lda_vis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import pyLDAvis.gensim\n",
    "pyLDAvis.enable_notebook()\n",
    "import gensim\n",
    "from gensim.models.ldamulticore import LdaMulticore\n",
    "from gensim import corpora\n",
    "from datautils.utils import lazy_load_processed_data\n",
    "from datautils.txt_corpus import TxtCorpus\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_LOC = os.path.join('data', '2020-06-20')\n",
    "\n",
    "#class TxtCorpus:\n",
    "#    def __init__(self, data_loc):\n",
    "#        self.data_loc = data_loc\n",
    "#        self.dictionary = gensim.corpora.Dictionary(lazy_load_processed_data(data_loc))\n",
    "#        #self.dictionary = gensim.corpora.Dictionary.load('lda.model.id2word')\n",
    "#\n",
    "#    def __iter__(self):\n",
    "#        for tokens in lazy_load_processed_data(self.data_loc):\n",
    "#            yield self.dictionary.doc2bow(tokens)\n",
    "#\n",
    "#    def get_dict(self):\n",
    "#        return self.dictionary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "lda_model = LdaMulticore.load(os.path.join('saved_models','lda.model'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "corpus = TxtCorpus(DATA_LOC)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we show different kinds of visualizations on the trained LDA model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>1. Show document wise, number of topics and which topic is dominant and percentage contribution</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_dominant_topic(lda_model, corpus):\n",
    "    \"\"\"\n",
    "    TODO: add doc string\n",
    "    \"\"\"\n",
    "    df = pd.DataFrame(index=range(50000), columns=['doc_id', 'dominant_topic', 'weight', 'topic_keywords'])\n",
    "    for doc_id, topic_vector in enumerate(lda_model[corpus]):\n",
    "        #print(topic_vector)\n",
    "        num_topics = len(topic_vector)\n",
    "        # sort based on topic weight\n",
    "        topic_vector = sorted(topic_vector, key=lambda x: (x[1]), reverse=True)\n",
    "        #print(topic_vector)\n",
    "        dominant_topic, weight = topic_vector[0]\n",
    "        #print(dominant_topic, weight)\n",
    "        words_vector = lda_model.show_topic(dominant_topic)\n",
    "        #print(words_vector)\n",
    "        topic_keywords = ', '.join([word for word, _ in words_vector])\n",
    "        df_row = [doc_id, dominant_topic, weight, topic_keywords]\n",
    "        df.loc[doc_id] = df_row\n",
    "        #print(doc_id)\n",
    "    df.dropna(inplace=True)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>doc_id</th>\n",
       "      <th>dominant_topic</th>\n",
       "      <th>weight</th>\n",
       "      <th>topic_keywords</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.912894</td>\n",
       "      <td>patient, infection, study, respiratory, pneumo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>0.762738</td>\n",
       "      <td>cell, expression, study, il, response, effect,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>0.43741</td>\n",
       "      <td>protein, cov, virus, cell, bind, sars, structu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>11</td>\n",
       "      <td>0.425117</td>\n",
       "      <td>lung, cell, patient, pulmonary, tissue, respir...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>12</td>\n",
       "      <td>0.381466</td>\n",
       "      <td>cell, expression, study, il, response, effect,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72214</td>\n",
       "      <td>72214</td>\n",
       "      <td>16</td>\n",
       "      <td>0.397957</td>\n",
       "      <td>sample, virus, pcr, test, study, positive, det...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72215</td>\n",
       "      <td>72215</td>\n",
       "      <td>4</td>\n",
       "      <td>0.534901</td>\n",
       "      <td>patient, ct, case, study, image, disease, resu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72216</td>\n",
       "      <td>72216</td>\n",
       "      <td>7</td>\n",
       "      <td>0.669211</td>\n",
       "      <td>patient, covid, study, health, case, care, dis...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72217</td>\n",
       "      <td>72217</td>\n",
       "      <td>5</td>\n",
       "      <td>0.998787</td>\n",
       "      <td>la, en, el, le, que, los, se, et, de, une</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72218</td>\n",
       "      <td>72218</td>\n",
       "      <td>8</td>\n",
       "      <td>0.514748</td>\n",
       "      <td>rna, sequence, gene, protein, genome, virus, n...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>72219 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      doc_id dominant_topic    weight  \\\n",
       "0          0              0  0.912894   \n",
       "1          1             12  0.762738   \n",
       "2          2             10   0.43741   \n",
       "3          3             11  0.425117   \n",
       "4          4             12  0.381466   \n",
       "...      ...            ...       ...   \n",
       "72214  72214             16  0.397957   \n",
       "72215  72215              4  0.534901   \n",
       "72216  72216              7  0.669211   \n",
       "72217  72217              5  0.998787   \n",
       "72218  72218              8  0.514748   \n",
       "\n",
       "                                          topic_keywords  \n",
       "0      patient, infection, study, respiratory, pneumo...  \n",
       "1      cell, expression, study, il, response, effect,...  \n",
       "2      protein, cov, virus, cell, bind, sars, structu...  \n",
       "3      lung, cell, patient, pulmonary, tissue, respir...  \n",
       "4      cell, expression, study, il, response, effect,...  \n",
       "...                                                  ...  \n",
       "72214  sample, virus, pcr, test, study, positive, det...  \n",
       "72215  patient, ct, case, study, image, disease, resu...  \n",
       "72216  patient, covid, study, health, case, care, dis...  \n",
       "72217          la, en, el, le, que, los, se, et, de, une  \n",
       "72218  rna, sequence, gene, protein, genome, virus, n...  \n",
       "\n",
       "[72219 rows x 4 columns]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = show_dominant_topic(lda_model, corpus)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>2. Freqency Distribution of word counts in documents</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "72219\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean:  2453.6794472368765\n",
      "Median:  1682.0\n",
      "Standard Deviation:  5418.437742922199\n"
     ]
    }
   ],
   "source": [
    "docs = lazy_load_processed_data(DATA_LOC)\n",
    "num_words_in_doc = [len(doc) for doc in docs]\n",
    "print(len(num_words_in_doc))\n",
    "plt.figure(figsize=(16,7))\n",
    "plt.hist(num_words_in_doc, bins=250, edgecolor='black', linewidth=1.2)\n",
    "plt.xlabel('Word Count')\n",
    "plt.ylabel('Num Documents')\n",
    "plt.title('Distribution of Document Word Counts')\n",
    "plt.show()\n",
    "\n",
    "#print Mean, Median and standard deviation of word counts\n",
    "print(\"Mean: \", np.mean(num_words_in_doc))\n",
    "print(\"Median: \", np.median(num_words_in_doc))\n",
    "print(\"Standard Deviation: \", np.std(num_words_in_doc))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>3. Visualize key_words in each