text_mining.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import necessary packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:08.125980Z",
     "start_time": "2018-11-12T15:29:05.480018Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "if (!(\"Notification\" in window)) {\n",
       "    alert(\"This browser does not support desktop notifications, so the %%notify magic will not work.\");\n",
       "} else if (Notification.permission !== 'granted' && Notification.permission !== 'denied') {\n",
       "    Notification.requestPermission(function (permission) {\n",
       "        if(!('permission' in Notification)) {\n",
       "            Notification.permission = permission;\n",
       "        }\n",
       "    })\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import re\n",
    "import string\n",
    "import numpy as np\n",
    "from sentiment_module import sentiment\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import nltk\n",
    "from nltk.corpus import gutenberg\n",
    "from pprint import pprint\n",
    "import seaborn as sns\n",
    "%load_ext jupyternotify"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:08.131645Z",
     "start_time": "2018-11-12T15:29:08.128480Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# df[df['review_id']=='cc0pfhAB-JIeBVeKJ1BM-A']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:08.410725Z",
     "start_time": "2018-11-12T15:29:08.134387Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# read pizza data\n",
    "\n",
    "df=pd.read_csv('top_pizza_reviews.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:08.416901Z",
     "start_time": "2018-11-12T15:29:08.413000Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(20877, 17)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:08.423901Z",
     "start_time": "2018-11-12T15:29:08.418970Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['business_id', 'name', 'biz_stars', 'state', 'city', 'neighborhood',\n",
       "       'address', 'postal_code', 'review_id', 'text', 'date', 'cool', 'funny',\n",
       "       'useful', 'stars', 'user_id', 'type'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:08.445933Z",
     "start_time": "2018-11-12T15:29:08.426303Z"
    }
   },
   "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>text</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>name</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Grimaldi's Pizzeria</th>\n",
       "      <td>5190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pizza Hut</th>\n",
       "      <td>4184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Domino's Pizza</th>\n",
       "      <td>3232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mellow Mushroom</th>\n",
       "      <td>2753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Blaze Fast-Fire'd Pizza</th>\n",
       "      <td>2214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Papa John's Pizza</th>\n",
       "      <td>2070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Papa Murphy's</th>\n",
       "      <td>998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Little Caesars</th>\n",
       "      <td>236</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         text\n",
       "name                         \n",
       "Grimaldi's Pizzeria      5190\n",
       "Pizza Hut                4184\n",
       "Domino's Pizza           3232\n",
       "Mellow Mushroom          2753\n",
       "Blaze Fast-Fire'd Pizza  2214\n",
       "Papa John's Pizza        2070\n",
       "Papa Murphy's             998\n",
       "Little Caesars            236"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# most popular restaurant (by count of review per restaurant name)\n",
    "df.groupby(['name'])[['text']].count()\\\n",
    "    .sort_values('text',ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:08.462845Z",
     "start_time": "2018-11-12T15:29:08.448067Z"
    }
   },
   "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>text</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>type</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Delivery</th>\n",
       "      <td>9722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dine-in</th>\n",
       "      <td>7943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fast Casual</th>\n",
       "      <td>2214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Take and Bake</th>\n",
       "      <td>998</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               text\n",
       "type               \n",
       "Delivery       9722\n",
       "Dine-in        7943\n",
       "Fast Casual    2214\n",
       "Take and Bake   998"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# most popular restaurant (by count of review per restaurant name)\n",
    "df.groupby(['type'])[['text']].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:08.489393Z",
     "start_time": "2018-11-12T15:29:08.465336Z"
    }
   },
   "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>address</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>name</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Pizza Hut</th>\n",
       "      <td>318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Domino's Pizza</th>\n",
       "      <td>233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Papa John's Pizza</th>\n",
       "      <td>149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Papa Murphy's</th>\n",
       "      <td>67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Little Caesars</th>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Blaze Fast-Fire'd Pizza</th>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Grimaldi's Pizzeria</th>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mellow Mushroom</th>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         address\n",
