kmeans_deepwalk2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "from scipy import sparse as sp\n",
    "from sklearn import cluster\n",
    "import sklearn\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('./deepwalk_results/pure_dataset/parsed_embedding.json') as f:\n",
    "    embeddings = json.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n"
     ]
    }
   ],
   "source": [
    "new_K = range(50, 450, 100)\n",
    "new_distortions = []\n",
    "for k in new_K:\n",
    "    kmeanModel = sklearn.cluster.KMeans(k).fit(embeddings)\n",
    "    new_distortions.append(kmeanModel.inertia_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n"
     ]
    }
   ],
   "source": [
    "distortions = []\n",
    "K = range(500, 5001, 500)\n",
    "for k in K:\n",
    "    kmeanModel = sklearn.cluster.KMeans(k).fit(embeddings)\n",
    "    distortions.append(kmeanModel.inertia_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "K = [*new_K, *K]\n",
    "dist = [*new_distortions, *distortions]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABR8AAAK9CAYAAACtshu3AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACJG0lEQVR4nOzdd3SU1fr28WuSQCBAQuggkSpFQkcU6V1EEP0hCnpodkFQbKASSlRARUURUM4BRCkKAioK0osI0psg0ps0QRJCIJDkef/YbyaMCUMSMvMkme9nrVnzlD2Te5JwjlzsvW+HZVmWAAAAAAAAACCT+dldAAAAAAAAAICcifARAAAAAAAAgEcQPgIAAAAAAADwCMJHAAAAAAAAAB5B+AgAAAAAAADAIwgfAQAAAAAAAHgE4SMAAAAAAAAAjyB8BAAAAAAAAOARhI8AAAAAAAAAPILwEQAApGrFihVyOByaPXu23aU4NWvWTM2aNXOeZ8Uas4qePXsqf/78XvlaDodDQ4cOven3GTp0qBwOh/7++++bL+omHDp0SA6HQ1OmTLG1jn8rW7as7rvvPrvLyFQ9e/ZU2bJls/TXzonfdwAAvInwEQAAH+JwONL0WLFihddqSgp6rvcYOXKk12rJbD179pTD4VBwcLAuXbqU4v7evXudn/P9999P9/vHxsZq6NChXv15wbN27dqloUOH6tChQ3aXkmn++usvDR06VFu3brW7FAAAYIMAuwsAAADe8+WXX7qcT506VYsXL05xvWrVqtq9e7c3S1PXrl117733prheu3Ztr9aR2QICAhQbG6sffvhBXbp0cbk3bdo05cmTR5cvX87Qe8fGxmrYsGGS5DIjFDenTJkyunTpknLlyuX1r71r1y4NGzZMzZo1s21GYGb766+/NGzYMJUtW1a1atVyuTdx4kQlJibaUxgAAPAKwkcAAHzIY4895nK+bt06LV68OMV1SV4PH+vUqZNqHdldYGCgGjZsqBkzZqQIH6dPn6727dvr22+/tak6pMbhcChPnjx2l+ET7Ah4AQCAd7HsGgAAuJWYmKi3335bpUuXVp48edSyZUvt27cvxbjffvtN99xzj0JCQhQUFKSmTZtqzZo1XqkxISFBr7/+ukqUKKF8+fKpY8eOOnr0aIpxs2bNUt26dZU3b14VKVJEjz32mI4fP+68//3338vhcGj79u3Oa99++60cDocefPBBl/eqWrWqHn744TTV161bNy1YsEDnz593XtuwYYP27t2rbt26pfqa8+fP64UXXlBYWJgCAwNVsWJFjRo1yjlL7NChQypatKgkadiwYc7l2//ee/H48ePq1KmT8ufPr6JFi+rll19WQkKCy5iLFy/qpZdecn6typUr6/3335dlWS7j4uLi9OKLL6po0aIqUKCAOnbsqGPHjqXpeyBJn3zyiapVq6agoCCFhoaqXr16mj59eqqfvWfPnipYsKBCQkLUq1cvxcbGuoyJj49XZGSkKlSooMDAQJUtW1avv/664uLinGMGDBigwoULu3yO559/Xg6HQx9//LHz2qlTp+RwODR+/Hjn9/bfez4m7aGZlu/n2bNn9Z///EfBwcEqWLCgevTooW3btt1wH8kpU6booYcekiQ1b978utsg/PLLL6pfv77y5Mmj8uXLa+rUqal+D939/tzIuHHjVK1aNQUGBqpUqVLq06ePy++vZGbbhoeHa9OmTbr77ruVN29elStXThMmTHCOWbFihe644w5JUq9evZyfKen78O99F5O+9++//74+/fRTlS9fXkFBQWrTpo2OHj0qy7IUGRmp0qVLK2/evLr//vt17tw5l7q+++47tW/fXqVKlVJgYKAqVKigyMjIFD+nm/HFF18oICBAr7zySqa9JwAAORXhIwAAcGvkyJGaO3euXn75ZQ0aNEjr1q3To48+6jJm2bJlatKkiaKjozVkyBC98847On/+vFq0aKH169en6evExsbq77//TvGIj4+/4Wvffvtt/fjjj3rttdfUr18/LV68WK1atXLZZ3HKlCnq0qWL/P39NWLECD355JOaM2eOGjVq5AxVGjVqJIfDoVWrVjlft3r1avn5+emXX35xXjtz5oz++OMPNWnSJE2f7cEHH5TD4dCcOXOc16ZPn64qVaqoTp06qX4vmjZtqq+++krdu3fXxx9/rIYNG2rQoEEaMGCAJKlo0aLOsOyBBx7Ql19+qS+//NIlJE1ISFDbtm1VuHBhvf/++2ratKlGjx6tzz//3DnGsix17NhRH374oe655x598MEHqly5sl555RXn10ryxBNP6KOPPlKbNm00cuRI5cqVS+3bt0/T92DixInq16+fbr/9dn300UcaNmyYatWqpd9++y3F2C5duujChQsaMWKEunTpoilTpjiXl19bS0REhOrUqaMPP/xQTZs21YgRI/TII484xzRu3Fjnzp3T77//7ryW9PNcvXq1yzVJN/x5puX7mZiYqA4dOmjGjBnq0aOH3n77bZ04cUI9evS44feoSZMm6tevnyTp9ddfd/5Mq1at6hyzb98+de7cWa1bt9bo0aMVGhqqnj17unzGtPz+uDN06FD16dNHpUqV0ujRo/V///d/+uyzz9SmTRtdvXrVZew///yje++9V3Xr1tW7776r0qVL69lnn9WkSZMkmZB++PDhkqSnnnrK+Zlu9L2eNm2axo0bp+eff14vvfSSVq5cqS5duujNN9/UwoUL9dprr+mpp57SDz/8oJdfftnltVOmTFH+/Pk1YMAAjRkzRnXr1lVERIQGDhx4w8+eFp9//rl69eqlgQMH6r333suU9wQAIEezAACAz+rTp491vf8cWL58uSXJqlq1qhUXF+e8PmbMGEuStWPHDsuyLCsxMdG67bbbrLZt21qJiYnOcbGxsVa5cuWs1q1bu63h4MGDlqTrPtauXesc27RpU6tp06Yparzlllus6Oho5/VvvvnGkmSNGTPGsizLunLlilWsWDErPDzcunTpknPc/PnzLUlWRESE81q1atWsLl26OM/r1KljPfTQQ5Yka/fu3ZZlWdacOXMsSda2bdvcfrYePXpY+fLlsyzLsjp37my1bNnSsizLSkhIsEqUKGENGzbM+fnfe+895+siIyOtfPnyWX/++afL+w0cONDy9/e3jhw5YlmWZZ05c8aSZA0ZMiTVry3JGj58uMv12rVrW3Xr1nWez5s3z5JkvfXWWy7jOnfubDkcDmvfvn2WZVnW1q1bLUnWc8895zKuW7du163hWvfff79VrVo1t2OGDBliSbJ69+7tcv2BBx6wChcu7DxPquWJJ55wGffyyy9bkqxly5ZZlmVZp0+ftiRZ48aNsyzLss6fP2/5+flZDz30kFW8eHHn6/r162cVKlTI+fub9DOZPHmyc0xav5/ffvutJcn66KOPnNcSEhKsFi1apHjP1MyaNcuSZC1fvjzFvTJlyliSrFWrVjmvnT592goMDLReeukl57W0/v6k5vTp01bu3LmtNm3aWAkJCc7rY8eOtSRZkyZNcl5r2rSpJckaPXq081pcXJxVq1Ytq1ixYtaVK1csy7KsDRs2XPez9+jRwypTpozzPOl7X7RoUev8+fPO64MGDbIkWTVr1rSuXr3qvN61a1crd+7c1uXLl53XYmNjU3ydp59+2goKCnIZ9++vfT1lypSx2rdvb1mW+d8/h8NhRUZG3vB1AADAYOYjAABwq1evXsqdO7fzvHHjxpKkAwcOSJK2bt3qXD589uxZ54zFixcvqmXLllq1alWalno+9dRTWrx4cYrH7bfffsPXdu/eXQUKFHCed+7cWSVLltRPP/0kSdq4caNOnz6t5557zmUvv/bt26tKlSr68ccfXT5f0ky4CxcuaNu2bXrqqadUpEgR5/XVq1erYMGCCg8Pv2FtSbp166YVK1bo5MmTWrZsmU6ePHndJdezZs1S48aNFRoa6jILtFWrVkpISHCZmXkjzzzzjMt548aNnT87Sfrpp5/k7+/vnHGX5KWXXpJlWVqwYIFznKQU41544YU01VGwYEEdO3ZMGzZsyFDNZ8+eVXR0tEst/57F99JLL0mS8+dZtGhRValSxfn9WrNmjfz9/fXKK6/o1KlT2rt3ryTz80ya9ZqR2q79fi5cuFC5cuXSk08+6bzm5+enPn363PC90+L22293/hmUzGesXLmySw038/uzZMkSXblyRS+88IL8/JL/qvDkk08qODjY5c+KZBoqPf30087z3Llz6+mnn9bp06e1adOmDH/Ohx56SCEhIc7zO++8U5LZtzYgIMDl+pUrV1y2T8ibN6/z+MKFC/r777/VuHFjxcbG6o8//shwTe+++6769++vUaNG6c0338zw+wAA4Gt8uuHMqlWr9N5772nTpk06ceKE5s6dq06dOqXrPSzLci63OXz4sIoUKaLnnntOb7zxhmeKBgDAy2699VaX89DQUElmuaUkZ4DjbllpVFSU83XXc9ttt6lVq1YZqvG2225zOXc4HKpYsaIOHTokSTp8+LAkqXLlyileW6VKFZcl1Y0bN9aECRO0b98+7d+/Xw6HQw0aNHCGkk8++aRWr16thg0buoQzN3LvvfeqQIEC+vrrr7V161bdcccdLjVea+/evdq+fbtzT8d/O336dJq+Zp48eVK8R2hoqPNnJ5nvTalSpVzCW0nOpb5J37vDhw/Lz89PFSpUcBmX2vc0Na+99pqWLFmi+vXrq2LFimrTpo26deumhg0bphjr7ncuODjYWUvFihVdxpUoUUIFCxZ01iyZn2dSWLl69WrVq1dP9erVU6FChbR69WoVL15c27Ztu24QfK20fj9LliypoKAgl3H/rjWj/v29Sa2Gm/n9ud6fldy5c6t8+fIu31tJKlWqlPLly+dyrVKlSpLM/o133XWXm09zff/+nElBZFhYWKrXr/38v//+u958800tW7bMGVgniYqKylA9K1eudG7twD6PAACkj0+HjxcvXlTNmjXVu3fvFJvIp1X//v21aNEivf/++6pevbrOnTuXYtNrAACyM39//1SvW/+/iUfSrMb33ntPtWrVSnVs/vz5PVKbJzRq1EiS+UfKAwcOqE6dOsqXL58aN26sjz/+WDExMdqyZYvefvvtdL1vYGCgHnzwQX3xxRc6cOBAisYw10pMTFTr1q316quvpno