topic</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAKICAYAAABt8wwaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdeZxN5eMH8M8zY4skJEvFKMr0bYxvSFlSkl1KRVlSqm8krRKljDZUfq1C+8iaKDFENfZSlJgYlCWyZQvZmfP74865zj33rPeee885937er9e85t5znvM8z73j1Oc+9znPEZIkgYiIiIiIAlLc7gARERERkZcwIBMRERERKTAgExEREREpMCATERERESkwIBMRERERKTAgExEREREpMCDHkRBCsvCzOUZtTxJCrHWorpJCiDeEEDuFEEeEEEuEEA2dqJvIKxLhfBVCpAohhgohvhFC7Cvs8x1O9JHIaxLknL1aCPGhEGJd4f9f/xRCZAshqjrRT7KuiNsdSDLXqJ5/AWAlgCzFtuMxansQgFIO1fUpgOsA9AOwFcAjAL4RQlwlSdJqh9ogclsinK9FAfQBsALALABdHaiTyKsS4ZztCqAmgP8DsBbARQAGA1guhKgtSdJOB9ogCwRvFOKewk+yiyVJ6uZ2X6wSQjQAsBRAF0mSJhZuKwZgHYBlkiR1crN/RLHix/MVAIQQKZIkFQghrgCQB+BOSZImud0voljz4zkrhKggSdJu1bZLEfh/7DOSJL3sTs+SD6dYeJgQ4h4hRJ4Q4rgQYrcQ4mMhxPmqMjuFEB8IIR4UQmwUQhwTQiwTQjRRlQv7+kcIUVoI8VrhcceFEDuEEFOEEOUNunUTgKMApsobJEk6AeAzAG2FEKlRv3AiH/Lo+QpJkgqce5VEicOL56w6HBduWw/gIIALonvFZAcDskcJIR4G8BGAXwHcjMDXNzcBmCeEOEtVvCWA3gCeAtClcNscIUR1g/pLAJgHoBeADwC0BfAwgEMAzjHo2n8A/F4YipVWAygJIM3stRElGg+fr0SkwU/nrBCiTuEx+XaOo+hwDrIHFU5ZGAxgjiRJ3RXbNwD4BkB3AO8pDqkAoL48N0kIMQ/AnwCeBnC/TjM9AdQF0EqSpDmK7VNMulcOwH6N7fsU+zeY1EGUMDx+vhKRip/O2cK+jgKwHUC2nWMpOhxB9qYrEAia45QbJUn6FsAuAE1V5RcqJ+5LkrQfwByEX7Cg1ALAn6oT1woBQGviurBZD1Gi8PL5SkThfHHOCiEEAkH9SgBdJUk6FGldZB8DsjeVK/y9Q2PfTsV+2S6NcrtgPF+pPIC/7HcN+zTaB4Cyiv1EycTL5ysRhfPLOft/CIxmd5ckaX6UdZFNDMjeJIfMShr7KgHYq9pWUaNcRQDbDNrYg8gm/K8GULPwax+lywEcAbA5gjqJ/MzL5ysRhfP8OSuEeAHAowB6SZL0WaT1UOQYkL3pNwRO4JAF/YUQNyBwUi5QlW8ihKikKFcWgYsKfjBoYy6ANCHEjTb79hWAswDcomivGIDbAcySJOm0zfqI/M7L5ysRhfP0OSuEeBKBiwb7SZL0vt3jyRm8SM+DJEk6IYQYAuBNIcTHACYDqArgJQBroJo3hcAn1W+EEM8DOA1gIAJ/25cMmvkYwL0ApgohXgawDEAZAK0BvCxJ0iadvi0VQkwHMLLwSt+/APQFUBnA85G8XiI/8/L5CgBCiOsR+Lr3osJNDYQQpwCcliTpC1svligBePmcFUL0APAKgOkAlgghrlbs/keSJEfuiEvmGJA9SpKkt4QQhwA8jsCyMgcB5ADoL0nSUVXxOQB+QeCkqoLAzQBaSpK02aD+Y0KIZgCGAHgQga+V9gBYBOCASfe6AhgKYDgCS8+sANBCkqQ8O6+RKFF4/HwdCqCB4vmjhT/HAZSw8vqIEo2Hz9nWhb87FP6o+9HK9MWRI3gnPZ8TQuwEMFOSpPvc7gsRGeP5SuQvPGeTF+cgExEREREpMCATERERESlwigURERERkQJHkImIiIiIFBiQiYiIiIgUzJZ54/wLIvcJG2V5zhK5z+o5y/OVyH2a5ytHkImIiIgoahnZGcjIznC7G45gQCYiIiIiUmBAJiIiIooDP4yuRjsK3P6S9g72xj281TQRUaxllSn8bXZX6ASQVSY5XicRhcnrked2FxzDEWQiIiIiIgWOIBMREZGuZ5c8iy//+FJzn96I4V2z78KKv1eEbCtVtBSWdlmqWT4jOyNYl9bX+1ZGJvWmBRgda6ctuWxejzxbx1076VrsP74/4jaN3peFdyxE2eJldftqtZ2vNnyFZxY/Y1qH1vGR/r3sHqt8/3t/2xuLty0O2d+mehsMv3a4pXat4AgyEVGsZR3gtAPypYzsDN1wbHSMOhwDwOGThw3nthrNfY30OD2bDmyKqC2j/VrbM7IzQsJxpPTatBOOjfapw3Gs9VvQL6r3Xx2OAWDWplmOzvE2u9W059doTBuQg83D2rrdDaJY4jrI5B/yHGR53nXIPp0PCVplAWDQLqBICfOyJcoAA7Zo98Oovdh9aEmIdZCVYSOSEUH1MbHY9/TipzFjwwxbfVTWaTQiaqcfu47sQvMpzQ37YdSmWT/tHNfnuz4YecPIiOuKpJ92j3X6/bfTtgaug0xEFJHpDxUGrjLAu1fHr125TfnHL5QBVDl6rhmaFduU5VsN0w/HynJVrwaOHdD+u/jpPfOwSIKS1jFW6rF7nNPhWGn30d2a27WOq1iyouX2I2Hn9WmFY7t1xJLZ+3/bpbeFlHOL7+cgc/SYiBwXEsRUIevvfPOVGv74Fhh3q0a9Fkcs9YKdnVFZt9ntl7r81b21yz2eH/q855zA+/J3vnZ5q+1RXJ0sOImiKUXd7oama6pcgx+2/4BmnzXzTKhMJoOvGYzP13+uuz9efxPPBOS0ATkAtAOv1j55m94x6uM+XrIZQ2asBgCsfr4lShUrYtiu3vYXZq7Bh4s3mbZNRAlA66v44DaDkFyjufYxdtpUt6vXHy+6ZUz4tivvAn4ZG7rNzuuRy55TRWOfxgcZeVt2e6DHjPB6yJA8eje742zLx0xZP8W0TN2KdfHzrp9x5adXuh4+3R6hjLVEf32x5pkpFmZBU71/87C2eKp1LUt1pw3IwZAZq/HzoBsBAB8u2mRyhH49Hy7ehM3D2gb7owzqyS5tQE7whyhhhIXUCINvtO0qt3k95GXe4U67P2oE800Lw7f13xj7viSIC0tfaLnskm1LTMtcU/maaLpDFqgvWmxRrQXm3jbX9Q8kfuOZEWTZ/32zHo/feGnwuVHY6t30Egyfvda0TmW4jnbE18m6Eg0/NFDC0b3IS+ciNHJX+Rqhz9V/p9GNA79Llo9fn3yqXIly2HdsHwYsGoBhTYZZOuaN698wHbV859d3AADC1rXHsZGIgVF+/88uejZ+6PKDy73xN8+MIMve+u73sG0pwv0TScbwR0TkUTVu0N4uh+SdiReIYmVB5wUAgJyNsfl/3qoeq2JSLwUwHEfPUwFZb0R249A2ce6JNuUIadqAHCzduNflHsWP/MEgbUAONu4+zKkURG5ItFFrO1NGrK6EYbUeSigNxjewXLZ//f4A3Juje8/X97jSrlfmJD9U5yEA5utJ691UJl48FZBlyjDmNfL846KpKbjjvaWe7GOsbNj9LzYPa4tmI+YH34dkev1EcaEMgs+XU2w3uHgvEYRdZGcQeqfeF15u8D869Ra+V988F3nfklTjCwJTUjKyM9Byasuw/UZ3QdO7aYayjBPkuo6cOmLYplL3y7sb7v9uy3cxCZNyX5fvWu543WqNJjYKee6VcAwAD2Q+EHys7pfyeamipeLWJy2em4M897Fr0eL1MxdVeHWe7+8vtQbgzRAfK5dUONvtLhAlB3nubMFp7dUZEknWAeCjlsCWpeYjwVkHgPG3AXlTAj/K7WaWvBldP5PQqOaj8Mi8R5C7JRfb/91uOWTJt2LWKj+uzTinuxly62e7fbRzjJPUbTr1oUF+XQdPHNRsw8prHdV8FHp/2zusbO/M3niwzoMh26zeDU/9+szefy/MD/dcQL60Yum4tqc1Cvr73//GtQ9ERCH8spybFr0+3/R24EdLzznW6++qvz6q7T6RqTevP/PBQhlkVvVYZXihXV6PPBw5dSQ49WFh54UoWyL8tsjK8kas7lf28bG6j6HnFT1Nj3n5x5cxce1EAEDt82pjfNvxMemjXj+Njov0Ji3T/5iOQUsGhdVhpb7GFzRGXo88dMnpgrw9gfKftPoEdSvWdaR/6mOdei+cDtWeC8gyo5FZ9T4rayJH055RmQ0ve2N+NBERUSzZDSAli5SM+0hgJO093eBpPN3g6Rj0Rl+s35cONTqgQ40OUdUxoe0Eh3pjzAujxVo8GZDtronstXJElCD0bgaSaBfLERFRCE8GZPIfvVF9fqgg38sqAzR+DGiedeZ5cJ+Fr+6/HnDm8bT7gY7v22tbdx+nDRARxYqQJMlov+FOIooLOwuB85x1gnIOsF5INQqo0Sw7dvxfYOgFkR1LXmH1nOX5SuQ+zfOVI8jkKWkDcnB/k4vxTNv0iI9XivcIdtqAHI6aJxo5hI6/HTiwFXjQwtqckQZXq6PTnOJBRBRTDMjkKZ/3boh61fSvdLbCzYBaugRPqYTVdYp5GadwZJiIyFX8vzl5SrThOJasjA7nZYUvpk9ERMnL6jJmsWpTyasrRngRAzJ5wqj5GzD867UAtEeAtZbYU5ZT7te7QNBs+oXefq26jdrX6v83a3bh/rHLNcukDcjBHy+3QY2nZxn2jwgAcCR5bnFPRJFRBuGs77Mw9fepLvbGn3iRHnmK3iht2oAcDGhdC72aXhJ8DmiHXL3j1aH0h4HNULnMWYb1Wak70vbVAVxvP3iRXvy5daMOo7vmfdAc+GvZmef91gNnV4xPv8guXqSXIJy6PXYsbrPtp/Y9TvN8TYl3L4giJYdjAHi9c52o6rrlvxfgmqG5IdtiNWKbNiAH5c8uprmdKIQ6kGeVOfOjDMdZBxiOiXwkr0cew6nPxGSKBa/kp1grXsT6Z7sJP24B4G4g/XnQjSHPm6dXxLf5u1zqDZly8yI5XqBHROQ6jiBTwuvSoCqAwAix+idehs7KD3nOcExEyaLbrG7oNqtbRMeOWjkKdcbWwTu/vhN1P46dOoZ64+phzKoxlsrvPro76jaddM/X96DxpMaYv3V+3NvO3ZKL2tm1MW/rvLi37RZepEcUY5uHtUXagBwMbJMetp2IKFHI81zHtRmHzAqZYSspWJ0Hq7UCw5iVYzBm5Zlga1SHuh11fe+seAfvrHhHsx691R/srgrh1CoSWvX0ze0bVZ3RtP9w7sNxadcLIg7IdlYEsLLv+ZlrcE/DNFxUriQW/b4b3T/8KeQiJoaJxGZlpYhoyCFVvc3qfq0yZqtofPdEU1xS4WzNMkREiUo9Wjzj5hlo/2X74POM7AzL4eqCsy/ABy0+QOtprUO2527JRbOqzQyPHb1yNEb+OlJ3fywDXl6PPKzdtxa3z7g94jr0Qra6nVhRtl80pSi+6PAF2n3RLmR/IofkiFexMAqtZlfsZ930H9zdME1znzKEMCATAeAqFkR+k5SrWKgDnVZ4inZNYCvHW+mHFa///Do++u2jqOoAAlMjlu9abqsep9dOtruKhVn7bqztHEPeWcVCDscAcP1l5xuOrE34aQvDMRERkY/ohaZow5Td4/0Y3tbtWxd87Hb/Y/V39IOIp1goR3rLnFUUKwe3iKiej++pr3sTiMFfrUb295vR5aqqkXaTiIicoF6jOWy/1hJ1hdtergycOGJc3qgds7Ja+3fmAaMbA3eMB2q1Cz1GLm+07jTFXL8F/fBa09fc7obn3DbjNlfbtzK1Q10+EQNzVBfpad1pzK6MrDm69WZ/vznieomIyAHTHwJWfBp4/Nw+ICXVXrA0C9ZWyipDrROM2ql3D9DuDefaIl1zNs+xHJC3HNqCL3//El/88UWMe+UdZxU5y+0u2A7LiSSigPza3HXo1+KyiBtVzik+dOxU8PHMvo0535iIyEvkcKwMqFkHrN1t0M4dCY3KBm+Y4mBILnEOMGBreDvLP2ZAdtnOwztx4+c3mhdMcD91/cntLiS1iAJyvxaXRbzSwOZhbXHlC99o3l73igtsjDQQqcTqw1VSfmhzOozYaTdsW4J97Z1VBug2Dahxg9s9iQ8rf7+dFr+eLTgFpDi0Oqk6HCvN6ge04Vf/brB7oR/FzsgbRqLGuTXc7oZropqDbHWf+vkvz+p/MlQvvUVkFf+9OCQ4khfHkLx9BfDedfFpywvGdUy84B+N0Y0Dv3XnGheOWD9fPrbvm9zOT+8zILtsRfcVKOLUhyGfqj++PpZ1XWZeMEb6fNcnIecWW5Xc//rIF7RGcNXLA8q0QnLagBw82fIyvDpnXVgZ5bFnFU1F/gutHO07WSSH425TgRrNXe2KKTvTBohI14uNXgx5rhwZTvZwDATu/Efu4b9A8jx5xRS9EWIrF4v2ub4G+lwf/lWRXlhOam4GP6+HYyf4LVjLo6p+m/5y/JDbPSATHWp0sH1MPKdXpJdPNy8UA90v745P13zqSttAYFoLp7G4tA4yUTSufOEbFC/Cf7pEcXPft+HbvByOAWBCJ7d7kNTs3rIZALqkd4lVdyLSKs2dbxT71+8ffOx2UHW7fTdxBJl8Z9/hE47NN+aoMZEFHzSPbSBu9Aiw5E1g6r3ArR+G759yd+C3cj1jM5GMII8vXH+2SAn7x1IYOVxNv3k6OnwZOlqsNbd14FUDMSF/QvDYiW0n4orzrsCJ0ydQd1zd2HfYgPxaZnWchU/XfIqJaycCML5hxqAlgzD9j+lh9ZQsUhIvNX4Jzavpf2OmHMU1CqlG7c/aOAtPLXpK83XcdultGHzN4KjbN+uDnzEgky+YTbOIhNYt0ZOW7rqwFpfw2rMeeKe+9TqM1qFVqtMFuHmUtbKyQbvMA46dm16MbgTs/M1aHU7c/MLKsVbayjoAvH0lsHeDveOttu+kG58PBOS8z7UD8urCtW/vGO9cmztWApUzQ7f9/k3g96BdzrWTpJQBy0o41jruzpw7Q/YVTy2O5d2Wx3VUUz3doM20NqbHmPXvyKkjeGz+YyFtWGnbKivHfL7+c3y+/vOYtJ8oGJDJV5JyybV4CFnj1mYgUt5IQq7njSuAf7aeqS/sLmsad13T2q7bX0Uf1XdFe7Ei0PdnoLzO8kRaN7nYsQoY00S7D72WRNdXddlIb5yhfp1ZZQJhMsPgrltyONa6c9ymhUD1ay30wWAOsrLuaKnbUa61bNZOVhngqv8BbV7VP15tTOFrf24fMO8lYNGI6PpPYeTg1TWnK4QQGNdmnK3jbpl+C4qnFsekdpM095sd7xS5vv4L+2P+1vl4relruPZC/XPHyfaVdbX7oh0OnzyMd5u/i/Ry+vOjY9X+LdNvwb8n/8XkdpNRrkQ5x9rwKiFJktF+w51E8fTU1FWYvGyr5ooWauqRYb1QLR9bt1pZXHfZ+ejbrEbYPr1640jYKBv9OWsnlMjOrgj0W69fxixEOXVDCbP9Vl9bNO07dbxcrml/4PpnrNejfI2PrQbKXBi6/4sHgJWTrPVBXZ9uGYP32u77ZHe0Xav8I6uAstUC+7RuNS3XGbsLD62eswn1/1jlaGOifu1OCUnzfOUIMvnG8FtrY/ittcO2m4VWO2t226mXFLTCsZd5/QIzJa1wrGS0XrU6HAPALWPOBGQzVu+W90lb4G7VB8qIP0DYDdRGfTP7EOKjfwdEFFcMyETkH3anJygD0LiOzvcnVqyO2sZjfrAVmxe73QMictjbK97Ge6veC9ueLN8OMCATkT/ZDYd/fBf4fXVv5/sSK6nFIj/27POd64ce3jSFKCEl88V5MgZkIkou9e9zuwfWXXxd5Meed6lTvdC50NIjo9dE5KgCqSD4OFlGi7UwIBOR/2Tcpr0cmBXjbgUeWelsf2Ll97nutq+cxhHrVSyIyBMyx2aaF0oCDMhE5B8iBZAK9NfLNfJoHvBGBrB/c0y65qj7vg3cnMMLEi0AJ9rr8ZBkHm1MRBeW1rjIN4nwfr1E5B+D90d+7LlVnetHrF2oc9MVJXlEt9mzse0LESWlc4qd43YXXMURZCLyJ6PlzcbfDnSdEtmx0x8COrwTff+cotXXD2888/jafvHtDxEZMrrArVTRUljaZanuMUaj8GbrTKvr0OuH+li9cmv2rgnbZ7V/Zm1GUsdDdR7CA5kP2D420m82GJAtkG8YwXVxKSEZXWylddc5t1mZFxvtsVYCstWbTFi9rbbW3QY5/5fIV8xWfzh88rAn+hHv9jKyMwyD6pzNc9BvgfGHfb1w/OOOH3HfXP2Lr83a1pNQATmWtyG+oOxZMamXiCKgdetkrf16+3b+BoxuZP9YeX+8VnDIOgDsWHnmtsjqfUTkGQMXDQw+1gtk8QiuVkajlfRGlC8vfzkmt5tsud1oXrMyHNsNs3I4NhpVjyQkJ1RAjhWOHFNCi+Udz6zWHa+7rskqXRFdwLRzbLRBtnKm8+0xXBM5bubGmaZl4nUhY7wvmLTSnl5IjeYW5fKx1c6pptuvSD+UOHqR3qnTEuo8b29ZoosHzjLcf7pAwiVPG5cBzkyDsGrMwo1o9eYiW8dY0W/KSrR6Y6Hj9RIREZF3dbikg9td8LVoQv3MW8w/nNgNylGPIMvTGpQB1WjOrl55dVlJAqoPNK9THYzVz62W1+urUV0AMHz2WoxasAEli6XiyInTluolIiKixPJi4xcxfcN0APanOTiJy+05w5EpFloBN21Aju6cYCtzheVwrCynVaf82EogzcnboVlGr6/quo0cOXE65PienyxD7tq/TY8jIiKixKD+Sl9+fP1F1+OtZm+51a24iGQqg9HFdWby9+VH1bYZV+Ygn13cuFm9sCuPPEd6MV6f8b/YPsYqdX8+urs+0gbk4M+9R1CtfMmYtUtERETeIY/g1h9fH8dOHQMAzNs6L+LVFLzObCk4o/D6z/F/Im5379G9wcdf3fxVxPXocSUg/zakpRvNBtmdrxyNVX/9w4BMRBSN4HJ3vLiQ/GNZ12XBx9GspuAXkbyuz9t/HvHob+MLGgcfVy9TPaI6jDhykV7lMuFLoDWpWcGJqomIiMhARnaGY18xO1mXG/V7lTI8/nnwT91y2//dHo/uOOah7x5yuwsx48gI8o4DR8O2Lfp9txNVO+7qi8tj0v+udrsbRET+YXTnQSKyZcyqMXi58cua+1pObemrEea8Pc71NZrR9ViMzDu6zJsfLN2417wQERFREmpTvY2vApqXWB0Z1wrHyukC0dQdbws6LzDcb6Xfy7stt1S+QCoI26b8t1pvXD3dY4326YnpHOSrqpeL6Di9i/GMVqoY3P4/GDJjtaV6m746DwuevD5k36WDZmP9i60j6i8RUcLavc7tHlAcDb92uNtd8LVIg+yo5qNC5ilruaj0Rdh6aGvEfYs1vX6b3ayjeGpxS/X8etevmts/bPkh7p1zL46fPu7oBwlHAvK8ftdpXvj22QPXRFynMiSrt2u5p1EahsxYbVq+X4vL8NrcdaYX6mntV27j+sZE5CtGt8dWT594px6w53fz49XHmV1MZ3W/URtW67JahihOOtbsiCENh+juNwqS8kipF0eSrfTbSh2A8etLEdqTHq6qdJVpCH+87uOW+qEkJEky2m+4EwiExgvLnoXFTzWz3TgRWSJslDU9ZykJKYNn+RpAentg8euK/TpB17ReBwOyss1rnwQWvmrcnrcDstVz1pHzVXlTiv9++l+cKjgVsr/Hf3qgX71+pscrWQ02VgKbsi71DTSiaZvIIZrnqyMjyPw/MhGRDyiDYvMsa+XiETLlNm79EMi4LfC42aDQfWp1ugC/TgCGXggM/Eu/ziQaPdYLq9mrs5G9Otvx4Klsb1bHWRj560jkbAz99rXvf/taOl69nSGZ3ObKOshERERh5HCs9PR24OUq4dtvHhUIyMcPxb5fPqIVLM1GebVGeM3I5YqmFMUv3QM34RrWZBiGNRlm6TbLGdkZKFO8DBbfsVizXobk+Np/fD+unXQtAG+M4Mv/DoY2GYp2F7dzpQ9Jt4oFEVHSsjp1wkuKlbJ/zKIRgd+VvDdfM5am3TQt7m3K4TgS6nAMeCOcEQEMyEREiU89ZUL+8YJDOyM/Vn5d6tfy3fOB373CA1giq1m2pttdAHDmYqodh3folnkg84F4dSepZWRnYNXuVW53w5einmLB1RyIiHxAK0xmlQEuqAvcn+tOnwDg2D/utU0xIa9XW7lUZd0yD9VJ3DuwUWLgCDIRUTLJOhD4GfR34Pm2n93tT4Va0R3/3L7A7xfPD/xOwovz3KC3CsVtMzTmkRP5EC/SI6LEcWgnMOKywFJmfV0Ofl5XpHhhWJanXPg0UKakBn6fOu5uP5JQdqts9Pi6B5dqM6C8YFF+nCJSsPKulSH75TJm9VgpCwC9v+2NxdsCU4y6zuoatt/s+EnrJuGlpS9ZLg8ALy59EZPXTQ4+H9F0BFqktTA9TvnaGl/QGKOajzI9Jh4YkIkocZSuFPh93UB3+5FIarUD1s60F6KdDNxemStNYU5LpwEAH7T4AA0qN3C5N96mDIEFUgFmbpyJgYsGhpVRB9EuOV2Qt0d/ZRJ1+YdzH8a8rfMi7mepoqU0P/CYrUyidcwTC54AFugfs/nAZrT/sn3ItsXbFiMjOwNvNXsLD+c+bLf7jmJAJqLE4teR0FjSmnYQMhfZ4D27Y/yZsov+D2hicEcqeURarl/vIjq947LKAFWvBnrOsXac+vghZc88p5jrOacnADAcW6ScliKHY7M7yMnhWB0y9ZbDe6vZW8HHn6z+BCOWj8D4NuNRu0JtS308fPKwYXta5H0ruq9AkZQzsfKvQ3+h9bTWukv2yeFYqy23wzHAOchERMlDawULK2Hy1g8Cv78bYr4KxmDFRXfKcvLcZz0DtgR+b1kafpxVhReHEXmNVkA0m7ZgNGobyyksWnV3vqxzSJ9kyufKcAwAF5a+MPh404FNusdZad8NHEEmIu8wC0WjGgK7VpCAKvwAACAASURBVJvfGtksVGWVAR5ZBZStBqzNASZ10T/eyu2Rtfb9vQZ49xqNY1wY3Yy2zYzbAz9WCGFy+2edfSUMpmXY6f9jq62XJQDA1kNbMXDRQKzcvTK4LSM7A/dn3I+HrzQfydMd/fRI0KHoDbp6UMj8YjW9v7U89/qmL2+Ke9CPFgMyEXlH8dLGd0bbpRN+LmoAbP3RXltvWvvKMSKcN+ueMheal6Ego5G89/Pex/t57wMIDzJ/H/nbUt1eDkB+YfXuhuQsBmQi8o6Bf5mvqpBaLHzbvXPPPLYbTmM5osu5sBQHZiHUaH8kATY4//W8DExoO8GwnDokW2mPoZq8gAGZiPxBDr7P7nawTgbYhMC1j11hFI7JOfzA4A4GZCLyFuVKCH6XVQZIawzcneN2TxKP3XnnREQ2cBULIvImrZDslxCkXLFh82LjVR8oMje/G7gg8Nk9/vl3kWCsLP3F0U/31CxbEwAwZ/OcmLfVeFJjze3yv4Mrz79Sc3+jiY1i1qdoMSATkff5NVyqlzbz6+vwojpdA0vKpRZ1uydJRxl65XnG6h8AmNJ+iltdTAhmaySb7WtUJRA+x64Z62zHFOQ+Hjhu/CE1u3W25nEHTxwMK+uVixI5xcJA2oAzX4uue7E1Nu7+F63fXITNw9qG7F/zfCtc/tzXABCyT36srE99bF5WS2RkzcGNl1fE+3fVCyn7btcr8dTUVTh07FRYXUQJTZ5mMaoh0Pt7t3sTPeUNM/x8W2eiQkbh7dvbv0XFkhXj3aWEVKtcLazdtzbq0Gj3VtV2yEu52V3u76wiZ+HoqaO6typ3OygzIJswCrzK/ZuHtQ0J1HbqNgrSbTIqI21Ajm77RAlNuawbQyWR53AKRWzJo/BaYbFjzY4Y0nCI4fHxCpp67Rj9+/ip608AYhveo8GA7LD+n6/CK7eZr6+qDNRawddu2CZKODe9DXzV1/1pCYP/AYacGz7ya9QvrVFirrRAcfTF71/glpq3uN2NpGf15hjRLNUXSR/UyhYvG3UfIu2j0U1G3MSA7LAq555luaxydPriCqWQ+8R1YfuIktaVdwUCshnNi/kcXOFACO16B2wFxt4EbF9hvV9W70ZHFAV5RI4BmShyDMgOe7R5Tc3tRiPCkUzPIEoKVoJtJOHX7jF65f8335n6gTOB+pwqwOP59o+nhMS70RG5gwE5Cm90rqMbbNWh9/Z6F2HK8q3B5+rjRnTK1D1W3kZERMkj0rmjDNRE0ROSJBntN9xJRHEhzIsE8Zz1K44gJxKr56zh+cq1hIniQvN85QgyERGRR0xaNwkvLX0pZJuVq/ztrCCQkZ2Bxhc0xgO1H0D32d0BAL/e9StSRWpIPUbHW22LyK94oxAiIiKPUIdjq/J65CGvR17w5hBmFm9bjO6zu2Ns68BNJOqMrRMMvqWKlgKgHYTlbakiFR+1/ChsO1Gi4BQLIu/jFItkwCkWicTVKRZmx6n3r9q9Cl1ndQ3ZplWHUb2cDkI+xikWRJSE1OsPHzsADKt6Zv+NQ4BGj9qvT3ZxU+CuryI/3mzFCzvrJyvr5nrLZFHtCoG1+8fcOEa3zNzNcwEwAFPy4BQLIkoeWWVCwzEAfDPY2s1I3q6rXW7jAmvHy7eZ1tpuhZ0bpjAcUwQaVmmou++JBU/EsSfeYHT7ZEp8HEEmouSgNRJ74jDwcpUz+/WC5bvXAHv/CDwuUQYYsEWjXoPj56nmlarvyGcUfptnAd9m6e+X7VhpXoYoSgyMlCwYkIkoeagDbLFSgW1yQJ16H3DrB+HH/b1G+3h5m1lIXvCKfh+Ux2tp/Ji1gDzmWvMyRFHiFAtKFpxiQUQky5sSvs3S9AmLUxr0ylk+3sG+EBGRLgZkIkoOHd7R32clVJaqYK2dr/qq6rYxd1iP1Qv5iGJEvboFUaLjFAsiSg7/7R7d8Yd3Wwuiv4wFbno7urYi1bCveRmiKGVkZ2hOtWgzrQ1mdZzlQo/iQ/3hYGmXpcE1o/Vs/3c7Wk5tGbbd6lQVvQ8kVo7nDV2ik5AB+cDRk8gcMjf4fPOwtiH70wbkhG0jIvIFo4sBW7wY375QTOX1yNNcScHqnfSsHBdJf/Tau6j0RVHV72Var/fqCVcDsHfHQfU+vWOfWPBEcGk9veP1jj188nCwb3aOo1AJGZAzh8w1DcAnThWgWBHOMCEii9ye26t3MR+nVyS0vB55eHLBk/h689cAgLebaX87YSf0aJVVb9OrT97eaUYn5O8L3NBmxHUj0KJaC8vt+41emJW3X/fZdZjfab7msY9e+Sjuzbg3ZFvvb3tj8bbFhm0arTutNyot0wvujSc1xoHjBxiSLUrKhLh5WFuGY6JkY3zXUHPqpdrctPC18G1uB3iKmVebvhq8lfR1F13ndncAAJ+1/yzYp0QOxzKjDxV7j+7VPUYdjgFgVPNRwcfDfhpmuy9Vzq5i+S6JSovvWBxWjvQlTEp8bvpvSBuQg7QBOQAQ8lgmb1Nv1ytjVpaIfGTIufr7rIzCKpdqs6PSFZEdp0UOwbkvOFcnEblmfP74uLbHkWPrEmaKxfMdrsDzHQL/I9KbYyxv0wu9aQNy8MdLbVAkVRjWQ0RJxGydYjO9lpgf/2uE/5Pk9AqimPu8/eemZX7e9TPqVqzreNtmc5WVjpw64nj7XmXnfYlUwgRkp8jhmIgSkNYFbsqQaWU5NaMyke7/8kHjdrXsWKWoN4GmV7xZG3hklXm5WFJ/8Eik95dsu6zcZaZlHvjmASzvtjxse6RTGfJ65GF8/vjgFAy5nnIlymFB5wWaxzQY3yDqdiORqHOaGZCJKDnII8F6o65G6xyr75YXaduRHq9V15gm0dXjVfv/NP+gQeQxJYuUDNumDKkruq9AkZQimvv0dE3viq7pXXHwxEE0mtgIALDv2D5LgTQRA2u8JcwcZCdsHtY2ZO4xp1cQJRi90DXob+DJP8yPrX9fZPVb2Xd5B+O6I6nXr9x+TVkH3O8Decb0DdNNy3x+U+g0DDkAv9b0NeT1yAsJx3adU+yc4AWR6vqVFt6xMOI2IvXVhq/i3ma8cARZofrAHMx57FpcVrG0210holiJJvi0HRH4iUXbncY6U4/fJfJrI18atHgQOlxi/AH2/JLna25vmaa/HFsklGtRq5UtXjbq+tV1t7u4HYY2GapZdteRXXhm8TOaxwHQDPSRrL5hdT1vszYkSKidXRsPZD6Ah+o8pFlGiQFZYc6j16LF66GfwDiKTERERH5id16wXuieuXEmZm6cGVLXsp3L0HNOz6j7GGm/Ip1fXTu7NgBYCsdAkgVk5eoVysdyCG7x+sKQQPzH3/9yqgURkdG8ab2LHnVvbKIxQqwuV6wk8PQOa/14cgPw6iX6bXz5oP4qIRytJou0Aqcc1Iwu4uu/sD9euTZ0iUizgNf+y/bY/u92/Nzt57B9ZsvCqe92qDcia3YjGCP1K9UPlovlahJy3YOuHoTOl3UObs8cm4kCqSCs/JCGQzD4+8GOtZ+QAVkv0BoFXa2l32qcf7ZjfSIi8rW0xsDdiv9OymF191qgQq3w8lllgGf3AKlFA88ndAbWf62zksiB0OPMyOWzypwJx1kHgI/bAH8uCS1787uBgKxs48Rh4OUqvBiQLNG75bdMaxk4+ZjZm2Zj9qbZuvv1nDh9wnC/USA1uyW43yjDMQCsvGul5uvqWLMjBn8/WPMDQN1xgSX47AR5XqRXaN2LrcO28SYhREQIhMi7c8K3AcDIBuHlAaDePWfCMQB0mex8v+SVR+S+3DNLu5w6BBcr5XxfKKFpBaufu/1sGlTVcjrmmIa0GTfPwKtNX7VVr1aZhlUa2j4+EQK1lhOnT9g+JiFHkCNRvEgKvn28aUgoblbrfHx0d30Xe0VE5FPt3oh9Gw8tA4anxb4dSkrqIBnJNAK7AVrWKq0VWqW1st2e0pgbx1guqzXq/OtdvyJVpEbVh3hTjvZHO+2DAVmhxvlnc74xEZEer925L7WY9bJe6zuRx6jnFdcZWydku19FOk+aAZmIEhvnmEZP726Dfgidyj5m3gncMjp8OxEFyUFy4tqJePnHl313pzyz+d1WcQ4yERHpC1mVwscfNrIOnAnHRGTqzlp3xqWdXt/0ikm9ypAcScBnQCYiosTm52BPlACevfpZANoXAS7ZviRsm5Fjp45ZLhvNSDKnWBARkX1+mqKgXs5tyt2udYXIi6JdUs7sTnqdLuuEF5a+oNlWevl05O/Nt1V3ZoVMrNy90rRf0WBAJiIiffINP2IViLXqPXFEf96zU+3o3sjEZBtHoykBFU0pipMFJ8O2W11SzkoY1SqnvjDQiWO06oiEkCTJaL/hTi/QW6vY6moU0R5PFAfCRlnPn7PkU5/eAmzIDTz2Y0iUQ+5z+4CUmC9dZfWc5flKFCM2Vq/QPF85gkxEROa6f+F2D6Ljx1BPRK7xfUBWjvRGcuc79Ugx755HRERE5F+Rrn2s5PuATERERETJTT0n+Y3ro7ubJwMyESUevQvKov2a/fX/AAf+OvM8pQjw3N7o6iQiIkc5cWMTBmQiIjN6gbvgFPBzNlC3R3z7Q0REIZy+259jAdlo7q6VFSG0judKEkQUEfVIsZNLlEUyCq28G13E7cZoVJyIiMJEHZDbvLUIa7YfNCyTNiBHN+waBWt5H4MyEblm/O1nHkcbRtd/DVzayv5xfropBxFRAogqIN+XvTwkHGsFWavhWF0m/dmvcfTkaQDAyr/+QeaF50bTVSKiyPw+17m6og3HHC0mIoqLlGgO/jZ/V/CxXgi2MvqrVSb/hTP/I+nwjr37dBMReUrWgejDbfHSzvSFiIhMRTyCfLrgzA2AIpkCYWW94c3D2nJdYiK/05seYOWOZlrHNu0PXP9M9P0y83Fr4M/vzfsTr1HdgX+ZlyEi1zix9m4y8fr7FfEI8iVPz3KkA1bDdaPhuY60R0RxklXGeO7s8+UiO3bBK7Gdkyu3rQ7HREQuyMjOwKETh9zuhi71+sOJwvVl3qyOEG/bfzTGPSEix6gDbKSrSqiPy/scmHrvmTpiMXpr1Fer7UWz4sQ79YE9663VqVWfesUMreO6TQNq3GDeFyKiGPHqyLEs6oB8ZdWyTvSDiBKVZogzCYp6+zNuA7b8ACz7IPp+eZVWOI6UXlAf17FwPy/6I6LIJeroMeBAQP5ly/6ojucSbkQJJtarLrQd4f2AHM06zHrH2n0v9Y7jknEUY8rQ5PVRwkhJkFA7uzYAe6+x34J+mLN5juXjWk2NYOWbQvLf4ZNWn6Buxbq2jwO89fd7dN6j+G7LdwDs9evWr27F+v3rbR/n+hQLSk5Ga1wbLQ3oZPv8cEYJT2/0PhieYzRNhSiBqUdNrYyiapUxukhNXb7hxIZhZfTCnvrYu7++GwCw5M4lOKfYOZrHvLb8NWSvzrbUR6PXYtY/q+XMjrHy3uX1yNP9W1kJyhEH5MplzsKOA9HPC2ZQSSxW/578mycBp4IXRzyJfMVLo45OkwNW0ZSi+KX7L2Hbtdz4+Y3Bx8r35s+Df6LdF+2QkZ0R9p7JzwctHoTpG6bj+zu/R+li5ks96gXAjOwMNJrYSPNvM3vT7GA4/qnrTziryFnBfRIkCAjNvhm1p8csaKtNyJ+geWyBVIDMsZma752y/tbVW+OVa18Ja3PtvrWoVa6WYdsRr2Lxw8BmwceXDZpt+3gGJCLSJa8kwXAcOY4ME8WMMhwDxgFx5+GdmmWqnVPN0T4ZhVV5m1Yo7b+wf7CMMhwDCAvH8Tb0p6EAwl9TikjBXZffBcA4aCvDsbKe22fcrlU8hCNTLI6fKjC9Y55RIDba/9u2A2j39mIG6hhRryKifp/TBuRg9fMt0f2Dn0Lmm2uVs1KnWXt2+ma1TIEk4eKBs3TLuD3dg1SUobhyJvDAQuMyRERxIAexZV2XOVanPA3AaCSUtD1Z/0mMXTNWd3+072dUAVl9Iw+7N/VQHh/JDUGMjlHv+7BHfdyQfn7ExwOJN+rdcFguNg1tC1H4ATFtQI5mIPzPc4ELCuTtWuWU+4zeJ2U5I+p6tNpUB1u9/l88cBY+vbcBmtQ8T7NtvRvS8CY1Udr2C3DBlZEfzxFQorhau2+t7sjawKsGokt6F91j9UbxrIQU9chnNHXFQ4kiJSyXVb6WeK34YKedaydfCwC4s9adsepOxHp/2xuAe3/3qG41DQTChVlwNAtMy55pHvHxFLnvBzQLhmPA/O9kpVysPHxDTc3teiFdJj+Ww7H6GKWPFm8K29alQVX7naWA96+3f8y2X8zLEFFMGH3tPPSnoXEJeF4Px4lm/7HAN8NPN3ja5Z6EW7xtsavtO7aKRTShqULp4hEdH21QY/D2LvXf5vEbL8Vb3/0efG5ndLd9ZhXTMmnlS+H5mWvQs3H1kO0v35K4azzGTM85wEctIzvWyogzp1cQxUTpYqXx/Z3hd5BsPa01/jpkfKtzuxdfafH6rYejEc/XFElbo1aOQu/M3jHoTeSaXtgUC/5a4Fr7UY8gk3/JUxLknxrPOHP7cDd1qndR2LbXO9UxPW7+k9eFPOf0iihUvfrM46wywKrPtMutnGRcj9Z+hmOimNEKxwAwu6P9C/EjlWjh2GzaiNvk/r3767su9yTcOze8A8C9947rICc5o4vt/Oiz5VvDtvXMXoaxPa+ydLxyDjO/YYiCcq3dafcHfrRk3qFfxxcPBH6M6tZs2+Y+znUm8oQHMjXOdw9x44K6phc1xfQN0/Ha8tcwpOEQS8cMWjwILzZ+McY9c9+LS2P7GjmCnKRm/7bT7S4YUgf12kPmhjy3E14Xrt9tqRwDscOyDgD/7R7ZcZHsIyJHSZAw7fdpuGv2XXFp76E6D8WlHSfUG1cv5LnRKGenyzoZljFcQ7laYA3lab9PM+2THNqnb5iOzLGZYfuvnXQtPl3zadj22ufVjrh/dspEYuVdKzXrlyBh8rrJAGL3rQNHkJNU6ysqhW1zavTYqaXR2r+9GDP6NgYAHDx6Ei3+o91n9cV56gv31KtbWFlBgxzS4Z3Aj11mQTieITrSWzxb3cfQTx7h1WkAXiKPIh8/fTzs/dK6cxsAPHv1szh04hBmb5od1Xus1Z5e/wqkAsttjW87Ho0mNsLBEwdt90/5ms36ZzWAK49LESm65bTacJKQJMlov+FO8rdjJ0+j1rNfB5/LYdJoKTXldqN1r2XtalfBO13+q7lPadittXFH/YtC6laWffiGmnj8xksN29Lqp1Y5rdcp27rvCJq8Ms9ro8l2VmrnOes2u/OkTT8MlLFWzk55ZR8Z0GPB6jnrmfNVGT761OmDXpm9NPfbWbbNbnk/zT+W7+QG2O+33fcn0uOU5VfetTIkbOqRIKF2dm3b/VO+H1VLV0VOR+cHmvov7I/Zm2bb6pdFmucrAzKRgtFNQ1zEgEzkL74NyEa37TXar1XWbnk/BWRKKJrnK+cgE6l4LBwTEcXU3V/f7XYXiDyHc5Ap6XHOMREls6euegqdZnTS3c+5yZSMGJAp6XHEmIiSWXq59ODjj3/7GPdccU/wOcMxJSvOQSbyPs5BJvIX381BbjC+AY6cOqK5L69HHjLHZqJAKtCcJ2wnRBsdzznI5BJepEfkUwzIRP7iu4AsU4bda6pcg/dufM/F3hDFBQMykU8xIBP5i28DMlES4ioWRERERERmGJCJiIiIiBQYkImIiIiIFBiQiYiIiIgUGJCJiIiIiBQYkImIiIiIFMyWeSMiIiIiSiocQSYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQCYiIiIiUmBAJiIiIiJSYEAmIiIiIlJgQI4jIYRk4WdzjNqeJIRY60A9NYQQM4QQW4QQx4QQu4UQuUKIG53oJ5FXJML5qlHvkMJ+f+t03URuS4RzVghRwqDvtZzoK1lTxO0OJJlrVM+/ALASQJZi2/EYtT0IQCkH6ikNYCeAyQD+AnAugN4A5ggh2kmSNMuBNoi8IBHO16DC/7n2A7DbyXqJPCSRztkxAD5RbdvsYP1kQkiS5HYfklbhJ9nFkiR1c7sv0RBCFAOwFcBCSZJud7s/RLHg9/NVCDEfgbDQAMC/kiQ1d7dHRLHlx3NWCFECwFEAz0qS9KLb/UlmnGLhYUKIe4QQeUKI44VTGT4WQpyvKrNTCPGBEOJBIcTGwmkPy4QQTVTlwr7+EUKUFkK8VnjccSHEDiHEFCFEeTv9lCTpBIBDAE5G+lqJ/M7L56sQoieASwE868RrJUoEXj5nyX0MyB4lhHgYwEcAfgVwMwJf39wEYJ4Q4ixV8ZYITHN4CkCXwm1zhBDVDeovAWAegF4APgDQFsDDCATdcyz0L0UIUUQIUVkI8SKACwG8a/0VEiUOL5+vQogKAF4F8IQkSQftvTKixOTlc7bQo0KIE0KIw0KIb4QQ6ukjFGOcg+xBhVMWBgOYI0lSd8X2DQC+AdAdwHuKQyoAqC9J0s7CcvMA/AngaQD36zTTE0BdAK0kSZqj2D7FYjffAtCn8PFBALdJkrTY4rFECcMH5+vrAFZIkjTR2isiSmweP2cLAGQDmAVgO4CLAfQHsEAIcZ0kSd9bepEUNY4ge9MVAMoBGKfcKEnStwB2AWiqKr9QPnELy+0HMAfhFywotQDwp+rEteMVAPUR+MSdC2CKEKJFhHUR+Zlnz1chRHMAt+HMh1ki8vA5K0nSCUmS7pYk6TNJkhZLkjQWQGMAewA8b6cuig4DsjeVK/y9Q2PfTsV+2S6NcrsAXGDQRnkEVqGIiCRJWyRJWi5J0gwAHRH4muqVSOsj8jFPnq9CiBQAowt/dgkhzhVCnAsgFUCRwufF7NRJlCA8ec7qkSTpHwBfIzAoRXHCgOxN+wp/V9LYVwnAXtW2ihrlKgLYZtDGHhif3JZJgaVQfgZQw4n6iHzGq+drMQCXAHgEwH7FTz0ERsj2I/A1MFGy8eo5a0QA4LJjccSA7E2/IXAC36HcKIS4AYGTcoGqfBMhRCVFubIIXFTwg0EbcwGkCQdu8CGEKAKgEYAN0dZF5ENePV9PALhe4ycfwC+Fj7+yUR9RovDqOaupsL1WAH6Mti6yjhfpeZAkSSeEEEMAvCmE+BiBm3JUBfASgDVQzZtC4JPqN0KI5wGcBjAQgb/tSwbNfAzgXgBThRAvA1gGoAyA1gBeliRpk9ZBQoihAEog8B+GXQCqIHCRQm0E5joSJRWvnq+SJBUAmK/eLoQ4iMA6yGH7iJKBV89ZABBCPAPgIgRC+g4A1RG4SK8sgOfsv1qKFAOyR0mS9JYQ4hCAxxFYVuYggBwA/SVJOqoqPgeBEaFXEAiseQBaSpK02aD+Y0KIZgCGAHgQga+V9gBYBOCAQdeWA+gLoBsCS9XsALACQENJkvjplpKSh89XItLg4XM2H4El4W5H4P+xBwqP6SpJ0i82XyZFgXfS8zkhxE4AMyVJus/tvhCRMZ6vRP7CczZ5cQ4yEREREZECAzIRERERkQKnWBARERERKXAEmYiIiIhIgQGZiIiIiEjBbJk3zr8gcp+wUZbnLJH7rJ6zPF+J3Kd5vnIEmYiIiIhIgQGZiIiIiEiBAZmIiIiISIG3miZPyK+VjvS1+XFvUyne7RMREZE3cQSZXKcOqvGSvjafoZgiNrJXbvAn1u0QEVF8md0ohFfYUsKTA7qHwzJXsfCozXl7kDNyFfqMbuZ2V8hbuIoF+dqIzu0AAE9MnulyT+KCq1gQEREREZnhHGRyjNFIrNY+O3OAlXOUzY7TmrLh4dFhSlDqqRHqUWatqRPKMsr9WiPURvWP7JWLPqObmfaBiIi0MSBTiPxa6SjTvj2qvPpKzNuSQ+u6/16JgqNHLR2TXysdaZMm4qw6dTQv7FMH8c2dOuPoqlWuXARIyUsOplqhVckosMr7tIK0lfq1nhMRafli+BBs/GVZ8HnNq67B7z/9EHwuT7mQpaQWwWMTvoxb/9zAgJxA3A6B6WvzDS+4i7Zv6tenVZ96W9pnk127CJCSm5XR2q/e/BU3PVLHsfolSYIQdqasExEBG39ZFjLfWB2IAZjuTzQMyOQ4dZB1KqCm56+J6LhKzw7CzhdedKQPRFaZjdjKUyC0RoMjNSHrR3QdcjUAoHS5ElHXR0Qkh+FkCMVKDMg+ojf3Vrldb66v2XO99rSmMBiNBJe4/HIcWxMeZB0Z2bY4MsYRY/ICK4FXLvP58OWaUzDsksMxEZGTWvV+FP+5rrnb3YgrrmLhE3IwVf7IlM/tru2rrlepRu53tsNm9WlTQ55v7nyHreOjpfyAIP+UufnmuPaBqGL1c2zN+b3tqXq26i9aPJVziokoLirXrIWvR73hdjfijiPIPhLvOcZFq1QJa7/S4OcsHSv39ejKlbHomm6bQPho9YEvE/tCAoo/dThVT5O47al6GNVnnu4qEmYrWHw9Jg8bVuzWrf9/bzYNu0kJV6ggomiM6NxOc93jLi++hhGd22HasCx0HJBlWDaRMCD7hHo6RTyD8snt24Nhueydd5qWV1+sx9UjKNFYCaO9R14f8fGtHsiIpY3x7wAAIABJREFUqg9a+xigiUjPE5NnYkTndhjRuR1aPPBw8LnWfvl5omNA9hllUI5H8JTDbiRtHZxpfgKd3LoVBSdOAAD2vv8Byt9/n+12ZHJf1WsmVxs/Dn927abd/rZtwcf7x49H2a5dI26fiIjIr9Sh1+x5omNATjJrM8OXlLISgCMJydv6PQmRmqq5r+DIEay7sm7Itr9HjMDfI0YAiHzU+eKvpmPjTR1MR9rXX30NTv/zT8i2nS+8GFztgqPeREREyUtIkuGt4HmfeI+wcte5LXffg8NLl4btP/3PP1h/9TXB7UarWKSvzcf2J/uH3SjE7TWWk5ydhW15zhK5z+o5y/OVyH2a5ysDMpn6o/mNOPnXXwzI7mFAJvIXBmQi/9A8X7nMG5liOCYiIqJkwjnIpMuNFTOIiIiI3MYpFkTexykWRP7CKRZE/sEpFkREREREZhiQiYiIiIgUGJCJiIiIiBQYkImIiIiIFBiQiYiIiIgUGJCJiIiIiBQYkImIiIiIFHijENIk3yRElp6/BhBCf7/qZiLK/elr85FfKz1YRvlY67nW8eqyyv1pkybirDp1jPuvqsOo70RERJTcGJApjNkd9PJrpaPGvHkoWrlS8PnaKzJQ67e84HOjQGqlfaMAbfe50Ws7tXevYXkiIiJKPpxiQWGshEU5HMvlpVOnHGlbL0wfX78+quOVlK+vSPny1jpGRERESYMjyKRJOY2h3D33oOJT/UP22x0Vtktd//7xE1BpSJalY8/r86CtuomIiIiUGJBJl3LO8LHf8lDt00/D9sW67UjsGfkuKvTtG5O6iYiIKPFxigVZcmTZ8pDnsRqFjTa8WjmeI8hERERkREiSZLTfcCclJq0AabTKBABcMGIEzmnbRnN/6RYtcGju3JARabv1q4+1swqGVnmz9j1GmBcJ4jlL5D6r5yzPVyL3aZ6vDMgUc2arYpApBmQif2FAJvIPzfOVUyyIiIiIiBQYkImIiIiIFDjFgsj7OMWCkt7IXrkAgD6jm7ncE0s4xYLIPzgHmcinGJBdsDV/H75689fgczeD2ei+83H6ZIHr/YiFkb1yE+41gQGZyE84BzkR/PXwI8ivlc6lyohi7KL0cp4Jbr3evs4zfSEiSga8UYjPHJo71+0uEBFFTZ4yoX6s/iBgtE+vnJW6tMrojWZrbVfX5ecPMCM6tws+rte+I5p266lb7onJMwEAp0+exBvdbgnuk7dTKOV7C/B98hMGZJ+puWQxfm/U2O1uEPnakYMn8HH/xWHb7YQcq2HKLEiN6TsfpwqnT0TSD602tOqQ+2U1cMaa3LbZFAtlOT1adfy7/1hYGWV9APDRk+H/BqwY2SsXD757PUSKCD7341SRL4YPwcZfloVsWz5jGpbPmKYb5H6Z/RXmffJe2HZleKYzlO+JOizHk17b/JvpY0D2mSLly3M9YaIolTynmKXRRTNmYXhkr1xAAH1G6Y9SnjpZEFaH3bCl9Vr0Rj39FuIidXbZEqZler5qf7Dhg8cXAkAwHAMI++DhF3I4VoekEZ3b6QZeORzbOYbc1/zeB/Hth++63Q1f4RxkIqII6AVN9XZlOFazMvIbiYzrLtTcnsjhWP5QoMfKSLQVx4+cCmnPrF2vkkcUtQJt2cpVDI9lCI4d+YOG06PNmS3a4InJM/m3s4EjyESUlJwONXJwsttOk86XRt22VCDh3QfnRV2PXynDr9FycH1GN8PX7/0WLNPz1SY4q3TRiNqsnnle2LbTJwuQWtT/404933iPI8JxJgfiitUvQbdhb/L9V1C/D8rn8mPlBwpl2QnPPIEdf6zT3GeGATlGlKtMmE2JMCurt2KF1akW6ls969VX9aOPUKrhNYbHajl98BDWX3WVaTmzlTc4dYTiRStERRKYb32ybnDawgePL9QcLTYbtV0xdwtqX6894mvVuw/OC2ln+usr8Ne6/VHV6UfqoKz13rf63xUAgGP/nsSH/RaZ/n02rNitub1N79pR9tYb3JwXS6GU4U0OfQzJ5tSBWWnHH+tC9m1fn48ql1pbBcz/H3U9KqKwl5rqfEdUjEKqOhzHq10iNzgx3aDSJWVCN9hZsRpAuSqlwi4mO3b4pK06lkz9I2xbMoZjJSt/2xJna48cqz8ofT0mz3bdRHbMevs1ze0Mxtao36fV878FED7y/MTkmZj47JOW6+UIchyszayDWit/NS2Xvvo37e2KsP1n97twZNkyzXJmtvd/Kqy+eLAyms4ATfGmHGGMdrpFwWnt+z3IF2+dVboYer7aGIcPHMcnTy0Jtnvncw1CRjv3bvsXk174yVYIa3RrDfz6zZbgcz/Ohx3ZKxdptc/D5lV7Ql57/vc7kDs2P6QcANSsVxEt7vtP2PZqGeXxZ95e3TYAoHT5Eji095hmmfrtqmPZzE0Y2SsXFaqWxu4th9Cx35WY9tovuvUp+TE8M4S5L3/xfADeHs1/v889OLhH+9sUAKhyaTrufOHVOPbImplvDg95flnDJpaPZUCOIVGsGKQTJyAdP65bJp7B8MBXX7k6jcGobU6voHiSg6tyqoXmChQaz8Muwis81uiiPaO5sX1GN8P7jy7U3W+lH8r+OxX646XP6GYY2TsXm1ftCduX3rAy0htWtlZHr9xgONabf/xx/8XBcKxV5qp21XFVu+oY2SsXu7ccCpbR+psBZ97jypeUQccn65r2k8iIlz+sGIVjANi+Ph+vd+mAxyZMj1OPrKl9QytUvSIzomMZkGOo1qqVlgNwMgTE/FrpSfE6yR/0Qo/eczt12d1//xvXRnysXrloXk+8Ga30YbkOC6/vnlesLesW6XvuR290uwWPjvvC7W4kNfVFZl6lF+DlvhecPh2ztj/oey/ue/tDLPtqquVj5Pc10g8enIMcJ8k8hUAZivNrpWNz5ztc7A0RESnviDeiczuMfqA7Vn03J2bLjJE5L7/nRiEz1iPfT0yeiUN792BE53aof9Otto+V/z0vnTbJ1rEcQY6x9LX5uuE43qG5yivDzQvFiPJ9OLpypaXVMYiIKHaUI5eH/9mPb9572+UeJR516NVbjkwZ5NS8PPUiXpRTN9Tvm5LWexXp+8eA7AHxComiePG4tKMnfW0+pFOnsPaKjOA2BmUiIvfYCQ9ujiL6lVPvL8UfA3IcJeoc3ANfWJ+/JooU0VyPOVHfGyIiokTg5SkgscCAHAda0yzk55WyBrvRJUftGjo0ouPUQTn/8v8gfc1qx/pFRERE0dny20pMeeGZkG3N7+uDzBtbB58nYnhmQI6zA19+iTI33xx8XvYOf1ywtvvtt1Ghb9/YNlJQENv6iYiIyBZlOE6maSBcxSJOLhgRuFPO9gEDXe5JZPaMfFdz+4mNGx1rI7VsWcfqIiIiouj8Ojcn+DiZwjHAgBw357RtG3wcvDBN5855XqJeok0pv1Y6NrRpi5L16xnWkV8rvXB5t84h2wuOHAmp89Ifvnegx0QUD8qbnxBRYvruw1Fud8E1nGLhptRU0yJGS8Gp953/+OMo/7/7o+6W3f5U+/RTS0vWHV25SrccL9Ajcp/RHQHVbn2yLmaPyYtxj7TZ6ScRRe7RcV/gjW63GJZJxPnHAANyXEUSAp0IjtHWobXqhLpeK7eR1grHDMb+oxdOTp8qwPjnluKulxu60CuKt0qXlLF8Zzoi8qfUokWDj9V3pVs04RP8NP1zN7oVFwzIZJlTQZsS0+iH5rvdhaQjT3EwGk3N/34HcscGzr3/vdEURUtof3O1e8sh5zsIYNFn67Eq9y8A5rdmnvj8j9i3/TCKlyyC+/5P+/bbseonUaIyGuG1cmMS5Q1l9Mon4iiykCTJaL/hTiKKC2GjbMzPWX697S69eb+dnq6PClVLWyqr/PuZzSNW/63V5Y3+LWjVrVVfr3eu0/yAFU0/XWb1nOX/Y8k33uzWEadOngg+T6CL9jTPVwZkIu9zPSBPffVn1GleFZf8t0JUAfmTp5agco0yaHn/Fbpl3nt0Af73RlPd/d9l52PHH/+g2wvX6JZZ8vkfWLt0B+59rYlumckv/YSa9SriypbVrHXeI7RGja2MJCvL6pWz87c1a1Nr/8x3VuLP3/Za6rtR/T74kMaATOQfmucrp1gQkSE5qHy9QfuCrLU/7MB32Wemz+gFGgAoUjQFf/z8N/74OTzgyGVKnlPMNDQpHyvL/LZgGxZMXGdYRlnHnq3/4ocvNng9bIVR97fP6Ga2VpTYu+1flL/gbKe7FUbdz3YPZXLlCyLyBQZkItI1slcuql1RHu0eygzZplTrmsqodU3lqL4C1wuy6pFCs6/7F0xcZynsatXjt5CsRet1nDpZgDF954dsk2J8Tx7576L3b+LzYctx2wDj5SGJiNyUVOsgy+vxEpF1ynAcDbsB+rqul9luIyVFYGSvXBSc1v7melnOJtt1+tnIXrkY03c+ql1RHn1GN/PMh4CTJ0673QUiIkOeGEHefGcXpE2cEPN2uIoCkTv6jG6GKcOWB0Ny75HXIyU1dNqXla/ezcr0fvf6sHLKUPjTjE2W2/Kjem3Tgo/tzEuOFa8EciIiuzwRkI+uWOF2F4goxm5XfKWuNRVAKzQraR1jtkqDeppGhYtKY/fWQwkb3Bq0vzjk+UXp5Vzph9050UREXhPzKRbytAblj3qfupydOqzs3z9pkmn9+ydONKyDKFnt+evfmNSrDlCj+sxzvA11EO70TH3H23CD+r3TC6Nb8/dZKme3jFUje+Xi5PHTYducqpuIKFZiOoL8z9SpAPSnNijvsGY2/cHKndpk6nBb9o47UPaOOwxD784hz4fUw4BMdGYksGylUmjd6wpMyPoxrIwkSfj24zXB59NfX4Eb7r4cZ5ctHtw2slcuzq9WGrcPrK/51b/czsheuaieeR7+2XUU+3ceRuPbayLzhotC6ukzuhnGP7cU//x9JKwvyrqXzdyEn2ZqzzmWy6U3rIz873eE9cfrtEZo9Va2UE83MQqWbR+sjZx3V+lOUdE6Vq+s3NZ7jyyw8IrsKVo8FSePn9Ztm4goWjFfBzm/VjqQmor01b8ZljEKwFYCtJXycujVCtRWthG5xPV1kJ2YzyrXYXSXNLO25H2dnqmPCheVDtuvLGPWX7lc7WYXokmnS0167x2JsuJGguM6yET+4d6NQpSjsXrBNZqArDXay4BMCcT1gEzewYDsCwzIRP6heb7GZZm39LX5IdMpnKQMvcp2iIiIiIgiEdd1kI3Cq3TyZEzqJSIiIiKyI6YX6VkdLT630+1Ym1E7+NxO4E0pVcq0Ha1VL+y2Q0TkBZxeQUQUe3GZg0xEUeEcZCJ/4RxkIv/QPF89caMQIiIi8p8RnduFPH9i8kyXeuJvIzq343vnMXGdg0xERESJ44nJMxnsKCFxBJmIKEmY3VyEiJynHGXvNeZTlDq3rOY+5WN+6HAfAzIRUZLoM7oZVi/ajvnj17rdFaKkMKJzOzz4wUScVbp08Lky/D4xeSaOHz6Md3p2Zij2GAZkIqIk8p8mVRiQyRL1/GKAI5uRkMMxANz2zAucb+wTngjITt61Tr2kmxeWcuNd+YiIyE+0QpxWYCZ9p06cAMD3za88EZCdEu/1jfe+9z7K/+9+wzKb7+wSl74QUWKSby2tnD9cqkxx3D28UUgZIHxOsd3bUo/slYvLGlTCuh93au5X1nXy+Gm898gC3f16fdcrZ9SWsoy6PmU5vdfL23PbYxToOPppXZFixQBw1N2vPBGQnQy08Ryp/fv//s80IKdNnBCn3hBRolIHPK1A6ZR1P+4MtjWyVy5uefxKVLn03LA233tkQViftIKoVt9/mrEJV7WvHtymdQzDrvs48hkfxUqWdLsLpIHLvEXI6l0CKXbya6Xz70BJQS8oxkOVS88NPv56TJ5u23qhVWv7spxNEffHLByr+xav94lIz4jO7XBwz+7g47zcOSH7hQjcp+LUycCUjPVLl8S3g6TJ1RFkZbjRGvmV5+4azSvWu420Vp1aYUpdZt/Hn2DX8OG65ay29++iRdh6//9027HaJ70AmFKqFC77ebnmPiKiWDhy8ETw8V0vN9QsE80Ib7RhVms6h7yd7OPUgOjJ76E8Gn/3iHdR/sKqmuXkMo+Mmxa/DpIuVwOyXuhUUl/gJo8aytvU+4yCqEhNRa3Vv4VsU9s1fLjhNA11n/XKnt2kienr06pD/frU7QLAyW3b8McNzXX7mCx44SORe0qXK+FIPX//eQhThi4DgJCpHdGQQ/qHTyyKun/JqOLFNbFr4+9udyOhWPmwwQ8k3uK7KRbRhCJlOJYpw6sbX9cbBWE9RS+4IFbdISKyxKmpC1OGLkPF6uc4NsqrrOfY4ZPoMVR7pJv0dRv6OoDwOcick0zJxBMX6Rkpc9NNjtXF+aqhzEbcgcimqSit+++VKDh6NGTbZct+QopiXUjZjucG45/PPgvZVmXoyyhzyy2GfTB6DVqj+KJIEdT6LU/3GCt1E8WT1rSF86uFn0PqY2JFpAhIBZLmvkiC7m1P1Yu2S7rOLuvMSHeykb/yNwrFBadP4/UuHUK28W5wlCg8H5BTzz3XvJBFDDvh7KzRrAyPNebl4o/rmxnWoTt/WiMc65U9ffBQ2Lb0tfnY/mR/HJgxw7TPx9aswaaOtwIALp45AxvbtYd06pRmn+U+pJQqhaoff4TNnTojv1Y6zmnbBheMGGHaFlEsjeyVi9LlS+DQ3mMAgNsH1g/uk+fejuyVi9LlSuDQvmOadWQPXIJ/9x8PqRMA2vSujeqZ51nuy4PvXh9sDwJAYVZOSRE2X9WZfmQ2uwgnj5/GmiXbdecSW1WiVFFenOcAs4CbkprKEEwJS0iS9ihAIcOdTjEKWOX+v707j4+ivv8H/pokBBIS7iOEcIYreICA4RYIrVqk1qu19vL8Kh4/rfpVQUAWBImi0tqieGv1W2tba23Rai3hCIRLBIpyBxCSkHBfgZBrfn8ss8zOzrk7uzOz+3o+HnmQnfnMZ967OPjKJ5/5zK9+hY5PTjHd3sxool4dgLkgbbWt2XrFc+ewbcBAw7nOdjx85OCL83Hktdcsf5bKdlr1qW230kekx1mpzY73EUVWUkdMrlmKHfno8YJJReg/KhvjftFPsy0Q25vS7Dhn9fFzeGfyyoj7Udbl4M15Zq9ZXq9EzlO9Xj03BzkSRlMspBBUt39/LMoBEFrTtgEDY3buDo88DACoP3xYtyY9XhyVz1ZZpUSLF98fxa/7FxZohmNpf6xDoR3nbN6qqa21l207Zks/RJS4HJtiYWV5NjtIy8UZzTHtt/m/2HbJparHG/Vndfk5szVF285Roy2dc0d+vurUByUrfRo9cMVOqd27ae7T+uGg/JFH0PnFF6NVEhHZ6JPfbMCVd13kdBlE5GGOBWQz4UmrjdXtVs4pNGliOtiZWQ4ukj709tsVovusXYsd+fmm+peHx2b9+6PL668hpW1bW25+bHPrrRH3EU2NZ844XQIR6Vi7aA/WLbrwAJLeQzo6WA0ReZ3rb9Kj6Epu4b9hTuuhLJJw5uKe274dTfv2NdV258hRrpjO4IYaiOT4kAtz8if2QP7EHsYNiYhMSKg5yGROu3vvNd1Wb/R494+us6McIiIiophiQKaQUdP2Dz0YcZ/JrVsDUA/Qu68JXlez98oVmm0PL1wYcS1m6D318Mgbb8akBnIXaRkz5RcREcU/TrGgAL3RYL0bCtWO67OqJLDdaI5yStu2SLvkEpzdvFm1bbtJk3RrtOsGz97Ll2HnFWNUz9H2rjvD7peIiIi8xRXrIBORLq6D7AJOrDFMnsV1kIm8g+sgE1HscVoCERF5DQMyEREREZGMJ+cg2/GYZSLSJj2mt+Rvu7Dh3/sC2+XTC5Qjw8qpB/L98u85RYGIiNzOkwGZiKKvdMMhbFtVGQi0h/YFPz1x3C/6of+obAAXVnyQh9/7FxZg37dH8M/fbUq4UGxmvvK5M/V445Hlqu2Ux2tNUzHzuZqd4pJof0dERHo4xcIF1FaHsLtv6auuvFx3P5Hk81c34455owKv23fNDNovhWOA4Spajh6o1g24RuFXOXKv9ffUJa9NeAUSEcUpT48gV69ciX133hV4rZx2oTUVQ759a7889F23Fvsn3Ysz69ej38YNEJo103xynFhXh22XXBp43XzECHR9K3idXGX/WvVFm3Tu5iNGIOe3v8H2y/PRpHPnoDZ66/9SYjMTeld/shvr/7U3+sUkqA9mrgGg/nchhV/lyL1yv/J45ag0f7ghIgrl2RHko2+/jX133oW8bVvR6Zk5AMIPedsvz0ddVRUAYNvAyzQDLoBAOM7bthVpAwaguqRE9byN1dXY2i8Puf/6DHnffqNbX962rVELz/2+/QZd33oTSZmZnLdNtpGmVAy6sqvuyCRFTuuzNfuZ8++GiMg6zwbkqmefCwS+VjfcEHH46/WfLwN9pHTsCEB9xFceZrt/+Cc0699ftb/tg4cgb9tWpPboASQnOxZOheRkR85L8e/+hQVITfP0L6EIwKLfb3K6BCIi1/Hu/90EK89OsKb3sqWm2/b420eWRq6r5s5FxylTAISOKGuFaK0R7cyCAuS8vCCkrd7rcIO61pQTre1mz603Wm+mT932ooitef3125AttObCdr2obYwrIasmPjDA6RKIiFzHsyPIeVu3OHJe5U1tWoEtKTNTdbv8uUl527ai7R13WDq3kJqKvl+vBwCcKirCkddeM318LEmfS+b3xgd+4NjaLw8Nx46FtN05+ooLQfv8dBR5H2HXcD4cZxcWovsf/y/QZ+3u3RH1S0Dzlk0D0ywWTCrCLTOG6raXt6XY4mdOiaqupgbvPvaAbf29eMu1tvVF7ufdEWQHGI2iyiWlp5vqs8Pjj+HIW2+ZOrf8vHnbtmJrvzwcfHE+2t59d9B2rWNiRe1zkurdMXxESE31hw6ptlVjZQRbuU3qt3TCNRxJNmA0b/W2Z0eaPoZzYJ1x/8KC4Bv5XikABAZmSgwv3Dwx6Pvv/88D+PL13+PRDxcBADZ9+S/8540Fgdfytm1zuuK2F17W7AtAyHEUf+I+IDeePYuktDTb+lMGq2P/93+29R3v9IKvUo+PP8ae66833S+R12xbdSDq5wgKyfcGB+O7fzsGTZryHgWKP2ohVh5y7ejrhZsnMiTHubgPyNsvGxQyqhqJ08uXI+OKKwKvK5+eHVF/8SySz7pJVkdT/WdNn6bbpuz/PYhTX34Zdh1E0bLiLzujfo73pq0KfM+RfEpkj364KOyQrAzCkfRF3hHXAbnft99g20UX27rG7/677wl6bWVUNBIpbb11s5OQmhrV/qXPvPXPf666v66sDLu+930AQKubbkSn2bODjiOKhf1bjqJLf+cewnHy8FkADMeUeOwc3WUYTkyeDMhav1JXbhfOL6928Ll5OL18OXou+qfhMXqvpe935A9F+4ceQuuf/8xUH0bb45GQkhK4mdBuevOOJVI4TqTPnNznHy9tZDi1mXIOtdHnqzbnmn8nZAWnUiQmTwZkqzo8/hg6PP6Ybf31WbvGtr7iVeOZM1Hpd99ttwNg8CV3C7lBThbInLhJTu+cXguLyicBmm1v5RjyPuUc4erjwSsoZXrst7IUewkRkCk6qletVt0uTTtRW0Uj0pU1qlevjngtZyIjekEqnBFJrWOiHdgEQYAoirpttB5VTeRVnfv1R/m24KVgF97zy6DXPQflhxz36qRfqfandkPe1hVLkTdqbGSFkqsxIDtELazJt8VihNRKDfLQK6c1B1urvR31mlkCTuv8PRctwu6JnE9G0ac10hmrMCqd956XxiAlVX21Cq1RbiIv++nM5wIrTUi0bqwzml8sHafWjgE5vgkGowv6Qw9EFAtWHhvJa5YCwddM6LXS1ur5lbTOEc4PEeHUHY33qsHsNcvrNYa4NBtpUL1eOYJMRES2+GDWGhytqDbdXitIxzDIEhGpYkAmIkpQ709fZdzIAikcWw22yva8mY6InJbkdAFERBQdRjcbnjjkXyf52ocG2nYus+FYr73VlSqIiOzGEWQiojijXGbOSJc85x5mQhQrnH9MVjAgExHFITPLyLlhji9HiYnIjRiQiYjilBsCsBEv1EhEiYdzkImIiIiIZBiQiSimFkwq4q/V49BNTwwBYH7KxM98wwAAr9y3JGo1ERGFiwGZiGLqvlfGxfxpchR9HXu0CHyv/NwP7TsVsq11VjoAoLFRDNlX+vVB/t0RkaP4JD0i9+OT9MLERyjHnl6wVfu7sNLeTGiWH2O1vY34JD3yLLXHastXAFn022exvaRYc78HqV6vDMhE7mc5IC+YVIShP+qJIT/oHhQSlGFg/5aj+MdLGzX3A4DYKOJlxa/BlSHk/oUF+MOTJTh1tAYAkN4yFbc/OyroGDOPE5avjatXt1Ffq/5eiq8//y6kf6O+Jt4/AN0uaRuyX++z1FvPlwE9YTEgkye9cPNEXDRmPK6+72HdNvJALAVqD4dkPmqaKJGs+WQ31nyyOxDQVv+9NGi/Mti9M3mlaqB7+b4lhqNyCyYVoXVW86DQ+Mr9S3DvgnGBNlYe/iCvo2rPyeCdItCseRPc+cLoQFtl3cOvy8Xw63INA+qCSUVo1yUDN0/ND6pNeYzys5Q/gU5rOTVOESAiL/p22WLNgKw1uqy23es4B5kojsmD3rDrcnX331Y4EkBwsJOPLqsdI/cz39Cgvhobwh8ck59DPrcVACAgEI716jEivU8pHBv1Jd/3i6eHh+wPCfIR1EZE5IRHP1wEISkJL9w8UTf0SvuN2nkZR5CJSNO1Dw5UHaE10rxV0yhWZR+7AuyA8V3w12e/YiAmIs975IN/BL7Xmj7h4ekUpjEgEyUwszcxvXLfkkDbO+aNQlpmarRL06Ssue+wLNvPsX/rUUuPXx71497YtHh/4DWnVxBRPFBOn7js6onY8Hn8h2NQHHVpAAAgAElEQVSAAZkooZkd8bz3Zf9c4sZGEa8o5iQ7QXn+7asrw+qn+vg51dFuK+FYTj7S7vRnRERk1Qs3T8QtTz+P7D79Aq/lCm6fhA2fLwq6Ue+FmyciLTMT973xQczrjSYGZCIyLSnJf7NvbU09Uptp//Px2kPLYlVSRN6ZvNK2ICvdrLe15IAt/RGRc+TBMBGmE0iUI8Zq7/3RDxdhx+qVQe3iLRwDvEmPHLK1X57TJSS8+18pCMwvriw9gT9MLQmZGqB86p30vTIct2yfFtj35qPFqDvXoBo85eFx1/qDYdcunWv3xkOG0xnk+1998EJwV66qsej3m7BgUhEG/6B72HUBQNEftnL0mMjjHv1wUUIFYznpveu9/z7DRppq52UMyBS3GMINCBdC4kfz1uPUkZrA44Il8hCpt97viUNnAyOoNdV1mmsCF/1ha+D1F69/E9ZcXXlN/1q42dTKE1L9fYd2DNnfND0FCyYV4btvjuD+hQUY9qOelmuS3PfKOONGRETkenxQCDlia7885G3batzQ5eeIEVc/SU8vOCcafhZ0Hh8UEieUD8VQ7pO75el5yO6Tp9sGSKwpGx7BB4WQ+8hHeZVhVm/fuV27sHviDwOvu771FpqPGB5ynPR9nARl8gCGY6L4p7b8WRw+YS6hMSCTY+QjvMrpEMrRX+Xr3RN/qBl687ZtReOZM9g+aDCDMUWdfJpIVs+WDlZCRLGkFnzlIbnHwCHYs/GrWJdFNmFAJscEhVdBCITgvTf/VLX9uV2laNor9Glw5KxEHzFN9PdPRH7KFSBumOILetIcR5K9hTfpkSskZWQEvj+7aRNSOnYMabN74oV/ePK2bcXWfnnY2i8PZzdsiEmNREREVshXeYjnxzLHI44gkyvVV1WFbEsbMCDoddD0DEFA3tYtMamNiIhISS/8ykMyeQNHkMl1tOYNd//wT6rb+23aCChWY0lKT7e9LiIiIonRChV/eXpqLMshm3GZN3KE8qa77Zfno/HUqcC26pIS7LvjzsD+nosWBc0/Vt7Upxaq9VbB8BhXL/NGRCG4zJuHmRkJVmunnGPMJd48Q/V6ZUAmcj8GZCJvYUAm8g7V65VTLIiIiIiIZBiQiYiIiIhkGJCJiIiIiGQYkImIiIiIZBiQiYiIiIhkGJCJiIiIiGQYkImIiIiIZBiQiYiIiIhkGJCJiIiIiGQYkImIiIiIZBiQiYiIiIhkGJCJiIiIiGQYkImIiIiIZARRFJ2ugYiIiIjINTiCTEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOATEREREQkw4BMRERERCTDgExEREREJMOAHEOCIIgmvvZG6dx/EgRhm439XSwIwseCIBwRBOGsIAjbBEG4167+iZwWD9erIAiFOrUft6NWIreIh2v2fF89BEF4XxCE/YIgnBEEYbsgCD5BENLs6J/MSXG6gAQzXPH6YwCbAPhk285F6dzTADS3oyNBEIYD+BLAFwBuB3AKQF8AqXb0T+QS8XC9LgDwd8W2lgA+A/APG/onchPPX7OCILQAUASgHsAUAOXwv68ZAHoAuDXSc5A5giiKTteQsM7/JLtCFMVfOF2LWYIgJAPYDmCdKIq3OF0PUax48XpVIwjC/wB4DcD3RFFc7HQ9RNHixWtWEIRrAXwCYIwoistl238D4F4AmaIo1jpVXyLhFAsXEwThdkEQNguCcE4QhEOCILwtCEIHRZtKQRDeEAThPkEQdguCUCMIwjpBEEYr2oX8+kcQhExBEJ4/f9w5QRAOCILwF0EQ2uqUdRWAXAAv2vZGieKAS69XNbcC2A9gSVhvlChOuPSalX4Te1Kx/TiY2WKKH7ZLCYLwIIC3AGwEcB38v765FsASlXlIV8H/k+UTAH52ftsXgiD00Om/Gfz/g5wE4A0A1wB4EP7pEi10Sht1/s/M8/9I1J3/B+RFQRCaWnmPRPHCxdersp9cACMBvCeKYqPZ44jijYuv2X8B2AvgeUEQ+p0P2VcCuB/A7zh6HEOiKPLLoS/4L4L3VbanAjgC4HPF9u8BEAHcLdtWCaAGQJZsW2v4f/p8XbbtTwC2yV7fd76vqyzW/A6ABgDHADwFYCyAyedr+MDpz5Rf/IrWlxevV5VaZ57vp4/Tnye/+BXtL69eswA6AVh9/njp6/c4Py2WX7H54giyO10MoA2A9+UbRVH8D4AqAGMU7ZeLolgpa3cM/hvolDcsyF0J4DtRFL+wWFvS+a/XRVGcJYriUlEUCwHMBfBTQRB6WuyPyOvcfL0GCIIgAPglgFWiKO4Itx+iOODaa1YQhOYA/gL/KPPPz9cyGcBt4NTGmGJAdqc25/88oLKvUrZfUqXSrgpAZ51ztAVQZr00HDn/55eK7f8+/+fAMPok8jI3X69yo+G/C/7dCPsh8jo3X7OT4A/eV4mi+EdRFJeLovgs/Cta/FoQhH5h9ElhYEB2p6Pn/8xS2ZeFCyFV0lGlXUf4l4fRchj6F7eWb8//qVz+RDj/J+c1UqJx8/Uqdyv8S1x9GGE/RF7n5mv2EgBVoijuV2xfe/5PBuQYYUB2p2/gv4B/Kt8oCMJ4+C/KZYr2owVByJK1aw3/TQWrdM7xbwDdBUH4vsXaPgVQB+Bqxfar4A/H6y32R+R1br5epXOkAbgJwCeiKPIBIZTo3HzNVgLoKAhCV8X2oef/1AvlZCMGZBcS/XepzgQw8fyyM1cLgnA3/DcBbIFi3hT8P6l+KQjCjwVBuAH+CzMFwByd07wNf5j9SBCEyYIgjBcE4QZBEF7XuzNXFMUDAF4A8JAgCLMEQfieIAhT4Z8j9brKT71Ecc3N16vM9fDPaeT0Ckp4Lr9m34L/psDPBUH4hSAI4wRBmALgGQAlAL4K4y1TGPgkPZcSRfElQRBOAXgE/mVlTsI/evu4KIpnFc2/APA1gOcAZAPYDP/8pb06/dcIglAA/z8S98H/a6XDAIoBnDAobyr8azLeA38wLgcwG/4LmCjhuPx6BfzTKyrPn5so4bn1mhVFcYcgCMPgf/pfIfzzoffDv4rFXFEU+XS3GOGT9DxOEIRKAItEUbzL6VqISB+vVyJv4TWbuDjFgoiIiIhIhgGZiIiIiEiGUyyIiIiIiGQ4gkxEREREJMOATEREREQkY7TMG+dfEDlPMG4SwGuWyHlmr1ler0TOU71eOYJMRERERCTDgBxnyiYXo2xysdNlEBEREXkWn6RHROQwn8+n+1ptu1YbIiKKHEeQVVQ8vRqHXvuvqbaHXv0vDr68Kar