vectors_busines_cases.ipynb

{
 "cells": [
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-11-14T20:03:05.033022Z",
     "start_time": "2024-11-14T20:02:57.401207Z"
    }
   },
   "cell_type": "code",
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "from src.utils import get_md_files_df, to_vectors, euclidean_to_poincare, create_gephi_graph_data_file_and_save, \\\n",
    "    poincare_distance\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "import pandas as pd\n"
   ],
   "id": "6e614329384e63c9",
   "outputs": [],
   "execution_count": 1
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-11-14T20:03:05.117823Z",
     "start_time": "2024-11-14T20:03:05.047548Z"
    }
   },
   "cell_type": "code",
   "source": [
    "df = get_md_files_df()\n",
    "df"
   ],
   "id": "35b31daaba8890c8",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                                          file_content relative_dir_path  \\\n",
       "0    # This is a content repository for markdown ba...                 .   \n",
       "1    \\n* [Current State](#HyperthermJSMOpportunity-...                 .   \n",
       "2    # Welcome to Our Knowledge Base\\n\\nWelcome to ...                 .   \n",
       "3    What Do Solutions Engineers Do?\\n-------------...                 .   \n",
       "4    What Do Solutions Engineers Do?\\n-------------...                 .   \n",
       "..                                                 ...               ...   \n",
       "681  Zero Trust Security Boot Camp\\nLearn what zero...         knowledge   \n",
       "682  Zero Trust Security Boot Camp\\nLearn what zero...         knowledge   \n",
       "683  note**Strategic Opportunity:** We can maximize...        leadership   \n",
       "684  By  and  \\n\\n## Introduction: *The Dawn of a N...        leadership   \n",
       "685  # Emergent Categories in the Viable System Mod...        leadership   \n",
       "\n",
       "                                             file_name  \n",
       "0                                             about.md  \n",
       "1                        hypertherm-jsm-opportunity.md  \n",
       "2                                             index.md  \n",
       "3            onboarding-aryaman-chakravarthy-intern.md  \n",
       "4                             onboarding-rajat-jain.md  \n",
       "..                                                 ...  \n",
       "681   Zero_Trust_Security_Boot_Camp_Course_Brochure.md  \n",
       "682  Zero_Trust_Security_Boot_Camp_Course_Brochure_...  \n",
       "683  aiaccelerated_development_atlassian_github_int...  \n",
       "684  the_value_proposition_of_the_enterprise_ai_arc...  \n",
       "685                             viable_system_model.md  \n",
       "\n",
       "[686 rows x 3 columns]"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>file_content</th>\n",
       "      <th>relative_dir_path</th>\n",
       "      <th>file_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td># This is a content repository for markdown ba...</td>\n",
       "      <td>.</td>\n",
       "      <td>about.md</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>\\n* [Current State](#HyperthermJSMOpportunity-...</td>\n",
       "      <td>.</td>\n",
       "      <td>hypertherm-jsm-opportunity.md</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td># Welcome to Our Knowledge Base\\n\\nWelcome to ...</td>\n",
       "      <td>.</td>\n",
       "      <td>index.md</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>What Do Solutions Engineers Do?\\n-------------...</td>\n",
       "      <td>.</td>\n",
       "      <td>onboarding-aryaman-chakravarthy-intern.md</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>What Do Solutions Engineers Do?\\n-------------...</td>\n",
       "      <td>.</td>\n",
       "      <td>onboarding-rajat-jain.md</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>681</th>\n",
       "      <td>Zero Trust Security Boot Camp\\nLearn what zero...</td>\n",
       "      <td>knowledge</td>\n",
       "      <td>Zero_Trust_Security_Boot_Camp_Course_Brochure.md</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>682</th>\n",
       "      <td>Zero Trust Security Boot Camp\\nLearn what zero...</td>\n",
       "      <td>knowledge</td>\n",
       "      <td>Zero_Trust_Security_Boot_Camp_Course_Brochure_...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>683</th>\n",
       "      <td>note**Strategic Opportunity:** We can maximize...</td>\n",
       "      <td>leadership</td>\n",
       "      <td>aiaccelerated_development_atlassian_github_int...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>684</th>\n",
       "      <td>By  and  \\n\\n## Introduction: *The Dawn of a N...</td>\n",
       "      <td>leadership</td>\n",
       "      <td>the_value_proposition_of_the_enterprise_ai_arc...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>685</th>\n",
       "      <td># Emergent Categories in the Viable System Mod...</td>\n",
       "      <td>leadership</td>\n",
       "      <td>viable_system_model.md</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>686 rows × 3 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 2
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-11-14T20:03:07.504528Z",
     "start_time": "2024-11-14T20:03:05.469720Z"
    }
   },
   "cell_type": "code",
   "source": [
    "vectors = to_vectors(df, rebuild_cache=True)\n",
    "vectors"
   ],
   "id": "313031b0d9899f11",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.00180037,  0.01510528, -0.00583903, ...,  0.01206134,\n",
       "        -0.06565797, -0.03174061],\n",
       "       [ 0.04978242, -0.03718074, -0.01629087, ...,  0.03849672,\n",
       "        -0.07964627,  0.00194491],\n",
       "       [-0.02221449, -0.03632433, -0.03482333, ..., -0.00974571,\n",
       "        -0.01893803, -0.04727153],\n",
       "       ...,\n",
       "       [ 0.06985775,  0.05197032, -0.03591631, ...,  0.04963623,\n",
       "        -0.01358351, -0.03230008],\n",
       "       [ 0.05722989, -0.0302305 , -0.03664086, ...,  0.0304099 ,\n",
       "        -0.00978195, -0.03099633],\n",
       "       [ 0.01339594,  0.00665453, -0.01832449, ..., -0.00352714,\n",
