU0n3F4c9ROMYuLp9wSmkl8CKw9zDzrWU5JUk7M4drDfM38zfzN/M3ScqK+VvrdFwOZ/7WU/kbmMNJqkGnNz7OoHCVxHDmUTiB/AyYGhETSj47oez28zdFxIeBA4ErKSQXH25q5DvboyyeDw43UvHk+yDw5uE+b7LbgAMj4vGI+GZE/AkUbrUHXpdSWkbhSqDy2C8CrgD+JqX0QksjLkop/ZrCvj5hhNH+lZ2TnzuBzcWuF46nkHy00kpgv4jYm8K+vKjs8z8HBlNKjwMvDNclRUTsCbwTeKjBGIbd5iW+V7LP7lM27zHFeS8pGfymsv38BAr78zPAExHx3Yh4X9k8vlz6nZGCTSltAT5NIQG6qPhjYaQyYvown91bHL6T4vpdWUymy50E/GCk2CRJuzCHaw3zt9Yyfyswf5Ok7mT+1jodmcOZv/VG/gbmcJJq09/uAJroQ8D7U0rbI2Ix8AEKt/YD/CyldHLpyBHx65RSiohLU0pXVrlKJ2svpZTeWn00oNDNRcullDZFxJHACRSudvuHKPSLvx+FhAcKJ+hvA39b8tWTgPUUTkDtNOJ6Syk9GRFjI+J1FBLLFcC/AcdSSH6+nnE8qYbhiynsx8cCHysbbx5wVfH/RcX39xXfv6mYKCTghymlHzUUYIVtnlK6vjjK6Smle8u+tkdx3gdTSCpuL/ls9XD7eUScBBxNIVn6akQcmVK6tPjxp1NK/69k3Ip9zhe9m1f3t9urjBsMvx1Kh30iIs4B3khhXy715Yi4kkJSfVyVeUmSamcOlxHzN/M3zN/M3ySpNczfMtThOZz5W70Bdmb+BuZwkqro9MbHR4DTygdGxFso9A99ezF/GQv8mlcTn12klFLx76Wl7/OkeDXLYcBj7Zh/8aqvpcDSiHgIOIPCre77R8TpxdEmRsSUlNLKiJgIXECh24SfRsS3U0rLWx13RLwR2AYMd8VMqbsp7E/ri0nwPRS6VjgGuCfjsJ5n1+4DxgOlXR4sopDQLCwm8EDhAdXAnwIzIiIBY4AUha41oEKS0YgK2/z6Eb7yUkrprVF4wPwtFLpV+FqVeSQKz4ZYFhG3A9+l0Nd7XaLwwPI5FJKQuyJiEYUyovyKsSGPAEex89VhRwKPlrz/akrpKxFxKvB3EfGmlNLLxc8+TSFBvYBC//pH1huzJPUwc7gWMX/LlPnbq/Mwf5Ok3mP+1kKdmMOZvzWuk/I3MIeTVJtO73b1X4A/KF4VAUBEHA1cTeEhwAcXXxOBSRHx+nYFOloRsRvwv4A1baoAmhoRU0oGvZVC4/VrUkqThtZ1McahPtC/CnwxpbQW+CTwjRZfzUZE7Eeh//hrakhmfw58gkISRPHvXwC/TSn9R5ZxpZQ2Aesj4p3FOMdTuKrnrpJxngI+B3yz7OunAX+XUnp9cb0fSCFpenuWMVbY5r+p5bsppf+kkBB8qrjvVprHxNi5y4qa51E2nQCupdDVw1PAl4GvAP8XOD4i3lsy7kkRcRiFH0JnFhOmoaTyCgrdvpQvz2IK3UGcUTZ8O4Xypi8iTqw3bknqYeZwrZm3+VuGzN92zMP8TZJ6k/lb6+bfcTmc+VvjOil/K07LHE5STTq68bF4Mns/MCciVkfEIxSu2JgN3Fw2+s28ekIu729+lyu3Wqy8v/kvlXz2vYhYDjwMvAY4pT0h8lpgYUQ8WoznUGA1u67n7wPzImIOcBCFLiBIhYcN/zuFZKLZhtbnI8CPKfSdflkN3/s5hVv77wZIKa2ncFXTvzYpzr8A/kcUukn4F+CylNLq0hFSSvPLh1Ho4mG49Z71MxKG2+aX1vrllNL9FPqUHzruyvucvwDYDfhKRPyquB4+CFzYQKznUHhI/VA3D9+k0H3HMcDJwMcjYmVEPAqcCWwobt+PANdFxK8obOfvpJIHY5f5PPDJiNip3CyWQ39D4UHZkqQamMO1jPlb9szfzN8kqSeZv7VUp+Rw5m/Z6KT8DczhJNUoctizgSRJkiRJkiRJkqQO1NF3PkqSJEmSJEmSJEnKDxsfJUmSJEmSJEmSJGXCxkdJkiRJkiRJkiRJmbDxUZIkSZIkSZIkSVImbHyUJEmSJEmSJEmSlAkbHyVJkiRJkiRJkiRlwsZHSZIkSZIkSZIkSZmw8VGSJEmSJEmSJElSJv4/AwI9G/MFmv0AAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 2304x648 with 72 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(\n",
    "    nrows=comparisons,\n",
    "    ncols=pattern_sizes * 3 - 1,\n",
    "    figsize=(8 * (pattern_sizes + 1), comparisons),\n",
    "    gridspec_kw=dict(width_ratios=([1,0.6,0.2] * (pattern_sizes - 1) + [1,0.6]), wspace=0.05, hspace=0.25)\n",
    ")\n",
    "\n",
    "with h5py.File('search-targets.h5', 'r') as f:\n",
    "    for c in range(comparisons):\n",
    "        for s in range(pattern_sizes):\n",
    "            res = np.array([x + results[s][c][1][i] for i, x in enumerate(results[s][c][0])])\n",
    "\n",
    "            colors = np.zeros((7, 4)).astype(np.float)\n",
    "            colors[0,:] = np.array([25.0, 154.0, 242.0, 0.0]) / 255.0\n",