1VH9VhYpZq6J6DgA4vbICFTOjfx4iCubz+UwFXrPtiIgoMhxBPk85LeHc7hOBbTmFow3bm2lrdp+0TUgWIDaIum3ValHbpnXunMLRqu+l5TU9kDk6BwAg1jag/KkS1T7kx7S742I069Nasw0RERGRFzAgIzhMWgmfam2l0GkHsUE0dQ61cG21Br32QmpyoG+9kM9wTInmxIkTmD9/fuC1IAiYMWNGSDtOjSCiSCwuysX4glKny0gonGIhoxb+lNtqdh4DAGQ/NVyzrV3TFOwK2nrsDPREicTn82H+/PkQBAE+nw8tWrSA8smkb731Fnw+H37+858HgjEDMhGR+yV8QLY64nr4zW8AAEnp2oPvjWfrIy/MZaTPRzmifuzjXUH7iRLB2bNnAfjDrjRi/Mgjj4SE33379sHn86F3796B9vI/iYjInTjFIsG1KOga0fHVaw7YVAmRdzz77LNOl0BEcWxxUW7ge62pFfI2ADB40J/QqtXlQfv1juWUDX0MyAlObyRcSe2GPiJSt2zZMgAcLSYia6Tg6w+wYkgQBoCzZ/fL2gDLiwdj/dc/DQm9akFYrT8KxYBMlknzlo+8vxUAp1eQn1oQTORweMUVV2DJkiUJ/RkQkTXB4RgABIwvKA0JtWlpXYKC7xWj12NxUW5QIFY7TsLRY2MJPwdZa26tFiFZ9ZHdCSOlTbPA92e/OexgJeQ2+fn5TpcQM8nJyYZtBCGx/60gInfgiHF4Ej4gBxGNm3SeMwqAteXftHhxukLW4/75TYfe2AyAo8d0wYQJExLmQRbTp08HYO4JeMptS5YsQX19/N3IS0T26NH9fsM29fWnAyPG0pcaTq8IH6dY4MLc2rIp6oFVGQID7VUCbufZI0O2tb/rEhx6Y3NMA7GZB4VE4tyu47b2R+Q10g8Dej8Q+Hw+7N69O6TNuHHjAt///ve/x+HDh0OOk/+p/N5oHxF51569C9Cz5yO6bZYtH4CCcTsgCBd+m6UXfhsaziI5OQ0Ap1eYxYB8Xk7haNTuP4WDCzaGbNdqbzaENu3VKqR9i4KuaHFlN9tDc07haDScrMWBZ9bY2q/yHG4e/a6YtUp1nWqzuDY0mWUmmPbs2VO33QMPPGDbuYgoccjDsZGlyy5GVtaPolhN/GFAlkntkmkpGNnxtDqz26ycM7lFqql24YbAk//5LqLjo63xDH997QVGawJr7ZeP2lq9MdAoZDKEakuU6TNETpJurDt3rgpNm3YEoD0yLIoNgZCsN3os9VlZ+QkuG/iu/UXHKc5BJstO/mef0yXocmtwJ/scOnTIMFhrbX/88cfh8/lwzz33BO0fPZr/3RhhQCaKvv55z2LFyhEhq1LItWw5CEVL+gTaZGTkYXzBLsO+27QZFY2S4xJHkMmSqhfXAwCyHhvicCWUyBYsWIBu3brh9ttvD9puZTS6U6dOQaOi48ePj0Kl7mZlVNipcMyR68QWuuxZ/OvU6SZ06nRT0Dbl+x8y+C+qxybS5xRtDMhkSG2+cUrbtJieP6dwNMqnroDYcGGpEeVIsbxOrVHk8qdKINY2hGw3GnUum1yMDg8MRGpOppXSKYqU4Zjiz3/+8x+nSyCHMfDZh5+lNQzIZKjNT/ri6J+3A3Bu+sLBBRsBQUBOofavh8ysaS3WNlh+D2WTi9F59kgIKZyR5Hbdu3fH3r17nS7DdmorZiQnJweWm5McOHAAr776qurxWq+Nlqozs5SdlWPVRoQzMjLwv//7v6b6mTRpErKysjRrIKJgXNotPAzIZCh9UAekD+rgaA21+085Es65okV8uOqqq/DFF1/gpZdewoMPPhjYvnr1agBAu3btnCrNNHmwrKmpQWFhIbZs2YL+/fsH2qxZswZXX301hg0bFnKcWig2M33B6GZKebtLLrkEN954o+Z51fqbO3cuTp8+rduGUywSjzLUqY1+SvNzldMwFhflonevJ9G1651B7ffs/T12756v26fW+bWO2bWrEN/tex0AMHbMZiQnp2v2GWvy98DRY+s4JEYJR2sNa7V2DMfxYfhw/7J/R48eDQpun3/+OQDzS605SR4SmzXzP9Hyz3/+c1Cb6667Ligcx4pUmxSO5duMVhqZMmWKZjtKXJ063WjcCP4QKAj+sb6vv/5ZIBTu3PVMSDt5OJa2KYliPRYX5aJZs04hoVLtoRtSOAaApcsucdVo7fiC0sAXWceATAklp3B00FQMraAsbXfzes9kjVZQ82owc1vdbquHvK1/3nOmgt34glIUjPNPATx2fE1QINy3/62gdsovIDQkFy3pCwAYOWJF4Dg18lFrZZ/Hj681/T7JvTjFgjTVlZ/Gobe+QfZ07RGphpO1qJq/Htkzwn8whxOUIVntaYl6+ylyubm5KC0txaFDh9C+ffuony9Rfl3v5GoTRE7p0GECDh78LGjbzp1z0LXLHZb7atKklal2Y8d8o7p9/de3cNQ2DjAgU5CyycVAkgA0isHboL9qhLKNMlQa7XdKcmYqGk7Vau6XP1bcDfXGk1/+8pfw+XxYsGBBSLg6fpyPMg+Hz+fDsGHDcPXVVwdti9W5iZySnNzMsM3iol4ARMN2dXWh//507vyzkG1Ll11sqjbyJgZkCtUo6oZh6XVy66bo9ER+0DZ5kIxlqFSrT6IW1OWMagyE5CnFyJnLkGxkxYoVge9feeUVXH/99YarDvh8Ptx5551obGzE22+/HdX61IJcSkoKpk2bFtXz2k35Pj766CMACArHseTl6SoU/9TWU1abLyzd9Kfc16/v06ptKX4xIFPY5OFYKWfuaJRNuTCPVwqZ4bDzsdk5haNx/O+7cHptJTpNyUdyZqqpvjh6bEwtHFVVVWHhwoWabeSh6s033wxpZ1fgEkX9UaP6+nrXBzyj+m688UZs3rzZcr/z58/Hww8/HEFl/qcTPvfccxH1oSYtLQ1nz561vV9KLOE8bMRM28NHlqBd23Fh10XuxoBMYdMNvELs6rCq1XW90Oq6Xk6XEXfCDZd6xxmtgKB02223qW6fOXMmunbtijvuUJ+P6OZgLOnZs6fmkml624yC9YkTJ1SXgDPqX74vPT0dubm5lv++jDzxxBMh9Xvh74q86/jxr0y33bTpLo4ixzEGZAqb1VHVVtf14qoQFHNSoNIKx17xq1/9yrCNVni0ut3sfrlf/vKXhm3CqYOBODGJYgO+3fJI4PX6r3+K/nnzkJbWxXJf8mkT4wtKQ9ZDlmvVagiA0OkXTVM7YNSoVap9AkDr1sNw7NjqwD7yPi7zRjGTMaxT4HtOWSAiIjVr1k5E0ZI+qKpaFNh2/Pg6lKwaG/Y6w/Jl3Xbvnq+5PrDa8m2CkIRztQdRXvGnkD4zMvzLwknhOLfnI6D4wBFkCpvqTXgiXD29ghKP9Ct6n8+HJ554AmlpaUH71L4nIucMzV9k3AjBI7X98+ahf9481X1mt2nNVS4YtxOLi3KxbdtUdM7+qaLW4KXlKH4wIFNY5MufyWWOyUHLH/QIasPRYnKaFJCfffZZzf1EREQSweDubuMFA4ko2qyMyfOaJXKe2WuW16vLSKPIvXs9iWZpnbF58/2BfZxbHLdUr1cGZCL3Y0Am8hYGZA9TznNmMI57DMhEHsWATOQtDMhE3qF6vXIVCyIiIiIimYS6SU+6oUztpjG9fUREicbupxkSkbPkU0c4bcQYR5CJiIiI4pzW2s9qwl1vOp4k1AiyHo4cExERERGQACPIB1/eiOp1lbb1Vz5tpW19EREREZH7xO0Icv2hM6h8YT0AoHbfKRz7aCdSu7VA7Xcng9opH3ShNZIsb8f5ykQUjmeeeQa1tbWB13fffTeys7M127/99tv47rvvAq+15gMr5wtbfUIgnyhIFLnFRbnIzr4ZFRUfBm1XTmvQmr4QTrvFRbmaj8y2Os/42y2PoLLyE83zJ9q85bgNyJUvrEfzIR3R+qY+gW3KMAxcCLlq+5THyQOx9BQ5KyE5XoK11mcVi/clP7fXP0dKHFqh87XXXkO3bt1w++23mzpGeiKglXPp3Wx36tQpvPDCC6ZqJSJjFRUfok/vaejSJfSalgs3NEfTRf1fxEX9XwzUk2iBWCluAzKAoHAcKWUYkx6jnIjkn0XFrNVoPFPnYDVE7hbO6KxWqJUCst5IslZAViOFYyvHEJE+vXCsd/MbQ6m7xOUc5IZTtcaNLJJGjOVfBGQ9NiSi4/lZUqKINBwb7QOA5ORk08dIr3/yk59oHkNE0bG4KDfoy6gdxV5cjiAnN29ie5/8dT4Rud306dMtH9O/f/8oVEJEWjp2uAYXX/ySYTtpNFkekjnCHDtxOYKMJCtP5iWn1B8+63QJRFHFh20QkVLVwU8ttZevX8zR5NiJz4B8np2/uo/aNAARKH+qBCc+3R3W4Sc+24Pyp0og1jeGdfzBVzahYkYJavedNG5ss8rnv7Ktr0MLN6HqxfW29UdERGS3lJSMsI8tGLdddbsyNDNE2yMup1gAF26iC1r14JlRKHtyReB149l6VMxcFXSc2koTan3J94VD2dep4nKcKi433afyePn6zHrHiw0iyqeuCNl+8OVNpo6PlNYPGuF+tsrjuMoFuVF1dTWaN2/udBlE5LAxV2zSnFcsnz6xdNklaGg4AwDo3WsKdu6aq9vv4qJc9OjxIPbsUZ+6cezYanz77cM4V3sw0D41tR3695+Htm2u0Oyzd68n8d2+19GmzYjACheJIm4DMqAekOTbktJSTIcoO8NW2ZQLIS61awu0+mFPHFyw8cJ+g+Xj5CGwWV4btBjXJSjg6h1f+ezaoNcZI7ORPqC96ePdRP45dHzwMhx8ZRPEugsj6edKj6NpbisnSiMCcGFViXnz5nGaBREBgOp0CeXc4rFjNmN58RDU1R0LhGO1+cfyvvbseQnDh32J9PSeIQH86w0/Dzm2tvYwNm68XbPfxUW52LnrGQBAj+4PWnmLcUEQRVFvv+5OskYe6ITUZHSeNUK3jVpIlfZnju6Mltf0tHy8lRrNHC8fhbd6vnBqNfNglzgcRbYyqZ7XrMtYnYc8c+ZMiKKIyy67DD/60Y9M9ae3lNyOHTvwxz/+MWRfaWkp3nvvPdVjfvOb3+D48eOW6qYgZq9ZXq9EzlO9XuN6DrKbqYVjI/LgpxaOgbgJhKZovddE+gzI/eSBVmtdY7kZM2YAADZs2BC0729/+5upsC3f949//EM1HANAbm6u6jE+ny8QjomIElVcT7Fwq9QumREd36Kgi02VhGrSMR11VWei1j9RIpIHYbOPfjZav1jvOLPHaJ1H72EkRESJgAHZAR3uHxjR8S2u7G6q3ZmvDyJ9UAdLfQupoQ8aIKLIWQ2c4QZUu87DgBzfVq+5CikpLTBk8F8029TWHsKKlSOQ1+9ZdOp0g2Gfy5YPQMcOE9Gv3xzddqdPb8X6r2/BmCs26rYjchIDsgeZXXKu7kC16vbjn+zC6VUH7CyJiIg8QGtJMPmNWso2W7Y+hi1bHwtp5ydicVGvwKvyij+hvOJPqm23b5+BsvL3dc9N5BYMyAlGGa7b3XExmvVprbmfiIjig9lAqrZfa21dKRybCbll5e+HtJOWPGNIJrdhQPYgO9Ze5o1sZLeFCxdi0qRJTpdBRCoiHa2Vlv2K9Px290sULQzIHlQ2pRg5cxlwyT04X5UoPtXXn8De7xbiwIGPVPcPzV+ENWsnYnFRLnrlPo5u3e7R7Y9hmLyCAdmLuHImuczEiROxaNEip8tIWPwBheykDLEdOkzAZQP/gDVrrwlpm5GRFxgF3lX6HHaVPgeA84rJ+xiQPUR65DVg/LS7E1/sRcurulvqn/OPKVxDhgzBkCFDnC6DiGzSp89T6JJzq+n2ypv8tOYVMziTVzAge4wyJOtRBmStgH3srztQ/VVVSBs9aithlE0uBgSgzY/7mlpeTllP8/wsZAzLxrG/7URt2SnOkyYicogyHB8+ssT0sSNHrMDKklF2l0QUUwzIHmQ2xKoRUpIg1jcCMPfYZiXD84rA0T9vx9E/bzfVZ0qbZqg/WgMAqF5bieq1lYY1UHRYXQ/3r3/9K7755htTbZX7tR6Ccemll+KGG24wfEiGfPucOXNQV1cXtG327Nmor68POf+MGTMgCMFPFbXydDpOZSCvkqZBqI3sbt/hQ98+vqBtp09vR0ZG38DrTZvuUu23oeEMkpPTg7aphWO985esKsCI4UVW3g5R1DEgx5CZAGp21FTermLWKiRnpqLjw4MNj+s8e2TQcWmXtEfr63sFtdGrwe5R3azHLw98X/n8V4AIZD2m/av6cD4fMiYPfu3bt0dubi5Wr14d2KcWDG+66SbcdNNNePXVV3HggPV1tdX6vOGGGzTP+fHHH6v2U1dXF7JNHo6vueYafPrppwCAmTNnWg65S5cuBQBMnjzZ0nFEbiMPqUrKgLxm7QTVY5WWLrtE81xWzk/kNoIo6t7xxdvBXMyuZduM5jOH05+EQdUWgnGTAFuvWbMjp1bbmW0rH/H1+XzIyclBWVmZaj9mH+Gs1lavD44eUxjMXrP8fyyR81Sv16RYV0H2ySkc7coA6ta6yLpohcOpU6eaajdz5syg13fd5f8176ZNm4K2T5gQPNpFpGX58uVhPY5b/kVE8Y8BmYhirkmTJhEdr5xukZ+fH1F/4YwsExFR/OIcZJfSuhkunJFZszfjydsJyQI6zwm+0aJ86gqIDcG/EeRIcfzwYgjcs2eP6vbXX38d5eXlMa6G4pWVqUEUvyJ9GiF5CwOyC0lBVQqfFbNWo/FMXUThWDq2bHKx6pxj5bayycU48dketJzQI7BNbBBD2tg9f5mcIf1Pv1WrVvj1r3+tui/W5L/OVgYU6fW7776repza92qv5UaMGIGSkpKQ7cnJyWZLJiIH7f1uIbp34+PuyR6cYuFS8tCZ/dSwsPporK4L6UselI2cWl6mWZPaa/ImeQhVhmMvsxrsr7zyyqDjpD+nT59uY1VEFC2lpfOcLoHiCEeQXarxbD2S0iL766l4erXptmrhl0/WIzf8OvnEiRMAQmuZN8//P8N77rkn1iWRi2itPGLmtwdmt5H7cek4shsDsgu1+mEuKmauQuvre6H50E62B9XMMTk4tazMuKGCso5mvVvbVRK5QE1NDZo1a+Z0GSHmz5+vur26uhoA0KlTJ1P9mF0GjvNNvUk+8u/z+TBgwABcf/31mn+Hyr9r6YEzv/3tb/HQQw9Fv2AypBV6lY+11jtG2VZt/rDadmU/gwb9UfW4Ht3vRyBheZkAABuQSURBVM+ej2jWzvnK3sWA7EIZI7Nx/J+lqDt4BpUvfGX7VIZIwrHa3GXyNikoFBYWau7TOs5oW7QC5pQpUzB37lzdNpGcm8HY+66//noA2v8NP/roo0Gvp06dCp/Ph2PHjsWiPDKwbPllAEID5urVVwa9lvbbGUiPH18LABia/ykyMvoF9a+0Z+8C1YBsVy3kHM5BdqHK59YB8I8kZz2q/VQ5I1aCtTLomgm+B+autVwTuZNbf9WsVUPTpk11jxk0aFDINp/PhzvvvNPwnJxznBgyMzOdLoF01NefVN0+bNi/o37u9V/fAgCBcAxoPxlQDad7xAeOILtQ1uOXB1aIkNMLvA0nziG5pXpokK80oTUSLG9XMXNVSJuUdmmoP3w28Lp82kqI9Y2a9TSea9DcR+5kZaRYb3s45wjnGL2+rr32Wlx77bUh27t06WJYw9NPP23YP8WnHj16aC4dSLElfyy1EyOxvXs9abqtWo0cPfY+BmQXUguxyiXVlOFZPpqrXLWiel2l4eOfcwpHo2JGiWaAzvrfIah8YX1gv9A0GTmzQ2/k03vNVS+IyM2kee3kDvKQDAAjR6xAs2bm7jmIVNeuxr9tAoCCcdtRtKRv4PXuPb+NVkkUYwzILqMVUJWrSlgJm80vz0Lzy7M090t9Zc8codtP1qODNY8Npy4it+GT8xLbwYMHnS4hrtgx+iufY7yyZFTQNjcQBH+MKt39InJ7PoI9e15yuCKyCwOyR5zdcsTpEojiEsNwYnrnnXdw2223OV0GmTS+oBRVBz/FN988GJPzWQ33e/cuQO75m/XcFOApfAzILiONFKvdJMfRWaLouuqqqzB8+HCny6AY2Lt3b9Br6QclhmZ9RkuvNTbWYsnSPM32kYRHO8JxY2MNkpIuLGdp9oa6I0eWa+6TpoKs//qnEddH7sGA7EIMwkSxwxHkxKS1/Fv37t1D2qkda9QmHsnD5PiCUjQ21mDJ0ouCQm9SUmrQtIhwA7FWcB2a/6nqduV8ZQDo1OkG9M8LfrrekqUXBb0ec8VGLFs+0LAvM44fX4duXf/H0jHkXoIoinr7dXcSUUwIFtrymiVyntlr1jPXazjrDNsxB3nDxltx8uRGjLlik6n2x0+sx8aNt2HY0H9r3tBXsmossrKuQ88evzbsb9nyyzDmig2mzu3UihsUMdXrlSPIRERE5EqXDXzXUvtWLQdj7JjNum1GDF9quj8r4ZjiCwMyEZHLSL+yHzt2LMaOHetoLVbs2bMH777rDzSJMu2ASMLR4/jCgExERLaQwjFRIggeNbYyE468gI+aJiKiIOGO/nIdaUok4wtKZV+7nC6HbMaATEREAcuWLYvoeIZjIooHDMgJomxyMY7/c7fTZRCRBcnJyTE/55IlS2J+TnI/+dJt8qkFRsuh8eY18irOQSYi25lZOxYAGhsbMWvWLM398u1q+5cvX46ioiLTNZhtk5ycjOnTp5s+Jj8/HxMmTNA9j3Tc4sWLUVwc+iAgtX779eunum/q1Klo0qSJ7vnU3HfffejQoYOpY8z+HRr9vempra3FM888Y7lPn8+HefPmobq6Omj/iBEjcOWVVxqel8IjrQ8MBAffvH6hf4fy9sr1k4m8gAGZiGyjF4rUHsyQlJSku1+vP1EUNcOxVn+SDRs24JNPPtE8tn///pp9qlm7di3Wrl1r+P6Tk5PR0NBgut/27dur7pszZw6aNGmCqVOnmu4LAF5++WXccsst6Nu3r+ljosXqfytmji0pKUFJSQmneUSR1YDLQExeFZcBuWxyceBpdAeeWYOGk7UAtJ9QJz3Wufngjmj94z6qbarXHMCxjy9MwtfqS6xtQPlTJQCAdrdfjGZ9W+ueEwA6PjQITTo1130f8vZG7wMAMkZ1RquJPYP2C8mC6b6IrJIHP2VAOXLkCH73u99phmCjQKO2XxCEsEcvpXBsJUhJbadPnx4y9UHat3LlSowcOVKzj4aGBsvnnDFjBgRBCNleV1enW6fWvg8++ED3BxGrN9rJ223cuBF///vfDY9ZvvzCY3u1RqXDCckMxkRkl7idg1w2uRhlk4sD4VivDQCk5mSien1VUHiUtzv28S6ktG4WCLxa7cqfKkFyZiqSM1Jx+O1vNNsBQPog/686q377tWo7ZXsz70NyekV5SLuWE3qgbHIxktJSIDRJMtU3kVlPP/00APWQ0rZtW91j1UKZ1aAml5qaCgCB6Rtq58rMzLTcL6A+L1jq88svv9Q9Npz3ogzHkfbnBtLIf7j15+Xl2VgNEVGouBxBlpgZHVW2kcKmcrtRX4de/a9mf6dXHUDG8AuPvJS3afOTvhDrGlE+faVqv2q1mHkfZvtiQCa3mDBhAj777LOg0KQ2hcCMJ598Ej6fD42NjZptHn30UdP97d+/P6w6osXMqLuXyUeQ1d7nzTffHPuiiCihxHVA1mM1GBoF1XN7TmjuO/7JrqCArCSN5qrvNK6LyE3CDW75+fn47LPPgrbp3YhmxznNevPNN2N2LrkuXbqEfey5c+cwd+5cG6shIkocCRuQJWYCZk7h6KBpDHpB2WxgNdsuZ659c4TTB3e0rS+iaJCPGJoJovE8igoAPXv2NG6kEO+fCRFRLCR8QDZ7k5r8ZjmtoJzUvAmypw/T7Ufr2FiMBCc1S/i/boqBSAKacg6ymZu01Nrs3bsX77zzTth1aOnevTtuu+022/vVUlJSgnHjxplu/8orrwS+501sREThi9ub9KIlp3C0ZqhurFa/q5yIjMkD75NPPhm0TWnVqlW6fX300Ud2loYZM2YA8AfvWNJaqUJLVVUVAO8HYb254+QN0vrHi4tycfbsd6ba23fuXmhoOGNbf/rnyg36oviRsAHZ7uXNIukvktFj6bz1h86G3QeR05SBTlqFAgB+97vfhbT/4osvdPs7deqU5XPq0VtJwglS7YMGDXK2kCgxengMud/4glIH10AWsXTZJTE5k7Pvk6IpYQMyADTp2DwwZUL+pWSmjdm2yS2bhrTLnjkiovchNElC5Qtfma6RKBrMzB02Cjxqy7wdOXLEUj9WziEtTac0Z86csPpfunSp7rnDoXe+a6+9VnW7Wv1Ww6beZxApvf9WZs6cGbXzknvZGTIZWskOgiiKevt1d8aL+mM1qHxuHdrffSma9mip2ubMfw/h2J93IGNkNlr+oIdufzU7j+HIu1uQ9diQQCBWqnxuHZrmtkLrG3tHXL+k4cQ5HChci9bX9ULzodqrZpDnWBm+dPyatRJQ5a+1jtPabxSSzfZrtk4zx4wdOxZjx47VPM5sSJW3t1q/HTc3Wunb6nuS7Nq1C++//76lGsL9b8UBZq9Zx6/XaFtclIsRw4uQltbN6VKiSppewWDuSarXK+/aApDSupnhahHpl7ZH+qXtTfXXrHdrdJ6t/TQtAMh6/HLT9ZmV3LKprateEIVDK9SlpKRg2rRpIW0BoGVL9R9M9frVOo+VwDZr1izV+a7hBLBohTK192n2hwPJtGnTkJKSYnisXh9269Wrl+p5brzxRlxySWx+PU7hUZtrO3DAW2jbdkxEfamFS2l/Wlq3kLnMyvbKuvTCqtp7kLdX2z/g0tfQrt14zT4pvnAEmcj9PDWCTETxO4JsdaTU7Ajy4qJc3YCsPKdWezM1Li7KRVJSE4wbuy2wrbHxHJKSmga1UQvM4Z6TXE31ek3oOchERERkjRMh0Cg8myUdIw/HAILCsdr5GHwTD6dYEBERUcKwGnb3fvcKDhz4a5SqIbdiQCYiIiKSkY9Od+92L4bmf4olSy9ysCKKNQZkIiIiShg7dj6NPr2na+7nfGICOAeZYozrMxMRedvy4sFOlxCR/fvfcboE8gAGZCIiIjJlfEEp6uqOY3FRLoqW9NF8xHJFxYeB7SWrCvDV+ptw5szukHbbd8wMtFtclItvv30YolgfVm1VVYtQtKRP4PXKkjE4dWpLSP3SuRYX5WJlyeiQ9yBvI2+rRv4+AWD1mqtw8uR/w6qf3IXLvGk4t+cEDr2q/h+5/LHSyhHRjg8OQpPs5kH71R5DLR2n11fn2SMhpCSFHJdTODqkrdSP1ghtcsum6DQlX7UPZX/KerXOJd/f5pZ+OPrBNs12eiPHdj/2Ow5xmTcib4nbZd4kwYFRwPiCXRr7ghmtNaxspzXdQbl9ZckVqKkpNzyn5NixVfh6wy9020jn6NnzEfTofr/m0m9mz0mupXq9MiBrUAZbvUCr3GbUxkr/WmFVK1SqBV3xXAPKZ5SEhGhlSNaqy47PwWg76WJAJvKWuA/IRHGE6yBHg97osJn9VvqSZE8fZqkuoWmyqXZmzq3V3sr7JCIiInIzrmJhM7XpD1rtJFJ7s4EyqXmT8IoLE4Muucnq1asxbJjxD4lEREThYkDWoTf3tv7QGct9GU0tcOPUg/b3XIqmPVo6XQZRwOeff47PP/8cPp/P6VKIiChOMSBryHrsclTOW6cZWlPap6tuP11SEbJNPqqsNhrb9pf9ceS9LSHb3eDQq/91ZXCnxMVgTERE0cY5yBr0wrGe4//QvnNVbPDfj6HsN+2itgCAxnMNls8XTa1+aP0590RERERexxFkDWkXtTW8SU1rvrFasM4c2wXlU1dons9KX7GSMTIbx/9ZantdRsvGEalRjhyrjST7fL7Al5m2SllZWZg0aZLl/sz0r9ZmyJAhmDhxYtC2BQsW4NChQ2H1x9F1IiJ7cJk3FXpLs6ltr9l+DMf+vhOdnsi35/xTitHu1ovQrF8bW/qzw4l/7UH1+ipkT7Pn5qiKmauQMSIbLb7fzZb+4hyXeZORQqBeQJTvM7tN61xdunRBbW0t7r33XtP1aG2Tvy4sLERNTY1q8FVu27NnD3r06GHLe6KY4DJvRN7BZd6ssDKq2axva9vCMQDkzB3tqnAMAC1/0MO2cAwA2TOGMxxTVAwZMiTo9VNPPaXazmyQ3L9/v2Y41urLzOju5MmTTZ0fQFA4tnJOIiIKD6dYaDj48kZ0uG9gyPaOD17mQDVEZJZyykJSUmTjALfddpthG6vhdOXKldi4caNhf3r9MhATuYvWU//ImxiQVUjzgZVzZZNbNUWT7AyHqiIiJ3Tv3t2WfuSBduDAgfjFL36B+fPna7aTT7VgGCYiii0GZA28cYyIzDIKsHPmzDHVTq1PKzceElF4FhflJvzILz+DYJyDTEQJ6b333rOtr5dffll3f11dXdh9awVho3MSkTnS1IhIjS8o9WzA3L59htMluA4DMhElnHbt2qG0tDQofM6aNQtnzlh7QibgD7AHDx4MWaFCbYUJ5XFKe/fuNX3Dn9E5iYjMKit/3+kSXIdTLIjI9d544w2UlZUFbYtkfu4DDzwQOFZ+fHq6+hMyjahNg1DW1aFDB91l34AL853NhmSjcxIRUXi4DjKR+3EdZCJv4TrIOpSrPahNcTA7VUFreoSZ481OrdDqK5Jzyx09uhIbNv7KUj/Kz7Ch4SyWLrvYci2RfgZxQvV65QgyEREROUIroBndMGYU7PSWXLNrzrEd9Goxe9NcOH246TNwK44gE7kfR5CJvIUjyDqU4UwZ4OT7jUZvjcKflYAZ6Sip1X706jR6D2ZG3a18Dgm+hjOfpEdERETuoRbI5Nv27l2geZxWmAve7s6fQeTTIYw+A6PR3qSkpqp9XDbwDxFUSAzIRERE5Eqlu1+M6PjFRb1sqsReDQ1nAQCXD/lIs43Z0dxxY7eobm/TZqT1wiiAAZmIiIhiTi8AJsqv+lu0GOh0CaSBN+kREbnIgQMH8OqrrwZt4/JtROp27pqLffvecLqMsPFmOfdiQLZJ2eRiANYfUS0dJ7FyfNnkYj4SmyiOaAVhtTWTiRIZgyVFGwOyxzEkE8UHvQd+MBwTXSAPxxkZeRiav0i3jZslylQSL+IcZIflFI6OKOAyHBPFF7OPpU4EHDknI2rhmMgODMgexnBMRE6or693ugSKA0YPuDASD6OvRUv6Ol0CaWBAJiIi07Zv347Zs2c7XQYliNathjpdQlSJIn/YdKuEmYMszdUte3IF0Bi8cLjaSKzeTXdGN+Qpb7zTa2uVsu/ULpnocL/xMjFqNUVal9H853BvXCTyok8//RTr1q0L2mZ2esDrr78ehYqi44MPPnC6BIojao9Clo8eDxr0R0vH1tUdx/LiwbbVEi3jC0oD71PvvIuLctGq5WAMHvznmNQVy8/A7RImIAOhga3xbD0qZq6y9UY35Tmk13adQ96HVuiVazxdh4rZqwEAKe3TkPXoEABA5byvUH/kbMT1ECU6vRAs7bNy012s5yBXVFTgtddeC9rWt29f3HLLLUHbSktL8d5772nWpHaTod57Ue5rbGzErFmzNPs36tfMfnIXKSRqTafQe1KePFwatTHSJedW7C97V7M/s4961tun9bQ8o/cRK0a1JGJoTqiADAQHzKS0FP+o8uRiW0OyvB+pf6dI4Vj53rIeGxL1c3PkmOKdMohNmDAB+fn5qvv0jpNvi2W4k59rxIgRKCkpAeCfRqEkD8d2mz17dmBec35+PtauXQufz4e+ffsG1SLdtHf69GlkZGRErR6KHa0gaxTItI4bNWoVmqZ2sFRDnz5PoU+fpxwJqeMLSlFffwrLlqv/Jrhf36fRufPPYlaL00HdTRIuIEebWiiUQvKxj3eh9fWxf+xlSvv0mJ+TKN7pLcumtc2tpFqvvPJKwzbRCPJSOJb6nDBhgu45nn/+edURaLvrotgId3TSTIiOVh12jqimpGRGrdZofgbxjgE5hqrXHIhpQJZGrrMeDW8+llpfEjvnX1s5pvL5r1B/+KyptlZrUWvfefZICCkq97I2iv757Cb7pvjVty/vQreD2WCrtfSbfHoGEVGkuIoFGZKCY/u7Lw2ZW60krevc7s6LLfVtZj3ossnFgXCcUzgaTTqma9YS2JYkIKdwNNr8uI9u3dL2TtOGIqdwNJJbpAIADr2+WbWtFI5zCkcj84oc3b4p/lRXVwe+V87V9RL5aLBTI6/hnlftuMGDIx8MICICOIJMBtRWojAzr7pZ79Zh961GrG8M2d/x4cGBfuqP1SCldTPNftIHd0T64I6q881Pl1SEHNPpSf2lhdpPGoCm3VsAAFpO6IGWE3rYPped3GvevHlOl2AbeTj26k1uc+fOBQD88Ic/dLgSIooXHEGOgcYa//y67BnDY3peo9FeO/qORLs7zI0yA0D5tJW6+yufXae7X0/GiGwAwIlP9xi2lT5LKRwTxQMpJHslIEv11dTUAADOnTvnYDVEFI8SfgT58NvfAgAyhmfb0p/aCGKFbxUA/6oZdEGzPv5RZil0pl/WAW1u1p/PaTXsn/hsD04tLzPV9lRxGU4V+9uGM2+aKB7EeqWISKZ3FBYWeibUE5G3JNQIcsaI7MCvwQF/yKnZfhQA0OpHwUubyEdfT/xrT+B7M8GobHIxGk7XQqxrNNX+3J4Tge+P/8PcHaTHPt4FAKirOoO6qmrNdvL3UTa5GEf+sAWVL6w3/V6iTT73+MyGg7bVJfUjheMWBV3Q+emRpuqQH0+UqKIRjq0ufWemLwZj7xlfUMrVEsj1EmpI83RJRdC6xxKt0cKsx4agct5XOLWsDKeWBY8s6t2kVja5GAdmrwlsa9KpOTo+NCikrVofp0sqAnNilbWptRdrG1A1/2vd9yKfM3x2yxHVup1m9sEqZqZ21FX4f2DQXIHCRB0HZq9Bw+naiOqg+OXkTW12Cuc9PP7443juuecMHxRi5TPSan/jjTfio48+slwjEVGkEiogS8yGm5S2aZptldvN3GgWbh3htrfr2Fiy48EqVS/5f2CwGo7lOk0bqlpHSts0PoGQgng5LKvVbvRe0tPTQ4679tprDfs2GvHVam8mIHv18ydykt5jpVevuQotWw5GXr9nYlyVeyRkQCYLBABi6GY7ph+ItQ0QUpNNtbXyxEMhWYDYEFq0Vs3VayvRPD/LVB1Zjw3RrePc3pO8gS8BzZkzB1OnTnW6jLDYucyamTZWp1kw/BKgH+bsPo/E7vMtLspFcnI6xo4JXT7Ubaqrd6G6ehcDMpGWnLmhU1IAIHN0Z5wqLg9pr7smMYJHssufKtE+r85UEaMHgHSeM8rSHOJjf9uJY3/baUsd7ScNMHVO8j75iGddXZ2t82vJPH62ZCcpFCf6I5c5R5wBmUxQzg+WXre8pqdm23D6NXN8TuFoNJ6pR8Us/8ogrX6Uq7oCSU7haBz76w5Uf1Vl2K9aHdlPDUNSehPTx6R2bYEO9zEcJxqfz4eFCxeisrLS6VISDoNxYmFgo1gTRFHl9+cX6O70ErWHUhB5hGChbdxcs0RK77zzDvbu3Rt47eKQbPaa5fVqo8bGOiwvHmQ4hWHdV9dDFBuRf/knmm3smtKxYuUIjBxRDEFI1p1isWRpP4wbuy3i82lZWXIFamou/NZ3fEGp6ntUjpzrfQZqo+xq7c30efDQF9i8+b6gbUOH/gsZzfuEfd7q6l1YveaqwOu2bcdi4IA3Q9pB43pNmIBM5GEMyETewoBsE6M5wXpTIYzCn1afUlu986n1Ld+27qsbcPLkppDjlQHZSk3hWrN2Ak6f3o6LL3oJHTteE3RevfdvZX99/QksWz5I8zNXTl1Ra5fd6cfIyys0bKfcrvZ3VbSkN0SxMaSdxntSvV45xYKIiIhcyeycYKMwbBT47Hby5KawalpclGv7DYmnT28HgEA4ls4Z6TxreY0pKS1Dal61+nsh7aTzKt+j8litdgAwbOgXmnVIlOFY3qdZCfWgECIiIkpMbpzHrBbiYnEeOxiFzTNn9th+TsC/BF0sbqLkCDIRERGRTfaXvWu6rVdXy5CP8Mq3qTHzHs1+DuMLSlFffxLLll+m+VuBdV9db6lPLQzIRERERDbp0P4q7Ngxy1RbN45qm2V2aojRe1y2fKBqO62Am5LSImjqjfK83bvdi/9uvjfiz5ZTLIiIiCjubfrvPTE5T9Om5h48BURvGoLS0aMrotp/JGG0vv5UxOeVh+n27a8Muz85BmQiIiKKay1bDsLhw/+JuJ9xY7eGbNMa6ZRvV2vTs+fDgRvZzPQXiQ0bb7XtHHu/W2iqnRRely67JGTftu3TdI+14zOItA8u80bkflzmjchbuMybDcws4WZ22TUAKCv7A7bvmKnaj9nzqbUzs6ZwVtZ1qKz8e8gybzU15VhZcoXu+eygrLl/3jxs2fqY6fcvr2nvdwtRWjovZH+zZtkYOSL0CbNq/XbufAv69Z1t6txmarSyBKDZZd4YkIncjwGZyFsYkIm8Q/V65RQLIiIiIiIZBmQiIiIiIhkGZCIiIiIiGQZkIiIiIiIZo5v0iIiIiIgSCkeQiYiIiIhkGJCJiIiIiGQYkImIiIiIZBiQiYiIiIhkGJCJiIiIiGQYkImIiIiIZP4/cMRYFAQx64EAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x720 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from wordcloud import WordCloud, STOPWORDS\n",
    "import matplotlib.colors as mcolors\n",
    "\n",
    "cols = [color for name, color in mcolors.TABLEAU_COLORS.items()]\n",
    "\n",
    "cloud = WordCloud(stopwords=STOPWORDS,\n",
    "                  background_color='white',\n",
    "                  width=2500,\n",
    "                  height=1800,\n",
    "                  max_words=10,\n",
    "                  colormap='tab10',\n",
    "                  color_func=lambda *args, **kwargs: cols[i],\n",