       "name                            \n",
       "Pizza Hut                    318\n",
       "Domino's Pizza               233\n",
       "Papa John's Pizza            149\n",
       "Papa Murphy's                 67\n",
       "Little Caesars                34\n",
       "Blaze Fast-Fire'd Pizza       13\n",
       "Grimaldi's Pizzeria           11\n",
       "Mellow Mushroom                9"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# review per business\n",
    "df.groupby(['name'])[['address']].nunique().sort_values('address',ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Remove punctuation line"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:08.494685Z",
     "start_time": "2018-11-12T15:29:08.491654Z"
    }
   },
   "outputs": [],
   "source": [
    "count = lambda l1,l2: sum([1 for x in l1 if x in l2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:09.016634Z",
     "start_time": "2018-11-12T15:29:08.496919Z"
    }
   },
   "outputs": [],
   "source": [
    "df['puncount']=df['text'].apply(lambda x: count(x,set(string.punctuation))/len(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:09.036151Z",
     "start_time": "2018-11-12T15:29:09.019288Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4734     ********** $9.99 For A 14\" Circle of Crap ****...\n",
       "4389     delivery has taken a half an hou already! not ...\n",
       "5545     fast and tasty...............delivery driver B...\n",
       "19541    STAFF+TIME------------------------------------...\n",
       "6447     Have to say they're fast ...\\nThe delivery guy...\n",
       "Name: text, dtype: object"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sort_values(by='puncount',ascending=False)['text'].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:09.222626Z",
     "start_time": "2018-11-12T15:29:09.038905Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "count    20877.000000\n",
      "mean         0.031995\n",
      "std          0.014593\n",
      "min          0.000000\n",
      "25%          0.022915\n",
      "50%          0.029740\n",
      "75%          0.038128\n",
      "max          0.243590\n",
      "Name: puncount, dtype: float64\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a1d6e4eb8>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAD8CAYAAACcjGjIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAFF5JREFUeJzt3H+s3fV93/Hne/YgJFuCCc0Vsmmvo3rdTGiV9A5Qq3V3cQWGdDHSYHLEFkMsWWrpj62WVrNMQkqCRLRlLMmaRFbwYioUoKybrUFLPcJZNakQfhYHKPWt8eDGLDSxob2hJbvpe3+cz01P7udc3+vzPb7nXn+fD+nofr+f7+fzPZ/3+Vp+3e+PcyMzkSSp198a9QQkSSuP4SBJqhgOkqSK4SBJqhgOkqSK4SBJqhgOkqSK4SBJqhgOkqTK2lFPYFAXXnhhjo+PDzT2u9/9Lu94xzuGO6FVoK11Q3trb2vd0N7aF6v7ySef/HZm/shi+1m14TA+Ps4TTzwx0NhOp8Pk5ORwJ7QKtLVuaG/tba0b2lv7YnVHxP9Zyn68rCRJqhgOkqSK4SBJqhgOkqSK4SBJqhgOkqSK4SBJqhgOkqSK4SBJqqzab0ivRuN7HhjZex+7/UMje29Jq49nDpKkiuEgSaoYDpKkiuEgSaoYDpKkiuEgSaoYDpKkyqLhEBH7IuK1iPhGT9u/j4g/johnI+K/RcT5PdtuiYipiHgxIq7qad9a2qYiYk9P+8aIeCwijkTEvRFxzjALlCSdvqWcOXwF2Dqv7RDwvsz8SeBPgFsAImIzsB24pIz5QkSsiYg1wG8CVwObgY+UvgCfBu7IzE3ASWBno4okSY0tGg6Z+QfAiXltv5+Zs2X1UWBDWd4G3JOZb2XmS8AUcFl5TWXm0cz8HnAPsC0iAvggcH8Zvx+4tmFNkqSGhvHnMz4G3FuW19MNiznTpQ3glXntlwPvBl7vCZre/pWI2AXsAhgbG6PT6Qw04ZmZmYHHNrH70tnFO50hnU5nZHWvBG2tva11Q3trH1bdjcIhIj4OzAJ3zzX16Zb0P0PJU/TvKzP3AnsBJiYmcnJy8nSm+wOdTodBxzZx4yj/ttINkyOreyVoa+1trRvaW/uw6h44HCJiB/ALwJbMnPsPfRq4uKfbBuB4We7X/m3g/IhYW84eevtLkkZkoEdZI2Ir8BvAhzPzzZ5NB4HtEXFuRGwENgFfBx4HNpUnk86he9P6YAmVR4DryvgdwIHBSpEkDctSHmX9KvCHwE9ExHRE7AT+M/B3gUMR8UxEfAkgM58D7gOeB34PuDkzv1/OCn4ZeAh4Abiv9IVuyPx6REzRvQdx51ArlCSdtkUvK2XmR/o0L/gfeGbeBtzWp/1B4ME+7UfpPs0kSVoh/Ia0JKliOEiSKoaDJKliOEiSKoaDJKliOEiSKoaDJKliOEiSKoaDJKliOEiSKoaDJKliOEiSKoaDJKliOEiSKoaDJKliOEiSKoaDJKliOEiSKoaDJKliOEiSKoaDJKliOEiSKoaDJKmyaDhExL6IeC0ivtHTdkFEHIqII+XnutIeEfG5iJiKiGcj4gM9Y3aU/kciYkdP+09HxOEy5nMREcMuUpJ0epZy5vAVYOu8tj3Aw5m5CXi4rANcDWwqr13AF6EbJsCtwOXAZcCtc4FS+uzqGTf/vSRJy2zRcMjMPwBOzGveBuwvy/uBa3va78quR4HzI+Ii4CrgUGaeyMyTwCFga9n2zsz8w8xM4K6efUmSRmTQew5jmfkqQPn5ntK+Hnilp990aTtV+3SfdknSCK0d8v763S/IAdr77zxiF91LUIyNjdHpdAaYIszMzAw8tondl84u+3vO6XQ6I6t7JWhr7W2tG9pb+7DqHjQcvhURF2Xmq+XS0GulfRq4uKffBuB4aZ+c194p7Rv69O8rM/cCewEmJiZycnJyoa6n1Ol0GHRsEzfueWDZ33POsRsmR1b3StDW2ttaN7S39mHVPehlpYPA3BNHO4ADPe0fLU8tXQG8US47PQRcGRHryo3oK4GHyra/iIgrylNKH+3ZlyRpRBY9c4iIr9L9rf/CiJim+9TR7cB9EbETeBm4vnR/ELgGmALeBG4CyMwTEfFJ4PHS7xOZOXeT+xfpPhF1HvC75SVJGqFFwyEzP7LApi19+iZw8wL72Qfs69P+BPC+xeYhSVo+fkNaklQxHCRJFcNBklQxHCRJFcNBklQxHCRJFcNBklQxHCRJFcNBklQxHCRJFcNBklQxHCRJFcNBklQxHCRJFcNBklQxHCRJFcNBklQxHCRJFcNBklQxHCRJFcNBklQxHCRJFcNBklQxHCRJlUbhEBH/OiKei4hvRMRXI+JtEbExIh6LiCMRcW9EnFP6nlvWp8r28Z793FLaX4yIq5qVJElqauBwiIj1wK8CE5n5PmANsB34NHBHZm4CTgI7y5CdwMnM/HHgjtKPiNhcxl0CbAW+EBFrBp2XJKm5ppeV1gLnRcRa4O3Aq8AHgfvL9v3AtWV5W1mnbN8SEVHa78nMtzLzJWAKuKzhvCRJDawddGBmfjMi/gPwMvCXwO8DTwKvZ+Zs6TYNrC/L64FXytjZiHgDeHdpf7Rn171jfkhE7AJ2AYyNjdHpdAaa+8zMzMBjm9h96ezinc6QTqczsrpXgrbW3ta6ob21D6vugcMhItbR/a1/I/A68NvA1X265tyQBbYt1F43Zu4F9gJMTEzk5OTk6U266HQ6DDq2iRv3PLDs7znn2A2TI6t7JWhr7W2tG9pb+7DqbnJZ6eeBlzLzzzLz/wG/A/wMcH65zASwAThelqeBiwHK9ncBJ3rb+4yRJI1Ak3B4GbgiIt5e7h1sAZ4HHgGuK312AAfK8sGyTtn+tczM0r69PM20EdgEfL3BvCRJDTW55/BYRNwPPAXMAk/TveTzAHBPRHyqtN1ZhtwJ/FZETNE9Y9he9vNcRNxHN1hmgZsz8/uDzkuS1NzA4QCQmbcCt85rPkqfp40y86+A6xfYz23AbU3mIkkaHr8hLUmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqrG0yOCLOB74MvA9I4GPAi8C9wDhwDPjnmXkyIgL4LHAN8CZwY2Y+VfazA/h3Zbefysz9Tea1mMPffIMb9zxwJt9Ckla1pmcOnwV+LzP/PvBTwAvAHuDhzNwEPFzWAa4GNpXXLuCLABFxAXArcDlwGXBrRKxrOC9JUgMDh0NEvBP4OeBOgMz8Xma+DmwD5n7z3w9cW5a3AXdl16PA+RFxEXAVcCgzT2TmSeAQsHXQeUmSmmty5vBe4M+A/xIRT0fElyPiHcBYZr4KUH6+p/RfD7zSM366tC3ULkkakSb3HNYCHwB+JTMfi4jP8jeXkPqJPm15ivZ6BxG76F6SYmxsjE6nc1oTnjN2Huy+dHagsatVp9NhZmZm4M9stWtr7W2tG9pb+7DqbhIO08B0Zj5W1u+nGw7fioiLMvPVctnotZ7+F/eM3wAcL+2T89o7/d4wM/cCewEmJiZycnKyX7dFff7uA3zmcKN78avOsRsm6XQ6DPqZrXZtrb2tdUN7ax9W3QNfVsrM/wu8EhE/UZq2AM8DB4EdpW0HcKAsHwQ+Gl1XAG+Uy04PAVdGxLpyI/rK0iZJGpGmvz7/CnB3RJwDHAVuohs490XETuBl4PrS90G6j7FO0X2U9SaAzDwREZ8EHi/9PpGZJxrOS5LUQKNwyMxngIk+m7b06ZvAzQvsZx+wr8lcJEnD4zekJUkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVDEcJEkVw0GSVGkcDhGxJiKejoj/UdY3RsRjEXEkIu6NiHNK+7llfapsH+/Zxy2l/cWIuKrpnCRJzQzjzOHXgBd61j8N3JGZm4CTwM7SvhM4mZk/DtxR+hERm4HtwCXAVuALEbFmCPOSJA2oUThExAbgQ8CXy3oAHwTuL132A9eW5W1lnbJ9S+m/DbgnM9/KzJeAKeCyJvOSJDXT9MzhPwH/Bvjrsv5u4PXMnC3r08D6srweeAWgbH+j9P9Be58xkqQRWDvowIj4BeC1zHwyIibnmvt0zUW2nWrM/PfcBewCGBsbo9PpnM6Uf2DsPNh96eziHc8inU6HmZmZgT+z1a6ttbe1bmhv7cOqe+BwAH4W+HBEXAO8DXgn3TOJ8yNibTk72AAcL/2ngYuB6YhYC7wLONHTPqd3zA/JzL3AXoCJiYmcnJwcaOKfv/sAnzncpPTV59gNk3Q6HQb9zFa7ttbe1rqhvbUPq+6BLytl5i2ZuSEzx+neUP5aZt4APAJcV7rtAA6U5YNlnbL9a5mZpX17eZppI7AJ+Pqg85IkNXcmfn3+DeCeiPgU8DRwZ2m/E/itiJiie8awHSAzn4uI+4DngVng5sz8/hmYlyRpiYYSDpnZATpl+Sh9njbKzL8Crl9g/G3AbcOYiySpOb8hLUmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpIrhIEmqGA6SpMrA4RARF0fEIxHxQkQ8FxG/VtoviIhDEXGk/FxX2iMiPhcRUxHxbER8oGdfO0r/IxGxo3lZkqQmmpw5zAK7M/MfAFcAN0fEZmAP8HBmbgIeLusAVwObymsX8EXohglwK3A5cBlw61ygSJJGY+BwyMxXM/OpsvwXwAvAemAbsL902w9cW5a3AXdl16PA+RFxEXAVcCgzT2TmSeAQsHXQeUmSmls7jJ1ExDjwfuAxYCwzX4VugETEe0q39cArPcOmS9tC7f3eZxfdsw7GxsbodDoDzXfsPNh96exAY1erTqfDzMzMwJ/ZatfW2ttaN7S39mHV3TgcIuLvAP8V+FeZ+ecRsWDXPm15iva6MXMvsBdgYmIiJycnT3u+AJ+/+wCfOTyUXFw1jt0wSafTYdDPbLVra+1trRvaW/uw6m70tFJE/G26wXB3Zv5Oaf5WuVxE+flaaZ8GLu4ZvgE4fop2SdKINHlaKYA7gRcy8z/2bDoIzD1xtAM40NP+0fLU0hXAG+Xy00PAlRGxrtyIvrK0SZJGpMm1lZ8F/iVwOCKeKW3/FrgduC8idgIvA9eXbQ8C1wBTwJvATQCZeSIiPgk8Xvp9IjNPNJiXJKmhgcMhM/83/e8XAGzp0z+BmxfY1z5g36BzkSQNl9+QliRV2vXITouN73mA3ZfOcuOeB5b1fY/d/qFlfT9Jw+GZgySpYjhIkiqGgySpYjhIkiqGgySpYjhIkiqGgySpYjhIkiqGgySpYjhIkiqGgySpYjhIkiqGgySpYjhIkiqGgySpYjhIkiqGgySpYjhIkiqGgySpYjhIkiqGgySpsnbUE5gTEVuBzwJrgC9n5u0jnpKGYHzPAyN772O3f2hk7y2tdivizCEi1gC/CVwNbAY+EhGbRzsrSWqvFREOwGXAVGYezczvAfcA20Y8J0lqrZVyWWk98ErP+jRw+YjmorNE7yWt3ZfOcuMIL3EtFy+laVhWSjhEn7asOkXsAnaV1ZmIeHHA97sQ+PaAY1etX21p3dCe2uPTVVMr6l5AW2tfrO4fW8pOVko4TAMX96xvAI7P75SZe4G9Td8sIp7IzImm+1lt2lo3tLf2ttYN7a19WHWvlHsOjwObImJjRJwDbAcOjnhOktRaK+LMITNnI+KXgYfoPsq6LzOfG/G0JKm1VkQ4AGTmg8CDy/R2jS9NrVJtrRvaW3tb64b21j6UuiOzuu8rSWq5lXLPQZK0gpx14RARWyPixYiYiog9fbafGxH3lu2PRcR4z7ZbSvuLEXHVcs67qUHrjojxiPjLiHimvL603HNvYgl1/1xEPBURsxFx3bxtOyLiSHntWL5ZD0fD2r/fc8xX1cMfS6j71yPi+Yh4NiIejogf69l2th/zU9V+esc8M8+aF92b2X8KvBc4B/gjYPO8Pr8EfKksbwfuLcubS/9zgY1lP2tGXdMy1D0OfGPUNZzBuseBnwTuAq7rab8AOFp+rivL60Zd03LUXrbNjLqGM1j3PwHeXpZ/seffehuOed/aBznmZ9uZw1L+DMc2YH9Zvh/YEhFR2u/JzLcy8yVgquxvNWhS92q2aN2ZeSwznwX+et7Yq4BDmXkiM08Ch4CtyzHpIWlS+2q2lLofycw3y+qjdL83Be045gvVftrOtnDo92c41i/UJzNngTeAdy9x7ErVpG6AjRHxdET8r4j4R2d6skPU5Jit5uMNzef/toh4IiIejYhrhzu1M+p0694J/O6AY1eaJrXDaR7zFfMo65As5c9wLNRnSX/CY4VqUverwI9m5nci4qeB/x4Rl2Tmnw97kmdAk2O2mo83NJ//j2bm8Yh4L/C1iDicmX86pLmdSUuuOyL+BTAB/OPTHbtCNakdTvOYn21nDkv5Mxw/6BMRa4F3ASeWOHalGrjuchntOwCZ+STda5p/74zPeDiaHLPVfLyh4fwz83j5eRToAO8f5uTOoCXVHRE/D3wc+HBmvnU6Y1ewJrWf/jEf9U2WId+wWUv3JtNG/uaGzSXz+tzMD9+Yva8sX8IP35A+yuq5Id2k7h+Zq5Puja5vAheMuqZh1d3T9yvUN6Rfontjcl1ZXhV1D6H2dcC5ZflC4Ajzbmyu1NcS/62/n+4vOZvmtZ/1x/wUtZ/2MR95wWfgA7wG+JPyAX28tH2CbooCvA34bbo3nL8OvLdn7MfLuBeBq0ddy3LUDfwz4LnyD+0p4J+OupYh1/0P6f7G9V3gO8BzPWM/Vj6PKeCmUdeyXLUDPwMcLsf8MLBz1LUMue7/CXwLeKa8DrbomPetfZBj7jekJUmVs+2egyRpCAwHSVLFcJAkVQwHSVLFcJAkVQwHSVLFcJAkVQwHSVLl/wNrErqrV3s7YwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(df['puncount'].describe())\n",
    "df['puncount'].hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:09.228129Z",
     "start_time": "2018-11-12T15:29:09.225107Z"
    },
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# # drop\n",
    "# df = df[df['puncount'] < 0.1]\n",
    "# print(df['puncount'].describe())\n",
    "# df['puncount'].hist()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Language detect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:09.251633Z",
     "start_time": "2018-11-12T15:29:09.230920Z"
    }
   },
   "outputs": [],
   "source": [
    "from langdetect import detect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:29:09.257926Z",
     "start_time": "2018-11-12T15:29:09.254233Z"
    }
   },
   "outputs": [],
   "source": [
    "def detectexception(text):\n",
    "    try:\n",
    "        return detect(text)\n",
    "    except:\n",
    "        return 'NaL' # return not a language"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:31:32.754407Z",
     "start_time": "2018-11-12T15:29:09.260340Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 2min 17s, sys: 5.41 s, total: 2min 23s\n",
      "Wall time: 2min 23s\n"
     ]
    }
   ],
   "source": [
    "# detect language (run for a while)\n",
    "df['lan']=df['text'].apply(lambda x:detectexception(x))\n",
    "#df=df.rename(index=str, columns={\"lan1\": \"lan\"})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:31:32.799817Z",
     "start_time": "2018-11-12T15:31:32.757859Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 30.4 ms, sys: 3.77 ms, total: 34.1 ms\n",
      "Wall time: 31.7 ms\n"
     ]
    },
    {
     "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>text</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lan</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>en</th>\n",