/Kdy5kev97OxQtWpV7dmzR/Pnz9fChQv17bffaty4cYqIiEixn+ONfueSpGWmYqNGjTRx4kQdOHBAq1evVuPGjeVwONSoUSOtXr1apUqVUmJiostswuvJCt/PtHxvMuv3x07X+5w3+vznz59X06ZNFRwcrOHDh6tChQrKkyePNm/erNdeey3NDXf+rVq1ajp//ry+/PJLPf300ypXrlyG3gcAAF/k0+Fju3bt1K5du+vej4uL0xtvvKEZM2bo/PnzCg8P16hRo9SsWTNJ0u7duzV+/Hjt3LnT+a/D/IcIAMDXJM2ECw4OzvDMxZuVNPsyiWVZ2rdvn2rUqCFJKlOmjCRpz549atGihcvYPXv2OO9LZsbVrbfeqtWrV+vAgQPOUKpJkyYaMGCAZs2apYSEhDQ3m7lWt27dNGnSJPn5+bk0Rvm3ChUqKCYm5obfz7SEbzdSpkwZLVmyRBcuXHCZ/Zi0PDXpe1OmTBklJiZq//79LrPi9uzZk+avlS9fPj388MN6+OGHdeXKFT344IN6++23NWjQIJfl8GmpOTExUXv37nVpxnLq1CmdP3/e5eeZ9PNbvHixNmzY4Gw60qRJE40fP945c69u3bpp/vo3qm358uWKjY11mf2YWof41GTGzzStvz+pufbPSvny5Z3Xr1y5ooMHD6Z4z7/++ksXL150mf34559/SpKzi3VmfKa0WrFihc6ePas5c+a4/Bk9ePDgTb1vkSJFNHv2bDVq1EgtW7bUL7/8olKlSt1suQAA+AT2fHSjb9++Wrt2rWbOnKnt27froYce0j333OP8C84PP/yg8uXLa/78+SpXrpzKli2rJ554gpmPAACfUrduXVWoUEHvv/++YmJiUtw/c+aMx2uYOnWqLly44DyfPXu2Tpw44fxHxnr16qlYsWKaMGGC4uLinOMWLFig3bt3p+jY3LhxYy1btkzr1693hle1atVSgQIFNHLkSOXNmzdDYVXz5s0VGRmpsWPHqkSJEtcd16VLF61du1Y///xzinvnz593dgBPCreSunVnxL333quEhASNHTvW5fqHH34oh8Ph/B4mPX/88ccu4z766KM0fZ2zZ8+6nOfOnVu33367LMtK0UE5LTWn9rU/+OADSXL5eZYrV0633HKLPvzwQ129etW5zLtx48bav3+/Zs+erbvuustlH8Gb0bZtW129elUTJ050XktMTNSnn36aptcnhXg38zNN6+9Palq1aqXcuXPr448/dplN+b///U9RUVEp/qzEx8frs88+c55fuXJFn332mYoWLer8M5IZnymtkmZGXlv7lStXNG7cuJt+79KlS2vJkiW6dOmSWrduneJ3GgAApM6nZz66c+TIEU2ePFlHjhxx/qvmyy+/rIULF2ry5Ml65513dODAAR0+fFizZs3S1KlTlZCQoBdffFGdO3fWsmXLbP4EAAB4h5+fn/773/+qXbt2qlatmnr16qVbbrlFx48f1/LlyxUcHKwffvjhhu+zefNmffXVVymuV6hQQQ0aNHD72kKFCqlRo0bq1auXTp06pY8++kgVK1Z0Nv3IlSuXRo0apV69eqlp06bq2rWrTp06pTFjxqhs2bJ68cUXXd6vcePGmjZtmnN5rmRCjbvvvls///yzmjVr5tKEJ638/PzS1KjilVde0ffff6/77rtPPXv2VN26dXXx4kXt2LFDs2fP1qFDh1SkSBHlzZtXt99+u77++mtVqlRJhQoVUnh4eLoa4XTo0EHNmzfXG2+8oUOHDqlmzZpatGiRvvvuO73wwgvOma21atVS165dNW7cOEVFRenuu+/W0qVL0zyjr02bNipRooQaNmyo4sWLa/fu3Ro7dqzat2+fYr/JG6lZs6Z69Oihzz//3LnMdv369friiy/UqVMnNW/e3GV848aNNXPmTFWvXt25f2TScvo///wzTfs9plWnTp1Uv359vfTSS9q3b5+qVKmi77//3vmP0zeaBVirVi35+/tr1KhRioqKUmBgoFq0aKFixYqluYa0/v6kpmjRoho0aJCGDRume+65Rx07dtSePXs0btw43XHHHXrsscdcxpcqVUqjRo3SoUOHVKlSJeeepp9//rly5colyfwZLliwoCZMmKACBQooX758uvPOOz2yYujuu+9WaGioevTooX79+snhcOjLL79MsWQ/oypWrKhFixapWbNmatu2rZYtW6bg4OBMeW8AAHIqwsfr2LFjhxISElLsiRMXF6fChQtLMv+KHRcXp6lTpzrH/e9//1PdunW1Z8+eNG/ADgBAdtesWTOtXbvWOasvJiZGJUqU0J133unSCdedGTNmaMaMGSmu9+jR44bh4+uvv67t27drxIgRunDhglq2bKlx48a5LHvt2bOngoKCNHLkSL322mvKly+fHnjgAY0aNUoFCxZ0eb+k2Y5VqlRx/v9+0vWff/45TfsD3oygoCCtXLlS77zzjvMfOYODg1WpUiUNGzbMpQvwf//7Xz3//PN68cUXdeXKFQ0ZMiRd4aOfn5++//57RURE6Ouvv9bkyZNVtmxZvffee87u0UkmTZqkokWLatq0aZo3b55atGihH3/8MUUTkNQ8/fTTmjZtmj744APFxMSodOnS6tevX4a7Bv/3v/9V+fLlNWXKFM2dO1clSpTQoEGDNGTIkBRjk8LHpCBZMl2aGzRooCVLlmTqz9Pf318//vij+vfvry+++EJ+fn564IEHNGTIEDVs2PCGy8tLlCihCRMmaMSIEXr88ceVkJCg5cuXpyt8TM/vT2qGDh2qokWLauzYsXrxxRdVqFAhPfXUU3rnnXecgWKS0NBQffHFF3r++ec1ceJEFS9eXGPHjnXp9p0rVy598cUXGjRokJ555hnFx8dr8uTJHgkfCxcurPnz5+ull17Sm2++qdDQUD322GNq2bKl2rZtmylfo3r16lqwYIFatWqlDh06aOHChS4dtgEAgCuHlVn/DJjNORwOl27XX3/9tR599FH9/vvvKTa2zp8/v0qUKKEhQ4bonXfecVkqdOnSJQUFBWnRokVq3bq1Nz8CAAAAsqh58+bpgQce0C+//JJqh+/sqFmzZvr777+1c+dOu0sBAABZGDMfr6N27dpKSEjQ6dOnr/uv4Q0bNlR8fLz279/vXJKUtMH2tRudAwAAwHdcunTJZSZcQkKCPvnkEwUHB6tOnTo2VgYAAOB9Ph0+xsTEuOxTdPDgQW3dulWFChVSpUqV9Oijj6p79+4aPXq0ateurTNnzmjp0qWqUaOG2rdvr1atWqlOnTrq3bu3PvroIyUmJqpPnz5q3bp1iuXaAAAA8A3PP/+8Ll26pAYNGiguLk5z5szRr7/+qnfeeYfluQAAwOf49LLrFStWpNiQXDJ7S02ZMkVXr17VW2+9palTp+r48eMqUqSI7rrrLg0bNkzVq1eXJP311196/vnntWjRIuXLl0/t2rXT6NGjVahQIW9/HAAAAGQB06dP1+jRo7Vv3z5dvnxZFStW1LPPPqu+ffvaXVqmYtk1AABIC58OHwEAAAAAAAB4jp/dBQAAAAAAAADImQgfAQAAAAAAAHiEzzWcSUxM1F9//aUCBQrI4XDYXQ4AAAAAAACQrViWpQsXLqhUqVLy83M/t9Hnwse//vpLYWFhdpcBAAAAAAAAZGtHjx5V6dKl3Y7xufCxQIECksw3Jzg42OZqAAAAAAAAgOwlOjpaYWFhzpzNHZ8LH5OWWgcHBxM+AgAAAAAAABmUli0NaTgDAAAAAAAAwCMIHwEAAAAAAAB4BOEjAAAAAAAAAI8gfAQAAAAAAADgEYSPAAAAAAAAADyC8BEAAAAAAACARxA+AgAAAAAAAPAIwkcAAAAAAAAAHkH4CAAAAAAAAMAjCB8BAAAAAAAAeAThIwAAAAAAAACPIHwEAAAAAAAA4BGEjwAAAAAAAAA8gvARAAAAAAAAgEcQPgIAAAAAAADwCMJHAAAAAAAAAB5B+AgAAAAAAADAIwgfAQAAAAAAAHgE4SMAAAAAAAAAjyB8BAAAAAAAAOARhI8AAAAAAAAAPILwMYcYOlSKjEz9XmSkuQ8AAAAAAAB4E+FjDuHvL0VEpAwgIyPNdX9/e+oCAAAAAACA7wqwuwBkjsGDzXNEhPTnn1LXrtKmTeZ8+PDk+wAAAAAAAIC3OCzLsuwuwpuio6MVEhKiqKgoBQcH211OpnvoIWn2bMnhkCyL4BEAAAAAAACZKz35Gsuuc5gRI8yzZUm5cxM8AgAAAAAAwD6EjznM9OnJx1euXL8JDQAAAAAAAOBphI85SGSkNGSIVL68Oe/QIfUmNAAAAAAAAIA3ED7mEEldrYcPlx5+2FwrVsycE0ACAAAAAADADnS7ziESEpKby8yda65t2CD997/J9wEAAAAAAABvott1DnTsmBQWJvn7S9HRUlCQ3RUBAAAAAAAgp6DbtY+75RapeHEz23HrVrurAQAAAAAAgK+yNXwcMWKE7rjjDhUoUEDFihVTp06dtGfPHrevmThxoho3bqzQ0FCFhoaqVatWWr9+vZcqzh4cDumOO8zxhg321gIAAAAAAADfZWv4uHLlSvXp00fr1q3T4sWLdfXqVbVp00YXL1687mtWrFihrl27avny5Vq7dq3CwsLUpk0bHT9+3IuVZ32EjwAAAAAAALBbltrz8cyZMypWrJhWrlypJk2apOk1CQkJCg0N1dixY9W9e/cbjveFPR8lacEC6d57pUqVpBtMJgUAAAAAAADSLD35Wpbqdh0VFSVJKlSoUJpfExsbq6tXr173NXFxcYqLi3OeR0dH31yR2US9eub5zz+lqCgpJMTeegAAAAAAAOB7skzDmcTERL3wwgtq2LChwsPD0/y61157TaVKlVKrVq1SvT9ixAiFhIQ4H2FhYZlVcpZWtKhUpow53rTJ3loAAAAAAADgm7JM+NinTx/t3LlTM2fOTPNrRo4cqZkzZ2ru3LnKkydPqmMGDRqkqKgo5+Po0aOZVXKWx76PAAAAAAAAsFOWCB/79u2r+fPna/ny5SpdunSaXvP+++9r5MiRWrRokWrUqHHdcYGBgQoODnZ5+ArCRwAAAAAAANjJ1j0fLcvS888/r7lz52rFihUqV65cml737rvv6u2339bPP/+sekmbGyKFpG/Nxo321gEAAAAAAADfZOvMxz59+uirr77S9OnTVaBAAZ08eVInT57UpUuXnGO6d++uQYMGOc9HjRqlwYMHa9KkSSpbtqzzNTExMXZ8hCytbl3zfPiwdOaMvbUAAAAAAADA99gaPo4fP15RUVFq1qyZSpYs6Xx8/fXXzjFHjhzRiRMnXF5z5coVde7c2eU177//vh0fIUsLCZEqVzbHLL0GAAAAAACAt9m+7PpGVqxY4XJ+6NAhzxSTQ91xh7Rnjwkf773X7moAAAAAAADgS7JEwxl4TlLTGfZ9BAAAAAAAgLcRPuZwSU1nNmyQ0jDRFAAAAAAAAMg0hI85XK1akr+/dOqUdOyY3dUAAAAAAADAlxA+5nBBQVJ4uDmm6QwAAAAAAAC8ifDRByTt+0j4CAAAAAAAAG8ifPQBSfs+0nQGAAAAAAAA3kT46AOu7XhN0xkAAAAAAAB4C+GjD6heXQoMlM6fl/bts7saAAAAAAAA+ArCRx+QK5fpei2x7yMAAAAAAAC8h/DRR7DvIwAAAAAAALyN8NFH0PEaAAAAAAAA3kb46COSwsfNm6X4eHtrAQAAAAAAgG8gfPQRlStL+fNLsbHS7t12VwMAAAAAAABfQPjoI/z9pTp1zDH7PgIAAAAAAMAbCB99CPs+AgAAAAAAwJsIH30I4SMAAAAAAAC8ifDRhySFj9u2SXFx9tYCAAAAAACAnI/w0YeUKycVKiRdvSrt2GF3NQAAAAAAAMjpCB99iMMh1atnjll6DQAAAAAAAE8jfPQx7PsIAAAAAAAAbyF89DGEjwAAAAAAAPAWwkcfkxQ+7tolXbxoby0AAAAAAADI2QgffUypUlLJklJiorRli93VAAAAAAAAICcjfPRBLL0GAAAAAACANxA++iDCRwAAAAAAAHgD4aMPSgofN260tw4AAAAAAADkbISPPqhuXfO8d690/rytpQAAAAAAACAHI3z0QUWKSOXKmWNmPwIAAAAAAMBTCB99FPs+AgAAAAAAwNMIH30U4SMAAAAAAAA8jfDRR9WrZ55Zdg0AAAAAAABPIXz0UXXrSg6HdPSodOqU3dUAAAAAAAAgJyJ89FEFCkhVqphjll4DAAAAAADAEwgffRj7PgIAAAAAAMCTCB99WFL4yL6PAAAAAAAA8ATCRx+W1HRmwwbJsuytBQAAAAAAADkP4aMPq1VLCgiQzpyRjhyxuxoAAAAAAADkNISPPixPHql6dXPMvo8AAAAAAADIbISPPo59HwEAAAAAAOAphI8+7tp9HwEAAAAAAIDMRPjo466d+ZiYaG8tAAAAAAAAyFkIH31ctWpm78foaGnvXrurAQAAAAAAQE5C+OjjcuWSatc2x+z7CAAAAAAAgMxE+Aj2fQQAAAAAAIBHED7Cue8j4SMAAAAAAAAyE+EjnOHjli1SfLy9tQAAAAAAACDnIHyEKlWSChSQLl2Sdu2yuxoAAAAAAADkFISPkJ+fVLeuOWbpNQAAAAAAADIL4SMkse8jAAAAAAAAMh/hIyQRPgIAAAAAACDzET5CUnL4uGOHdPmyvbUAAAAAAAAgZyB8hCSpTBmpSBHp6lVp+3a7qwEAAAAAAEBOQPgISZLDIdWrZ45Zeg0AAAAAAIDMQPgIJ/Z9BAAAAAAAQGYifIQT4SMAAAAAAAAyE+EjnJKWXe/eLcXE2FsLAAAAAAAAsj/CRziVLCndcotkWdLmzXZXAwAAAAAAgOyO8BEuWHoNAAAAAACAzEL4CBeEjwAAAAAAAMgshI9wkRQ+btxobx0AAAAAAADI/ggf4aJuXfO8f7907py9tQAAAAAAACB7I3yEi0KFpAoVzDGzHwEAAAAAAHAzCB+RAvs+AgAAAAAAIDMQPiIF9n0EAAAAAABAZiB8RAr16plnZj4CAAAAAADgZhA+IoU6dSQ/P+n4cenECburAQAAAAAAQHZF+IgU8ueXqlY1x8x+BAAAAAAAQEYRPiJV7PsIAAAAAACAm0X4iFSx7yMAAAAAAABuFuEjUpU083HDBsmy7K0FAAAAAAAA2RPhI1JVs6aUK5d09qx06JDd1QAAAAAAACA7sjV8HDFihO644w4VKFBAxYoVU6dOnbRnz54bvm7WrFmqUqWK8uTJo+rVq+unn37yQrW+JTBQqlHDHLPvIwAAAAAAADLC1vBx5cqV6tOnj9atW6fFixfr6tWratOmjS5evHjd1/z666/q2rWrHn/8cW3ZskWdOnVSp06dtHPnTi9W7huuXXoNAAAAAAAApJfDsrLOjn5nzpxRsWLFtHLlSjVp0iTVMQ8//LAuXryo+fPnO6/dddddqlWrliZMmHDDrxEdHa2QkBBFRUUpODg402rPif73P+mJJ6RmzaTly+2uBgAAAAAAAFlBevK1LLXnY1RUlCSpUKFC1x2zdu1atWrVyuVa27ZttXbt2lTHx8XFKTo62uWBtEma+bhpk5SYaG8tAAAAAAAAyH6yTPiYmJioF154QQ0bNlR4ePh1x508eVLFixd3uVa8eHGdPHky1fEjRoxQSEiI8xEWFpapdedkt98u5c0rXbgg/fmn3dUAAAAAAAAgu8ky4WOfPn20c+dOzZw5M1Pfd9CgQYqKinI+jh49mqnvn5MFBEh16phj9n0EAAAAAABAemWJ8LFv376aP3++li9frtKlS7sdW6JECZ06dcrl2qlTp1SiRIlUxwcGBio4ONjlgbSrV888Ez4CAAAAAAAgvWwNHy3LUt++fTV37lwtW7ZM5cqVu+FrGjRooKVLl7pcW7x4sRo0aOCpMn0aHa8BAAAAAACQUQF2fvE+ffpo+vTp+u6771SgQAHnvo0hISHKmzevJKl79+665ZZbNGLECElS//791bRpU40ePVrt27fXzJkztXHjRn3++ee2fY6cLCl83LpVunpVypXL1nIAAAAAAACQjdg683H8+PGKiopSs2bNVLJkSefj66+/do45cuSITpw44Ty/++67NX36dH3++eeqWbOmZs+erXnz5rltUoOMq1hRCgmRLl+Wfv/d7moAAAAAAACQnTgsy7LsLsKboqOjFRISoqioKPZ/TKOWLaVly6TPP5eefNLuagAAAAAAAGCn9ORrWaLhDLI29n0EAAAAAABARhA+4oYIHwEAAAAAAJARhI+4oaTwcedO6dIle2sBAAAAAABA9kH4iBsKC5OKFpXi46Vt2+yuBgAAAAAAANkF4SNuyOFg6TUAAAAAAADSj/ARaUL4CAAAAAAAgPQifMQNDR0q7dtnjjdudL0XGWnuAwAAAAAAAP9G+Igb8veXpk0zx3/8IV24YI4jI6WICHMfAAAAAAAA+LcAuwtA1jd4sHmOiJAsS9q0SVq92pwPH558HwAAAAAAALiWw7Isy+4ivCk6OlohISGKiopScHCw3eVkK7ffLu3ebWY6JiQQPAIAAAAAAPii9ORrLLtGmj3xhHlOSJBy5yZ4BAAAAAAAgHuEj0iz06eTj69cMXs+AgAAAAAAANdD+Ig0iYyURo2S6tQx59WqmT0fCSABAAAAAABwPYSPuKGkrtbDh0uTJ5trf/whvfQSASQAAAAAAACuj/ARN3Rtc5kaNaRmzcy1gABzPSHB7goBAAAAAACQFdHtGuk2b570wANSaKh07JgUFGR3RQAAAAAAAPAWul3Dozp0kMqVk/75R/rqK7urAQAAAAAAQFZF+Ih08/eX+vY1xx9/LPnW3FkAAAAAAACkFeEjMqR3bylfPun336Vly+yuBgAAAAAAAFkR4SMypGBBqUcPczxmjK2lAAAAAAAAIIsifESG9etnnufPl/bvt7cWAAAAAAAAZD2Ej8iwypWle+4xez5+8ond1QAAAAAAACCrIXzETenf3zxPmiRFR9tbCwAAAAAAALIWwkfclDZtzAzICxekKVPsrgYAAAAAAABZCeEjboqfX/Lej598IiUm2lsPAAAAAAAAsg7CR9y07t2lkBBp3z7pp5/srgYAAAAAAABZBeEjblr+/NITT5jjjz+2txYAAAAAAABkHYSPyBR9+5ol2IsXS7t22V0NAAAAAAAAsgLCR2SKsmWl++83x8x+BAAAAAAAgET4iEyU1Hhm6lTp3Dl7awEAAAAAAID9CB+RaZo2lWrUkC5dkv77X7urAQAAAAAAgN0IH5FpHA6pf39zPHasFB9vbz0AAAAAAACwF+EjMlW3blKRItLRo9K8eXZXAwAAAAAAADsRPiJT5ckjPf20OR4zxt5aAAAAAAAAYC/CR2S6556TAgKkX36RNm+2uxoAAAAAAADYhfARma5UKemhh8wxsx8BAAAAAAB8F+EjPCKp8czMmdKpU/bWAgAAAAAAAHsQPsIj7rzTPK5ckT77zO5qAAAAAAAAYAfCR3hMv37mefx4E0ICAAAAAADAtxA+wmM6d5ZKlpROnpS++cbuagAAAAAAAOBthI/wmNy5TedryTSesSx76wEAAAAAAIB3ET7Co55+WgoMlDZulNautbsaAAAAAAAAeBPhIzyqaFGpWzdzPGaMvbUAAAAAAADAuwgf4XH9+5vnb7+Vjh61txYAAAAAAAB4D+EjPK5mTalpUykhQRo3zu5qAAAAAAAA4C2Ej/CKpNmPn38uXbpkby0AAAAAAADwDsJHeEXHjlLZstK5c9K0aXZXAwAAAAAAAG8gfIRX+PtLffua4zFjJMuytx4AAAAAAAB4HuEjvKZ3bykoSNq5U1q+3O5qAAAAAAAA4GmEj/Ca0FCpRw9zPGaMvbUAAAAAAADA8wgf4VX9+pnnH36Q9u+3txYAAAAAAAB4FuEjvKpKFaltW7Pn49ixdlcDAAAAAAAATyJ8hNf172+eJ02SLlywtxYAAAAAAAB4DuEjvK5tW6lSJSk6Wpoyxe5qAAAAAAAA4CmEj/A6P7/kvR8/+URKTLS3HgAAAAAAAHgG4SNs0aOHFBIi7d0rLVxodzUAAAAAAADwBMJH2CJ/funxx83xmDH21gIAAAAAAADPIHyEbfr2lRwOadEiafduu6sBAAAAAABAZiN8hG3KlZM6djTHH39sby0AAAAAAADIfISPsM3QoVKhQuZ46lTpn3+S70VGmvsAAAAAAADIvggfYRt/f2nyZKl4cSk2Vvrvf831yEgpIsLcBwAAAAAAQPYVYHcB8F2DB5vniAjzPHasCSGHDpWGD0++DwAAAAAAgOyJmY+w1eDBySHjkSMEjwAAAAAAADmJw7Isy+4ivCk6OlohISGKiopScHCw3eXg/wsIkBISzPG5c1JoqL31AAAAAAAAIHXpydeY+QjbRUaa4NHhMOdt29pbDwAAAAAAADIH4SNsldRcZvhwaflyc23DBunpp+2tCwAAAAAAADeP8BG2uTZ4HDxYatpU6tHD3Pv8c2nYMHvrAwAAAAAAwM0hfIRtEhJSNpd5773k/R7XrrWnLgAAAAAAAGQOGs4gy5k4UXrqKSl/fumPP6RbbrG7IgAAAAAAACSh4Qyytccflxo0kGJipBdesLsaAAAAAAAAZBThI7IcPz9p/HjJ31+aPVtasMDuigAAAAAAAJARtoaPq1atUocOHVSqVCk5HA7Nmzfvhq+ZNm2aatasqaCgIJUsWVK9e/fW2bNnPV8svKpmTal/f3Pct6906ZK99QAAAAAAACD9bA0fL168qJo1a+rTTz9N0/g1a9aoe/fuevzxx/X7779r1qxZWr9+vZ588kkPVwo7DB0qlS4tHTggvfOO3dUAAAAAAAAgvQLs/OLt2rVTu3bt0jx+7dq1Klu2rPr16ydJKleunJ5++mmNGjXKUyXCRgUKSB9/LD34oDRqlPToo1KVKnZXBQAAAAAAgLTKVns+NmjQQEePHtVPP/0ky7J06tQpzZ49W/fee+91XxMXF6fo6GiXB7KPTp2k9u2lq1el556TfKs3OwAAAAAAQPaWrcLHhg0batq0aXr44YeVO3dulShRQiEhIW6XbY8YMUIhISHOR1hYmBcrxs1yOKRPPpHy5pWWL5emTbO7IgAAAAAAAKRVtgofd+3apf79+ysiIkKbNm3SwoULdejQIT3zzDPXfc2gQYMUFRXlfBw9etSLFSMzlCsnDR5sjgcMkP75x956AAAAAAAAkDYOy8oaC1kdDofmzp2rTp06XXfMf/7zH12+fFmzZs1yXvvll1/UuHFj/fXXXypZsuQNv050dLRCQkIUFRWl4ODgzCgdXnDlilSrlrR7t/TMM9L48XZXBAAAAAAA4JvSk69lq5mPsbGx8vNzLdnf31+SlEUyVHhI7tzShAnm+LPPpN9+s7ceAAAAAAAA3Jit4WNMTIy2bt2qrVu3SpIOHjyorVu36siRI5LMkunu3bs7x3fo0EFz5szR+PHjdeDAAa1Zs0b9+vVT/fr1VapUKTs+AryoSROpRw/TdOaZZ6T4eLsrAgAAAAAAgDu2ho8bN25U7dq1Vbt2bUnSgAEDVLt2bUVEREiSTpw44QwiJalnz5764IMPNHbsWIWHh+uhhx5S5cqVNWfOHFvqh/e9954UGipt3SqNHWt3NQAAAAAAAHAny+z56C3s+Zj9TZwoPfWUlD+/9Mcf0i232F0RAAAAAACA78ixez4CkvT441KDBlJMjPTCC3ZXAwAAAAAAgOshfES24+dnms/4+0uzZ0sLFthdEQAAAAAAAFJD+IhsqUaN5FmPffpIly7ZWg4AAAAAAABSQfiIbGvoUKl0aengQentt+2uBgAAAAAAAP9G+IhsK39+6eOPzfG775rmMwAAAAAAAMg6CB+RrXXqJN13n3T1qvTss5Jv9W4HAAAAAADI2ggfka05HNInn0h580orVkhffWV3RQAAAAAAAEhC+Ihsr2xZKSLCHL/0kvTPP7aWAwAAAAAAgP+P8BE5woAB0u23S2fOSIMG2V0NAAAAAAAAJMJH5BC5c0vjx5vjzz+X1q2ztx4AAAAAAAAQPiIHadJE6tnTNJ155hkpPt7uigAAAAAAAHwb4SNylHfflUJDpW3bpLFj7a4GAAAAAADAtxE+IkcpWtQEkJI0eLB07Ji99QAAAAAAAPgywkfkOL17Sw0aSDEx0gsv2F0NAAAAAACA7yJ8RI7j5yfVrCk5HNK330o//eR6PzJSGjrUltIAAAAAAAB8CuEjcqRSpUzjGUnq21eKjTXHkZFSRITk729fbQAAAAAAAL4iwO4CAE8YPFi6ckV66y3p4EHpnXekwEATPA4fbu4DAAAAAADAsxyWlTQ/zDdER0crJCREUVFRCg4OtrsceFi3btKMGcnnBI8AAAAAAAA3Jz35GsuukaNNm2b2gEzSvr19tQAAAAAAAPgawkfkaG+9JSUmmuYzktSkibR/v701AQAAAAAA+ArCR+RYSc1lhg+X/vlHKlFCunhRuuMO6eRJu6sDAAAAAADI+QgfkSNdGzwOHiyFhEhbtkgFC5ogsnZtKTra7ioBAAAAAAByNsJH5EgJCSmby5QoIW3YIAUFmZmPnTpJcXG2lQgAAAAAAJDj0e0aPmfTJqlZMykmRurcWZo5U/L3t7sqAAAAAACA7IFu14AbdetKc+dKuXJJs2dL/ftLvhXBAwAAAAAAeAfhI3xSq1bSV1+ZLtiffiq9/bbdFQEAAAAAAOQ8hI/wWV26SGPGmOPBg6WJE+2tBwAAAAAAIKchfIRPe/556Y03zPEzz0jz5tlaDgAAAAAAQI5C+AifFxkpPf64lJgoPfKItGqV3RUBAAAAAADkDISP8HkOhzRhgtSxoxQXZ5537LC7KgAAAAAAgOyP8BGQFBAgzZwpNWokRUVJbdtKhw7ZXRUAAAAAAED2RvgI/H9580rffy9VqyadOGECyL//trsqAAAAAACA7IvwEbhGaKj088/SrbdKf/4ptW8vxcTYXRUAAAAAAED2RPgI/Mstt5gAslAhaf16qXNn6epVu6sCAAAAAADIfggfgVRUqSL99JMUFGSCyN69TTdsAAAAAAAApB3hI3Add94pzZ4t+ftLX30lvfqq3RUBAAAAAABkL4SPgBvt2kmTJpnj0aOl996ztx4AAAAAAIDshPARuIHu3ZNDx1dflb74wt56AAAAAAAAsgvCRyANXn5Zeuklc/z449KPP9pbDwAAAAAAQHZA+Aik0bvvSo89JiUkSA89JK1bZ3dFAAAAAAAAWRvhI5BGfn5m/8d77pEuXZLat5d277a7KgAAAAAAgKyL8BFIh1y5TAfs+vWlc+ektm2lY8fsrgoAAAAAACBrInwE0ilfPrPnY+XK0tGjJoA8d87uqgAAAAAAALIewkcgA4oUkX7+WSpVStq1S+rYUYqNtbsqAAAAAACArIXwEcigMmVMAFmwoLRmjfTII1J8vN1VAQAAAAAAZB2Ej8BNCA+Xvv9eypNH+uEH6emnJcuyuyoAAAAAAICsgfARuEmNG0szZyZ3w37zTbsrAgAAAAAAyBoIH4FMcP/90mefmeN33pE+/tjeegAAAAAAALICwkcgkzzxhPTWW+b4hRfMbEgAAAAAAABfRvgIZKLXX5f69jX7PnbvLi1ZYndFAAAAAAAA9iF8BDKRwyF99JHUpYt09ar0wAPSpk12VwUAAAAAAGAPwkcgk/n7S1OnSi1aSDExUpMm0t69KcdFRkpDh3q9PAAAAAAAAK8hfAQ8IDBQmjtXKllSio2V6teXTpxIvh8ZKUVEmKASAAAAAAAgpyJ8BDwkOFjaskUKDZXOn5fq1JGiopKDx+HDpcGD7a4SAAAAAADAcwLsLgDIyYoXlzZskGrWlE6eNEGkZRE8AgAAAAAA38DMR8DDKlSQVq82x5Zlnhs2tK8eAAAAAAAAbyF8BLxg/nzX85YtpX79zH6QAAAAAAAAORXhI+Bh1+7xeOGCdMcd5vonn0i1a0u//WZvfQAAAAAAAJ5C+Ah40L+by+TPL61fL/3nP+b+n39Kd98tvfGGdOWKvbUCAAAAAABkNsJHwIMSElJvLjN1qjRokBQeLiUmSu+8I9WvL23fbk+dAAAAAAAAnuCwrKQWGL4hOjpaISEhioqKUnBwsN3lAJo9W3rmGensWSlXLjNb8uWXJX9/uysDAAAAAABIKT35GjMfAZt17izt3Cl16CBdvSoNHCg1aSLt22d3ZQAAAAAAADeH8BHIAkqUkL77Tpo0SSpQQPr1V6lmTWncOMm35iYDAAAAAICchPARyCIcDqlXL2nHDqlZMyk2VurTR7rnHunYMburAwAAAAAASD/CRyCLKVNGWrpU+ugjKU8eadEi05jmq6+YBQkAAAAAALIXwkcgC/Lzk/r3l7Zske64Q4qKkv7zH7M/5JkzdlcHAAAAAACQNoSPQBZWpYrZ/zEyUgoIkObMMbMgv/vO7soAAAAAAABujPARyOICAqQ335TWrzfB4+nTUqdOUs+eZkYkAAAAAABAVpWh8PHUqVP6z3/+o1KlSikgIED+/v4uDwCZr3ZtaeNG6dVXTXOaL76QqleXli2zuzIAAAAAAIDUBWTkRT179tSRI0c0ePBglSxZUg6HI7PrApCKwEBp1CipQwepRw/pwAGpZUvp+eelkSOloCC7KwQAAAAAAEjmsKz0988tUKCAVq9erVq1anmgJM+Kjo5WSEiIoqKiFBwcbHc5QIbFxEivvCJNmGDOK1UysyHvusveugAAAAAAQM6WnnwtQ8uuw8LClIHMMoVVq1apQ4cOKlWqlBwOh+bNm3fD18TFxemNN95QmTJlFBgYqLJly2rSpEk3XQuQ3eTPL40fLy1YIJUqJf35p9SwofTGG9KVK3ZXBwAAAAAAkMHw8aOPPtLAgQN16NChm/riFy9eVM2aNfXpp5+m+TVdunTR0qVL9b///U979uzRjBkzVLly5ZuqA8jO7rlH2rlTevRRKTFReucdqX59aft2uysDAAAAAAC+LkPLrkNDQxUbG6v4+HgFBQUpV65cLvfPnTuX/kIcDs2dO1edOnW67piFCxfqkUce0YEDB1SoUKF0fw2JZdfI2WbPlp55Rjp7VsqVS4qMlF5+WaIPFAAAAAAAyCzpydcy1HDmo48+ysjLbtr333+vevXq6d1339WXX36pfPnyqWPHjoqMjFTevHlTfU1cXJzi4uKc59HR0d4qF/C6zp2lRo2kp56SfvhBGjhQ+v57sxdkxYp2VwcAAAAAAHxNhsLHHj16ZHYdaXLgwAH98ssvypMnj+bOnau///5bzz33nM6ePavJkyen+poRI0Zo2LBhXq4UsE+JEtJ330lTpkj9+0u//irVrCm995707LMSzekBAAAAAIC3ZGjZtSQlJCRo3rx52r17tySpWrVq6tixo/wzuL4zLcuu27Rpo9WrV+vkyZMKCQmRJM2ZM0edO3fWxYsXU539mNrMx7CwMJZdwyccPiz17CmtWGHOW7eWJk2SSpe2syoAAAAAAJCdebzb9b59+1S1alV1795dc+bM0Zw5c/TYY4+pWrVq2r9/f4aKTouSJUvqlltucQaPklS1alVZlqVjx46l+prAwEAFBwe7PABfUaaMtHSp9NFHUp480uLFUni49OWXUiY0rAcAAAAAAHArQ+Fjv379VKFCBR09elSbN2/W5s2bdeTIEZUrV079+vXL7BqdGjZsqL/++ksxMTHOa3/++af8/PxUmqlcQKr8/Mzy6y1bTBfsqCipe3ezP+SZM3ZXBwAAAAAAcrIMhY8rV67Uu+++69JxunDhwho5cqRWrlyZ5veJiYnR1q1btXXrVknSwYMHtXXrVh05ckSSNGjQIHXv3t05vlu3bipcuLB69eqlXbt2adWqVXrllVfUu3fv6zacAWBUqSKtWWM6YAcESHPmmFmQ331nd2UAAAAAACCnylD4GBgYqAsXLqS4HhMTo9y5c6f5fTZu3KjatWurdu3akqQBAwaodu3aioiIkCSdOHHCGURKUv78+bV48WKdP39e9erV06OPPqoOHTro448/zsjHAHxOQID05pvS+vUmeDx9WurUyewLGRVld3UAAAAAACCnyVDDme7du2vz5s363//+p/r160uSfvvtNz355JOqW7eupkyZktl1Zpr0bIgJ5GRxcVJEhOmCbVlSWJg0ebLUsqXdlQEAAAAAgKzM4w1nPv74Y1WoUEENGjRQnjx5lCdPHjVs2FAVK1bUmDFjMlQ0AO8KDJRGjZJWrZLKl5eOHpVatZKef16KjbW7OgAAAAAAkBNkaOZjkr179+qPP/6QZLpOV6xYMdMK8xRmPgIpxcRIr7wiTZhgzitVkr74QrrrLnvrAgAAAAAAWU968rWbCh+zI8JH4Pp+/lnq3Vv66y/TJXvgQGnIECkdW7kCAAAAAIAcziPh44ABAxQZGal8+fJpwIABbsd+8MEHaa/WywgfAff++ccsvZ42zZzXrClNnSrVqGFvXQAAAAAAIGtIT74WkNY33bJli65eveo8BpAzhYZKX31lumA/84y0bZtUr54UGSm9/LLk7293hQAAAAAAILtg2TWA6zp5UnrqKemHH8z53XebvSCzwfauAAAAAADAQzze7bp37966cOFCiusXL15U7969M/KWALKgEiWk776TJk2SChSQfv3VLMMeN07yrX+2AAAAAAAAGZGhmY/+/v46ceKEihUr5nL977//VokSJRQfH59pBWY2Zj4CGXP4sNSzp7RihTlv3dqEkqVL21kVAAAAAADwNo/NfIyOjlZUVJQsy9KFCxcUHR3tfPzzzz/66aefUgSSAHKGMmWkpUulMWOkPHmkxYul8HDpyy+ZBQkAAAAAAFKX5oYzklSwYEE5HA45HA5VqlQpxX2Hw6Fhw4ZlWnEAshY/P6lfP6lNG6lHD2n9eql7d2nePGnCBKloUbsrBAAAAAAAWUm6ll2vXLlSlmWpRYsW+vbbb1WoUCHnvdy5c6tMmTIqVaqURwrNLCy7BjJHfLw0cqQ0bJg5LlZM+vxz6f777a4MAAAAAAB4UnrytXTv+RgfH68nn3xSw4cPV1hY2E0VagfCRyBzbdliZj/u3GnOe/QwS7NDQuytCwAAAAAAeIZHu10HBARo9uzZSkhIyHCBAHKO2rWljRulV1+VHA7piy+k6tXN/pAAAAAAAMC3pTt8lKQWLVpo5cqVmV0LgGwqMFAaNUpatUoqX146elRq1Up6/nkpNtbu6gAAAAAAgF3S1XAmSbt27TRw4EDt2LFDdevWVb58+Vzud+zYMVOKA5C9NGokbdsmvfKKaUAzdqy0aJGZDXnXXXZXBwAAAAAAvC3dez5Kkp/f9SdMOhyOLL0kmz0fAe/4+Wepd2/pr79Ml+yBA6UhQ6Tcue2uDAAAAAAA3AyP7vkoSYmJidd9ZOXgEYD3tG1rmtA8+qiUmCi9845Uv760fbvdlQEAAAAAAG/JUPgIAGkRGip99ZU0a5ZUuLBZkl2vnjRypMS/UwAAAAAAkPNlOHxcuXKlOnTooIoVK6pixYrq2LGjVq9enZm1AcghOnc2syA7dJCuXpUGDZIaN5b27rW7MgAAAAAA4EkZCh+/+uortWrVSkFBQerXr5/69eunvHnzqmXLlpo+fXpm1wggByhRQvruO2nSJKlAAWntWqlWLWncOCn9O88CAAAAAIDsIEMNZ6pWraqnnnpKL774osv1Dz74QBMnTtTu3bszrcDMRsMZwH6HD0u9eknLl5vz1q1NKFm6tL11AQAAAACAG/N4w5kDBw6oQ4cOKa537NhRBw8ezMhbAvAhZcpIS5ZIY8ZIefJIixdL4eHSl18yCxIAAAAAgJwkQ+FjWFiYli5dmuL6kiVLFBYWdtNFAcj5/Pykfv2kLVtMF+yoKKl7d+n//k86c8bu6gAAAAAAQGYIyMiLXnrpJfXr109bt27V3XffLUlas2aNpkyZojFjxmRqgQBytipVpDVrTAfsYcOkuXOlX36RPv9c6tTJ7uoAAAAAAMDNyNCej5I0d+5cjR492rm/Y9WqVfXKK6/o/vvvz9QCMxt7PgJZ15YtZvbjzp3mvEcPszQ7JMTeugAAAAAAQLL05GsZDh+zK8JHIGuLi5MiIqT33jP7P4aFSZMnSy1b2l0ZAAAAAACQvNBwpnz58jp79myK6+fPn1f58uUz8pYAIEkKDJRGjZJWrZLKl5eOHpVatZKef16KjbW7OgAAAAAAkB4ZCh8PHTqkhISEFNfj4uJ0/Pjxmy4KABo1krZtk5591pyPHSvVqiWtW2drWQAAAAAAIB3S1XDm+++/dx7//PPPCrlmI7aEhAQtXbpUZcuWzbTiAPi2/PmlceOk+++XeveW9u6VGjaUXntNGjpUyp3b7goBAAAAAIA76drz0c/PTJR0OBz698ty5cqlsmXLavTo0brvvvsyt8pMxJ6PQPb0zz9m6fW0aea8Zk1p6lSpRg176wIAAAAAwNd4bM/HxMREJSYm6tZbb9Xp06ed54mJiYqLi9OePXuydPAIIPsKDZW++kqaNUsqXNgsya5XTxo5UkplFwgAAAAAAJAFZGjPx4MHD6pIkSIu186fP58Z9QCAW507Szt3Sh06SFevSoMGSY0bmyXZAAAAAAAga8lQ+Dhq1Ch9/fXXzvOHHnpIhQoV0i233KJt27ZlWnEAkJoSJaTvvpMmTZIKFJDWrjXNaMaNk9K+kQQAAAAAAPC0DIWPEyZMUFhYmCRp8eLFWrJkiRYuXKh27drplVdeydQCASA1DofUq5e0Y4fUvLkUGyv16SO1bSsdPWp3dQAAAAAAQMpg+Hjy5Eln+Dh//nx16dJFbdq00auvvqoNGzZkaoEA4E6ZMtKSJdKYMVKePNLixVL16tKXXzILEgAAAAAAu2UofAwNDdXR/z+1aOHChWrVqpUkybIsJdD5AYCX+flJ/fpJW7ZI9etLUVFS9+7S//2fdOaM3dUBAAAAAOC7MhQ+Pvjgg+rWrZtat26ts2fPql27dpKkLVu2qGLFiplaIACkVZUq0po1UmSkFBAgzZ0rVasmzZtnd2UAAAAAAPimDIWPH374ofr27avbb79dixcvVv78+SVJJ06c0HPPPZepBQJAegQESG++Ka1fL4WHm5mPDzwg9expZkQCAAAAAADvcViWb+2KFh0drZCQEEVFRSk4ONjucgB4UFycFBEhvfee2f8xLEyaPFlq2dLuygAAAAAAyL7Sk6+lOXz8/vvv1a5dO+XKlUvff/+927EdO3ZMe7VeRvgI+J41a6QePaT9+815377SqFFSUJC9dQEAAAAAkB15JHz08/PTyZMnVaxYMfn5XX+1tsPhyNJNZwgfAd8UEyO9+qo0frw5v+02aepU6a677K0LAAAAAIDsJj35Wpr3fExMTFSxYsWcx9d7ZOXgEYDvyp9fGjdOWrhQKlVK2rtXathQev116coVu6sDAAAAACBnSnfDmcTERE2aNEn33XefwsPDVb16dd1///2aOnWqfGz7SADZUNu20s6d0qOPSomJ0ogRUv360vbtdlcGAAAAAEDOk67w0bIsdezYUU888YSOHz+u6tWrq1q1ajp06JB69uypBx54wFN1AkCmCQ2VvvpKmjVLKlxY2rZNqldPGjlSYvI2AAAAAACZJ13h45QpU7Rq1SotXbpUW7Zs0YwZMzRz5kxt27ZNS5Ys0bJlyzR16lRP1QoAmapzZzMLskMH6epVadAgqXFjsyQbAAAAAADcvHSFjzNmzNDrr7+u5s2bp7jXokULDRw4UNOmTcu04gDA00qUkL77Tpo8WSpQQFq7VqpVS/r0U7MsGwAAAAAAZFy6wsft27frnnvuue79du3aadu2bTddFAB4k8Mh9ewp7dghNW8uxcZKffua/SGPHrW7OgAAAAAAsq90hY/nzp1T8eLFr3u/ePHi+ueff266KACwQ5ky0pIl0pgxUp485rh6demBB6Thw1N/TWSkNHSoV8sEAAAAACDbSFf4mJCQoICAgOve9/f3V3x8/E0XBQB28fOT+vWTtmwxXbCjoqR586QhQ6SBA13HRkZKERGSv78tpQIAAAAAkOVdP0lMhWVZ6tmzpwIDA1O9HxcXlylFAYDdqlSR1qwxHbCHDZPi46VRo6QjR6Tp05ODx+HDpcGD7a4WAAAAAICsyWFZlpXWwb169UrTuMmTJ2e4IE+Ljo5WSEiIoqKiFBwcbHc5ALKBLVuk7t1NZ2zJzI5MTCR4BAAAAAD4pvTka+kKH3MCwkcAGREXZ2Y6vvtu8rU2baQXXzTPfunaxAIAAAAAgOwrPfkaf10GgDQIDJTy5zfHSUHjokVSu3ZStWrShAnSxYv21QcAAAAAQFZE+AgAaXDtHo8JCdILL5jruXNLf/whPfusFBZmmtIcO2ZrqQAAAAAAZBmEjwBwA6k1l/nwQ3N+5YqZ/Vi+vPTPP6YpTdmyUteu0m+/2Vo2AAAAAAC2I3wEgBtISEi9uczgweZ6/frSn39K8+ZJTZua8TNnSnfdJTVoIH39temWDQAAAACAr6HhDABksi1bpDFjpBkzzMxIySzJ7ttXevJJKTTU3voAAAAAALgZNJwBABvVri1NmSIdPiwNGSIVLSodPSq99ppUurT03HPSnj12VwkAAAAAgOcRPgKAh5QoIQ0dKh05Ik2aJNWoIcXGSuPHS1WqSO3bS4sXS741/xwAAAAA4EsIHwHAw/LkkXr1krZulZYtkzp2lBwO6aefpDZtpOrVpYkTpUuX7K4UAAAAAIDMRfgIAF7icEjNm0vffWca1Dz/vJQvn/T779JTT5l9Id98U/rrL7srBQAAAAAgcxA+AoANKlaUPv5YOnZMGj1aKlNGOntWevttc/zYY9KmTXZXCQAAAADAzSF8BAAbFSwoDRgg7dsnffut1KiRFB8vTZsm1asnNW5srsfH210pAAAAAADpR/gIAFlAQID04IPS6tXShg1m5mNAgPTLL1Lnzmam5OjRUlSU3ZUCAAAAAJB2hI8AkMXUqyd9+aV0+LD0xhtS4cLm+OWXpdKlpX79zExJAAAAAACyOsJHAMiiSpWS3npLOnrUdMOuVk2KiZE++USqVMl0zV6+XLIsuysFAAAAACB1hI8AkMXlzSs98YS0Y4e0eLF0770mcPzhB6lFC6lWLWnyZOnyZbsrBQAAAADAFeEjAGQTDofUqpX044/SH39Izz0nBQVJ27dLvXtLt94qDRkinTxpd6UAAAAAABiEjwCQDVWuLH36qXTsmDRqlNkL8swZafhwqUwZqWdPaetWu6sEAAAAAPg6wkcAyMZCQ6VXX5UOHJC+/lpq0EC6ckX64gupdm2pWTNp3jwpIcHuSgEAAAAAvojwEQBygFy5pC5dpF9/ldatk7p2lQICpJUrpQceMA1qxoyRoqPtrhQAAAAA4EtsDR9XrVqlDh06qFSpUnI4HJo3b16aX7tmzRoFBASoVq1aHqsPALKjO++Upk+XDh6UBg40syMPHJBeeMEsz37xRXMPAAAAAABPszV8vHjxomrWrKlPP/00Xa87f/68unfvrpYtW3qoMgDI/kqXlkaMMPtCTpggVakiXbggffSRVLGi9OCD0qpVpnM2AAAAAACe4LCsrPHXTofDoblz56pTp043HPvII4/otttuk7+/v+bNm6et6eiqEB0drZCQEEVFRSk4ODjjBQNANpOYKC1aZMLHn39Ovl67tpkV+fDDUmCgXdUBAAAAALKL9ORr2W7Px8mTJ+vAgQMaMmRImsbHxcUpOjra5QEAvsjPT7rnHmnhQun336Wnn5by5pW2bJF69JDKlpUiI6XTp+2uFAAAAACQU2Sr8HHv3r0aOHCgvvrqKwUEBKTpNSNGjFBISIjzERYW5uEqASDru/12sxT76FHpnXekUqWkkyeliAjp1lulxx+Xduywu0oAAAAAQHaXbcLHhIQEdevWTcOGDVOlSpXS/LpBgwYpKirK+Th69KgHqwSA7KVwYWnQIOnQIdOk5o47pLg4adIkqUYNqVUraf58s2QbAAAAAID0yjZ7Pp4/f16hoaHy9/d3XktMTJRlWfL399eiRYvUokWLG34d9nwEgOuzLGndOunDD6Vvv00OHW+7TerXT+rZU8qf39YSAQAAAAA2y5F7PgYHB2vHjh3aunWr8/HMM8+ocuXK2rp1q+688067SwSAbM/hkBo0kL75RjpwQHrlFSkkRNq7V3r+edNB+5VXpMOH7a4UAAAAAJAd2Bo+xsTEOINESTp48KC2bt2qI0eOSDJLprt37y5J8vPzU3h4uMujWLFiypMnj8LDw5UvXz67PgYA5EhlykjvvisdOyaNHWtmP0ZFSe+/L5UvLz30kPTrr2a2JAAAAAAAqbE1fNy4caNq166t2rVrS5IGDBig2rVrKyIiQpJ04sQJZxAJALBH/vxSnz7SH3+Y/R9btTLLsWfPlho2lO680+wXeeWK3ZUCAAAAALKaLLPno7ew5yMA3LydO6UxY6QvvzQNaiTTMbtPH+npp00jGwAAAABAzpQj93wEAGQd4eHSxInS0aNSZKRUooT011/SG2+YfSGfflratcvuKgEAAAAAdiN8BABkWNGi0ptvmgY0X34p1akjXb4sff65VK2a1LattGBBctdsAAAAAIBvIXwEANy03Lmlxx6TNm6UVq2SHnxQ8vOTFi2S7r3XBJETJkgXL9pdKQAAAADAmwgfAQCZxuGQGjeWvv1W2rdPGjBACg42zWqefVYKC5MGDjTLtQEAAAAAOR/hIwDAI8qVk0aPlo4dM81pKlSQ/vlHGjXK3HvkEem33+yuEgAAAADgSYSPAACPKlBA6tdP2rNH+u47qXlzKSFB+vpr6a67pAYNzPHVq3ZXCgAAAADIbISPAACv8PeXOnaUli2TtmyRevY0e0WuW2dmQZYvb2ZFnjtnd6UAAAAAgMxC+AgA8LpataTJk6UjR6ShQ6Vixczy7IEDzb6Qzz1n9okEAAAAAGRvhI8AANsULy4NGWJCyMmTpZo1pdhYafx4qWpV0yl78WLJsuyuFAAAAACQEYSPAADbBQaaZdhbtkjLl0v33286Zy9YILVpI4WHSxMnSpcu2V0pAAAAACA9CB8BAFmGwyE1aybNmyft3Sv17y/lzy/t2iU99ZRZkv3GG9Jff9ldKQAAAAAgLQgfAQBZUoUK0kcfmb0gP/hAKltWOntWeucdqUwZ6bHHpI0b7a4SAAAAAOAO4SMAIEsLCZFefFHat0+aM0dq0kSKj5emTZPuuENq1Ej69ltzDQAAAACQtRA+AgCyBX9/6YEHpJUrzYzH//xHypVLWrNG6txZqlhRGj1aOn/e7koBAAAAAEkIHwEA2U7dutLUqdLhw9LgwVKRIub45Zel0qWl5583e0YCAAAAAOxF+AgAyLZKlpSGD5eOHJH++1/TFfviRWnsWKlyZaljR2nZMsmypKFDpcjI1N8nMtLcBwAAAABkLsJHAEC2lzev9Pjj0vbt0pIl0n33mcDxhx+kli2lWrWkbdukiIiUAWRkpLnu729L6QAAAACQoxE+AgByDIfDhI0//CDt2SP16SMFBZlQct48cxwRIb36qhmfFDwOH26WbwMAAAAAMpfDsizL7iK8KTo6WiEhIYqKilJwcLDd5QAAPOyff6T//U/65BOzPDuJn5+UmCgNGcKSawAAAABIj/Tka8x8BADkaKGhphHN/v3SN99Id99tricmmudPPpGeecZ00U66BgAAAADIHISPAACfEBAgPfSQdM895jxpj8dz56TPPpOaNZNuvdUElZs2mT0jAQAAAAA3h/ARAOAzrt3jMT4+ebl1nTpSSIh0/Lg0erRUr57plj10qNk7EgAAAACQMYSPAACfkFpzmSFDzPnmzVL//qYpTZcupnv23r3SsGFSlSomnHz/fenoUVs/AgAAAABkOzScAQD4hKFDzVLr1LpaR0ZKCQnJMyEvXJC+/16aPl1atMjMkkzSuLHUrZvUubNUpIg3KgcAAACArCU9+RrhIwAAbvz9t/Ttt9KMGdKqVcl7QQYESK1bS127Sp06SQUK2FomAAAAAHgN4aMbhI8AgIw6dkz6+msTRG7alHw9Tx6pQwcTRLZrZ84BAAAAIKcifHSD8BEAkBn+/NOEkDNmuDalCQmRHnzQBJHNm5sZkgAAAACQkxA+ukH4CADITJYlbdliQsiZM83syCTFi5sGNl27SnfdJTkc9tUJAAAAAJmF8NENwkcAgKckJkq//GKCyFmzpLNnk++VKyc98ogJIqtXt69GAAAAALhZhI9uED4CALzh6lVp8WITRM6bJ8XEJN+rVs10zH7kEal8edtKBAAAAIAMIXx0g/ARAOBtsbHS/PkmiPzpJ+nKleR7d95pgsguXaQSJeyrEQAAAADSivDRDcJHAICdzp+X5swxQeSyZWaptiT5+ZkGNd26mYY1BQvaWSUAAAAAXB/hoxuEjwCArOLkSbM35PTp0rp1yddz55batTNB5H33SUFB9tUIAAAAAP9G+OgG4SMAICs6cMB0y54xQ9q5M/l6/vzS/febILJ1aylXLvtqBAAAAACJ8NEtwkcAQFa3c6cJIadPlw4dSr5euLDUubPpmN24sVmqDQAAAADeRvjoBuEjACC7sCzpt99MEPn119KpU8n3SpeWHn7YBJF16kgOh311AgAAAPAthI9uED4CALKj+HhpxQozG3LOHCkqKvlepUomhOzaVapc2bYSAQAAAPgIwkc3CB8BANldXJy0YIGZEfn999Lly8n36tQxIeTDD0thYfbVCAAAACDnInx0g/ARAJCTXLggffedCSIXLTIzJJM0aWKCyM6dpSJF7KsRAAAAQM5C+OgG4SMAIKf6+29p9mwTRK5alXw9IMB0yu7WzXTOLlDAvhoBAAAAZH+Ej24QPgIAfMHRo6ZJzYwZ0ubNydfz5pXuu88Eke3aSYGB9tUIAAAAIHsifHSD8BEA4Gv27DEh5IwZ0p9/Jl8PCZEefNAEkc2bS/7+9tUIAAAAIPsgfHSD8BEA4KssS9qyxXTMnjlTOn48+V7x4lKXLiaIvPNOyeGwr04AAAAAWRvhoxuEjwAASImJ0urVZjbkrFnSuXPJ98qVkx55xASR4eH21QgAAAAgayJ8dIPwEQAAV1evSosXmxmR8+ZJFy8m3wsPNx2zu3Y1oSQAAAAAED66QfgIAMD1xcZK8+ebIHLBAunKleR7d91lQsguXaQSJeyrEQAAAIC9CB/dIHwEACBtzp+X5swxQeTy5WaptiT5+UktWpgg8sEHpYIF7awSAAAAgLcRPrpB+AgAQPqdPCl9840JIn/7Lfl67tzSvfeaIPK++6SgIPtqBAAAAOAdhI9uED4CAHBzDhww3bKnT5d+/z35ev78UqdOJohs3VrKlcu2EgEAAAB4EOGjG4SPAABknh07TMfsGTOkQ4eSrxcuLHXubDpmN2pklmoDAAAAyBkIH90gfAQAIPNZlrRunQkhv/5aOn06+V7p0tIjj5gZkbVrSw6HfXUCAAAAuHmEj24QPgIA4Fnx8aZBzYwZpmFNVFTyvUqVzGzIrl3NMQAAAIDsh/DRDcJHAAC85/JlacECE0T+8IM5T1KnjgkhH3nEzI4EAAAAkD0QPrpB+AgAgD0uXJDmzTNB5KJFUkKCue5wSI0bmyCyc2epSBFbywQAAABwA4SPbhA+AgBgvzNnpNmzTRC5enXy9YAAqU0bE0Tef79UoIB9NQIAAABIHeGjG4SPAABkLUeOmCY1M2ZIW7YkX8+bV+rQwQSR7dpJgYHS0KGSv780eHDK94mMNLMphw71VuUAAACAb0pPvubnpZoAAABSdeut0iuvSJs3S7t3SxER0m23SZcuSd98Iz3wgFS8uPT449KhQ+Z+ZKTre0RGmuv+/rZ8BAAAAADXwcxHAACQ5ViWCSOnTzezIo8fT76XL5908aL05JPSZ59Jb71lgsfhw1OfEQkAAAAgc7Hs2g3CRwAAspfERLMv5PTpZp/Ic+dSjunWTZo4UQoK8n59AAAAgK9h2TUAAMgx/Pykpk3NLMcTJ6T586VHH3UdM326VLiwdO+90qefmuXZAAAAAOxH+AgAALKN3Lml9u2lypXNeUCAeQ4JkS5flhYskPr2lcqVk8LDpddeM7Mm4+PtqxkAAADwZYSPAAAgW0lqLjN8uHT1qnmOipL69JFGjpQaNzazJX//XXr3XalJE6lYMbM0e9o06exZuz8BAAAA4DsIHwEAQLZxbfCY1Fxm8GBz/umn0pUr0qpV0pkzZin2o49KhQpJ//wjzZghPfaYCSIbNpRGjJC2bzfNbQAAAAB4Bg1nAABAtjF0qOTvn3pX68hIKSHBjLlWQoK0bp30449mv8gdO1zvh4WZpdzt20stWtC0BgAAALgRul27QfgIAIBvO3JE+uknE0QuXWr2ikySJ48JIJPCyDJl7KsTAAAAyKoIH90gfAQAAEkuXZKWLzdB5I8/mmDyWuHhJoS87z7prruSG9wAAAAAvozw0Q3CRwAAkBrLMk1qkoLIX3+VEhOT74eGSvfcY4LIe+4xe0kCAAAAvojw0Q3CRwAAkBZnz0o//2yCyAULTNOaJH5+UoMGJohs397MkHQ47KsVAAAA8CbCRzcIHwEAQHrFx7s2rdm50/X+rbe6Nq3Jm9eeOgEAAABvIHx0g/ARAADcrMOHk5vWLFuWsmlNy5bJYeStt9pXJwAAAOAJhI9uED4CAIDMFBtrmtYkzYo8etT1fvXqyUEkTWsAAACQE6QnX/PzUk2pWrVqlTp06KBSpUrJ4XBo3rx5bsfPmTNHrVu3VtGiRRUcHKwGDRro559/9k6xAAAAqQgKMsHiuHFmRuT27dKIEVLDhmZvyB07pJEjpcaNpeLFpUcflWbMkM6ds7tyAAAAwPNsDR8vXryomjVr6tNPP03T+FWrVql169b66aeftGnTJjVv3lwdOnTQli1bPFwpAADAjTkcZqbjwIHSL79Ip09LX30lde1qumWfOydNny516yYVLWoCyVGjzB6SvrUWBQAAAL4iyyy7djgcmjt3rjp16pSu11WrVk0PP/ywIiIi0jSeZdcAAMAO8fHS2rVmefaPP6betCape3bz5jStAQAAQNaVbZZd36zExERduHBBhQoVuu6YuLg4RUdHuzwAAAC8LSDAzHQcOdIsxT50SPr0U+nee02TmiNHzNLt9u2lwoWlDh2kCRNS7iEJAAAAZCfZOnx8//33FRMToy5dulx3zIgRIxQSEuJ8hIWFebFCAACA1JUpIz33nJkFefas9MMP0jPPSKVLS5cumeY1zz5rZkTWrCm9/rq0Zo2UkGB35QAAAEDaZdtl19OnT9eTTz6p7777Tq1atbruuLi4OMXFxTnPo6OjFRYWxrJrAACQJVmWmRmZ1D173TopMTH5fuHC0j33mBmSbdtKbhaAAAAAAB6RnmXXAV6qKVPNnDlTTzzxhGbNmuU2eJSkwMBABQYGeqkyAACAm+NwSDVqmMegQWZW5MKFJohcuNCcT5tmHv7+0t13myDyvvuk2283rwcAAACyimy37HrGjBnq1auXZsyYofbt29tdDgAAgEcVLiw9+qg0Y4Z05oy0apX06qtStWpmCfbq1aa7dni4VK6c1KePtGCBWboNAAAA2M3WZdcxMTHat2+fJKl27dr64IMP1Lx5cxUqVEi33nqrBg0apOPHj2vq1KmSzFLrHj16aMyYMXrwwQed75M3b16FhISk6WvS7RoAAOQUhw4ld89etky6ZqcZ5c0rtWyZ3EG7dGnbygQAAEAOk558zdbwccWKFWrevHmK6z169NCUKVPUs2dPHTp0SCtWrJAkNWvWTCtXrrzu+LQgfAQAADnRxYsmgEzaK/L4cdf7NWokB5F33mmWbAMAAAAZkW3CRzsQPgIAgJzOsqTt212b1lz7X3yFC0vt2iU3rQkNta9WAAAAZD+Ej24QPgIAAF/z99/JTWt+/lk6fz75nr+/1LBhctOaqlVpWgMAAAD3CB/dIHwEAAC+LD5e+vVXE0T++KO0a5fr/bJlk4PIZs2kPHnsqBIAAABZGeGjG4SPAAAAyQ4eTG5as3y5a9OaoKDkpjX33kvTGgAAABiEj24QPgIAAKTu4kVp6dLkMPLfTWtq1kxuWlO/Pk1rAAAAfBXhoxuEjwAAADdmWdK2bclB5L+b1hQp4tq0pmBB20oFAACAlxE+ukH4CAAAkH5nzpimNT/+aJ6jopLv+ftLjRqZILJ9e5rWAAAA5HSEj24QPgIAANycq1dN05offzSNa3bvdr1frlxy05qmTWlaAwAAkNMQPrpB+AgAAJC5kprWzJ9vmtZcuZJ8LyhIatUquWnNLbdIQ4ea2ZKDB6d8r8hIKSHBjAEAAEDWlJ58zc9LNQEAACCHKldO6tvXLMc+d0767jvpySelUqWk2Fjp+++lp54y3bJr15ZWrJAiIqRhw1zfJzLSXKeRDQAAQM7BzEcAAAB4RFLTmvnzzczI335zbVojmQ7ar75qxr37rjR8eOozIgEAAJB1sOzaDcJHAAAAe5w5Iy1YkNy0Jjra9X6xYlK3bmaZdpMmUoEC9tQJAAAA9wgf3SB8BAAAsN/Vq9KaNSZoTEhIeT8gQLrzTnO/ZUtznDu39+sEAABASunJ1wK8VBMAAADglCuXtHq1CR5z5zZNarp0kQoWlJYskQ4cMOHkmjVmb8h8+Uzn7JYtTSAZHi75sXs5AABAlsd/sgEAAMDrkprLDB8uxcWZ52++MU1p9u834ePEidLDD0tFikgXL0o//SS99JLZJ7JkSalrV+l//5MOH7b70wAAAOB6WHYNAAAAr7o2eLy2ucz1ricmSjt2mBmRS5ZIq1aZLtrXqlDBzIhs1Upq3lwqXNg7nwUAAMAXseejG4SPAAAA9ho6VPL3T72rdWSkWYo9dOj1X3/lirRunbR0qQkjf/vNdd9Ih0OqXTt5v8hGjaSgoMz+FAAAAL6L8NENwkcAAICcJTrazIZMmhn5+++u93Pnlho2TN4vsm5d09AGAAAAGUP46AbhIwAAQM524oS0bFlyGHnsmOv9kBCpWbPkmZFVqpjZkgAAAEgbwkc3CB8BAAB8h2VJe/eaEHLpUhNKnj/vOqZUqeT9Ilu2NOcAAAC4PsJHNwgfAQAAfFdCgrR5c/J+kb/8YrptX6tq1eQl2s2amZmSAAAASEb46AbhIwAAAJJcuiT9+mvyEu1Nm8xsySR+ftIddyTPjGzQQAoMtK9eAACArIDw0Q3CRwAAAFzPuXPSihXJMyP//NP1ft68UuPGyUu0a9UyASUAAIAvIXx0g/ARAAAAaXX0aHIQuWSJdOqU6/3ChaUWLZKXaZcvT/MaAACQ8xE+ukH4CAAAgIywLOn335PDyJUrpQsXXMeULZscRLZoIRUrZkupAAAAHkX46AbhIwAAADLD1avShg3JnbTXrjXXrlWjRvJ+kY0bS/nz21MrAABAZiJ8dIPwEQAAAJ4QEyOtXp08M3LbNtf7AQGmYU3SfpH160u5ctlTKwAAwM0gfHSD8BEAAADecPq0tHx58n6Rhw653s+fX2rWLHmZdrVq7BcJAACyB8JHNwgfAQAAYIcDB5KDyGXLpLNnXe8XL54cRLZsKd16qz11AgAA3AjhoxuEjwAAALBbYqJZlp20X+SqVdKlS65jbrstOYhs3lwqVMieWgEAAP6N8NENwkcAAABkNXFxpmFN0n6R69ebgDKJwyHVrZs8M7JhQylvXvvqBQAAvo3w0Q3CRwAAAGR1UVHSypXJMyN37XK9HxhoAsikTtp16kj+/vbUCgAAfA/hoxuEjwAAAMhu/vrLhJBJMyOPH3e9X7CgWZqdFEbedhvNawAAgOcQPrpB+AgAAIDszLKkPXuSg8jly81MyWuVLp28X2TLllLJkvbUCgAAcibCRzcIHwEAAJCTxMdLmzcnd9Jes0a6csV1TLVqyftFNm0q8Z/BAADgZhA+ukH4CAAAgJwsNtYEkEn7RW7ebGZLJvH3l+rXT16ifdddUu7c9tULAACyH8JHNwgfAQAA4EvOnjVLs5OWae/b53o/KEhq0iR5mXaNGpKfnz21AgCA7IHw0Q3CRwAAAPiyw4eTg8ilS6XTp13vFymSvFdkq1ZSuXL21AkAALIuwkc3CB8BAAAAw7KknTuT94tcuVK6eNF1TPnyyUFkixYmnJSkoUPNEu7Bg1O+b2SklJBgxgAAgJwnPflagJdqAgAAAJDFOBxS9erm8eKLplHN+vXJsyLXrZMOHDCPiRPNa2rVMkHkqVPSl1+aa9cGkJGRUkSENHy41z8OAADIgpj5CAAAACBVFy5Iq1cnz4zcscP1vp+flJhoZkQOHmwCy7feMsFjajMiAQBAzsCyazcIHwEAAICMOXVKWrYsOYw8ciTlmJIlpQcflO6+W2rQQCpb1sywBAAAOQfhoxuEjwAAAMDNsyxp/34TQvbpY2ZApqZECRNCJoWRdetKefJ4t1YAAJC52PMRAAAAgEc5HFLFitKMGSZ4zJ3b7BnZpYuZ/bh2rbRli3TypDR3rnlIUq5cUp06roFk6dL2fhYAAOA5fnYXAAAAACB7ura5TFycef7mG6lwYem336SoKLNn5LvvSp06ScWLS1evmnsffWSCyrAw6dZbpYcflsaMMQ1vrlyx+5MBAIDMwsxHAAAAAOl2bfCY1Fwm6TkiIvm8USPzkMxS7YMHzazIX381z9u3S0ePmsc335hxefJI9eq5zo4sXty7nw8AAGQOwkcAAAAA6ZaQkHpX66TzhISUr3E4pPLlzePRR821mBhpwwbXQPLcOemXX8wjSfnyrmFk9epSAH+bAQAgy6PhDAAAAIAsw7KkP/90DSN//91cv1a+fFL9+slh5F13meXeAADA8+h27QbhIwAAAJC9REWZfSKTAsl166To6JTjKld2nR15++2SH7vcAwCQ6Qgf3SB8BAAAALK3xERp1y7X2ZF79qQcFxIi3XlnciB5553mGgAAuDmEj24QPgIAAAA5z9mzZkZkUiC5fr108aLrGIdDqlbNdXZkpUrmOgAASDvCRzcIHwEAAICcLz5e2rHDdXbkgQMpxxUubPaLTAok77hDyp/f+/UCAJCdED66QfgIAAAA+KZTp1zDyI0bpcuXXcf4+Uk1a7rOjixXjtmRAABci/DRDcJHAAAAAJJ05Yq0datrIHn0aMpxxYolB5F33y3VrSvlzev1cgEAyDIIH90gfAQAAABwPceOmRAyKZDcvFm6etV1TK5cUu3arrMjw8LsqRcAADsQPrpB+AgAAAAgrS5fljZtcp0defJkynG33OI6O7J2bSl3bu/XCwCANxA+ukH4CAAAACCjLEs6dMg1jNy2TUpIcB0XGCjVq+c6O7JECVtKBgAg0xE+ukH4CAAAACAzXbwobdjgGkiePZtyXLlyJoRMCiRr1JACArxfLwAAN4vw0Q3CRwAAAACeZFnS3r2uYeTOneb6tYKCpPr1k8PIu+6SihSxp2YAANKD8NENwkcAAAAA3hYVJa1fnxxIrltnrv3bbbe57h15++2Sv7/36wUAwB3CRzcIHwEAAADYLTFR2r3bdXbkH3+kHFeggJkRmRRG3nmnVLCg18sFAMAF4aMbhI8AAAAAsqJz58yMyKRA8rffzH6S13I4pKpVXWdHVqok+fnZUzMAwDcRPrpB+AgAAAAgO4iPN3tFXjs7cv/+lONCQ5Mb2TRoYPaRLFDA+/UCAHwH4aMbhI8AAAAAsqvTp00ImRRIbtggXb7sOsbPT6pe3XV2ZPnyZtbktYYONftJDh6c8utERkoJCWYMAAD/lp58LcBLNQEAAAAAblKxYtL995uHJF25Im3b5jo78sgRc23bNmn8eDOuaNHkMLJBA6lePRM8RkSY+9cGkJGR5vrw4d79bACAnImZjwAAAACQgxw/7jo7cvNmE1JeKyBAqlXLPK9bJw0YIL3/vvTWW8nBY2ozIgEAkFh27RbhIwAAAABfcvmyCSCTwshff5VOnrz++JYtpVdfNbMjCxXyXp0AgOyD8NENwkcAAAAAvsyypMOHXZdqb9qU+tiKFaU77kh+1K4t5cvn3XoBAFkP4aMbhI8AAAAAkCxpj8dcuaSrV6XwcCk2VjpwIOVYPz+pWjXXQLJ6dSl3bu/XDQCwT3ryNT8v1QQAAAAAyGKubS5z5Yp53rlT6tlT+vtvaeFCM6ZjR6lkSSkxUdqxQ5o0SXr2WbM0u0ABqX59qU8facoU6fffTadsAAAkul0DAAAAgE+6NnhMai6T9HxtF+y2bZNfc/y4tGFD8mPjRumff5LPk+TPL9Wp4zpDslw5yeHwzmcDAGQdhI8AAAAA4IMSElLvap10ntrsxVtuMY9Oncy5ZUn797sGkps3SzEx0qpV5pGkcGEzU/LaQLJkSY98NABAFmLrno+rVq3Se++9p02bNunEiROaO3euOiX9v9h1rFixQgMGDNDvv/+usLAwvfnmm+rZs2eavyZ7PgIAAACA58THS3/84RpIbttm9pP8t1tucQ0j69WTQkO9XzMAIH3Sk6/ZOvPx4sWLqlmzpnr37q0HH3zwhuMPHjyo9u3b65lnntG0adO0dOlSPfHEEypZsqTaXrsWAAAAAABgi4AA07QmPFzq1ctci4uTtm93DSR37TLLuI8fl+bNS349HbYBIGfJMt2uHQ7HDWc+vvbaa/rxxx+1c+dO57VHHnlE58+f18KFC9P0dZj5CAAAAAD2i4kxS7SvDSTpsA0A2UO2mfmYXmvXrlWrVq1crrVt21YvvPDCdV8TFxenuLg453l0dLSnygMAAAAApFH+/FKTJuaR5OxZ08Tm2kDyxAnTYTupy7YkBQZKNWu6BpKVK0v+/vZ8FgDA9WWr8PHkyZMqXry4y7XixYsrOjpaly5dUt68eVO8ZsSIERo2bJi3SgQAAAAAZFDhwqa7trsO2xs2SOfPS+vXm0eS/PmlunVdA8myZemwDQB2y1bhY0YMGjRIAwYMcJ5HR0crLCzMxooAAAAAAGmVng7bK1eaR5IiRVJ22C5RwpaPAQA+K1uFjyVKlNCpU6dcrp06dUrBwcGpznqUpMDAQAUGBnqjPAAAAACAhzkcpilNxYpS167mWny8tHu3ayC5fbv099/SwoXmkaR06ZQdtgsWtOWjAIBPyFbhY4MGDfTTTz+5XFu8eLEaNGhgU0UAAAAAALsFBJgmNNWrS717m2uXL6fssL17t3TsmHnMnZv8+ttuS9lhOyjIns8CADmNreFjTEyM9u3b5zw/ePCgtm7dqkKFCunWW2/VoEGDdPz4cU2dOlWS9Mwzz2js2LF69dVX1bt3by1btkzffPONfvzxR7s+AgAAAAAgC8qTR6pf3zySXLiQssP2wYPS3r3mMX26Gefvn3qH7Vy57PksAJCdOSzLsuz64itWrFDz5s1TXO/Ro4emTJminj176tChQ1qxYoXLa1588UXt2rVLpUuX1uDBg9WzZ880f830tAIHAAAAAORsf/+dssP2yZMpxwUGSrVqpeyw7efn9ZIBwHbpyddsDR/tQPgIAAAAALgey0rZYXvjRtNh+98KFEjZYbtMGTpsA8j5CB/dIHwEAAAAAKSHZUn79qXssH3pUsqxRYq4hpF33CEVL+79mgHAkwgf3SB8BAAAAADcrPh4adcu1yXb27dLV6+mHBsWlrLDdkiI92sGgMxC+OgG4SMAAAAAwBOu12E7tb91V6rkGkjWqkWHbQDZB+GjG4SPAAAAAABvuV6H7X/z95fCw10DyfBwOmwDyJoIH90gfAQAAAAA2CmtHbbz5EnZYbtSJdcO20OHmuBy8OCUr4+MlBISzBgAyEzpydcCvFQTAAAAAACQaUpzzz3mIbnvsL1unXkkCQ527bB94YL0wQfm3rUBZGSkFBEhDR/utY8FAKli5iMAAAAAAFlMYqK0f3/aOmwHBUmxsVKzZtJLL0krV0rvv2+Cx9RmRALAzWLZtRuEjwAAAACA7Cipw/a1geT27eb6vwUFSY0aSbVrm6XbtWtLt93mumQbADKK8NENwkcAAAAAQE5x+bK0bZsJIvv3NzMmrydfPqlmTddAMjxcCgz0WrkAcgj2fAQAAAAAwAfkySPdeae0aJEJHnPnlq5ckZ56SqpTR9q6VdqyxcyQvHhR+vVX80gSECDdfntyGJkUTIaE2PSBAOQ4hI8AAAAAAGRj1zaXGTzY9Xz8eDMmPl7688/kMDLpce6cCSa3b5emTk1+z3LlksPIpECyVCnJ4bDjEwLIzlh2DQAAAABANvXv4PFG169lWdKxY65h5Nat0uHDqY8vWtQ1jGQfScB3sewaAAAAAAAfkJCQesCYdJ6QcP3XOhxSWJh5dOyYfP3cOdcZklu3Srt3S2fOmOXdixYlj03aR/LaZdvsIwngWsx8BAAAAAAAbl26JO3YkRxGJu0jeelSyrEBAVLVqq7LtmvWlAoW9HbVADyFbtduED4CAAAAAHDz3O0jmZpr95FMminJPpJA9kT46AbhIwAAAAAAnsE+koBvIHx0g/ARAAAAAADvut4+komJKcemto9ktWpSnjxeLhrAdRE+ukH4CAAAAACA/W5mH8latcyDfSQBexA+ukH4CAAAAABA1sQ+kkD2QPjoBuEjAAAAAADZR0b3kbx22Tb7SAKZi/DRDcJHAAAAAACyv/TuI1mjRnIYyT6SwM0hfHSD8BEAAAAAgJyJfSQB7yB8dIPwEQAAAAAA38E+kkDmI3x0g/ARAAAAAADflhn7SFasKPn7e7NqIOsgfHSD8BEAAAAAAKTmZvaRrFVLCg9nH0n4BsJHNwgfAQAAAABAWnljH8mhQ80sysGDU96LjJQSEswYIKtIT74W4KWaAAAAAAAAsp28eaX69c0jibt9JHfsMI+pU5PHJ+0jee2y7Wv3kfT3lyIizPG1AWRkpLk+fLinPyXgOcx8BAAAAAAAuEkZ2Ufy2jBy7Vrp449N0Dh4sGvwmNqMSMBO/6+9ew+K6rz/OP5ZwF0guIBBQPCCl4j1RoxGhlxaWxmptU7S5g/HOKljax0b7ERN05o2go2/FieZODYtMem0qf2jVptMbSY3W6qAk4SYBLSiEipKIpNwKVUuonLb5/fHDqvrbjcac3YXeL9mzqyc58vuc2b4zmY+ec55uO06AMJHAAAAAAAQLDfyHMkRI6TeXikiwj2+erX0zDMS8QXCDeFjAISPAAAAAAAglG7kOZKSlJHh3txm1iz36+zZ7t22o3iYHkKE8DEAwkcAAAAAABBu+vqkDRuk3/zG/QzI/n5p5Eips9N/vcMhzZjhG0omJwd33hie2HAGAAAAAABgECkqcgeP1z7zcdMmKS/PvVLy2DH3cfy4dPGiVFXlPq6WnHwliBwIJadPl6KjQ3NdAOEjAAAAAABACPnbXGbgtaBAio313nTG5ZLOnHEHkVeHkqdPSy0t0j//6T4GRERIU6f6rpKcMOHKjtuAVbjtGgAAAAAAIIS2bHHfau1vV+utW923YG/Z8tnv09UlnTjhG0qeO+e/fuTIK2Hk1a/x8TdzNRgOeOZjAISPAAAAAABguDBGamz0DSRratw7a/szfrzvrdtTp7LBDa4gfAyA8BEAAAAAAAx3vb1Sba13KFldLTU0+K93OKQvfck3lExJ4dbt4YjwMQDCRwAAAAAAAP/On3eHkNducHPhgv/6pCT/G9zExgZ33gguwscACB8BAAAAAACun8slffSR763bdXXusWtFREhTpviGkhkZ7jEMfoSPARA+AgAAAAAA3LyLF6WTJ31DydZW//VxcdLMmd6h5KxZUmJicOeNm0f4GADhIwAAAAAAgDWMkZqbfQPJkyelnh7/vzN2rO8qycxMacSI4M4d14/wMQDCRwAAAAAAgODq7ZVOnfLd4Objj/3Xjxjhf4ObMWPY4CYcED4GQPgIAAAAAAAQHtra3BvaXL1Ksrpa6uz0Xz9q1JVAciCUnDFDuuWWoE572CN8DIDwEQAAAAAAIHwZ414Ree2t2//+t/8Nbmw2afJk31WSkyaxwY1VCB8DIHwEAAAAAAAYfC5dkmpqfFdJNjf7r4+N9b/Bza23BnfeQxHhYwCEjwAAAAAAAENHc7M7hLx2g5vLl/3Xp6dfWR05EEpOmybZ7cGd92BG+BgA4SMAAAAAAMDQ1tcn1dX5rpKsr/dfHxXlDiCvvXU7PZ0NbvwhfAyA8BEAAAAAAGB46ujwv8FNe7v/+sRE31WSM2dKcXG+tVu2SJGR0ubNvmNbt0r9/e6aoeBG8rWoIM0JAAAAAAAACCmnU7rrLvcxwBipocE3kPzwQ+n8eenQIfdxtcmTr4SSA682m1RQ4B6/OoDcutV9/sknrb++cMTKRwAAAAAAAOAa3d3+N7hpbPRfHxPjXin56afS4sXSj34klZW5w8cnn/S/InKw4rbrAAgfAQAAAAAA8Hm1tvoGksePu3fj9meoBY8S4WNAhI8AAAAAAAD4IvX3S6dPXwkl/+//JJfLvYN2d3eoZ/fFu5F8LSJIcwIAAAAAAACGpMhIaepU6YEH3DtnDwSPPT3u266HM8JHAAAAAAAA4Atw9eYy3d3u14KC4R1Asts1AAAAAAAAcJOuDh4HnvE48OpvF+zhgvARAAAAAAAAuEn9/f43lxn4ub8/+HMKB2w4AwAAAAAAAOC6seEMAAAAAAAAgJAjfAQAAAAAAABgCcJHAAAAAAAAAJYgfAQAAAAAAABgCcJHAAAAAAAAAJYgfAQAAAAAAABgCcJHAAAAAAAAAJYgfAQAAAAAAABgCcJHAAAAAAAAAJYgfAQAAAAAAABgCcJHAAAAAAAAAJYgfAQAAAAAAABgCcJHAAAAAAAAAJYgfAQAAAAAAABgCcJHAAAAAAAAAJYIi/CxuLhYGRkZio6OVnZ2tt57772A9Tt27FBmZqZiYmI0btw4bdiwQZcvXw7SbAEAAAAAAABcj5CHj3v37tXGjRtVWFioqqoqZWVlKS8vTy0tLX7rd+/