    "            colors[np.where(res == np.max(res))[0],3] = 1.0\n",
    "            edgecolors = np.ones((7, 4))\n",
    "            edgecolors[:,3] = 0\n",
    "            edgecolors[np.where(res == np.max(res))[0],3] = 1.0\n",
    "            \n",
    "            ax[c, s * 3].bar(technique_names, res, color=['#199AF2', '#000000', '#000000', '#000000', '#000000', '#000000', '#000000'], zorder=1)\n",
    "            ax[c, s * 3].scatter(technique_names, res, color=colors, edgecolors=edgecolors, linewidths=2, s=60, zorder=2)\n",
    "            \n",
    "            ax[c, s * 3].set_yticks(range(min(res), max(res) + 1, max(1, math.floor(max(res) / 2))))\n",
    "            if c == 0:\n",
    "                ax[c, s * 3].set_title(pattern_sizes_names[s], fontsize=20, fontweight='bold')\n",
    "            if c < (comparisons - 1):\n",
    "                ax[c, s * 3].tick_params(axis='x', length=0)\n",
    "                ax[c, s * 3].set_xticks([], [])\n",
    "            ax[c, s * 3].spines['top'].set_visible(False)\n",
    "            ax[c, s * 3].spines['left'].set_visible(False)\n",
    "            ax[c, s * 3].spines['right'].set_visible(False)\n",
    "\n",
    "            target_pattern = f['targets_{}'.format(pattern_sizes_shortnames[s])][c]\n",
    "            ax[c, s * 3 + 1].bar(np.arange(target_pattern.size), target_pattern, width=1.0, color='k')\n",
    "            ax[c, s * 3 + 1].set_ylim(0, 1)\n",
    "            ax[c, s * 3 + 1].set_xticks([], [])\n",
    "            ax[c, s * 3 + 1].set_yticks([], [])\n",
    "            ax[c, s * 3 + 1].spines['top'].set_visible(False)\n",
    "            ax[c, s * 3 + 1].spines['left'].set_visible(False)\n",
    "            ax[c, s * 3 + 1].spines['right'].set_visible(False)\n",
    "            if c == 0:\n",
    "                ax[c, s * 3 + 1].set_title('Targets', fontsize=16)\n",
    "            \n",
    "            if s < pattern_sizes - 1:\n",
    "                ax[c, s * 3 + 2].set_xticks([], [])\n",
    "                ax[c, s * 3 + 2].set_yticks([], [])\n",
    "                ax[c, s * 3 + 2].spines['top'].set_visible(False)\n",
    "                ax[c, s * 3 + 2].spines['right'].set_visible(False)\n",
    "                ax[c, s * 3 + 2].spines['bottom'].set_visible(False)\n",
    "                ax[c, s * 3 + 2].spines['left'].set_visible(False)\n",
    "\n",
    "fig.savefig(\n",
    "    'user-study-1-results.png',\n",
    "    bbox_inches=\"tight\",\n",
    "    pad_inches=0\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chi-Square"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Size: 3 kb\n",
      "Chi2 Stat 206.4444444444444\n",
      "P-Value: 8.055437295394612e-42\n",
      "\n",
      "Size: 12 kb\n",
      "Chi2 Stat 284.5644444444444\n",
      "P-Value: 1.6557386673466367e-58\n",
      "\n",
      "Size: 120 kb\n",
      "Chi2 Stat 319.0044444444444\n",