       "         0.05654123, -0.01874361]], dtype=float32)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 4
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-11-14T20:03:07.806654Z",
     "start_time": "2024-11-14T20:03:07.575119Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# PCA\n",
    "from sklearn.decomposition import PCA\n",
    "pca = PCA(n_components=2)\n",
    "pca_vectors = pca.fit_transform(vectors)\n",
    "# plot \n",
    "plt.scatter(pca_vectors[:,0], pca_vectors[:,1], s=10)\n",
    "plt.title('2D representation of documents')\n",
    "plt.show()"
   ],
   "id": "2fda33e0df55e2",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ],
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAi8AAAGzCAYAAADnmPfhAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8hTgPZAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB3DUlEQVR4nO2de3gV1bn/vwmQIJdcPARiIJGLyE2EDQIFoSigAa1K4EikKkKtVuSiYk/B87Qi2hboUQQR5ByPtv5UNFoJ9bQqYgCPCJoogSJCBEV2EANBQsIByyWZ3x90tnvPnvue+/5+nodHM3sua9bMrPVd73rfd6UIgiCAEEIIIcQnpLpdAEIIIYQQI1C8EEIIIcRXULwQQgghxFdQvBBCCCHEV1C8EEIIIcRXULwQQgghxFdQvBBCCCHEV1C8EEIIIcRXULwQQgghxFdQvBBCPEVKSgoeeeQRt4uhSkVFBYYNG4bWrVsjJSUF27dvN3T8n/70J6SkpODrr7+2pXyEBB2KF+JZKioqMHPmTPTp0wetW7dGQUEBJk2ahC+++CJu36uuugopKSlISUlBamoqMjIy0KNHD9x+++1Yv369C6UPLocOHcIjjzxiuMOO5q233vK8QFHi7NmzuPnmm3Hs2DE8+eSTePHFF3HxxRe7XazAcOrUKTzyyCPYtGmT20UhHqa52wUgRInFixfjww8/xM0334zLL78cNTU1ePrppzFgwAB89NFHuOyyy2L279SpExYuXAgAOHnyJPbt24c1a9bgpZdewqRJk/DSSy+hRYsWbtxKoDh06BAWLFiAzp07o3///qbO8dZbb2HFihWyAub7779H8+bebZq+/PJLHDhwAM8++yx+/vOfu12cwHHq1CksWLAAwPlBCSFyeLeFIEnPnDlzsHr1aqSlpUW2FRcXo2/fvli0aBFeeumlmP0zMzNx2223xWxbtGgRZs+ejZUrV6Jz585YvHix4XKcOnUKrVq1MncTBjl58iRat27tyLW8SsuWLd0ugipHjhwBAGRlZblbEEKSGYEQnzFgwABhwIABMdtGjhwp9OnTR3b/c+fOCb179xZatWolHD9+XPXc4nk++eQTYcSIEcIFF1wg3HfffYIgCMI//vEP4eGHHxa6desmpKWlCZ06dRL+7d/+TfjHP/4Rcw4AwowZM4SXXnpJuPTSS4X09HRhwIABwvvvvx+z3/z58wUAwq5du4TJkycLWVlZQv/+/SO/v/jii8KAAQOEli1bCtnZ2UJxcbEQDodjzvHFF18IEyZMEDp06CCkp6cLHTt2FIqLi+PuU8+5xHvftWuXcNVVVwkXXHCBkJeXJyxevDiyz8aNGwUAcf/++Mc/CoIgCP/7v/8r/Ou//quQn58fqaP7779fOHXqVOQcd9xxh+w5outv/vz5MWXbtm2bMHbsWKFt27ZC69athVGjRglbt26N2eePf/yjAEDYvHmz8MADDwjt2rUTWrVqJYwfP144cuSI9FHLUlZWJgwfPlxo1aqVkJmZKdx4443C559/rlr2kSNHqp7zs88+E66++mqhZcuWQseOHYXHHntMeO655wQAwv79+2P2XbFihdC7d28hLS1NuOiii4R7771XqKurizvnRx99JIwbN07IysoSWrVqJfTt21dYunRp5PeRI0fKluuOO+4QLr744sjf+/fvFwAI//Ef/yE8/fTTQpcuXYQLLrhAuOaaa4RwOCw0NTUJjz76qNCxY0ehZcuWwo033ih89913ced96623IvXWpk0b4brrrhM+++yzuGu3bt1aOHjwoHDTTTcJrVu3Ftq1ayc8+OCDwrlz52LKI/0nvg/ffvutMHXqVKFjx45CWlqakJubK9x4441x9UiCDy0vxFcIgoDDhw+jT58+uo9p1qwZJk+ejN/85jfYvHkzrr/+etX9v/vuO4wbNw633HILbrvtNnTo0AFNTU248cYbsXnzZtx9993o1asXdu7ciSeffBJffPEF1q5dG3OO999/HyUlJZg9ezbS09OxcuVKjB07FuXl5XHTXTfffDO6d++O3//+9xAEAQDwu9/9Dr/5zW8wadIk/PznP0dtbS2WL1+OH//4x6isrERWVhbOnDmDwsJCnD59GrNmzUJubi6++eYb/PWvf8Xx48eRmZmp+1widXV1GDt2LCZMmIBJkybhz3/+M+bOnYu+ffti3Lhx6NWrFx599FE8/PDDuPvuuzFixAgAwLBhwwAAr7/+Ok6dOoXp06fjX/7lX1BeXo7ly5fj4MGDeP311wEAv/jFL3Do0CGsX78eL774oubz27VrF0aMGIGMjAz86le/QosWLfCf//mfuOqqq/D+++9jyJAhMfvPmjUL2dnZmD9/Pr7++mssXboUM2fORElJiep13nvvPYwbNw5du3bFI488gu+//x7Lly/HlVdeiW3btqFz5874xS9+gY4dO+L3v/89Zs+ejUGDBqFDhw6K56ypqcHVV1+Nc+fOYd68eWjdujX+67/+CxdccEHcvo888ggWLFiAMWPGYPr06aiqqsIzzzyDiooKfPjhh5Epz/Xr1+MnP/kJLrroItx3333Izc3F7t278de//hX33XefZn3K8fLLL+PMmTOYNWsWjh07hj/84Q+YNGkSRo0ahU2bNmHu3LnYt28fli9fjl/+8pd4/vnnI8e++OKLuOOOO1BYWIjFixfj1KlTeOaZZzB8+HBUVlaic+fOkX0bGxtRWFiIIUOG4PHHH8d7772HJ554At26dcP06dORk5ODZ555BtOnT0dRUREmTJgAALj88ssBABMnTsSuXbswa9YsdO7cGUeOHMH69esRDodjrkOSALfVEyFGePHFFwUAwnPPPRezXc3yIgiCUFpaKgAQli1bpnr+kSNHCgCEVatWxV03NTVV+OCDD2K2r1q1SgAgfPjhh5Ft+Odo8ZNPPolsO3DggNCyZUuhqKgosk20vEyePDnmnF9//bXQrFkz4Xe/+13M9p07dwrNmzePbK+srBQACK+//rri/eg9V/S9/7//9/8i206fPi3k5uYKEydOjGyrqKiIsbZEE21hEVm4cKGQkpIiHDhwILJtxowZglLzA4nlZfz48UJaWprw5ZdfRrYdOnRIaNu2rfDjH/84sk20vIwZM0ZoamqKbH/ggQeEZs2aaVrd+vfvL7Rv3z7GsrBjxw4hNTVVmDJlSmSbaH1Sq3eR+++/XwAgfPzxx5FtR44cETIzM2MsL0eOHBHS0tKEa6+9VmhsbIzs+/TTTwsAhOeff14QhPNWxC5duggXX3xxnEUm+p6NWl5ycnJi6uehhx4SAAj9+vUTzp49G9k+efJkIS0tLWJtPHHihJCVlSXcddddMdepqakRMjMzY7aLVqtHH300Zt9QKCQMHDgw8ndtba2s9a2uri5iJSKE0UbEN+zZswczZszA0KFDcccddxg6tk2bNgCAEydOaO6bnp6OadOmxWx7/fXX0atXL/Ts2RNHjx6N/Bs1ahQAYOPGjTH7Dx06FAMHDoz8XVBQgJtuugnr1q1DY2NjzL733HNPzN9r1qxBU1MTJk2aFHOt3NxcdO/ePXIt0bKybt06nDp1SvZe9J4rup6i/YbS0tIwePBgfPXVV5r1BiDGonDy5EkcPXoUw4YNgyAIqKys1HWOaBobG/Huu+9i/Pjx6Nq1a2T7RRddhJ/+9KfYvHkzGhoaYo65++67kZKSEvl7xIgRaGxsxIEDBxSv8+2332L79u2YOnUqLrzwwsj2yy+/HNdccw3eeustw2UHzjsm/+hHP8LgwYMj23JycnDrrbfG7Pfee+/hzJkzuP/++5Ga+kOzfNdddyEjIwN/+9vfAACVlZXYv38/7r///jifm+h7NsrNN98ceZ8ARKxZt912W4zz9JAhQ3DmzBl88803AM5bgY4fP47JkyfHvF/NmjXDkCFD4t4vIP59HzFihK7364ILLkBaWho2bdqEuro6U/dJggOnjYgvqKmpwfXXX4/MzEz8+c9/RrNmzQwd/3//938AgLZt22ru27FjxxgnYQDYu3cvdu/ejZycHNljRCdOke7du8ftc+mll+LUqVOora1Fbm5uZHuXLl3iriUIguw5AESmD7p06YI5c+ZgyZIlePnllzFixAjceOONuO222yIdkd5ziXTq1CmuE8zOzsbf//532eOlhMNhPPzww3jzzTfjOpj6+npd54imtrYWp06dQo8ePeJ+69WrF5qamlBdXR0zjVhQUBBXfgCqHZ4obJSus27dOlPO1AcOHIib1pK7jtL109LS0LVr18jvX375JQDETT0mirTOxPcnPz9fdrtYl3v37gWAiIiXkpGREfN3y5Yt476h7OxsXWIkPT0dixcvxoMPPogOHTrgRz/6EX7yk59gypQpMd8TSQ4oXojnqa+vx7hx43D8+HF88MEHyMvLM3yOzz77DABwySWXaO4r54/Q1NSEvn37YsmSJbLHSBt5I0iv19TUhJSUFLz99tuyIk20IgHAE088galTp+Ivf/kL3n33XcyePRsLFy7ERx99hE6dOhk6FwBFUSj80xdHjcbGRlxzzTU4duwY5s6di549e6J169b45ptvMHXqVDQ1NWmewwoSuYegkJKSInu/UqufiFKdadWl+ExffPFFWQEhDXk3OuiQcv/99+OGG27A2rVrsW7dOvzmN7/BwoULsWHDBoRCoYTOTfwFxQvxNP/4xz9www034IsvvsB7772H3r17Gz5HY2MjVq9ejVatWmH48OGmytGtWzfs2LEDo0eP1mWeF0ek0XzxxRdo1aqVovUm+lqCIKBLly649NJLNa/Vt29f9O3bF7/+9a+xZcsWXHnllVi1ahV++9vfGj6XHpTuf+fOnfjiiy/wwgsvYMqUKZHtckkC9U5x5OTkoFWrVqiqqor7bc+ePUhNTU1IOIqISeaUrtOuXTtTIewXX3yx7LsgvU709aOnx86cOYP9+/djzJgxAM6/G8B5MS5ukyM7O1t2KkZt6swMYnnat2+vWh4jaL0b3bp1w4MPPogHH3wQe/fuRf/+/fHEE0/EpU4gwYY+L8SzNDY2ori4GFu3bsXrr7+OoUOHmjrH7NmzsXv3bsyePTvOjK2XSZMm4ZtvvsGzzz4b99v333+PkydPxmzbunUrtm3bFvm7uroaf/nLX3Dttddqjj4nTJiAZs2aYcGCBXGjZ0EQ8N133wEAGhoacO7cuZjf+/bti9TUVJw+fdrQuYwgduLHjx+P2S7eV/R1BEHAsmXLdJ9DSrNmzXDttdfiL3/5S0wq/cOHD2P16tUYPny46WcazUUXXYT+/fvjhRdeiCnTZ599hnfffRfXXXedqfNed911+Oijj1BeXh7ZVltbi5dffjlmvzFjxiAtLQ1PPfVUTP0999xzqK+vj0TIDRgwAF26dMHSpUvj6i76uG7dumHPnj2ora2NbNuxYwc+/PBDU/ehRGFhITIyMvD73/8eZ8+ejfs9+vp6EXMqSe/v1KlT+Mc//hGzrVu3bmjbtm3kfSfJAy0vxLM8+OCDePPNN3HDDTfg2LFjcSMraUK6+vr6yD6nTp2KZNj98ssvccstt+Cxxx4zXZbbb78dr732Gu655x5s3LgRV155JRobG7Fnzx689tprWLduHa644orI/pdddhkKCwtjQqUBRDKHqtGtWzf89re/xUMPPYSvv/4a48ePR9u2bbF//36Ulpbi7rvvxi9/+Uts2LABM2fOxM0334xLL70U586dw4svvohmzZph4sSJhs5lhG7duiErKwurVq1C27Zt0bp1awwZMgQ9e/ZEt27d8Mtf/hLffPMNMjIy8MYbb8j6M4jOzLNnz0ZhYSGaNWuGW265RfZ6v/3tb7F+/XoMHz4c9957L5o3b47//M//xOnTp/GHP/zBUNnV+I//+A+MGzcOQ4cOxZ133hkJlc7MzDS9lMGvfvUrvPjiixg7dizuu+++SKj0xRdfHONHlJOTg4ceeggLFizA2LFjceONN6KqqgorV67EoEGDIu96amoqnnnmGdxwww3o378/pk2bhosuugh79uzBrl27sG7dOgDAz372MyxZsgSFhYW48847ceTIEaxatQp9+vSJc3BOhIyMDDzzzDO4/fbbMWDAANxyyy3IyclBOBzG3/72N1x55ZV4+umnDZ3zggsuQO/evVFSUoJLL70UF154IS677DKcO3cOo0ePxqRJk9C7d280b94cpaWlOHz4sOK7QwKM4/FNhOhEDN1V+qe2b5s2bYTu3bsLt912m/Duu+8auqZSyPWZM2eExYsXC3369BHS09OF7OxsYeDAgcKCBQuE+vr6yH6ISlLXvXt3IT09XQiFQsLGjRtjzieGStfW1spe74033hCGDx8utG7dWmjdurXQs2dPYcaMGUJVVZUgCILw1VdfCT/72c+Ebt26CS1bthQuvPBC4eqrrxbee+89w+dSu3dpeK0gCMJf/vIXoXfv3kLz5s1jwqY///xzYcyYMUKbNm2Edu3aCXfddZewY8eOuNDqc+fOCbNmzRJycnKElJQUXUnqCgsLhTZt2gitWrUSrr76amHLli0x+4ih0hUVFTHbxdBmaf3L8d577wlXXnmlcMEFFwgZGRnCDTfcEJOkLvp8ekKlBUEQ/v73vwsjR47UlaTu6aefFnr27Cm0aNFC6NChgzB9+nTZJHWbN28WrrnmmkjSvssvv1xYvnx5zD4vvfSS0LVrVyEtLU3o37+/sG7dOtUkdXruUa2OCwsLhczMTKFly5ZCt27dhKlTp8akCxCT1EkRv4NotmzZIgwcOFBIS0uLvA9Hjx4VZsyYIfTs2VNo3bq1kJmZKQwZMkR47bXX4s5Jgk+KICSRFxshDpCSkoIZM2YYHnESQgjRB31eCCGEEOIrKF4IIYQQ4isoXgghhBDiKxhtRIjF0I2MEELshZYXQgghhPgKihdCCCGE+IrATRs1NTXh0KFDaNu2bUKrrBJCCCHEOQRBwIkTJ5CXlxezurocgRMvhw4dsmStE0IIIYQ4T3V1NTp16qS6T+DES9u2bQGcv3kr1jwhhBBCiP00NDQgPz8/0o+rETjxIk4VZWRkULwQQgghPkOPywcddgkhhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKwK3PAAhhCRKZbgO+4+eRJd2rREqyHa7OIQQCRQvhBASxaK3d2PV+19F/r5nZFfMG9cr8jeFDSHuQ/FCCAkkJRVh7Kg+jn75WSgeVKDrmMpwXYxwAYBV73+Fwj65CBVkawobQogzULwQQgLH+BWbsb26HgCwurwar5SHsXbGcM3j9h89qbpdTdgQQpyDDruEkEBRUhGOCBeR7dX1KKkIax7bpV1rxe1awoYQ4hwUL4SQQLGj+rih7dGECrJxz8iuMdumj+yKUEG2qrAhhDgLp40IIYGiX34WVpdXy27Xw7xxvVDYJzfOKVcUNtFTR6KwIYQ4S4ogCILbhbCShoYGZGZmor6+HhkZGW4XhxDiAtE+LwAQys9EqQ6fFz1oRRsxGokQcxjpvyleCCGBxEy0USJUhuvwVNlebKyqjWxjNBIh+qF4oXghhDiINIQ6mtJ7h9ECQ4gOjPTfdNglhJAEkMsNEw2jkQixHooXQghJAC1xwmgkQqyH4oUQQhJATZwwGokQe6B4IYSQBJDLDTOqZw5K7x2GuXTWJcQWmOeFEEISRCk3DCHEHiheCCHEAkIF2RQthDgEp40IIYQQ4isoXgghhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKyheCCGEEOIrmrtdAOJPKsN12H/0JLq0a41QQbbiNkIIIcRqKF6IYRa9vRur3v8q8vc9I7sCQNy2eeN6yR5PkUMIISQRKF6IISrDdTEiBUDc3+K2wj65ceJETvgoiRwiD8UfISTZoXghhth/9KShfaM7VyXh06Vda7RolsrOWAdy4q+wTy7FDHGNkoowdlQfR7/8LBQPKnC7OCRJoHghhujSrrXpfZWEz9w3dkb+n5YYZZTEHy1ZxC3Gr9iM7dX1AIDV5dV4pTyMtTOGu1wqkgww2iiJqQzXYc22g6gM16lui2bdrpq4bdNHdo34vURvk1oB9AifVe9/pXjtZEeP1Yv1R5yipCIcES4i26vrUVIRdqlEJJmg5SVJMeN0KzfyB4Br/+nbojV9ESrIxj0ju8qeIxrpdBM5j16rF+uPOMGO6uOK2zl9ROyG4iUgiE6cZxubNP1HzDrdbqo6Inu+TVVHECrIjvxTY964XhGRc7axKWbKSMTI1FQyoVf8sf6IE/TLz8Lq8mrZ7YTYDcVLAJBaUUSU/B/MON1Whuuw+9sG1X31RsFEi5z9R0/GlF1uuon8QLT469KuNdbtqmH9EVcoHlSAV8pjp45C+Zm0uhBHSBEEQXC7EFbS0NCAzMxM1NfXIyMjw+3i2E5luA5FK7co/l5677C4zkzrGOnx0g5SyuKJfeNEyNU9cjB7dHddHSlDfxOD9UfchNFGxCqM9N8ULz5nzbaDmPPaDsXfl0zqhwkDOsVtl1prJoTysKbyUNx+iyf2lZ3a0QujXwghRqa1SfJipP/mtJHP0fJvUPpdOv2w/+hJWfGi5JSnF6VkdYSQ5MDotDYhenAkVHrFihXo3LkzWrZsiSFDhqC8vFzXca+++ipSUlIwfvx4ewvoY0QnTjm0/B9CBdmYMKATQgXZiiLHCuc7Iz42hJDgoBShCDCsnySG7ZaXkpISzJkzB6tWrcKQIUOwdOlSFBYWoqqqCu3bt1c87uuvv8Yvf/lLjBgxwu4i+h5pBI8Zs6xcJMv0kV1RPKggzp9FaYpJCUa/EJKcaA1cGNZPzGK7z8uQIUMwaNAgPP300wCApqYm5OfnY9asWZg3b57sMY2Njfjxj3+Mn/3sZ/jggw9w/PhxrF27Vnbf06dP4/Tp05G/GxoakJ+fnzQ+L1aj5Pwp3S41BU8f2RXX9snF8g17sWFPbcz2uTQNE5KUmAkoIMmLZ3xezpw5g08//RQPPfRQZFtqairGjBmDrVu3Kh736KOPon379rjzzjvxwQcfqF5j4cKFWLBggWVlTnaUcrVIt0t9ZsTfnp86mNEvhBAA6rmJGNZPEsFW8XL06FE0NjaiQ4cOMds7dOiAPXv2yB6zefNmPPfcc9i+fbuuazz00EOYM2dO5G/R8kLsR6/QIYQkL1ZMaxMixVPRRidOnMDtt9+OZ599Fu3atdN1THp6OtLT020uGSGEELNwQEOsxlbx0q5dOzRr1gyHDx+O2X748GHk5ubG7f/ll1/i66+/xg033BDZ1tTUdL6gzZujqqoK3bp1s7PIRCecGiKEEOIWtoqXtLQ0DBw4EGVlZZFw56amJpSVlWHmzJlx+/fs2RM7d8YmRPv1r3+NEydOYNmyZZwO8ghyizoyXwMhhBCnsH3aaM6cObjjjjtwxRVXYPDgwVi6dClOnjyJadOmAQCmTJmCjh07YuHChWjZsiUuu+yymOOzsrIAIG47cQelRR29moiOFiJCCAketouX4uJi1NbW4uGHH0ZNTQ369++Pd955J+LEGw6HkZrqSK48X+OVTlgpb4MX8zXQQkQIIcGEaxv5AGknXBTKw5PFIVfKopS3wQv5GqIFHgDPlpMQQkg8nsnzQhJHbpqm9J/Zbd0QMEqZeN0WBFKBd3WPHNn9vGghIoQQYgyKF4+jNE1TWnkIU4Z2dqUjVkpQ5xZyAm9jVa3svlyqwF68Mr1JCAk2FC8eR62zddOK4KW8DUoCb1TPnLilCrxS5iDyQEllxCoI0MeIEGIfFC8eJ1SQjaJQXkynIEIrwnmU6mHWqO6YNao7LQEOIBUugLej0Agh/oZhPj7gyeIQikJ5MdvssiJUhuuwZttBXy1VL/rhRCPWT6ggGxMGdGIHaiOV4TpZcQ1orypMCCFmoOXFJzxZHMKUoZ1ttSIkElrstq+D1/xwkgk1gULroHeQfqNuf7OEJALFi4+w088kkeRzcqLHDSHhJT+cZEJJoEwI5fF5eATpN9o/PxPbq+sjf9M/ifgNThsRAOrJ59RQEj1FK7dgzms7ULRyCxa9vduychLvITdtNyGUhyUu5SIisch9o9HCBTj/zfppqpgQWl4IAOXRs5bZX49PAx03gw+n7byLXr8j5kAifoKWFwJA3ekVUHbk1evTQMfN4EPnaG+i9xulfxLxE7S8eBC3HOmURs9qjrxyGXflYMPoHUoqwthRfRz98rNQPKjA7eIQm5H7RqU+L8yBRPwG1zbyGF5bTFDvWkbRgmvdrpq45QPm0hnQE4xfsTmm0+qfn4m1M4a7WCLiFIw2Il6Haxv5lEQifuxC7yrS0ZE+oYJs+j94kJKKcJyj5vbqepRUhGmBSQKk0XiMziN+hj4vHmJT1RFD253ArCMv/R+8x47q44a2E0KIV6F4IapoOfLK4ccsvclAv/ws2e3ZrdOcLQghhCQIp408xFU92mNZ2T7Z7W5iJAzWaz475AeKBxXglfL4qaMVG79EY5PA50QI8Q20vHgIM1YOp9AzDaTks0MLjHdYO2M4ZlzdLW47nxMhxE/Q8uIx/JzsS69zL3GXbjltZLfzORFC/ALFiwfxaxSAWede4ix8ToQQv8NpI2IZXp72Ij/g9nOiQzchJFGYpI5YDpNf+QM3nhMdugnxL3a3GUb6b4oXQogj6M3WTAjxHk4MPIz035w2SiJoriduouTQ7WYSRmI9bGeChxcjSemw6zHsWjQvqOZ6TlH5ByWH4GVl+3D6XFMg3sdkJ6jtTLLjxUhSihcPEb1o3uryarxSHrZk0TyzayZ5XRhIG8qiUB6eLA65WCKihtoK5NL30evvHonHi2uzEWvwYoQip408gtqieYmippqVWPT2bhSt3II5r+1A0cotWPT27oTLYSVyDWVp5SE8UFLpUomIHuaN64X7Rl8i+5v4Pnr93SPymGlniD9wO0JRDooXj2DnonlGVbMX5zelKDWIpZWHPFVOEo/Schdd2rX2xbtH5PHi6JxYx7xxvVB67zAsmdQPpfcOw1yXpwMpXjyC0qJ5StuNYFQ1+2EEpdYgeqmcJB6199EP7x6Rx4ujc2ItepaJcQr6vHgEuUXzQvmZljntGll2wKsjKKkfRFEoD6WVh+L2c7ucRBul99Gr7x7Rh5+XNyH+guLFQ6ydMdy2aCNA/7IDco6Vbo+g5KIYROfcaAHjdjmJfuTeRy++e8QYfl3ehPgLJqkjingl4kMruZlXykmsg8+U+AG+p9ZipP+m5YUo4pURlFaOAa+U0+94qSF265l6qQ6It2FOG3eheCG6cathpx+E/bAhDk4dUIDZD3PauA/FC9GFmw07/SDshQ1xcOrA6HdKoWMOL2acTTYoXogmXmjYGcVgH2yIg1EHRr/ToFia3IDWYPeheCGauNGwiyPCs41NaNEsNSJY/NKR+AkrGmK/j+CD0BkZ+U69MCDxM7QGuw/FC9HE6YZdOiIU4cjQHhJtiIMwgg9CZ2TkOw2CpcltaA12F4oXokkiDbvREbnciFCEI0P7MNsQB2kE7/fOyMh3GgRLkxegNdg9KF6ILsw07GZG5Fpp4DkytA8zDXHQRvB+74z0fqdBsDSR5IbixSR+n+M3g5GG3eyIXGvkx5GhPZh9nzmC9x56v1O/W5pIckPxYgKpRaEolBdJVU/OY3ZELjciFJkQyouclw2tdSTis+L3Ebydy3H4Ab9bmoJAMg6ErYDLAxhEKVU9BUwsWin99RwfHW30wd7amDWM/OgU6kUSfU7R5/FbAzx+xeaYhVD752di7YzhLpaIJBtBcHa3EiP9d6pDZQoMShaF0spDqAzXWXadynAd1mw7aOk5nUQckUdjZEQuLr1ePKgAXdq1jls9etX7X/m2bryEmoXMCOLz8otwKamIXcEdALZX16OkIqzreL9/n8R9lKbW9b6DyQ6njQyiNpdvlZOi39W4OAov7JNryZy6Uke6fMNePD91cCJFTXqS1WdlR/Vxxe1a00d+/z6JN1Bq1+a+sRP7j57kO6UBLS8GCRVkoyiUJ/ubFQ2+khr3ywhv0du7UbRyC+a8tgNFK7dg3a6ahEfkSvW6YU+tb+rFqyRqIfMr/fKzDG0X8fv3SbyDWn/Bd0obihcTPFkcihMwVjX4Zsz4XjFh29WwhwqycXWPHNnfjE5vkHjmjeuF0nuHYcmkfii9dxjmJsGIr3hQAfrnZ8ZsC+VnalpdrJpmcwuvtBVEfuAQjV/eKbfgtJFJniwOYcrQzpY7KRo143vJhL2p6oji9kTrZ/bo7thYVRu3PejTG06RjFEna2cMl402UnM+9vM0m5faCnKeeeN6oUu71pj7xs643/zwTrkJLS8JYIeTohEzfjKZsI3Wi9ujSy+UwSn8fK/Fgwrw+wmXR4SLdNpz0du7Y/a3cprNyXpLprbCbxQPKkjKqdtEoeXFg+hNHmWnpcMMV/Voj2Vl+2S3W4GeevHC6NILZXAKv99rtJUFgK7EilYkd3ugpNLR0P+gZUIOGkwYaByKF4/iRzO+EwnL1OrFC+vseKEMTuH3e5UKr1E9lf2qpPeTyPcpFS6A/fXm5+muZMGPbb6bcNrIxyhZNPKyLogxR0vN03aaq910/vSCM6UXyuAUfr5XOeG1YU+8TxVgbQdfGa6LEy4iRuvNyHecyHSXn6cFSXCh5cXHyFk6+udnxjh/9c/PjMsiGv23mrnabNZUK0cQRsrghdGlF8rgFH6+VyWh0Da9GU6cboz8bbXlUE2gGKk3M9N1Ti2uSogTULz4nOgG6WxjU5zXulwW0WiUzNVeWL/JaMPphXV2vFAGp/DzvSoJhROnGzHj6m7oltPGFt8DpetOCOXZvugp4MziqoQ4AcVLABAbpDXbDpo6XjqnL9dolVYewrGTZ/DCz4YkVFa9mG04veD45oUyOIVf7zVUkI1eF7XF7m9PxP1Wd/IMJhR2su26UsE3IZSHJQYGBk453ypdx62AAOIdvLCWGcVLgDBrrpcep9Rovf/FUTxQUumIBSaRBtoLjm9eKINT+PVepw7rLJtfQyvLbqIkKvicmq5TOt+ysn04fa6J00dJilemEumw6yO0HOfknPKkWUSlf8uZ+dUaQasXoFTCz/4UxB+YzbJrBYnkiHJqSQe1DLDMEeNtovsKKx2uvZQviJYXh0jEzFYZrsPyDXtjoiGU1K7cqE56ba2yiOs3qUVF2D3S9rM/BfEPSll2vY5T03XzxvVCevNU2fxNzBHjTaSWkWgStZJ4KV9QiiAIgqNXtJmGhgZkZmaivr4eGRkZbhcHQGJmNrUXsfTeYba+MHc8/zHe/+KobdfVI+i8MLdKSDJTGa5D0cotcdvtbn+UyiLXHrCdOI/Ss4omkedm97tgpP+m5cVmEvHYlzs2GrvV7gs/GxKXUEstJb8dIZh+9acgJCh4xQqq1GZ4xQfDDoy2q3pyBSXSb3jlXQAoXmwnETObUvp/ESf8P/QsQGm08WAIJiH+wu2oMqU2o0u71oFtS8yIMj19QqL9htvvgogjDrsrVqxA586d0bJlSwwZMgTl5eWK+z777LMYMWIEsrOzkZ2djTFjxqju73Xscjx1Uu2qOReaceBSEnRPle2V3c4Mn4S4jx0L0epFqc3405avDe3vF8w6xqo5WQPW9RtuvgsitouXkpISzJkzB/Pnz8e2bdvQr18/FBYW4sgRhUUFN23C5MmTsXHjRmzduhX5+fm49tpr8c0339hdVFtIJDJAKf3/4ol9HU27r4aZFPFKwm1jVW3cx6m1yi8hJPgotRlyeXrU9ncTIxFAiSy9UdgnV3a7l/oNK7B92mjJkiW46667MG3aNADAqlWr8Le//Q3PP/885s2bF7f/yy+/HPP3f//3f+ONN95AWVkZpkyZYndxbcGsmU1pftFLERFmLEuhgmyM6pkju5ZM9HQap5cIIcD5NuPqHjnYWBXfZki3ezEq0WgEUCIWeyWB06JZsDKj2Cpezpw5g08//RQPPfRQZFtqairGjBmDrVu36jrHqVOncPbsWVx44YWyv58+fRqnT5+O/N3Q0JBYoW3CrONpIvOLTnjgm3XgmjWqu6x4if44vRSWR4hZKsN1Ef+1q3q057trktmju8uKl9mju2P26O6u+2AooRV4ITcgS8QxNllyZNkqXo4ePYrGxkZ06NAhZnuHDh2wZ88eXeeYO3cu8vLyMGbMGNnfFy5ciAULFiRcVi9jRvg46YFvRmDp+TiNfoR+zNdBgo30O1xWti9Q0TBOotVmeE20iJiNAFJqV/Xk6ZKrpy8On8Drn1QHpn20Nc/LoUOH0LFjR2zZsgVDhw6NbP/Vr36F999/Hx9//LHq8YsWLcIf/vAHbNq0CZdffrnsPnKWl/z8fE/leXE6B4GX8jJooVU30sZ/+siusvO241dsjls9e+2M4fYUmhAdqOXc8OK36Bf8ltPFytwrRgal0fW04H92+aJ99Eyel3bt2qFZs2Y4fPhwzPbDhw8jN1feqUjk8ccfx6JFi/Dee+8pChcASE9PR3p6uiXltQM3chD4abpFy6qkx6pTUhGWXT27pCIciBEG8SdqqQ68+C36Bb/lfpKzhESjdzrIqA+gWE9BbR9tFS9paWkYOHAgysrKMH78eABAU1MTysrKMHPmTMXj/vCHP+B3v/sd1q1bhyuuuMLOItqKWw6nXpnztGqEpNVY7ag+rrjdjo/TbyM/4jxqDppA8PwPiDrSQRgAw22I2UGp0+2jU9gebTRnzhzccccduOKKKzB48GAsXboUJ0+ejEQfTZkyBR07dsTChQsBAIsXL8bDDz+M1atXo3PnzqipqQEAtGnTBm3atLG7uJbilgXEC1kQnbQ49cvPwuryatntVhPkbJ7EGrQcNPvnZ1L0GiAogwXpIMyp1cSdbB+dxHbxUlxcjNraWjz88MOoqalB//798c4770SceMPhMFJTfwjheuaZZ3DmzBn867/+a8x55s+fj0ceecTu4lqKmxYQO7MgajUmTlucigcV4JXyWNOoHasDM3T7B4LSodiBloPm9up6VIbrPFFvXn+O0sHCqJ45mDWquyfLajdmB6VOtY9Ow4UZbUavw6lf0GN5eHJ9lewqtEsm9cOEAZ1sK5vd0UZrth3EnNd2xG23+768Bq1P6uhx0PTCO+P156hWj14rq5OYFZx+iMb0jMMu8c46EFagtr5Ii2ap6NKuNdbtqlE0mdttcSoeVGDrR+kVXyI3ofVJGy0HTcD9d8YPz1HNguW1sjqJWYdlu9tHp6F4cQC/eccrodSYzH1jp+axXsx6aRQv+BK5jZ8i2Yxi5RRK9KDlg721ulZmdwLxHg985/3nqCXwvFRW4jwUL0Q3ZkeL942+BA9c08Pi0qhjRUckd44gWdLMEFTrkx1TKOKgZcKATporszuBVgQU4K3nqGXB8lJZ/YzX/Z6UoHghutFjDpdDaYFJu7CiI1I7R1AsaWYIovXJiSkUM++MlZ2KVgQU4M3nKA4Wnirb6/n1i/xC9Hslneb3ky8RxQsxhNiYLN+wV3ZtIilONzJWdER+8AdwEy9anxLp6L04FSYVz0WhPIzonmO6vpXu8b7Rl+Dif2ntmecoR6ggG3+cNtiVTOX7j57E2camiE+fV+tIL1rWNz+1cxQvxBR6hMviiX0ddxCzoiPyYmfmNdy0Pkk7sUQtbV6bCpMTz6WVhyJ+M2ZGx0r3kpd1gblCuoCT75xSJ2/WMuGFqRk91jfAP+0cxQsxjJ6FxqaP7OqKZ7sVHZHXOjPyA3IWiWhnWMD46NFrU2Fa35eZ0bHcPfbPz4xxtvfTlIGdqHXyZureKyHpetptwD/tHMULMYzSy118RSfkZrbEVT3au9bwW9ERea0zI+dRskjIYXT06KWpMD2dx6aqI4bLGH2PZxub4qIE/TRlYCdanbzWuxVtZQHg2BS0lnVHz3vlp3aO4oUYRqlz90ryPSs6Ii91ZuQ8ekeOgLnRo1ccsc06xus9d6ggG2u2HZT93S9TBnai9e6o/S61slzdI0d2P6vrWY91R+u9cmOaPxEoXogpvN65W9EReaUzI+dR6jSkU0d+Gj0qIX5fm6qOyGarTjSCj1Ojyqh18mrvlpxlMDpCKhor69lIgIFa9JafhAtA8UISgJ07cRI1i58X8qhYjfh9nT7XZPkUJqdG1ZFOsemJNlKyDI7qmRMT4GB1PRsNMHArestquLYRIcRX+L3RNYNd9+xW+HEQn53SWkyl9w4DANvuW+26fqtjI/03xQtJSoLciBLiRbwSdWMnbi3EG5QFgCleKF6ICsnQiJIfkApVCldnqQzXKfru+NE6oEWi75fZ44PwXnNVaUIUYPbc5EIqVPvnZ2J7dX3kbwpXe9HK6Bq06KZEBYSVA6sgiBk1KF5IUsHsucmDnFCNFi4Ahaud6MnoGqTopkSFh9rAClD3mUlGkU7xQpKKD/baH7pI4nFjFKg3LwyFqz1o1X+QopsSseiWVISxo/o4GpvkPTik68hJhUiyinSKF5I0VIbrZDOyTgjl6fqog26GtQu3fIz0ClIK13jEDrVffpbp/B9K9Xrf6EtczcJtB2YtuuNXbI4TGlKk68hJhUiyivRUtwtAiFMofeTDu8dmwawM12HNtoOoDNdFti16ezeKVm7BnNd2oGjlFix6e7etZQ0KSiPS6Lq1CzGXSTT98zNj/g7S6N8qxq/YjLlv7MTq8mrMfWMnxq/YbOo8cvU/fWRXPHBNj8DVuZmkfyUVYU3hMqqncoZePdfQWxY/QssLSRr0NDByVoLCPrmBdfK125rkto+RXCZoWtCUketQt1fXo6QibMoC4/VM3FZhJunfjurjstuv7d0eYy+7KNIuSS0vQGybpbToZvRzDKJIp3ghgUOpc9JqYJSsBOnN5Q2UfjfDOjGd44U09NJM0G5khvaLYFLqUHdUHzc9fZQsmbiNCrV++VlYXV4dt310rw6YMKBT5G89oigZRTrFCwkUWh2yWgNjZOE/wN9mWKdCxpmG3l95hZQ61H75Wc4XxocYEWrFgwrwSnmspSuUnxknEvWKIi+IdCeheHEZP6tjr5Vdb4es9FEriZGrerS3ZX0ZN3FyOidZpg7kUHonu7RrbelCeFZ9i3o7VDfKZjVWlSuR86ydMVyXc3TQhYgZKF5cxE8jMileLHuiHbKalSBUkB1Z5RdIfFVft3F6OidZG1+ld3LuGzux/+hJS74Zq79FvR2qG2UzgpqosKpcVpyneFCB71Z09gIULy7h50yvXi27FTlc1KwE63bVRO57Wdk+Twg2s3A6xxnONjYp/mbFN2PXt2hFh2pV2cxYNtREhZXlcqMdFJdbABC4kHMjULy4hNtRGImgt+xOmosTzeESjZyVwKuCLRGkQg0AnlxfBSC5G0Wr0EqNDyT+vYudmNx2t5+fFW2cGcuG1rdqVZ250YZL68Pvg6hEoHhxCS9EYZjFbMixnR+Y3hwuVp/fD2JTDVGosVG0Fj2p8QF/fO9mSbSNMztg0PpWN+87quv6Wjjdhiu9U34fRJmFSepcQimBkxdfQGnSNq2yu5GYzO6GxM9iUwu1RtGJZHJBRE/kmhXfu5LvlRd8shJt49REiBpq32pluA6fHjgu+7vROnO6DVe7b6ORkkGAlhcX8UMUhpIFxUzIcaJWCrW5Xrt9OLzmI2LllJxWo5jo+b0abaIHs2VX6kAXT+yLFs1SLasLr7+XibRxZgcManWyZttB2WNG9cwxVWdOtuFq9x2EQZRRUgRBkF8Nyqc0NDQgMzMT9fX1yMjIcLs4vqYyXIeilVviti+e2FfVmU/puNJ7h5n+uOX8B+SmNezuKL3QEVs9Jaf0vIDEnhngzag0vSRadunx00d2xVyb7j2I76XcOY3UoVyd2NE2OYlcO2jne+U0RvpviheiyJptBzHntR2yv2k1THrFhh7s7Fz9hl2Nrx2NopVldbpztqrsdpbbC4Iluix2iQKr79NJUamG2fsKcrSRkf6b00ZEETVTpJaTmJH1gLQ+4iBNayR6LSun5KLLIpq/rWwUlcr6+ifnM7jaGXGSKFbVs135bbyWP8VOh3ar69AL0/WJPL9kzZkkheKFKCI