erU2bNqmwsFA1NTX6/e9/r7179+qnP/1pkGcOAAAAAAAAIJCQh4/bt2/X97//fa1atUrTp0/X888/r9jYWL344ot+69955x3dfffdevDBB5WRkaFFixZp+fLln7laEgAAAAAAAEBwhTR87OnpUWVlpXJzcz3nIiIilJubq4qKCr+/c9ddd6mystITNp45c0ZvvPGGvvGNb/it7+7uVkdHh9cBAAAAAAAAwHpRofzw1tZW9ff3KyUlxet8SkqKPvzwQ7+/8+CDD6q1tVX33HOPjDHq6+vT2rVr/+dt10VFRfr5z3/+hc8dAAAAAAAAQGAhv+36RpWVlemXv/ylnnvuOVVVVemvf/2rXn/9dW3dutVv/eOPP6729nbP0dDQEOQZAwAAAAAAAMNTSFc+JiUlKTIyUs3NzV7nm5ublZqa6vd3Nm/erIceekirV6+WJM2aNUtdXV1as2aNfvaznykiwjtPdTgccjgcnp+NMZLE7dcAAAAAAADA5zCQqw3kbIGENHy02+2aO3euDhw4oPvvv1+S5HK5dODAAa1bt87v71y8eNEnYIyMjJR0fRfc2dkpSRo3btxNzBwAAAAAAAAY3jo7OxUfHx+wJqThoyRt3LhRK1eu1Lx58zR//nzt2LFDXV1dWrVqlSTpO9/5jtLT01VUVCRJWrp0qbZv3645c+YoOztbdXV12rx5s5YuXeoJIQNJS0tTQ0ODRo4cKZvNZum1fVE6Ojo0btw4NTQ0yOl0hno6wLBGPwLhg34Ewgf9CIQHehEIH0O9H40x6uzsVFpa2mfWhjx8XLZsmf7zn/+ooKBATU1Nuv3227V//37PJjRnz571Wun4xBNPyGaz6YknntAnn3yi0aNHa+nSpfrFL35xXZ8XERGhsWPHWnItVnM6nUPyDxYYjOhHIHzQj0D4oB+B8EAvAuFjKPfjZ614HGAz13OvMkKqo6ND8fHxam9vH7J/sMBgQT8C4YN+BMIH/QiEB3oRCB/04xWDbrdrAAAAAAAAAIMD4eMg4HA4VFhY6LVrN4DQoB+B8EE/AuGDfgTCA70IhA/68QpuuwYAAAAAAABgCVY+AgAAAAAAALAE4SMAAAAAAAAASxA+AgAAAAAAALAE4SMAAAAAAAAASxA+DgLFxcXKyMhQdHS0srOz9d5774V6SsCgdujQIS1dulRpaWmy2Wz629/+5jVujFFBQYHGjBmjmJgY5ebm6tSpU141586d04oVK+R0OpWQkKDvfe97unDhglfNsWPHdO+99yo6Olrjxo3TU089ZfWlAYNKUVGR7rzzTo0cOVLJycm6//77VVtb61Vz+fJl5efn69Zbb1VcXJweeOABNTc3e9WcPXtWS5YsUWxsrJKTk/XYY4+pr6/Pq6asrEx33HGHHA6HpkyZol27dll9ecCgsnPnTs2ePVtOp1NOp1M5OTl68803PeP0IhAa27Ztk81m0/r16z3n6EcgeLZs2SKbzeZ1TJs2zTNOP14fwscwt3fvXm3cuFGFhYWqqqpSVlaW8vLy1NLSEuqpAYNWV1eXsrKyVFxc7Hf8qaee0rPPPqvnn39ehw8f1i233KK8vDxdvnzZU7NixQqdOHFCJSUleu2113To0CGtWbPGM97R0aFFixZpwoQJqqys1NNPP60tW7bot7/9reXXBwwW5eXlys/P17vvvquSkhL19vZq0aJF6urq8tRs2LBBr776ql566SWVl5fr008/1be//W3PeH9/v5YsWaKenh698847+uMf/6hdu3apoKDAU1NfX68lS5boq1/9qo4ePar169dr9erV+vvf/x7U6wXC2dixY7Vt2zZVVlbqgw8+0Ne+9jXdd999OnHihCR6EQiF999/Xy+88IJmz57tdZ5+BIJrxowZamxs9BxvvfWWZ4x+vE4GYW3+/PkmPz/f83N/f79JS0szRUVFIZwVMHRIMvv27fP87HK5TGpqqnn66ac959ra2ozD4TB//vOfjTHGnDx50kgy77//vqfmzTffNDabzXzyySfGGGOee+45k5iYaLq7uz01P/nJT0xmZqbFVwQMXi0tLUaSKS8vN8a4e2/EiBHmpZde8tTU1NQYSaaiosIYY8wbb7xhIiIiTFNTk6dm586dxul0evrvxz/+sZkxY4bXZy1btszk5eVZfUnAoJaYmGh+97vf0YtACHR2dprbbrvNlJSUmK985SvmkUceMcbw3QgEW2FhocnKyvI7Rj9eP1Y+hrGenh5VVlYqNzfXcy4iIkK5ubmqqKgI4cyAoau+vl5NTU1efRcfH6/s7GxP31VUVCghIUHz5s3z1OTm5ioiIkKHDx/21Hz5y1+W3W731OTl5am2tlbnz58P0tUAg0t7e7skadSoUZKkyspK9fb2evXjtGnTNH78eK9+nDVrllJSUjw1eXl56ujo8KzYqqio8HqPgRq+SwH/+vv7tWfPHnV1dSknJ4deBEIgPz9fS5Ys8ekZ+hEIvlOnTiktLU2TJk3SihUrdPbsWUn0440gfAxjra2t6u/v9/ojlaSUlBQ1NTWFaFbA0DbQW4H6rqmpScnJyV7jUVFRGjVqlFeNv/e4+jMAXOFyubR+/XrdfffdmjlzpiR3r9jtdiUkJHjVXtuPn9Vr/6umo6NDly5dsuJygEGpurpacXFxcjgcWrt2rfbt26fp06fTi0CQ7dmzR1VVVSoqKvIZox+B4MrOztauXbu0f/9+7dy5U/X19br33nvV2dlJP96AqFBPAAAAID8/X8ePH/d6hg6A4MrMzNTRo0fV3t6ul19+WStXrlR5eXmopwUMKw0NDXrkkUdUUlKi6OjoUE8HGPYWL17s+ffs2bOVnZ2tCRMm6C9/+YtiYmJCOLPBhZWPYSwpKUmRkZE+OyU1NzcrNTU1RLMChraB3grUd6mpqT6bPvX19encuXNeNf7e4+rPAOC2bt06vfbaayotLdXYsWM951NTU9XT06O2tjav+mv78bN67X/VOJ1O/qMRuIrdbteUKVM0d+5cFRUVKSsrS7/61a/oRSCIKisr1dLSojvuuENRUVGKiopSeXm5nn32WUVFRSklJYV+BEIoISFBU6dOVV1dHd+PN4DwMYzZ7XbNnTtXBw4c8JxzuVw6cOCAcnJyQjgzYOiaOHGiUlNTvfquo6NDhw8f9vRdTk6O2traVFlZ6ak5ePCgXC6XsrOzPTWHDh1Sb2+vp6akpESZmZlKTEwM0tUA4c0Yo3Xr1mnfvn06ePCgJk6c6DU+d+5cjRgxwqsfa2trdfbsWa9+rK6u9vofAiUlJXI6nZo+fbqn5ur3GKjhuxQIzOVyqbu7m14EgmjhwoWqrq7W0aNHPce8efO0YsUKz7/pRyB0Lly4oNOnT2vMmDF8P96IUO94g8D27NljHA6H2bVrlzl58qRZs2aNSUhI8NopCcCN6ezsNEeOHDFHjhwxksz27dvNkSNHzMcff2yMMWbbtm0mISHBvPLKK+bYsWPmvvvuMxMnTjSXLl3yvMfXv/51M2fOHHP48GHz1ltvmdtuu80sX77cM97W1mZSUlLMQw89ZI4fP2727NljYmNjzQsvvBD06wXC1Q9+8AMTHx9vysrKTGNjo+e4ePGip2bt2rVm/Pjx5uDBg+aDDz4wOTk5JicnxzPe19dnZs6caRYtWmSOHj1q9u/fb0aPHm0ef/xxT82ZM2dMbGyseeyxx0xNTY0pLi42kZGRZv/+/UG9XiCcbdq0yZSXl5v6+npz7Ngxs2nTJmOz2cw//vEPYwy9CITS1btdG0M/AsH06KOPmrKyMlNfX2/efvttk5uba5KSkkxLS4sxhn68XoSPg8Cvf/1rM378eGO32838+fPNu+++G+opAYNaaWmpkeRzrFy50hhjjMvlMps3bzYpKSnG4XCYhQsXmtraWq/3+O9//2uWL19u4uLijNPpNKtWrTKdnZ1eNf/617/MPffcYxwOh0lPTzfbtm0L1iUCg4K/PpRk/vCHP3hqLl26ZB5++GGTmJhoYmNjzbe+9S3T2Njo9T4fffSRWbx4sYmJiTFJSUnm0UcfNb29vV41paWl5vbbbzd2u91MmjTJ6zMAGPPd737XTJgwwdjtdjN69GizcOFCT/BoDL0IhNK14SP9CATPsmXLzJgxY4zdbjfp6elm2bJlpq6uzjNOP14fmzHGhGbNJQAAAAAAAIChjGc+AgAAAAAAALAE4SMAAAAAAAAASxA+AgAAAAAAALAE4SMAAAAAAAAASxA+AgAAAAAAALAE4SMAAAAAAAAASxA+AgAAAAAAALAE4SMAAAAAAAAASxA+AgAAIGwsWLBA69evD/U0AAAA8AUhfAQAAAAAAABgCcJHAAAAAAAAAJYgfAQAAEDYev311xUfH68//elPoZ4KAAAAPoeoUE8AAAAA8Gf37t1au3atdu/erW9+85uhng4AAAA+B1Y+AgAAIOwUFxfr4Ycf1quvvkrwCAAAMIix8hEAAABh5eWXX1ZLS4vefvtt3XnnnaGeDgAAAG4CKx8BAAAQVubMmaPRo0frxRdflDEm1NMBAADATSB8BAAAQFiZPHmySktL9corr+iHP/xhqKcDAACAm8Bt1wAAAAg7U6dOVWlpqRYsWKCoqCjt2LEj1FMCAADA50D4CAAAgLCUmZmpgwcPasGCBYqMjNQzzzwT6ikBAADgBtkMD9IBAAAAAAAAYAGe+QgAAAAAAADAEoSPAAAAAAAAACxB+AgAAAAAAADAEoSPAAAAAAAAACxB+AgAAAAAAADAEoSPAAAAAAAAACxB+AgAAAAAAADAEoSPAAAAAAAAACxB+AgAAAAAAADAEoSPAAAAAAAAACxB+AgAAAAAAADAEv8PKPdEgwvMJe8AAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1600x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16,8))\n",
    "plt.plot(K, dist, 'bx-')\n",
    "plt.xlabel('k')\n",
    "plt.ylabel('Distortion')\n",
    "plt.title('The Elbow Method showing the optimal k')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ayakovyna/miniconda3/envs/deepwalk/lib/python3.9/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n"
     ]
    }
   ],
   "source": [
    "kmeanModel = sklearn.cluster.KMeans(4000).fit(embeddings)\n",
    "node_cluster_assignment, counts = np.unique(kmeanModel.predict(embeddings), return_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "lower_counts = counts[counts > 10]\n",
    "upper_counts = lower_counts[lower_counts <= 100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of Sybils (cluster):  343\n",
      "Mean of nodes in the cluster:  16.002915451895042 Median:  14.0\n"
     ]
    }
   ],
   "source": [
    "print('Number of Sybils (cluster): ', len(upper_counts))\n",
    "print('Mean of nodes in the cluster: ', upper_counts.mean(), 'Median: ', np.median(upper_counts))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "deepwalk",
   "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": "3.9.17"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}