      "P-Value: 6.902620545376983e-66\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for s in range(pattern_sizes):\n",
    "    res_merged = np.sum(np.sum(np.array(results[s]), axis=1), axis=0)\n",
    "    chi2_stat, p_val = stats.chisquare(res_merged)\n",
    "    \n",
    "    print('Size: {}'.format(pattern_sizes_names[s]))\n",
    "    print('Chi2 Stat {}'.format(chi2_stat))\n",
    "    print(\"P-Value: {}\".format(p_val))\n",
    "    print()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Post-hoc pairwise comparisons with Holm-Bonferroni-corrected Fisher exact tests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====================\n",
      "Size: 3 kb\n",
      "====================\n",
      "\n",
      "Comparison : CAE* vs ED\n",
      "P-value    : 0.32178117350790286\n",
      "Corr. P-val: 0.32178117350790286\n",
      "Significant? False\n",
      "\n",
      "Comparison : CAE* vs SAX\n",
      "P-value    : 0.0019483958035769848\n",
      "Corr. P-val: 0.0038967916071539696\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs DTW\n",
      "P-value    : 2.2910768200020403e-08\n",
      "Corr. P-val: 9.164307280008161e-08\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs UMAP\n",
      "P-value    : 0.0005608619652869397\n",
      "Corr. P-val: 0.0016825858958608192\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs TSFRESH\n",
      "P-value    : 6.026919842904277e-22\n",
      "Corr. P-val: 3.616151905742567e-21\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs XCORR\n",
      "P-value    : 2.207381602327927e-10\n",
      "Corr. P-val: 1.1036908011639635e-09\n",
      "Significant? True\n",
      "\n",
      "\n",
      "====================\n",
      "Size: 12 kb\n",
      "====================\n",
      "\n",
      "Comparison : CAE* vs ED\n",
      "P-value    : 0.7362157398954523\n",
      "Corr. P-val: 0.7362157398954523\n",
      "Significant? False\n",
      "\n",
      "Comparison : CAE* vs SAX\n",
      "P-value    : 1.6938853864885319e-06\n",
      "Corr. P-val: 3.3877707729770638e-06\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs DTW\n",
      "P-value    : 3.9412794995340284e-07\n",
      "Corr. P-val: 1.1823838498602086e-06\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs UMAP\n",
      "P-value    : 9.10989617529731e-08\n",
      "Corr. P-val: 3.643958470118924e-07\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs TSFRESH\n",
      "P-value    : 5.70130130385976e-23\n",
      "Corr. P-val: 3.4207807823158555e-22\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs XCORR\n",
      "P-value    : 1.2575371517176866e-14\n",
      "Corr. P-val: 6.287685758588433e-14\n",
      "Significant? True\n",
      "\n",
      "\n",
      "====================\n",
      "Size: 120 kb\n",
      "====================\n",
      "\n",
      "Comparison : CAE* vs ED\n",
      "P-value    : 0.18313507356538222\n",
      "Corr. P-val: 0.18313507356538222\n",