       "      <td>20833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>it</th>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fr</th>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>de</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>es</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>af</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>et</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>da</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pl</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sk</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>so</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      text\n",
       "lan       \n",
       "en   20833\n",
       "it      16\n",
       "fr      14\n",
       "de       3\n",
       "es       3\n",
       "af       2\n",
       "et       2\n",
       "da       1\n",
       "pl       1\n",
       "sk       1\n",
       "so       1"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get language count\n",
    "df.groupby(['lan'])[['text']].count().sort_values('text',ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:31:32.828175Z",
     "start_time": "2018-11-12T15:31:32.802116Z"
    },
    "scrolled": true
   },
   "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>business_id</th>\n",
       "      <th>name</th>\n",
       "      <th>biz_stars</th>\n",
       "      <th>state</th>\n",
       "      <th>city</th>\n",
       "      <th>neighborhood</th>\n",
       "      <th>address</th>\n",
       "      <th>postal_code</th>\n",
       "      <th>review_id</th>\n",
       "      <th>text</th>\n",
       "      <th>date</th>\n",
       "      <th>cool</th>\n",
       "      <th>funny</th>\n",
       "      <th>useful</th>\n",
       "      <th>stars</th>\n",
       "      <th>user_id</th>\n",
       "      <th>type</th>\n",
       "      <th>puncount</th>\n",
       "      <th>lan</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>vQXhuCQ4Wc3WWesp9hy3VQ</td>\n",
       "      <td>Pizza Hut</td>\n",
       "      <td>1.5</td>\n",
       "      <td>ON</td>\n",
       "      <td>Toronto</td>\n",
       "      <td>St. Lawrence</td>\n",
       "      <td>77 Lower Jarvis</td>\n",
       "      <td>M5E 1R8</td>\n",
       "      <td>gJ32Kq6Wwq6Pr9_dwj_C_w</td>\n",
       "      <td>Gross pizza, better than pizza pizza, but stil...</td>\n",
       "      <td>2017-08-08</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>723Rler7IsBi80ps97vrxA</td>\n",
       "      <td>Delivery</td>\n",
       "      <td>0.055556</td>\n",
       "      <td>it</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>217</th>\n",
       "      <td>XT7-XSDM3kGS4KYS6AkZjg</td>\n",
       "      <td>Pizza Hut</td>\n",
       "      <td>2.5</td>\n",
       "      <td>QC</td>\n",
       "      <td>Lasalle</td>\n",
       "      <td>LaSalle</td>\n",
       "      <td>7551 Boulevard Newman</td>\n",
       "      <td>H8N 1X3</td>\n",
       "      <td>dhvokw5bM3MdQm85xR3fZg</td>\n",
       "      <td>Très mauvais service nous avons attendu au moi...</td>\n",
       "      <td>2015-12-03</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>dR0Eo3A6AKE_Hj5MxbVhmQ</td>\n",
       "      <td>Delivery</td>\n",
       "      <td>0.011811</td>\n",
       "      <td>fr</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>224</th>\n",
       "      <td>XT7-XSDM3kGS4KYS6AkZjg</td>\n",
       "      <td>Pizza Hut</td>\n",
       "      <td>2.5</td>\n",
       "      <td>QC</td>\n",
       "      <td>Lasalle</td>\n",
       "      <td>LaSalle</td>\n",
       "      <td>7551 Boulevard Newman</td>\n",
       "      <td>H8N 1X3</td>\n",
       "      <td>CmMgk7W1xu_fH-b22nLfXg</td>\n",
       "      <td>Je sais que je ne devais pas m'attendre à de l...</td>\n",
       "      <td>2015-11-28</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>x459culD4Wy_auLcAoQgrw</td>\n",
       "      <td>Delivery</td>\n",
       "      <td>0.023636</td>\n",
       "      <td>fr</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>225</th>\n",
       "      <td>XT7-XSDM3kGS4KYS6AkZjg</td>\n",
       "      <td>Pizza Hut</td>\n",
       "      <td>2.5</td>\n",
       "      <td>QC</td>\n",
       "      <td>Lasalle</td>\n",
       "      <td>LaSalle</td>\n",
       "      <td>7551 Boulevard Newman</td>\n",
       "      <td>H8N 1X3</td>\n",
       "      <td>UA5RCkj6LzxKaW-6cm2X0Q</td>\n",
       "      <td>J'aime bien Pizza Hut surtout pour leur croûte...</td>\n",
       "      <td>2017-07-10</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>MD-QMO8zcqd584wcpzo9sg</td>\n",
       "      <td>Delivery</td>\n",
       "      <td>0.017391</td>\n",
       "      <td>fr</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375</th>\n",
       "      <td>fsMgUPvNasNOMoz5yx680Q</td>\n",
       "      <td>Pizza Hut</td>\n",
       "      <td>2.5</td>\n",
       "      <td>QC</td>\n",
       "      <td>Laval</td>\n",
       "      <td>Laval</td>\n",
       "      <td>1700 Boulevard des Laurentides</td>\n",
       "      <td>H7M 2Y4</td>\n",
       "      <td>1FZ-vdLjcGACFgit1uTOHg</td>\n",
       "      <td>Plus mauvaise expérience client que j'ai eu de...</td>\n",
       "      <td>2017-04-19</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>_xJx7K2tq71hJUGxUOzUYw</td>\n",
       "      <td>Delivery</td>\n",
       "      <td>0.038339</td>\n",
       "      <td>fr</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                business_id       name  biz_stars state     city  \\\n",
       "122  vQXhuCQ4Wc3WWesp9hy3VQ  Pizza Hut        1.5    ON  Toronto   \n",
       "217  XT7-XSDM3kGS4KYS6AkZjg  Pizza Hut        2.5    QC  Lasalle   \n",
       "224  XT7-XSDM3kGS4KYS6AkZjg  Pizza Hut        2.5    QC  Lasalle   \n",
       "225  XT7-XSDM3kGS4KYS6AkZjg  Pizza Hut        2.5    QC  Lasalle   \n",
       "375  fsMgUPvNasNOMoz5yx680Q  Pizza Hut        2.5    QC    Laval   \n",
       "\n",