3dxyNWsOktzHT8xHbOdfrZCdgxbWschxWKotWo2hEfCmVaXV5NVaXV1sS9moXXnbQdjNsX+m98XJ9yeGmAAhi2gU3YLQRUWXeuF5YPLGv7G9mREX0dr1RSXJe/UDijolORkVZdS0rIhzMlmXR27tRtHIL5ry2A0Urt2DR27s1y1oUylP8Xc81zUacAPFRckbwcjSgWp0kcs9aqL03St/oul01lpfD76jlmSH6oeWFaHJph7a4ukcONlb9kMFWqyHXEwVhxNTs5LRGIqZuJcuElddK1OxtpixmR4sjuufIJg/Uc03AvAXECiuXF6YX5FC69w/21sb4qCnds9mpS633xshUsRfwks9QIgTlPoxC8UJUkXYCo3rmYNao7ro+EqXGX/zYlFKnKzXOVpt6rTZ1S+uqKJSHJ4tDpq+l1iglUhdGyiKW4cB35sSXVl0mEvaqhJVmebenF+Sev1ydTAjlYY1EJMrdcyKiTulZid+xnxI5uukzdFWP9lhWtk92u1H8HM2XKBQvRBG5TmDDnlrMGtVd9zmkjb/0Yyu48AKEj30f+dtJ07yVOTLk6kq0ODxZHDJ8LTsapejOUE9Z9KS3NyM+1K4ph1ELiJ86USW0nr+0TvYfPRknXoDYe05U1Ck9S3GRycI+ubLHec3vxW2fE6vaHbfvw20oXogiVncCch+bKFyMWHTEc1lhKrVqakCprkorD2HK0M6GEnbZ0SjJdYal9w5TLIue9PZmxMfZxiZTSdqMWED85jwqRe/z11Mn0fdsxfcsOufOfWOnbPm8lDBPCS+IWyvaHS/ch5tQvBBFrO4E1JwsjVh0rLZKWDE1oFYnm6qOxDRSdlkOlASdWmc4YUAnQ2W4b/QluPhfWtsqPhLFSouaG5h5/nru2arvuUUz+TiP/UdPetZPKBoz06Z23Eui34TfRXqiULwQRazuBLQ+qqfK9uKP0war7uNVU6kYWSPnnBo9v61HaJlplNQEnZnOUOla4rx89Dm92FH5oRNVwmynpHXPVn3PWuXzeh4SvfXgdX8Sv4v0RGGGXaKJlaMPLT8KrYycSll/l0zqp2hFcJIHSipVo2sAfVlHjWQB1cpuajb7qVwZBEDx+XmtcfczdmaBteJ79kqW2kRQqwc/LSMQpGgjZtgllmLlSEocHf77mp3YXXMi7ne7Qmed4sniEKYM7YxNVUdkIwoAfXPSRiwHWpYVsyM0aRkAKC7TAHjDAuYWVncgdlqOrPie/WzZElGrBz/5k3jd0mUXFC/EcUIF2fj9hL6yHaEdobNOEyrIVvXvOdvYpKuz09soKYWcK20HAL3m1ugyKK3IG40XG3ez6F1Dxq7pBa93Sl4vXyJ4fZBEKF4Cj1dNiomIELdHfXrqVK2Rk0ZqJNrZKTlQitut8hPS03AHpXGXCpJlZfsUVzH3og+Wk3i1jUkEPwySkh2KlwDjlsOZ3sYsERGiNOqzuyHVW6dq+U2kJNrZaY0SrVxkUO2e/Ni4y70vSmHics/JT9MLdmC0jfGT0HF7kJQIfqpns1C8BBS3RoRGGzMrTc92izWjdSo2fq9/cn4hQjUS6ey0RolWmsDl/GD82kgqvS9GVjFP5ukFo9+D16N35PDj1Jgf69kMFC8BxY0RoZsmdCeubTb/BgBN8WK2sxNHWIV9chVHiVabwKUNutHzlFSEsaP6OPrlZ6F4UIGpMiSK2vti5yrmQcLI98DpNWdIpnqmeAkobowIlRqz5Rv24vmp2vlb7Mg2uanqiGUfbSLrEynlgAHMCwkjIyyvmMDHr9iM7dX1AM4LulfKw1g7Y7jmcVabwdU63gkDOslOj8k9J6WVgJNh2sjI95Ds02tOkUz1TPHiccw22m44nClFt2zYU4vKcJ2t0RpKDemysn04fa7JsugPuToFzkfiSJ+R3EKNI7rnWDLdYmaE5bQJXPrullSEI8JFZHt1PUoqwqoWGDvM4Fodr55VzNVyFvndQqM3Gk5vG5PM02tOkkz1TPHiYRJttJ0cbWsln1NS/laZOdWcSfWeT0+DLa3TdbtqYkK+xWektFCjuM6RWGazeH2EJffuNnx/VnbfHdXHFcWLXWZwPR2vmthTW/vJK47LZqfn7LDoBSV6x+uOsEGpZz1QvHgUKzt1N/xNpCgpfys74XnjeiG9eapscjit8xlpsMU6VXtGdosLL4+wlOplxtXdZPfvl5+leC476zERca+29tMD1/QwXSarOsdEpufknl2Xdq0VBZDeNsYrU5dmkWsjvHg/fq9nvVC8eBSvj6yjUYvOANSVv1Jn+2p5GGcbmwyNGCvDdYq/afmlmBGKas/IDnEh7di8OsJSqpduOW3QPz8zZuoolJ+p+oztFmlmxb3W2k9msGp6zOz0HKD87Oa+sTOy8GIiuB29o7UkgJFV1le9/5Vno3rcrmcnoHjxKF4eWUtRKtN9oy9RzUwKKE/3lH9dh/Kv6+JGjEoNjNq0lVanrubsK5cDRNym9oysFhdKHZsXR1hq9bJ2xnBD0xleFWlWl8vK6bEd1ccVt2vVt1r74tWoFb3vk5o41BKOWgM0wLv1E1QoXjyKVxttOZTKqtd8Pm9cL3Rp1zou8ywQO2JUamCUpq30iCdA3dk3+lpA7KKE/fMz446JfkZWiQutjs1r74TWu1s8qMCQRc2rIs3Kcllpae2XnyUbmq82PSei5jtmtjx2ond6TO0bEv9f7jetXElSvFY/QYbixcN4tdGWI9GyKqW4B4Cy3YdxaYe2hv1LLv4XfeXQarDFa0mRmuYB4Np/NobR5070uflpClHE6nfXiyINsK5cVlpaiwcV4JXysKHpuWjUBhNesvwamR5bvmGv7Dn0JiTU00YA3qqfoEPx4nG82mjLkUhZ1T76JgF4/RP5JG9W+ZfMG9cLZbsPY+8RbfOwGnYICiP356VoCDPvg5Pl91pdWWlpNTo9J6V4UAH2Hz3pacuv3umxynAdNuypld3XSEJCuUhDL9dP0KF4MYCXGrugESrIVkzk9t5u+URggHX+JSUV4YSFi1geq9F7f35PC+5k+b1YV2rWKjNtj9HpOSPl8QJ6p8eUrCujeuZE7klv+xEtyEMF2Z6un6CTIgiC4HYhrKShoQGZmZmor69HRkaGZef1YmPnR7Qa4QdKKhUz0UqZPrIr5kY9g0TE5b+v+btqCv/pI7tCQLzPS7TZWloeq9GKhojONyNSeu8wXzSqTpbfb3Xl97Yn+r0F1JMzGv2Go31egPPTY6USnxe9z5uDU/cx0n/T8qKDZFovwk70NMJPFocwZWhn7D96Ege+Oymbs+Wng/Nx8xX5cXWfyLSV0iiu+IpOuGVwQeS80pGWkw2e2v350S8mGqU0+3IRX4nip7pSanvSm6fqckY3ei2r61otClD6/ZsRaXqmx/RaLvW0HxQ43oHiRQd+auy8ihEBKDYileE6WfHSLz8r8kysqn8lJ8fF/9pPtmxKf7uFn0Lrpah1cNKIL6XOzEin4qe6Ump7lpXtw7KyfZZZYeyw7mglr4z+/hMZIOqZHrNiCszvFrCgoRziYSErVqxA586d0bJlSwwZMgTl5eWq+7/++uvo2bMnWrZsib59++Ktt95yopiK+Kmxk1IZrsOabQdVE7g5gZoAVEIcMUXTPz8Tc9/YiTmv7UDRyi1Y9PZuy8q4dsZwLJ7YFz8dnI/FE/vGmZ+9jFxd+cGBUE92ZpFV738l+x4vens3ilZu0f1O+KmutNoYpToxgpJwSPS8enKjiPuo5VqyilBBNiYM6GTqOdtVR8Q8tlteSkpKMGfOHKxatQpDhgzB0qVLUVhYiKqqKrRvH5+NcsuWLZg8eTIWLlyIn/zkJ1i9ejXGjx+Pbdu24bLLLrO7uLL4KedKNF4aKZgVgNEjprONTXHhm1ZP3yXq5OgmXnewlEOp07q2d3u8+3l8xyW1dpodsfulrvSE6Ca6crpdlmU9gztxn0QWVnViKofWd+9hu+VlyZIluOuuuzBt2jT07t0bq1atQqtWrfD888/L7r9s2TKMHTsW//Zv/4ZevXrhsccew4ABA/D000/bXVRV5o3rhdJ7h2HJpH4ovXeYrY6ZVmBkpFBSEca/r/k7SirCThXPEOKISSkXjJ4Rnhm8YrUyQiKjS7MkUk9KndboXh107W/GoifiRl2ZQWx7ru1tfukBJUoqwnjns29lf7PCsjyqZ47ib9EDQDlrmIiahcOo1c0sfra+BxVbLS9nzpzBp59+ioceeiiyLTU1FWPGjMHWrVtlj9m6dSvmzJkTs62wsBBr166V3f/06dM4ffp05O+GhobEC66AV/wb9KB3pGB2ATe7yqOGkw2Il6xWXsauerq0Q1td1s5k6VRCBdmYftUlstYoufWU9FgjpJE60SRqWZa+F1f3yMHs0d0BKEcb6V1YVby3s41NjgVS+NX6HmRsFS9Hjx5FY2MjOnSIHUV16NABe/bskT2mpqZGdv+amhrZ/RcuXIgFCxZYU+AAoadRT2QBNzvKo4VTDUgyR5cZMcFbUU9qolbP1E4ydSpW5vuR+/aB+Og6M8i9FxurajF7dHfNAeBVPdrLihexnVBz7haxayrHL1ONyYLvo40eeuihGEtNQ0MD8vPzXSyRN1i3K17sSRs6IxkqE/1grepknGhAknV+26gVxQlrmh5rZzJ1Klr3qldQKn37zVJTEq6/RN4LtXZCr3O3nVa36PfRSP4aYj22ipd27dqhWbNmOHz4cMz2w4cPIzc3V/aY3NxcQ/unp6cjPT3dmgIHBKWPXLrujlJuk8YmITLHvHzD3pjU2kbCVaV/W9XJ2D19lyxTEdGYsaIo1ccHe2sxYUAnXde1StT6aUo3UazI95PI4o1aJPr9KLUTevyYnLK6aeWvSRYx7Sa2ipe0tDQMHDgQZWVlGD9+PACgqakJZWVlmDlzpuwxQ4cORVlZGe6///7ItvXr12Po0KF2FjVQ6G3A5HKbAEDJJwdR8slB2XModWjSj1mafVYUPWoNr1cSQCXTVISImdFyqEB+SYfSykOYMrSz7vpKJsuJ3SgJhLONTTF/J7p4oxpWfD9y7YTSvS2e2BctmqU69u7oyV9Dfzn7sX3aaM6cObjjjjtwxRVXYPDgwVi6dClOnjyJadOmAQCmTJmCjh07YuHChQCA++67DyNHjsQTTzyB66+/Hq+++io++eQT/Nd//ZfdRQ0MRkY+YobKst2HZZ0B5dATrioVRFpZQb3mIJtsHarZ0fKI7jmyyzkYnWJLJsuJnSiFVs99Y2fEj0gk0cUb1bDj+1ESRU6nNjAa3Zgs/nJOY7t4KS4uRm1tLR5++GHU1NSgf//+eOeddyJOueFwGKmpP4TADhs2DKtXr8avf/1r/Pu//zu6d++OtWvXupbjxY9ozRtLG5TiQQVo0SxVt3jRG64qRSkrqJ0OsnqsOUr7JFOHana07PUptkSev1+ZN64XmqWmYMXGL2O2y31T0rxGVtaFld+PWK7CPrmuDyrMvNtB95dzA0ccdmfOnKk4TbRp06a4bTfffDNuvvlmm0sVbORGPmrWDb0fpFyHJjVJayFtRO1ykNVjzZEuBOm2xcdNzIyWvTzFpuf5e83iZwVq/hhq35QVdeHE+khuPyM9iQOleEXMBwnfRxv5HTtHfVLPeDXrht4PUur0q9RQdm/fGnuPKFtkohtRO0bvSvfbLDUFdSfPoF9+Fj766ru4KY9kNfFGv4d6nW1FvDjFpsea53RIvBMWHi1/DKVvSq0ugNhIGqX7cGp9JC98o9J3HvihjtbtqvGkmA8aFC8uovWxW9nY6bFuRH+QSis6S5NFyTWUIy9th/e/OKpanuhG1I7Ru9L9iqZ0uUiL6GOTqbGxotPx2hSbnvc9EYtfZbgOr5aHUXfqDEb36qDpd+GU9UBtClftm1I67qmyvdhY9UO0oZIjvl6xaLQ983LaAuk7L/5/qCA7Ttis2XbQM8I+KFC8uITWx251Y6fXuiF+kEorOkfvr9SwaAkXuUbU6tF7IlYbr5h43Rqpe2Fka5bo7KtyRD9bsxY/6bf57udHVDNTO1nHahE5agJL6bho4QLIO+KL360c4rpLZtszr/tUKSG2o16b8goSjqwqnSwYWeNFbURh1wqmV/eIXWdEbSQmWkPU9jfagIzpmYP7Rl8SN/UUfU2r1pqRK78eJoTyPNFpO7VmSyJrA3mN6Dqb+8ZO9M/PjPld+v7qecelKFkbxczUcjhZx0r3pGUZkjtObV2iaKItDFKWle3DAyWVptszM8/IK3Alanuh5cUijCpstRHFU2V7ZX8zayqVlm1UzxzMGtVd81xa1hC56Z4JoTyskQmdBYD39tTivT21shFHdhBd/i9r/y8u+kLKFRdn4fahnW0tkx68MFL3+shWilK4vlYOEKMWPzXBIc1MLaJkBTLq6K4Xs1ZMOT+O6ASVSrzz2bcY3auDos+cXCg9cN4qo6eMRu7HS5FjXp7yCgIULxZgprNR8vMA4k21ImY6FLmybdhTi1mjuus6XsuXQa5h+e7kGc2pI6emJqLL/+G+o3FJuR6+oU9kXv+TA8dRtHKL6xkynWz0vBwtZASlOmvRLFXTAVnqA6N272rfoFJ2WqXV0JW2W4FZHyTpcdJ3Q+rzApyfNnv38yPon5+J+0ZfIjvdLEf0flqDGT3347XIwaAMDLwKxYsFmO1s5Dr+NdvkM9uO6pljqjGyuiOUG9lIG5YXfjYE1yzZpBptpKcMVo+i5JJyVYbr4sSi2xkynW70vBgtZJRE6syI1VQpKk8tO62fOzG5d0N0VpZm4d5eXY8rL2knex6t6MNEBzNS4WLFORMlKAMDr0LxYgGJNE7Sjl/pGDlLiZ7O3cqGU62Rl5Zl/ZyrIkIhu3Wa7JTNge/O+/fIld2uSCxpUi49fgdON4JuNHpeixYyyv/b+rWp48xYTcUOXSvaKPod9XMnJhdV8/on8tF6dSfPGJpKjiaRAZXS1JTbUzRBGBh4FYoXC7Cys9F7Lr2jRavKptbIS/MaiGWJFgqNTULc8WYz7lrpwa9XxDnRCEZ3dmz09JNI52XWMqkl9uTe0dJ7hwXmeaot7Cit0+Pfn9V1TrOWKLUBSKK5oqx4Xn4fGHgViheLsLKzsWrZeyvLptRAPLNpX9yyAnJlEcuwqepI3Jy4kYy74v5a19OL0jSAFCMrJeslegpr/9GTsoKMjZ480R1LIp2XkwkSC/vkWv4OuYXSwo6XdmiLuW/sjNlXj9NvIpYopWeVSOQgQ5y9D8WLhVipsNXOZXYF4ETKphQZobQe0qaqI9hUdf43cTHGUEG2rrKr5eiww5lVKu7+39avE14pWYvxKzZHGn65Eazb8/VeRtqxFIXyZPfT03mpWSbNjryTJcpEzodMyWdPyvSRXXGtRYM9pajHJcUhU+cLWu6joELx4kPccAA0GhkRbV2JnhrSKrvScgNaI7NE7z1a3O0/elL3SslmOriSinBcxIYcfursnApRletYSisPoSiUF/PMjHReRtcB08LPDrpKKD1fqQ+ZnnuMTphn1btipeU7WcSn36F48SFuOHQm2vBGj1yUQsSfXF8lK1yiG7t1u2rifrf63vV2PmY7uB3Vx02Xw0t5LEScNLErdSwjuudgytDOpusmWrwqjby7tGutmjMm+lx+dtCVYjQaa1TPHNWpIrtCxK2yfAdRfAYRihef4rRDpx7fkPtGXwIAinkexJFLYZ9cpDc/34Bd1aM91u2qQdHKLYrnFRs7peymShl7zaKn80nEtKzk7BhNIk7aTuK0iV2tY7Gq81ISSNG+HFp1HxSHazPPd9ao7qrixesiIGjiM6hQvPgYqYOr3R9XdIP8wd7aGDP99JFd8cA1PRTXRALON1rSDjh87JRipEj0cYCz5lytzieRsig5Oz58Qx/LnLSdwmkTuxMdi57OVU/dByHKxKx/ndJAR7Syen2hwqCIzyBD8eJj3BiJiw3yhAGdZM30Sg2X2GjJ+SuoEd0xOW3OVet8Ei2LnLOjeE05vDQPHz115YaJ3e6ORW8EmtPh8250oGafb/QzOtvYFJluk1pZvWA9VCII4jPIULz4BGkj5oWRuNLHHR0WDfwQbaQ3EkGkKJSHa/vkxozSvGLOtaIsUmdHNbwyDy8nmN14JnZ3LNLOVxr+C9hf915Id5/Iey59Rl5os0hwoHjxAXIdxqUd2sru6xWPeLnOxehCdKWVh2Qb70RH3VaNZp00LXtBuCl1PqX3DgukiV0agWam7uWsa3rwUrp7q95zL1kPif+hePE4Sh3G4ol9Zff3sjNcolEG0Y23V5JPJWoBMCKk3J6HV+t8JgzoFOgOyEzdS3P5vFIextoZwzWPU8sYLKbld7qurbB0ecV6SIKBfcuaEktQ6jDe+awG9/zTj0TECx7xleE6rNl2EJXhurjfrGik1LKpql1b/F1OCKrtr3a+RI9Z9PZuFK3cgjmv7UDRyi1Y9PZuzWNEfyM/+T+4jZnnKIeRupfL5bO9uh4lFWHNY9Xe8dXl1brfFa8RKsiOSyjohTaL+BNaXjyOUsewsaoWs0d395S5XsuqodcRUg1pfYiWC2n0k5xFxYjZ2oiFxkgZxH3PNjb5bv7fC1NXRnErvFwpl8+O6uOa00dWRTt5jUVv7475PopCeZjrkA+P247PxHooXjyOXFI2ES+Z6/U640nN7wBk1zsCEJc1VdpRKmXjVbq2XsuBEcdCI2VQ21fE6/P/bk9dGcEtB9HKcB0amwTZ3/rlZ2ke71a0k50dvFJmZCuX3FDCi/mRSOJQvHgYpaRsIl4y1xuxakjnz0MF2Th9riluRD93XC/FrKladSN3bblOYUIoLy5Pjt57MVIGPfsC9j7ToK6Sq3RfbjiIqgnUUH6mbqfdeeN6Ib15qmLOJMDad0Va7qt75GD26O6W1ZNbzrqMcAouFC8eRm3u22vm+kT9IZRG9EodpVrdiMhFN0WHcW/edxRrKg9hzT+tO3rXXzJShgPfnYx0rlpEP1OrR8FeCLu1A7VRtdM+OooZoHu3x+heHQxFGwHnUwyoiZd1u2oseTfkyr2xqhYbq2pVpz6NvJtu+Usxwim4ULx4GKUPO3qtH69ghT+EkRG9nkZPKbpp3a4a2U5Ga/0ladn0lGFZ2T4sK9uHUH6m7O+LJ/aNWy9HrkNOZKpGK+zWr/4AWqNqq310tOpJqaMce9lFmDCgk+HraU0fWWVBUBPWWlOfekWwW/5SfnUyJ9pQvHgYpQ/ebeGi1Ii7nfdEitLChmrHiCMyPfdixAG5sroeWRc0x/Hvz0W2SZ9lZbgOm6qOyHbIZufs1cJu9x89GSfkvGaRURMMekbVVr2T0k67KJSHJyWrVtvRUYrlf/2Tatn1sKywIGiVT23q04iAcqp9kL4zfnMyJ/qgePE4XnOQ1BNRpFVGO5LEya21ZGa6yWhHEz0NpWbiB4Dj35/DjKu7oVtOm7h71+PMK2Kkw1C7X69HPGm9a3rFQqI+OiUVYcVlLaIFTCIdpdo3If4tJ16ssCBoiXAr1xaz219K6Z3xUhtKrIHixQdY9cEnKhrMjLyk17QrSZzSWktS1Br76I7GSDnFMkidjuWoO3kGEwpjpxD0OvNGo7fDULrfCaE8xWk1L/gD6HnXnBhVq4lKuWgZMx2lnnfN7nst7JMre58TQnmRa3hlCkapHdMzjUiCA8VLkmCFaDA68pJec+Sl7fD+F0dj9rFypK+ngZLrBEb1zMGsUT9EVpg1j0d3XC9u/RqVkiRlgHyorB5nXil6OwylCKslxSHFpG1e8AfQ+67ZOao2E9EGGBtsGHnX7LxXpfoe3j0n8v9emIJRa8fonJtcULwkAVaFCxoZecldUypcRJxuXLQ6gUQawWhL0DVLNmHvkR/OpRQqq1Sv942+BFf1aB/nl2K0w1CL5HK7M1LCyLtm16haj6j8YG8tJgzoZNqqafRds+te9da3m1MwWu2YVyxDxBkoXpIAq0YkcgnzzPqWRGO0cTHTUUiPUesErGoE18+5StfCfEr1+sA1PQCc77DkOgwj9aB0v171B7BbWOmpOz3PW/R9MRuG7pUO10h9uzUFo9WOqd2DXyPqiDIUL0mAFQ2kYg6LPrkJnXtUzxxDDr5momOMTpkpOTCayatRPKhANTpMb71GdxiV4Tr89m+f49MDxyO/J+I7ZKYzcqIzsEtYSd8H6bShiN5oskRWf/aS9Sva+Rw4n2fGS+hpx+TemaDmOEp2KF6SACsaSDPmbT0N/6xR3VV/14rC0eMwbMbJOL15vDOrHZE4ifoR2Vk2JZxMt271KF/ufdiwpxYb9sgnZIvuDM82NmHuGzt1XceIVdNL1q/owcGysn2e6uj1tmPR74xWjiPiXyhekoREG0gz1hu1UGZAW0DpjcJR6yisEgd6rmWGRP2I7CybHH5Pt24kIZtIdGe4/+jJOOfnNTJ5dIxO+3ghGsYPz9ZIO6aV48gr90TMQfGSRCTSQJq13kQ7sE4Z2jnGJK11rF6/GbWOwkpxoHUtMxipV6tz1JjBqxEdeqex9CZkU0Ku82yf0dIT0z56MJP0b1PVEU/dj952TO17oROv/6F4IbpJ1HpjVDzpaWD652eqOrBaKQ4A69aTiUZvvarVR3Q+DjvxioNpNErTWHrfh2j03If0PfbStI8aZpP+LSvbh9PnmjwzfaQXtRxHXn1GRD8pgiDIr93uUxoaGpCZmYn6+npkZGS4XRySIHoyz5beO0zTkVfPyLwyXIeilVs0y1R67zBXGj+luhDztrhVDnEFcDdQemZFoTxVJ83KcB2eKtuLjVW1kW1u3ofdKNWT9F1WTcrn0nufCNL7cfpbIcYw0n/T8kI8TfSo9sB3J2VT8CutByTNxqpnikuPk7GZKRI7shsD7izS6SVLg5K1TMtJM1SQjT9OG5w0IbRGkv6lN0+V/c7cnho0g5feVWItFC/E84jCo6QiLPt7Tf0/ZLdvqjpiuNHSE11idIrEilBNpc5nR/VxXNqhbcKNstFO3AsOpoCxZ5FoNlw/Y2S676oe7WXFi1/9RJLlGScb8oubEOJBlNbiyc1sKbt9Wdk+zHltB4pWbsGit3frvo7oYFw8qAD3jOwa85tRZ0ylUE2l1PxKKHUcq8urDd+flEVv70bRyi2m6sptRGtZNBNCebL7+rXzNUJluA5rth2Me7/k6kktCZ1cne4/etLwe0uIXdDnhfgGtXl7qc+LHFpz9moLvpkxO6v50CyZ1A8TBnSS/U0JLf8fMz4Jen0hvI7WAqBB9mcR0ZN/x8i7LO4rTXPgpdwvJFjQ54UEErXIoVBBtqZvjNqcvVrDb9bsbHWopjil9fon1VhdXi17PaPl9Gros1H8GgFkFXpztBh5l8X95ry2Q/O8hDgNxQuxBbscIdU6JbFhrgzX6Z6zrwzX6XL4NYMdoZricXLixYwg8mLos1Ukk6+DXSLUDXGbLE7UJDEoXpIUKxY3VMLK9PHiNc82NqFFs9SYhRWV0Jvbxe5sunLlsCJU02zCQLvPZTV6w9vZ0dknQp0Wt04uPUH8DX1ekhAzDYTeY6z0oVASF3obNLWOTU9OF6v8PuzqYK08r9dEgJ73jR1dLHb5+TjlPxQU/ytiHvq8EEX0zo1Hd2biPkrHRO9rlZlZLVW/3ikdOQuNWNYD36ln09VrfdDT6ds1fWHleb00xaLnHU1kHR6vCTWrsNrPR6ynwj65jvgPBcX/ijgDxUuSoaeBkI60RvXMkT1m+Ya9uLRD25h9iywKU9VK1W+mQZMLW5Zy3+hLdK27BHDkbxd63lGzHV3Qn5lVItSNegqy/xWxHuZ5STK0Ggi5Ee2GPbVyh2DDntq4fUsrD8UJGDkrhlI+Cq1y6v1dih7hMn1kVzxwTQ/NcOo12w6ipCIsO/JnHgxltJ65iJ5OzExHp2St4TOLxa16MpKLhhBaXpIMLQdNpRFtr9y22F1zQtc1RnTPwZShnRXNzHpGdWqp+o02aJXhOkXhct/oS3Dxv7TWZQ7Xs86S1SZuOx2rncTISF6PE7EZR2O/TEu4/fzcrKdkC3En5qF4SULUGgilkevUKzvLpsqXQy0iyIivgjRVf3S0kRHUpqD0ThGp+eBEY6WJ207HaidRe+YAZN9DPZ2Y2j4lFWHsqD6OfvlZkbWf3JiWMCpEvPD83J6+8ZL/FfEuFC9JilIDoTSiLR5UgP1HT8ZtFwBbR79WNGRW5FvR8sExej4tlDr89OapioIrESdWK5EKB6W6k67qLO2o9Tx7uX3Gr9iM7dX1AM7nw3mlPIy1M4Y7HhZuRIjYmW/IKF4OnydEhOKFxKE0olXabsTM68aozop8K3rKN7y7vGOzGZQ6/GVl+7CsbJ9sR2jG3G/1FIWccJh/Qx/ZfaOFC2BNR11SEY5cX2R7dT1KKsIoHlTg2LSEESFpd74hM3D6hngdihcii5plRrrdiHXErVFdoo2xmg+OiJUCTOtcch2hUWFo9RSFknD44vCJuLob1TNH1hF8+Ya9eH7qYNNl2FF9XHG7OH2kFkJvVUe9qeqI4nZpSgKt6Uil52e3bwynb4