    "                  prefer_horizontal=1.0)\n",
    "\n",
    "topics = lda_model.show_topics(formatted=False)\n",
    "\n",
    "fig, axes = plt.subplots(3, 3, figsize=(10,10), sharex=True, sharey=True)\n",
    "\n",
    "for i, ax in enumerate(axes.flatten()):\n",
    "    fig.add_subplot(ax)\n",
    "    topic_words = dict(topics[i][1])\n",
    "    cloud.generate_from_frequencies(topic_words, max_font_size=300)\n",
    "    plt.gca().imshow(cloud)\n",
    "    plt.gca().set_title('Topic ' + str(i), fontdict=dict(size=16))\n",
    "    plt.gca().axis('off')\n",
    "\n",
    "\n",
    "plt.subplots_adjust(wspace=0, hspace=0)\n",
    "plt.axis('off')\n",
    "plt.margins(x=0, y=0)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>4. Distribution of documents based on topics</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>doc_id</th>\n",
       "      <th>dominant_topic</th>\n",
       "      <th>weight</th>\n",
       "      <th>topic_keywords</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.912894</td>\n",
       "      <td>patient, infection, study, respiratory, pneumo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>0.762738</td>\n",
       "      <td>cell, expression, study, il, response, effect,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>0.43741</td>\n",
       "      <td>protein, cov, virus, cell, bind, sars, structu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>11</td>\n",
       "      <td>0.425117</td>\n",
       "      <td>lung, cell, patient, pulmonary, tissue, respir...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>12</td>\n",
       "      <td>0.381466</td>\n",
       "      <td>cell, expression, study, il, response, effect,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72214</td>\n",
       "      <td>72214</td>\n",
       "      <td>16</td>\n",
       "      <td>0.397957</td>\n",
       "      <td>sample, virus, pcr, test, study, positive, det...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72215</td>\n",
       "      <td>72215</td>\n",
       "      <td>4</td>\n",
       "      <td>0.534901</td>\n",
       "      <td>patient, ct, case, study, image, disease, resu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72216</td>\n",
       "      <td>72216</td>\n",
       "      <td>7</td>\n",
       "      <td>0.669211</td>\n",
       "      <td>patient, covid, study, health, case, care, dis...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72217</td>\n",
       "      <td>72217</td>\n",
       "      <td>5</td>\n",
       "      <td>0.998787</td>\n",
       "      <td>la, en, el, le, que, los, se, et, de, une</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72218</td>\n",
       "      <td>72218</td>\n",
       "      <td>8</td>\n",
       "      <td>0.514748</td>\n",
       "      <td>rna, sequence, gene, protein, genome, virus, n...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>72219 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      doc_id dominant_topic    weight  \\\n",
       "0          0              0  0.912894   \n",
       "1          1             12  0.762738   \n",
       "2          2             10   0.43741   \n",
       "3          3             11  0.425117   \n",
       "4          4             12  0.381466   \n",
       "...      ...            ...       ...   \n",
       "72214  72214             16  0.397957   \n",
       "72215  72215              4  0.534901   \n",
       "72216  72216              7  0.669211   \n",
       "72217  72217              5  0.998787   \n",
       "72218  72218              8  0.514748   \n",
       "\n",
       "                                          topic_keywords  \n",
       "0      patient, infection, study, respiratory, pneumo...  \n",
       "1      cell, expression, study, il, response, effect,...  \n",
       "2      protein, cov, virus, cell, bind, sars, structu...  \n",
       "3      lung, cell, patient, pulmonary, tissue, respir...  \n",
       "4      cell, expression, study, il, response, effect,...  \n",
       "...                                                  ...  \n",
       "72214  sample, virus, pcr, test, study, positive, det...  \n",
       "72215  patient, ct, case, study, image, disease, resu...  \n",
       "72216  patient, covid, study, health, case, care, dis...  \n",
       "72217          la, en, el, le, que, los, se, et, de, une  \n",
       "72218  rna, sequence, gene, protein, genome, virus, n...  \n",
       "\n",
       "[72219 rows x 4 columns]"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>dominant_topic</th>\n",
       "      <th>topic_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>145</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>6529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>3068</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>1269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>6514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>14971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>3319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "      <td>2263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>3868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>11</td>\n",
       "      <td>11</td>\n",
       "      <td>1066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>3770</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>13</td>\n",
       "      <td>13</td>\n",
       "      <td>1765</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>14</td>\n",
       "      <td>14</td>\n",
       "      <td>5719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>15</td>\n",
       "      <td>15</td>\n",
       "      <td>1245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>16</td>\n",
       "      <td>16</td>\n",
       "      <td>4758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17</td>\n",
       "      <td>17</td>\n",
       "      <td>2524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>18</td>\n",
       "      <td>18</td>\n",
       "      <td>2063</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>19</td>\n",
       "      <td>19</td>\n",
       "      <td>808</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    dominant_topic  topic_count\n",
       "0                0         3739\n",
       "1                1          145\n",
       "2                2         6529\n",
       "3                3         2816\n",
       "4                4         3068\n",
       "5                5         1269\n",
       "6                6         6514\n",
       "7                7        14971\n",
       "8                8         3319\n",
       "9                9         2263\n",
       "10              10         3868\n",
       "11              11         1066\n",
       "12              12         3770\n",
       "13              13         1765\n",
       "14              14         5719\n",
       "15              15         1245\n",
       "16              16         4758\n",
       "17              17         2524\n",
       "18              18         2063\n",
       "19              19          808"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topic_df = df.groupby('dominant_topic').size().to_frame(name='topic_count').reset_index()\n",
    "topic_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,8))\n",
    "plt.bar(topic_df['dominant_topic'], topic_df['topic_count'])\n",
    "plt.xlabel('Topic ID')\n",
    "plt.ylabel('Num Documents')\n",
    "plt.xticks(np.arange(topic_df['dominant_topic'].min(),topic_df['dominant_topic'].max(), 1))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>5. TSNE Visualization of Document Clusters</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.manifold import TSNE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "topic_perc_df = pd.DataFrame(index=range(50000),columns=range(20))\n",
    "for doc_id, topic_vector in enumerate(lda_model[corpus]):\n",
    "    #print(topic_vector)\n",
    "    perc_df_row = [p for i, p in topic_vector]\n",
    "    #print(perc_df_row)\n",
    "    topic_perc_df.loc[doc_id] = sorted(perc_df_row)[:20]\n",
    "\n",
    "print('Done Loading Data')\n",
    "topic_perc_df.dropna(inplace=True)\n",
    "topic_perc_df = topic_perc_df.astype('float64')\n",
    "best_topic_ids = topic_perc_df.idxmax(axis=1).tolist()\n",
    "#print(best_topic_ids)\n",
    "tsne_model = TSNE(n_components=2, random_state=42, init='pca').fit_transform(topic_perc_df.values)\n",
    "print(\"Done TSNE\")\n",
    "\n",
    "plt.figure(figsize=(12,8))\n",
    "plt.scatter(tsne_model[:,0], tsne_model[:,1],c=best_topic_ids)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}