       "     neighborhood                         address postal_code  \\\n",
       "122  St. Lawrence                 77 Lower Jarvis     M5E 1R8   \n",
       "217       LaSalle           7551 Boulevard Newman     H8N 1X3   \n",
       "224       LaSalle           7551 Boulevard Newman     H8N 1X3   \n",
       "225       LaSalle           7551 Boulevard Newman     H8N 1X3   \n",
       "375         Laval  1700 Boulevard des Laurentides     H7M 2Y4   \n",
       "\n",
       "                  review_id  \\\n",
       "122  gJ32Kq6Wwq6Pr9_dwj_C_w   \n",
       "217  dhvokw5bM3MdQm85xR3fZg   \n",
       "224  CmMgk7W1xu_fH-b22nLfXg   \n",
       "225  UA5RCkj6LzxKaW-6cm2X0Q   \n",
       "375  1FZ-vdLjcGACFgit1uTOHg   \n",
       "\n",
       "                                                  text        date  cool  \\\n",
       "122  Gross pizza, better than pizza pizza, but stil...  2017-08-08     0   \n",
       "217  Très mauvais service nous avons attendu au moi...  2015-12-03     0   \n",
       "224  Je sais que je ne devais pas m'attendre à de l...  2015-11-28     1   \n",
       "225  J'aime bien Pizza Hut surtout pour leur croûte...  2017-07-10     2   \n",
       "375  Plus mauvaise expérience client que j'ai eu de...  2017-04-19     0   \n",
       "\n",
       "     funny  useful  stars                 user_id      type  puncount lan  \n",
       "122      0       1      2  723Rler7IsBi80ps97vrxA  Delivery  0.055556  it  \n",
       "217      0       0      2  dR0Eo3A6AKE_Hj5MxbVhmQ  Delivery  0.011811  fr  \n",
       "224      1       1      2  x459culD4Wy_auLcAoQgrw  Delivery  0.023636  fr  \n",
       "225      1       4      4  MD-QMO8zcqd584wcpzo9sg  Delivery  0.017391  fr  \n",
       "375      0       0      1  _xJx7K2tq71hJUGxUOzUYw  Delivery  0.038339  fr  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['lan']!='en'].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:31:32.850712Z",
     "start_time": "2018-11-12T15:31:32.831024Z"
    }
   },
   "outputs": [],
   "source": [
    "import wordcloud"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sentence tokenize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:31:40.330585Z",
     "start_time": "2018-11-12T15:31:32.852987Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    20877.000000\n",
       "mean         7.348853\n",
       "std          5.924358\n",
       "min          1.000000\n",
       "25%          4.000000\n",
       "50%          6.000000\n",
       "75%          9.000000\n",
       "max         96.000000\n",
       "Name: reviewsenlen, dtype: float64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# get sentences of review\n",
    "default_st=nltk.sent_tokenize\n",
    "df['reviewsen']=df.text.apply(lambda x: default_st(text=x))\n",
    "df['reviewsenlen']=df.reviewsen.apply(len)\n",
    "# review sentences stat\n",
    "df.hist('reviewsenlen')\n",
    "df['reviewsenlen'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:31:40.648350Z",
     "start_time": "2018-11-12T15:31:40.333376Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'review length')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# if sentence lenth are related to review stars\n",
    "sns.violinplot(x='stars',y='reviewsenlen',data=df).set_title('Review Sentence Length vs. Star')\n",
    "plt.ylabel('review length')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:31:40.667439Z",
     "start_time": "2018-11-12T15:31:40.651161Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[\"This review is about this one location's unacceptable behavior.\",\n",
      " \"To start, I like Papa John's.\",\n",
      " '* The pizza is reasonably tasty for my family (not as good as the local top '\n",
      " 'notch mom  & pop places, but decent none-the-less.)',\n",
      " '* The value is exceptional - they have the best specials, which we always '\n",
      " 'take advantage of.',\n",
      " 'Their online ordering and REWARD program is unparalleled (I belong to  every '\n",
      " 'chains reward and email clubs, they are all pathetic.',\n",
      " \"Not PJ's!\",\n",
      " 'Join and find out!)',\n",
      " '* My local PJ is a franchise that in my opinion is a model for any franchise '\n",
      " 'of any type.',\n",
      " 'They know us by face and by name and greet us so when we come in.',\n",
      " 'The food is well prepared and consistent.',\n",
      " 'They service is fast and extra friendly.',\n",
      " 'They rarely screw-up; and IF they do they are apologetic and fast to fix it '\n",
      " 'in a method that we prefer.',\n",
      " 'So... What about THIS LOCATION??',\n",
      " 'We were having a family gathering in Phoenix and someone said \"Let\\'s get '\n",
      " 'pizza!',\n",
      " 'We\\'ll have Domino\\'s deliver!\"',\n",
      " '\"Wait!\"',\n",
      " 'I said.',\n",
      " 'As the pater familias I decided to treat everyone and said \"Can we do Papa '\n",
      " \"John's?\",\n",
      " \"That's all we ever order and it's great.\",\n",
      " 'I think I have a free pizza or two waiting for me anyway?\"',\n",
      " '\"Well we usually get white pizza with spinach and veggies from Domino\\'s\" '\n",
      " 'the vegan Phoenician daughter chimed in.',\n",
      " '\"No problem.',\n",
      " 'Papa John\\' has that too.\"',\n",
      " '\"Do they have an online configurator so I can pick our toppings?\"',\n",
      " 'she continued.',\n",
      " '\"Yeah, it\\'s great.\"',\n",
      " '\"Do they still give that Garlic Butter sauce to dip your crust in?',\n",
      " 'That\\'s the main thing we always liked from there.\"',\n",
      " '\"Domino\\'s is super cheap...\" she challenged.',\n",