iXoXghstjZMDoxqpMrf6KNsVhu6XQHYF6AKdWzHrEkHZEbEYZ2TDGpCYffT7g85pkD8lFsG/bUojJcZ7oM/fKzZJdO6JefpXiMm34mWtORSs/PC74xhLgJxQuJw4mG0c5RnZnyGxFrUuECANf+0/nUynKKYmlT1RHdazXpFYaJRJQo1ZWWcJA+86t75MjWZSLTJMWDCvBKeawFKJSfGbG6yN2LHX4mV/VoL/vMrurRPuZvJauKWr4hr/g2EeImFC8Bwgprid8bRjPlNyJ27MwgrLYK8OlzTbqn2vQIQ7O+R2p1ZVQ4zB7dXVa8aOVq0XrH184YLhttJIddYcF6rWBK+z1wTQ/Hy0yIn6B4CQhWWUv8voqs0fIbFTtWORwbLafVU21mfI/01JUR4QDEW1/UymDkHS8eVKB5bcBeB3K9z8zos3U7lJkQL0DxEgCstJb4fRVZI+WvDNfh9U/ipzkA9dBtKxyOzdSz1VNt88b1Qpd2rXULDb2CS49wkFuCYtao7qpRUXZYBO12INf7zKxwegeANdsOMjqIJAUULy5ilcXBSmuJk9FAdnRIesuvFZ6qJiKssIJ4IZdGdB2sLq/G/qMnVYWjVcJWaQmKWaO6R36X1q2dFkE/hgVLy7xuV03Misx04CVBh+LFJay0OFhtLXGqMberQ1KyKIid4tnGJlXhokdEWGEFsaue9YhiM8LRKsGl9tzX7aqR/S7stgjqeZ5up+2XIpbZ735qhJjBNvFy7NgxzJo1C//zP/+D1NRUTJw4EcuWLUObNm0U958/fz7effddhMNh5OTkYPz48XjssceQmZlpVzFdwerGxo5RvBM5HuzqkOQsCgBUBQsA/HRwPm6+It/RBt/qetYris0KRysEl9LzlROV0d+Fm5YqL4cmqz1L8b/Rz8prIowQM9gmXm699VZ8++23WL9+Pc6ePYtp06bh7rvvxurVq2X3P3ToEA4dOoTHH38cvXv3xoEDB3DPPffg0KFD+POf/2xXMV3BDouDH03fdnRISsJQD/3ys3xRb0oYEcWJCMdEBZfSc2/RTH6Re/G7cOsd97plQ+mZfbC3FnNe2xH5W1yx2asijBAj2CJedu/ejXfeeQcVFRW44oorAADLly/Hddddh8cffxx5eXlxx1x22WV44403In9369YNv/vd73Dbbbfh3LlzaN48ODNcdlkc/JgR0+oOSc8aREoodZ5+wYgodtuSIffcK8N1svtGfxduvONeD02We5YTQnlYI1lJXU7Ee0mEEWIEWxTB1q1bkZWVFREuADBmzBikpqbi448/RlFRka7z1NfXIyMjQ1W4nD59GqdPn4783dDQYL7gDuF2x+E1rOyQEhGAVvlPRPvWmF0JO/o8eo83Kort9LkR0+Orrdotfe5e/S78EJosfZb7j56MEy9KeEWEEWIEW8RLTU0N2rePzSTZvHlzXHjhhaipqdF1jqNHj+Kxxx7D3XffrbrfwoULsWDBAtNldQs/TvP4AaUOULr6df/8zJhkalZ1kkpRTEbN82Z8LMx0/nb73CgtIilFFGqFfXI99114VVRJMfssExVh9KEhbpAiCIKgd+d58+Zh8eLFqvvs3r0ba9aswQsvvICqqqqY39q3b48FCxZg+vTpqudoaGjANddcgwsvvBBvvvkmWrRoobivnOUlPz8/YrUhyYlcgyrdZnWjWxmuiwlXlVJ67zBd11E6j5Hj3ehM1O5frexyQk26DpIXOke/ddLSepUT8dNHdsXcBHxevOzITPxHQ0MDMjMzdfXfhiwvDz74IKZOnaq6T9euXZGbm4sjR2JXVT137hyOHTuG3Fz1NWBOnDiBsWPHom3btigtLVUVLgCQnp6O9PR0XeUnyYPcKFRumsLKTkjL30aveT5RHwu3fJ/U7t9ohmMlJ2s3O0e/+ZQpWXetsmx53ZGZBBtD4iUnJwc5OTma+w0dOhTHjx/Hp59+ioEDBwIANmzYgKamJgwZMkTxuIaGBhQWFiI9PR1vvvkmWrZsaaR4hLiKlvldr3neDz4WcqiVT+k3ow7W7ByNoUfEm8Xrjswk2NgSXtGrVy+MHTsWd911F8rLy/Hhhx9i5syZuOWWWyKRRt988w169uyJ8vJyAOeFy7XXXouTJ0/iueeeQ0NDA2pqalBTU4PGxkY7ikmIpYi+EXIY8ZFQOs+6Xfr8xdxCqdxq925GkCUSUUYSozJchzXbDqIyXOdbkU2CgW3xxy+//DJmzpyJ0aNHR5LUPfXUU5Hfz549i6qqKpw6dQoAsG3bNnz88ccAgEsuuSTmXPv370fnzp3tKirxIH7zLxCJNtUnEm1U2CfXlyZ58f71RBsB8s6wWtjZORpZVDLZkPNv8YMjMwkmhhx2/YARhx/iTYw4AfpV5GixZtvBmARjIksm9cOEAZ1cKJEyVjyD6HNIlwiIJlEHUzXGr9gcE4HWPz8Ta2cMt+VafkPNiRzwhkM18T+2OewSYjdGnACDHOngF5O82WcgFTzRfhihgmzHo41KKsIxwgUAtlfXo6QiTAsM1P1bJgzoRNFCHMffKUVJ4NBap0VESeQoZWn1G3L+I14zyZt9Bove3o2ilVsw57UdKFq5BYve3h23T6ggO9Iphgqy0aVda2yqOoIn11fZ8ox3VB83tD3Z8IuYJskDLS/EU+htJJMh0sHriQzNPAM1y5p4rDQ3z1Nle7Gxqjayv97Ed0bol5+F1eXVstuJfxL1keSB4oV4Cr2NZLKMBI2GtTrpA2TmGSgJnuUb9mLDnh8EitwigtFY7bxcPKgAr5THTh2F8jM5ZRSF18U0SS7osEs8iZ5OWC6DqF3OnH7ADR8go89AKwuxEexwXma0ESHuYaT/pnghvsaIpSGokUlA4ksKJHptI/UqFTyjeubEWF304sS9EUKcg9FGJGnQO60S5MgkAJG8KlKc8AEyOrUlnX4AYFi80N+CkOSG4oUEnqCvwaK0kjXgjg+QnqkXqeDRsxL4wIuzMPySdpqJ7wghwYfihQSeIEcmyQmzaJ4q24vZo7s7dp/Rid5Wl1fjlfKwrkRvdi8iSAgJFhQvJPAEOTJJa52fjVW12FhV68g0WaKJ3hJZRDDI/kyEkHiYpI4EHj8kfDOLXgHmRAI/txK9SZPeTftjueq9Ri8uSAjxJ7S8kKQgqDkqjCxsaPc0mRuJ3uSmzdSsTUF33Cby0DIXPCheSNJgNCrGL0QLsw/21qK08pDsfnZPkyklegOAf1/zd1typ6hNm0mdsoPuuE3koWANJpw2IklPEKYRxLWAniwOofTeYRjVMyfmd6emydbOGI7FE/vip4PzsXhiXwgA5r6xE6vLqzH3jZ0Yv2KzpdfTEmTR4kbvulkkOAR9DbRkhpYXktQEcVQWKsjG81MHu2YqLx5UgOJBBY6s1Kw1bfbB3tpIFt4gO24TeYIcaZjs0PJCkpagj8qiV2Z2A6cceOeN64XSe4dh4MVZcb+VVh6KPM8gO24TeShYgwstL8RVzFoHrLAqcFRmL0468IYKsnHrkIvx6YHjcb9FP8+gOm4TebgadnCheCGuYXbKxqqpHqtHZYxoiMXplZr1Ps+gOm4TeShYgwkXZiSuYHYhQasXILRqZWqv+864KaycXKmZK40T4l+4MCPxPGanbMwcp9ZxWzEq83oIrpXCyowIEh14nYCjbEKSA4oX4gpmp2yMHqen4050GsHLvjNWCiuvW5dEOC1ESPBhtBFxhVBBNopCeTHb9DjSGYkYcSqa6Gxjk6HtTmJVbpOgR2YRQvwFLS/EFR4oqYzJBFsUytPtm6B3asBKi4jadEmLZvJjAKXtTmKVU7KXrUuEkOSD4oU4jlS4AOfzcUwZ2ll3R6hnasCqjltrusTLuSSsChX18j2KMNqLkOSB4oU4SmW4TnHtHatH8VZ03Hp8RryeS8KsE6tUDHjpHqVl84s/DiHEGiheiKOo+VrYMYpPNPrE7HSJ1/IPGHViVRIDXojkkZatKJQXJ4iNOiXTakOIv6B4IY6iJFAmhPJs6zQSiT7RM13i9VBpKVodtdb9WH1PRoSDXNkSseRVhuuwfMNebNhTG9lGqw0h3ofihTiK3PTDhFAelhSHXCyVMnqmS/zkzKpnesXJ+zE63WMkSkrLkie9toiXhSch5DwUL8RxvDL9oBet8vrBmRXQbyFy6n7MWKyUyiCdOpr+z3D6NdsOyj4zuWtHs6nqiOffS0KSGYoX4gp+SySmVl6vObMqodei4tT9mLHwKJVt7rhemDK0c0RgrttVE7OMhNSis6nqiGrZlpXtw+lzTZw+IsSjULwQYgF+sCYZsag4cT9WWXhE52hRYFrlg8TpI0K8i/tZtAgJCKGCbEwY0MnWzq4yXIcn11fhyfVVhrPbGslOLO5v5/0YLQ+gL9OvnqzCV/Vor6uMRjMRE0KcgZYXQnyC1MF0Wdk+w5ExokVFnDbR24nbhVELj9J0T7SPih6Ljtz0k9YxhBDvQPFCiA9QcjA1M7WxbldN5FxmBJDVWO3/pNdnRyqcoutF6RhCiDegeCHEZfTkOVGbvjASwuxGThorE8Bd1aM9lpXtk90ejV6LTrRwChVke95viRByHooXQlxEb54TtekLI1MbVuRwMSJGrE7bbyQSyoxFx29RcIQkKxQvhLiEESuIko+G0amNRCN8jIgRu6w8fojsIoTYC8ULIS6hFRUj7ZzlnG3NWBbM5nAxKkb0ONeqXUtOnERvnzCgk2aZCSHBhOKFEJdQsnY8s2kf9h75QdiM6pmDWaO6R6Y0ErU0mLVc2L1sgChMPthbG5MtV7TucOVoQogIxQshLhEqyJZdETlauADAhj212LCn1tLO2owIMjrlpOVcWxmui1hnwsdOKS6wuOr9r9ClXWvTU1BcMZqQ4EHxQogN6O0wR3TPUey0pbid8dXolJPa/kqLIiqxo/q47HYtqw+tNYQEE4oXQizGSIdpNAma2ytVG51ykttfa1FEOfrlZ2F1eXXcdrX6cyMsnBDiDFwegBAL0ZO+Phq5FPkZLZXHFF7I+Gp02QDp/kZT7k8f2RXFgwoMLyWgZ5kAQog/oeWFBB4nfR7MOLWK1on7Xq1E+Nj3aPjHOdn9Bl6cFTm/ny0HegTYhFAehnfPkY220vssrVr4kRinpCKMHdXH0S8/C8WDCtwuDgkgFC8k0Djt82C2w/zi8AmEj32vus+nB47j0wPHAfjbd0MpZ42cYJE71ojFx2xYODHP+BWbsb26HgCwurwar5SHsXbGcJdLRYIGxQsJLG74PJjtMJUcUpXwu++GFTlrjFyH0UbOUFIRjggXke3V9SipCNMCQyyF4oUEFrvzkihhpsNUckgd0zMH7+2plT3GbefdRHEqFT9T/juHkgjfUX2c4oVYCh12SWBx0+fBqFNr8aAC9M/PjD1HfiauuzxP8Rj6bhCv0S8/y9B2QsxC8UICi1wkj5d9HtbOGI7FE/vip4PzsXhiX5TOGK4oUCaE8jx7HyR5URLhtLoQq0kRBEFwuxBW0tDQgMzMTNTX1yMjI8Pt4hAP4PcMq1Kn4wmhPCwpDrlYIkLUYbQRMYOR/pvihRAf4HcBRgghWhjpv+mwS4gPiHY69YKQ8UIZCCHJC8ULIT5Cb94aO8UF1wsihLgNxQshJnHa+qA3b42d4oLrBRFCvADFCyEmkAqEolAenrTZiVZP3hq7xYVbuXOshtNehPgbihdCDCInEEorDwGArQJGT94au8WFn9YLUhIonPYixP8wzwshBlESCKWVhxRXj7YCPXlr7BYXfsmds+jt3ShauQVzXtuBopVbsOjt3QDUV/2uDNdhzbaDtj5DQog10PJCiEHUhIDbSw84sRihV9cLEi0tZxubFKfOlITnU2V7sbHqh2UYRvXMwaxR3T1zb4SQWCheCDFIqCAbRaG8yFRRNE4tPaDWqTohLry0XlBluC5OfMgh1occ0mM37KnFhj21nFIixKNw2ogQEzxZHEJRKHbdIS9NnxhdW8mPVIbrMO2P5ShauUVTuACICDnptNeonjmKx4hTSoQQb2Gb5eXYsWOYNWsW/ud//gepqamYOHEili1bhjZt2mgeKwgCrrvuOrzzzjsoLS3F+PHj7SomIaZ5sjiEKUM7e276JBmQOt1qES0spZYp4LylRQm/RVIRkgzYJl5uvfVWfPvtt1i/fj3Onj2LadOm4e6778bq1as1j126dClSUlLsKhohluGl6ZNkQc7pVo7FE/uiRbNURd+g6G1SP6FovBhJRUiyY4t42b17N9555x1UVFTgiiuuAAAsX74c1113HR5//HHk5eUpHrt9+3Y88cQT+OSTT3DRRRfZUTxCiI9RcrqNZvrIroYWBBStMVLfGS9NBRJCfsAW8bJ161ZkZWVFhAsAjBkzBqmpqfj4449RVFQke9ypU6fw05/+FCtWrEBubq6ua50+fRqnT5+O/N3Q0JBY4QkhnkbNEpJIlFCoIBt/nDZYNj8Mk9oR4i1sES81NTVo37597IWaN8eFF16ImpoaxeMeeOABDBs2DDfddJPuay1cuBALFiwwXVZCgkhluA6bqo4AAK7q0T5QHa5cOLiVoc3ilJKY9+WDvbUxkWWMQCLEfQyJl3nz5mHx4sWq++zevdtUQd58801s2LABlZWVho576KGHMGfOnMjfDQ0NyM/PN1UGQoKA1Jl1Wdm+wHW4doeDqzkEcy0nQtzHkHh58MEHMXXqVNV9unbtitzcXBw5ciRm+7lz53Ds2DHF6aANGzbgyy+/RFZWVsz2iRMnYsSIEdi0aZPscenp6UhPT9d7C4QEGiVnVna4+tHjEMwIJELcxZB4ycnJQU6Ock4EkaFDh+L48eP49NNPMXDgQADnxUlTUxOGDBkie8y8efPw85//PGZb37598eSTT+KGG24wUkxCkhY1Z9Ygdbh2rk+kxyH4bGOTJdcihJjDliR1vXr1wtixY3HXXXehvLwcH374IWbOnIlbbrklEmn0zTffoGfPnigvLwcA5Obm4rLLLov5BwAFBQXo0qWLHcUkJHCoObOaCfn14no/ausTWYGeemrRjPk9CXET2/K8vPzyy5g5cyZGjx4dSVL31FNPRX4/e/YsqqqqcOrUKbuKQEjSIefMCugP+S2pCGNH9XH0y8/C/qMnPbn6st0rZyvVYTTM/UKIu6QIgiC4XQgraWhoQGZmJurr65GRkeF2cQhxBTPRRuNXbMb26nrVfUrvHeb61FNluA5FK7fEbbe6bGJ4tDTaaPrIrpjrARFHSNAw0n9zYUZCAojRzL8lFWFN4QJ4w2/GiZWzxeuIa0RxGQhCvAXFCyEEO6qP69rPK9MlTqycHQ2XgSDEW9DrjBCCfvlZmvt4LVV+MqycTQiRh5YXQgiKBxXglfLYqaNQfiYevqEPp0sIIZ6D4oUQAgBYO2N4TLSRuLAhRQshxGtQvBBCIhQPKjC0GjMhhLgBfV4IIYQQ4isoXgghhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKyheCCGEEOIrKF4IIYQQ4isoXgghhBDiKwK3tpEgCACAhoYGl0tCCCGEEL2I/bbYj6sROPFy4sQJAEB+fr7LJSGEEEKIUU6cOIHMzEzVfVIEPRLHRzQ1NeHQoUNo27YtUlJSHLtuQ0MD8vPzUV1djYyMDMeum+yw3t2Dde8OrHd3YL3bjyAIOHHiBPLy8pCaqu7VEjjLS2pqKjp16uTa9TMyMvhiuwDr3T1Y9+7AencH1ru9aFlcROiwSwghhBBfQfFCCCGEEF9B8WIR6enpmD9/PtLT090uSlLBencP1r07sN7dgfXuLQLnsEsIIYSQYEPLCyGEEEJ8BcULIYQQQnwFxQshhBBCfAXFCyGEEEJ8BcULIYQQQnwFxUsCHDt2DLfeeisyMjKQlZWFO++8E//3f/+n61hBEDBu3DikpKRg7dq19hY0YBit92PHjmHWrFno0aMHLrjgAhQUFGD27Nmor693sNT+ZMWKFejcuTNatmyJIUOGoLy8XHX/119/HT179kTLli3Rt29fvPXWWw6VNFgYqfdnn30WI0aMQHZ2NrKzszFmzBjN50TkMfq+i7z66qtISUnB+PHj7S0giUDxkgC33nordu3ahfXr1+Ovf/0r/vd//xd33323rmOXLl3q6NpLQcJovR86dAiHDh3C448/js8++wx/+tOf8M477+DOO+90sNT+o6SkBHPmzMH8+fOxbds29OvXD4WFhThy5Ijs/lu2bMHkyZNx5513orKyEuPHj8f48ePx2WefOVxyf2O03jdt2oTJkydj48aN2Lp1K/Lz83Httdfim2++cbjk/sZovYt8/fXX+OUvf4kRI0Y4VFICABCIKT7//HMBgFBRURHZ9vbbbwspKSnCN998o3psZWWl0LFjR+Hbb78VAAilpaU2lzY4JFLv0bz22mtCWlqacPbsWTuKGQgGDx4szJgxI/J3Y2OjkJeXJyxcuFB2/0mTJgnXX399zLYhQ4YIv/jFL2wtZ9AwWu9Szp07J7Rt21Z44YUX7CpiIDFT7+fOnROGDRsm/Pd//7dwxx13CDfddJMDJSWCIAi0vJhk69atyMrKwhVXXBHZNmbMGKSmpuLjjz9WPO7UqVP46U9/ihUrViA3N9eJogYKs/Uupb6+HhkZGWjePHBrk1rCmTNn8Omnn2LMmDGRbampqRgzZgy2bt0qe8zWrVtj9geAwsJCxf1JPGbqXcqpU6dw9uxZXHjhhXYVM3CYrfdHH30U7du3pxXXBdhym6Smpgbt27eP2da8eXNceOGFqKmpUTzugQcewLBhw3DTTTfZXcRAYrbeozl69Cgee+wx3VN8ycjRo0fR2NiIDh06xGzv0KED9uzZI3tMTU2N7P56nwsxV+9S5s6di7y8vDghSZQxU++bN2/Gc889h+3btztQQiKFlhcJ8+bNQ0pKiuo/vY2IlDfffBMbNmzA0qVLrS10ALCz3qNpaGjA9ddfj969e+ORRx5JvOCEeIhFixbh1VdfRWlpKVq2bOl2cQLLiRMncPvtt+PZZ59Fu3bt3C5OUkLLi4QHH3wQU6dOVd2na9euyM3NjXPkOnfuHI4dO6Y4HbRhwwZ8+eWXyMrKitk+ceJEjBgxAps2bUqg5P7GznoXOXHiBMaOHYu2bduitLQULVq0SLTYgaVdu3Zo1qwZDh8+HLP98OHDivWcm5traH8Sj5l6F3n88cexaNEivPfee7j88svtLGbgMFrvX375Jb7++mvccMMNkW1NTU0AzluCq6qq0K1bN3sLney47XTjV0TH0U8++SSybd26daqOo99++62wc+fOmH8AhGXLlglfffWVU0X3NWbqXRAEob6+XvjRj34kjBw5Ujh58qQTRfU9gwcPFmbOnBn5u7GxUejYsaOqw+5PfvKTmG1Dhw6lw65BjNa7IAjC4sWLhYyMDGHr1q1OFDGQGKn377//Pq4tv+mmm4RRo0YJO3fuFE6fPu1k0ZMSipcEGDt2rBAKhYSPP/5Y2Lx5s9C9e3dh8uTJkd8PHjwo9OjRQ/j4448VzwFGGxnGaL3X19cLQ4YMEfr27Svs27dP+PbbbyP/zp0759ZteJ5XX31VSE9PF/70pz8Jn3/+uXD33XcLWVlZQk1NjSAIgnD77bcL8+bNi+z/4YcfCs2bNxcef/xxYffu3cL8+fOFFi1aCDt37nTrFnyJ0XpftGiRkJaWJvz5z3+OebdPnDjh1i34EqP1LoXRRs5C8ZIA3333nTB58mShTZs2QkZGhjBt2rSYBmP//v0CAGHjxo2K56B4MY7Ret+4caMAQPbf/v373bkJn7B8+XKhoKBASEtLEwYPHix89NFHkd9Gjhwp3HHHHTH7v/baa8Kll14qpKWlCX369BH+9re/OVziYGCk3i+++GLZd3v+/PnOF9znGH3fo6F4cZYUQRAEp6eqCCGEEELMwmgjQgghhPgKihdCCCGE+AqKF0IIIYT4CooXQgghhPgKihdCCCGE+AqKF0IIIYT4CooXQgghhPgKihdCCCGE+AqKF0IIIYT4CooXQgghhPgKihdCCCGE+Ir/D8pdJYFDz6AgAAAAAElFTkSuQmCC"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 5
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "### Hierarchical knowledge example after vectors transformation with Poincare method\n",
    "![Graph](data/hierarchical-knowledge.png)"
   ],
   "id": "528ebf9ce9091be2"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-11-14T20:03:08.096266Z",
     "start_time": "2024-11-14T20:03:08.089636Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# Convert to hyperbolic space\n",
    "hyperbolic_vectors = euclidean_to_poincare(vectors)\n",
    "hyperbolic_vectors"
   ],
   "id": "a85e10ac003a59c9",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.00074574,  0.00625681, -0.00241861, ...,  0.00499597,\n",
       "        -0.02719642, -0.01314739],\n",
       "       [ 0.02062055, -0.01540077, -0.0067479 , ...,  0.01594586,\n",
       "        -0.03299056,  0.00080561],\n",
       "       [-0.00920154, -0.01504603, -0.0144243 , ..., -0.0040368 ,\n",
       "        -0.00784439, -0.01958051],\n",
       "       ...,\n",
       "       [ 0.02893602,  0.02152681, -0.01487702, ...,  0.02056   ,\n",
       "        -0.00562647, -0.01337913],\n",
       "       [ 0.02370539, -0.01252188, -0.01517714, ...,  0.01259619,\n",
       "        -0.00405182, -0.0128391 ],\n",
       "       [ 0.00554878,  0.0027564 , -0.00759025, ..., -0.00146099,\n",
       "         0.02342014, -0.00776386]], dtype=float32)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 6
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-11-14T20:03:22.430868Z",
     "start_time": "2024-11-14T20:03:08.103253Z"
    }
   },
   "cell_type": "code",
   "source": [
    "dist_matrix = []\n",
    "for i in range(len(hyperbolic_vectors)):\n",
    "    dist_matrix.append([poincare_distance(hyperbolic_vectors[i], v) for v in hyperbolic_vectors])\n",
    "\n",
    "dist_matrix = np.array(dist_matrix)\n",
    "# hist dist matrix values \n",
    "\n",
    "plt.hist(dist_matrix.flatten(), bins=100)\n",
    "plt.show()\n",
    "doc_indices = df.index.tolist()\n",
    "create_gephi_graph_data_file_and_save(df, dist_matrix, doc_indices, dist_threshold=0.7)"
   ],
   "id": "15d04809735541b9",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 687/687 [00:00<00:00, 120001.95it/s]\n",
      "100%|██████████| 687/687 [00:00<00:00, 2317.30it/s]\n"
     ]
    }
   ],
   "execution_count": 7
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}