      "Significant? False\n",
      "\n",
      "Comparison : CAE* vs SAX\n",
      "P-value    : 8.133203561168418e-11\n",
      "Corr. P-val: 2.4399610683505253e-10\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs DTW\n",
      "P-value    : 3.6469170024051124e-06\n",
      "Corr. P-val: 7.293834004810225e-06\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs UMAP\n",
      "P-value    : 1.1526306164071256e-14\n",
      "Corr. P-val: 4.6105224656285024e-14\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs TSFRESH\n",
      "P-value    : 8.342504655932159e-18\n",
      "Corr. P-val: 4.1712523279660796e-17\n",
      "Significant? True\n",
      "\n",
      "Comparison : CAE* vs XCORR\n",
      "P-value    : 9.548435165528943e-19\n",
      "Corr. P-val: 5.7290610993173665e-18\n",
      "Significant? True\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pairwise_comparisons = [[0, 1], [0, 2], [0, 3], [0, 4], [0, 5], [0, 6]]\n",
    "pairwise_comparisons_pvalues = []\n",
    "pairwise_comparisons_pvalues_corrected = []\n",
    "\n",
    "for s in range(pattern_sizes):\n",
    "    print('====================')\n",
    "    print('Size: {}'.format(pattern_sizes_names[s]))\n",
    "    print('====================')\n",
    "    print()\n",
    "    \n",
    "    pvalues = []\n",
    "    \n",
    "    results_merged = np.sum(results[s], axis=1)\n",
    "    \n",
    "    for pairwise_comparison in pairwise_comparisons:\n",
    "        data = np.concatenate(\n",
    "            (\n",
    "                np.sum(results_merged[:,pairwise_comparison], axis=0).reshape((1, -1)),\n",
    "                np.array([[\n",
    "                    np.floor(np.sum(results_merged[:,pairwise_comparison]) / 2).astype(np.int),\n",
    "                    np.ceil(np.sum(results_merged[:,pairwise_comparison]) / 2).astype(np.int)\n",
    "                ]])\n",
    "            ),\n",
    "            axis=0\n",
    "        )\n",
    "\n",
    "        oddsratio, pvalue = stats.fisher_exact(data)\n",
    "\n",
    "        pvalues.append(pvalue)\n",
    "\n",
    "    significance, corrected_pvalues, _, _ = multipletests(pvalues, alpha=0.05, method='holm')\n",
    "    \n",
    "    pairwise_comparisons_pvalues.append(pvalues)\n",
    "    pairwise_comparisons_pvalues_corrected.append(corrected_pvalues)\n",
    "\n",
    "    for i, pairwise_comparison in enumerate(pairwise_comparisons):\n",
    "        print('Comparison : {} vs {}'.format(technique_names[pairwise_comparison[0]], technique_names[pairwise_comparison[1]]))\n",
    "        print('P-value    : {}'.format(pvalues[i]))\n",
    "        print('Corr. P-val: {}'.format(corrected_pvalues[i]))\n",
    "        print('Significant? {}'.format(significance[i]))\n",
    "        print()\n",
    "    \n",
    "    print()\n",
    "    "
   ]
  },
  {
   "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": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}