      " '\"Let me login and order.',\n",
      " \"They have great specials, I've got freebies coming.\",\n",
      " 'You can pick your own toppings.',\n",
      " \"I'll order delivery and pay online including tip for driver.\",\n",
      " 'We\\'ll be eating in under a half hour.\"',\n",
      " '\"OK, dad.\"',\n",
      " 'Online I go.',\n",
      " 'Couple big loaded pizzas for about 10 bucks each, couple free 2-liters of  '\n",
      " 'soda, an apple pie desert pizza at half price.',\n",
      " 'Paid online.',\n",
      " 'Done!!!',\n",
      " 'Problems begin.',\n",
      " '50 minutes later the delivery guy arrives.',\n",
      " 'Someone gets the door, then calls me \"to sign the receipt.\"',\n",
      " 'I come to the door to sign.',\n",
      " \"I only see the pizza's so I ask about  the drinks.\",\n",
      " 'The zombie with the hot bag blank stares at me with glazed-over eyes usually '\n",
      " 'seen only on reggae musicians.',\n",
      " 'I sign.',\n",
      " 'There is still no response.',\n",
      " 'I ask again  about the drinks and the desert pizza.',\n",
      " 'The expressionless zombie mumbles and someone  behind me yells \"we got it '\n",
      " 'all dad!\"',\n",
      " 'OK!',\n",
      " '\"Thanks!\"',\n",
      " 'I say to zombieman, who slowly turns and shuffles away with no response.',\n",
      " 'He was well tipped, but maybe I needed to offer fresh brains instead of cash '\n",
      " 'to show appreciation to him.',\n",
      " 'Problems continue.',\n",
      " 'We open up the pizza boxes.',\n",
      " 'The crusts are uber dark and burnt in many spot.',\n",
      " 'Cheese is mostly brown.',\n",
      " 'I pick up and bite into a piece of  Spinach Alfredo and it crunches.',\n",
      " 'It  is completely dry, sauce was cooked away.',\n",
      " 'Cheese was dried out too.',\n",
      " 'Both pizzas were like this.',\n",
      " '\"Gross!\"',\n",
      " '\"Ewww\"  \"This is inedible\" \"They burnt this this!\"',\n",
      " '\"Call them back and tell them.\"',\n",
      " 'were the comments ringing from all my spawn.',\n",
      " '\"Sorry guys!',\n",
      " \"I'll take care of it.\",\n",
      " \"I ordered these 'well-done' as I always order every pizza.\",\n",
      " \"It makes the cheese 'break' a little and makes the crust better for dipping.\",\n",
      " \"I've talked to store managers about this and know just what they do.\",\n",
      " 'The pizzas are amazing that way.',\n",
      " 'Someone screwed these up.',\n",
      " 'They are burnt.',\n",
      " 'I\\'m sorry.\"',\n",
      " 'I \\nlamented.',\n",
      " 'Problem gets worse.',\n",
      " 'I call in.',\n",
      " 'I explain the problem.',\n",
      " 'I ask what they did to mess these up.',\n",
      " 'Unapologetically, the fellow on the other end explains how they cooked the '\n",
      " 'pies and seems to know the correct  method for well-done pizza.',\n",
      " 'As a retired restaurateur, I can only figure either the pizza handled did '\n",
      " 'something different, or the ovens in that store are turned-up a bit for '\n",
      " 'faster cooking which fouls up a \"well done\" order.',\n",
      " 'Either way, they guy on the phone was apathetic but said \"We can remake '\n",
      " 'those for you.\"',\n",
      " '\"Ok.',\n",
      " 'Thanks.\"',\n",
      " 'I say \"These are inedible.\"',\n",
      " '...click... \"OK, guys!',\n",
      " 'They are making new pizzas!',\n",
      " 'AN HOUR LATER.',\n",
      " 'Let me repeat.',\n",
      " 'AN HOUR LATER, a snarly, untipped female driver brings  2 perfectly cooked '\n",
      " '\"well done\" pizzas.',\n",
      " 'By that time, my hungry spawn had torn out and devoured all the '\n",
      " 'semi-burnt-but-still-somewhat-edible-as-long-as-they-are-covered-with-ranch-dressing '\n",
      " 'center nuggets of the unholy abominations this location had first delivered.',\n",
      " 'The group had subsequently moved on to the livingroom to watch True Blood.',\n",
      " '\"Hey, the new pizzas are here!',\n",
      " 'They are prefect!',\n",
      " 'De-LISH!\"',\n",
      " 'But no one was interested any more.']\n",
      "[\"Mellow Mushroom isn't exactly the sort of place I'd ordinarily go.\",\n",
      " 'Woodstock murals, giant hippy-painted school-buses in the middle of the '\n",
      " \"dining, and awkward wrap-around booths just don't quite appeal to my more \"\n",
      " 'refined dining sensibilities.',\n",
      " 'But, my coworkers dig it.',\n",
      " 'And the one on Selwyn has a reputation for slow, inattentive service - which '\n",
      " 'is perfect when you and your coworkers are looking to extend your one-hour '\n",
      " 'workday lunch break into a 2.5 hour marathon.',\n",
      " 'During the carpool-carride over, the majority of us joked about how terrible '\n",
      " 'the service would be.',\n",
      " \"Shauna, our lone female (and lone optimist), righteously defended MM's \"\n",
      " 'honor.',\n",
      " '\"You guys are so harsh,\" she groaned.',\n",
      " '\"We got bad service one time... and now it\\'s like they\\'re the worst place '\n",
      " 'ever.\"',\n",
      " 'We were quick to point out to Shauna multiple specific examples of the bad '\n",
      " 'service we had received on several separate occasions: \\n'\n",
      " '\"Remember the time Mark had to go behind the bar and pour his own beer '\n",
      " 'because the waiter never came back?\"',\n",
      " '\"Remember the time we were seated for 47 minutes before anyone took our '\n",
      " 'drink order?\"',\n",
      " '\"Remember the time we had three different (lousy) waiters because none of '\n",
      " 'them knew whose section we were in?\"',\n",
      " 'Nevertheless, we keep coming back, hopeful each time that the experience '\n",
      " 'will change.',\n",
      " 'It never does.',\n",
      " 'Last Friday we were welcomed and seated promptly.',\n",
      " 'Out waitress approached, comely and upright, with jet-black hair, loose but '\n",
      " 'not sloppy clothing, and a lovely porcelain-skinned face - plain and '\n",
      " 'inexpressive... but not entirely lifeless.',\n",
      " 'She introduced herself with a breathless murmur: Kim, Amy, Amber...',\n",
      " 'I dunno.',\n",
      " \"I don't think she even knew.\",\n",
      " '\"Hi,\" we all said with genuine enthusiasm.',\n",
      " 'She just sorta stood there and teetered like a weeble.',\n",
      " '\"We need drinks,\" I declared.',\n",
      " '\"Any specials?\"',\n",
      " '\"Um...\" Kim/Amy looked bewildered.',\n",
      " 'Bart leaned forward and read off the specials on the chalkboard behind her.',\n",
      " 'The last one was a \"Mystery Beer: $3\"\\n\"What\\'s the \\'mystery beer\\'?\"',\n",
      " 'I asked.',\n",
      " 'Kim/Amy shrugged.',\n",
      " '\"It\\'s a mystery.\"',\n",
      " 'I was intrigued.',\n",
      " 'So I ordered one.',\n",
      " '(Turns out the mystery beer is just a random bottle of a crappy beer like '\n",
      " \"Grolsch or Longboard or something else they're trying to get rid of that \"\n",
      " \"shouldn't be sold for more than a buck...) \\n\"\n",
      " 'Bart cleared his throat.',\n",
      " '\"Is there a lunch special?\"',\n",
      " 'he asked, catching Kim/Amy completely off-guard.',\n",
      " '\"What day is it?\"',\n",
      " 'she asked us, sincerely.',\n",
      " '\"Um...',\n",
      " 'Friday.\"',\n",
      " '\"Oh, then yeah.',\n",
      " 'We have the special.\"',\n",
      " 'We all, then, politely extended a protracted pause.',\n",
      " 'When no further information was offered, we were compelled to pry: \"Oh yeah?',\n",
      " 'So... what is the special?\"',\n",
      " '\"It\\'s a slice and a salad,\" Kim/Amy replied, quite matter-of-factly.',\n",
      " '\"What kind of slices can I choose from?\"',\n",
      " 'Bart asked.',\n",
      " '\"You can get anything you want,\" She said.',\n",
      " '\"Oh!\"',\n",
      " \"Bart's face lit up as he started scanning through the specialty pies in the \"\n",
      " \"menu's center page.\",\n",
      " '\"So, I can get the Mighty Meaty?\"',\n",
      " 'As if he should have known, Amy/Kim halted him sternly, \"No, you can\\'t get '\n",
      " 'those.',\n",
      " 'You have to make your own.\"',\n",
      " 'Bart, a bit flummoxed, tried to reason this out.',\n",
      " '\"Oh, so I can\\'t choose a specialty pizza but I can get whatever toppings I '\n",
      " 'want?\"',\n",
      " 'Kim/Amy nodded.',\n",
      " '\"Yeah, you can get whatever you want.',\n",
      " 'I mean, you have to pay for toppings.\"',\n",
      " 'Bart, now more flabbergasted than flummoxed, persisted.',\n",
      " '\"Ok, so... toppings are extra?\"',\n",
      " '\"I mean... you get one,\" she said.',\n",
      " 'Now the picture was forming more clearly.',\n",
      " 'An audible \"ahhh!\"',\n",
      " 'of clarity was released by half the table.',\n",
      " '\"So, it\\'s a one-topping slice and a salad?\"',\n",
      " 'Kim/Amy nodded.',\n",
      " 'Now we were getting somewhere.',\n",
      " 'But then, as if she were somehow made uncomfortable by the sudden moment of '\n",
      " 'clarity, she decided to throw us an additional curveball.',\n",
      " '\"Or you can get two slices.\"',\n",
      " 'Bart, now just being a jack-ass, forced her hand.',\n",
      " '\"Oh, so instead of toppings I can just get two slices - but does cheese '\n",
      " 'count as a topping?\"',\n",
      " '\"No, a slice and a salad.\"',\n",
      " 'We all laughed.',\n",
      " 'And she laughed too.',\n",
      " 'Not sure if she knew it was an at-you-not-with-you type thing.',\n",
      " 'Orders were placed.',\n",
      " 'Most of us got the special, some got sandwiches.',\n",
      " 'I ordered pretzels for the table (garlic bread in the shape of a pretzel '\n",
      " 'with marinara for dipping).',\n",
      " 'The side-salads came out promptly.',\n",
      " 'Unspectacular - just some greens with onion and tomato sprinkled in.',\n",
      " 'Then, we waited.',\n",
      " 'Took about half an hour for the food to arrive (the appetizer brought at the '\n",
      " 'same time as the entrees... sigh).',\n",
      " 'And we did get one or two more rounds of drinks served in that time.',\n",
      " 'So, all in all, not too bad.',\n",
      " 'The food is... mediocre.',\n",
      " 'Greasy slices of pizza with run-of-the-mill ingredients.',\n",
      " \"The slices are large, so there's that.\",\n",
      " 'The guys who got sandwiches were impressed and noted that \"sandwiches are '\n",
      " 'the way to go here\".',\n",
      " 'Good to know.',\n",
      " 'Checks were a bit of a situation.',\n",
      " \"We had to each recount what exactly we'd ordered since Kim/Amy evidently \"\n",
      " \"didn't split it up as she took things down.\",\n",
      " 'No big deal.',\n",
      " \"It's Mellow Mushroom.\",\n",
      " '1.5 hours in and out.',\n",
      " \"Best experience I've ever had there.\"]\n"
     ]
    }
   ],
   "source": [
    "# look at some longest reviews\n",
    "for i in df[df['reviewsenlen']>80]['reviewsen'].head():\n",
    "    pprint(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:31:40.689210Z",
     "start_time": "2018-11-12T15:31:40.671070Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# get length of review\n",
    "df.loc[:,'reviewlen']=df.loc[:,'text'].apply(len)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:31:40.941816Z",
     "start_time": "2018-11-12T15:31:40.692214Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    20877.000000\n",
       "mean       527.401207\n",
       "std        483.445307\n",
       "min          2.000000\n",
       "25%        219.000000\n",
       "50%        378.000000\n",
       "75%        672.000000\n",
       "max       5000.000000\n",
       "Name: reviewlen, dtype: float64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# review length stat\n",
    "df.hist('reviewlen')\n",
    "df.loc[:,'reviewlen'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Word Tokenize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:08.691754Z",
     "start_time": "2018-11-12T15:31:40.944962Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    20877.000000\n",
       "mean       113.180773\n",
       "std        104.224505\n",
       "min          1.000000\n",
       "25%         47.000000\n",
       "50%         81.000000\n",
       "75%        144.000000\n",
       "max       1119.000000\n",
       "Name: reviewwordlen, dtype: float64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# get sentences of review\n",
    "default_wt=nltk.word_tokenize\n",
    "df['reviewword']=df.text.apply(lambda x: default_wt(text=x))\n",
    "df['reviewwordlen']=df.reviewword.apply(len)\n",
    "# review sentences stat\n",
    "df.hist('reviewwordlen')\n",
    "df['reviewwordlen'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:08.914480Z",
     "start_time": "2018-11-12T15:32:08.694152Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    20877.000000\n",
       "mean        15.791324\n",
       "std          8.867116\n",
       "min          1.000000\n",
       "25%         11.125000\n",
       "50%         14.333333\n",
       "75%         18.200000\n",
       "max        218.000000\n",
       "Name: avgword, dtype: float64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAEICAYAAAC0+DhzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAGGtJREFUeJzt3X+QXeV93/H3JxCoQ+wCwWxlhCMlkTPmRyobFZhxbW9CDIJ0ItzGiSg1MqaV7YE2njJTizRTHLu0uIniGWYcXNlWgcYGU2OM6uBghfqCacEGbAUQmLKAbBZpUI2wYcEhI/nbP+5RuOjclfaXdFe779fMnXvu9zzPOc995mo/Oj/2bqoKSZJ6/cygByBJmn0MB0lSi+EgSWoxHCRJLYaDJKnFcJAktRgO0kEkyXCS0UGPQ3Of4SBJajEcpFkqyaGDHoPmL8NB806SNUkeT/JCkoeTvDvJ4Ul+lOSknnavT/KTJMc2r/9dkm1Jtib5l0kqya8kWdz0/Zmm3WeTbO/Zzp8n+XCz/IYkG5LsSDKS5F/1tPtoki817Z8H3pfkNUmuSfJckoeBf3Sg5knzm+Gg+ehx4O3A3wf+CPhz4Gjgy8B5Pe1+F7ijqrYnWQ78W+A3gV8B3rm7UVU9CTwPvKUpvR0YS/Lm5vU7gDua5euBUeANwO8A/ynJGT37XAF8CTgS+DxwOfDLzeMsYNU037s0IYaD5p2q+h9VtbWqflpVXwQeA04FvsCrw+GfNzXoBsV/q6rNVfUS3VDpdQfwziT/oHn9peb1YuB1wF8nOR74x8BHqupvqmoT8FngvT3bubuqvtKM7SfNfq+oqh1V9RRw1QxNg7RXhoPmnSQXJNnUnAr6EXAScAzwv4DXJDktyS8CS4Gbm25vAJ7q2cxTvNodwDDdo4Q7gQ7do4t3At+sqp8229hRVS/09Ps+cNxetrvnfr8/ibcqTZkXvDSvND/0PwOcQfd/6buSbAJSVT9NciPdo4dngK/2/CDfBizs2dTxe2z6DuCP6Z4yugO4C/g08De8ckppK3B0ktf2bPeNwNM929nza5K3Nfva3NNe2u88ctB8cwTdH8D/DyDJhXSPHHb7AvB7wPm8ckoJ4EbgwiRvTvJzwH/o3WhVPQb8BPgXwJ1V9TzdgPlnNOHQnBb6P8B/TvL3kvwacBHdawvjuRG4LMlRSRYC/3pK71qaJMNB80pVPQysBe6m+8P7ZOB/96z/FvAi3dM5X+upf43u+f5vACNNf4CXezZ/B/BsVf2g53WA7/a0OQ9YRPco4mbg8qrauJch/xHdU0lPAl8H/vuE36w0DfGP/UiT19yJ9BBweFXtHPR4pJnmkYM0Qc3vQxyW5CjgE8D/NBg0VxkO0sR9gO61iseBXcCHBjscaf/xtJIkqcUjB0lSy0H7ew7HHHNMLVq0aNL9XnzxRY444oiZH9BBznnpz3npz3np72CYl/vvv/+HVfX6fbU7aMNh0aJF3HfffZPu1+l0GB4envkBHeScl/6cl/6cl/4OhnlJMqHfsve0kiSpxXCQJLUYDpKkFsNBktRiOEiSWgwHSVKL4SBJajEcJEkthoMkqeWg/Q3p6Vi05i8Gst8tV/7WQPYrSZPlkYMkqcVwkCS1GA6SpBbDQZLUYjhIkloMB0lSi+EgSWoxHCRJLfsMhyTrk2xP8lBP7YtJNjWPLUk2NfVFSX7Ss+7TPX1OSfJgkpEkVyVJUz86ycYkjzXPR+2PNypJmriJHDlcAyzvLVTV71XV0qpaCtwEfLln9eO711XVB3vqVwOrgSXNY/c21wC3V9US4PbmtSRpgPYZDlV1J7Cj37rmf/+/C1y/t20kWQC8rqrurqoCrgPObVavAK5tlq/tqUuSBmS63630duCZqnqsp7Y4yXeB54E/rKpvAscBoz1tRpsawFBVbQOoqm1Jjh1vZ0lW0z36YGhoiE6nM+kBj42NcenJuybdbyZMZbwHytjY2Kwe36A4L/05L/3NpXmZbjicx6uPGrYBb6yqZ5OcAnwlyYlA+vStye6sqtYB6wCWLVtWw8PDkx5wp9Nh7V0vTrrfTNhy/vBA9jsRnU6HqcznXOe89Oe89DeX5mXK4ZDkUOCfAqfsrlXVy8DLzfL9SR4H3kT3SGFhT/eFwNZm+ZkkC5qjhgXA9qmOSZI0M6ZzK+tvAt+rqr87XZTk9UkOaZZ/ie6F5yea00YvJDm9uU5xAXBL020DsKpZXtVTlyQNyERuZb0euBv41SSjSS5qVq2kfSH6HcADSf4a+BLwwarafTH7Q8BngRHgceBrTf1K4F1JHgPe1byWJA3QPk8rVdV549Tf16d2E91bW/u1vw84qU/9WeCMfY1DknTg+BvSkqQWw0GS1GI4SJJaDAdJUovhIElqMRwkSS2GgySpxXCQJLUYDpKkFsNBktRiOEiSWgwHSVKL4SBJajEcJEkthoMkqcVwkCS1GA6SpBbDQZLUYjhIklr2GQ5J1ifZnuShntpHkzydZFPzOKdn3WVJRpI8muSsnvrypjaSZE1PfXGSbyV5LMkXkxw2k29QkjR5EzlyuAZY3qf+yapa2jxuBUhyArASOLHp82dJDklyCPAp4GzgBOC8pi3AJ5ptLQGeAy6azhuSJE3fPsOhqu4EdkxweyuAG6rq5ap6EhgBTm0eI1X1RFX9LXADsCJJgN8AvtT0vxY4d5LvQZI0ww6dRt9LklwA3AdcWlXPAccB9/S0GW1qAE/tUT8N+AXgR1W1s0/7liSrgdUAQ0NDdDqdSQ96bGyMS0/eNel+M2Eq4z1QxsbGZvX4BsV56c956W8uzctUw+Fq4ONANc9rgfcD6dO26H+EUntp31dVrQPWASxbtqyGh4cnNWjo/oBee9eLk+43E7acPzyQ/U5Ep9NhKvM51zkv/Tkv/c2leZlSOFTVM7uXk3wG+GrzchQ4vqfpQmBrs9yv/kPgyCSHNkcPve0lSQMypVtZkyzoefluYPedTBuAlUkOT7IYWAJ8G7gXWNLcmXQY3YvWG6qqgG8Av9P0XwXcMpUxSZJmzj6PHJJcDwwDxyQZBS4HhpMspXsKaAvwAYCq2pzkRuBhYCdwcVXtarZzCXAbcAiwvqo2N7v4CHBDkv8IfBf43Iy9O0nSlOwzHKrqvD7lcX+AV9UVwBV96rcCt/apP0H3biZJ0izhb0hLkloMB0lSi+EgSWoxHCRJLYaDJKnFcJAktRgOkqQWw0GS1GI4SJJaDAdJUovhIElqMRwkSS2GgySpxXCQJLUYDpKkFsNBktRiOEiSWgwHSVKL4SBJatlnOCRZn2R7kod6an+c5HtJHkhyc5Ijm/qiJD9Jsql5fLqnzylJHkwykuSqJGnqRyfZmOSx5vmo/fFGJUkTN5Ejh2uA5XvUNgInVdWvAf8XuKxn3eNVtbR5fLCnfjWwGljSPHZvcw1we1UtAW5vXkuSBmif4VBVdwI79qh9vap2Ni/vARbubRtJFgCvq6q7q6qA64Bzm9UrgGub5Wt76pKkATl0BrbxfuCLPa8XJ/ku8Dzwh1X1TeA4YLSnzWhTAxiqqm0AVbUtybHj7SjJarpHHwwNDdHpdCY92LGxMS49edek+82EqYz3QBkbG5vV4xsU56U/56W/uTQv0wqHJP8e2Al8viltA95YVc8mOQX4SpITgfTpXpPdX1WtA9YBLFu2rIaHhyc95k6nw9q7Xpx0v5mw5fzhgex3IjqdDlOZz7nOeenPeelvLs3LlMMhySrgnwBnNKeKqKqXgZeb5fuTPA68ie6RQu+pp4XA1mb5mSQLmqOGBcD2qY5JkjQzpnQra5LlwEeA366ql3rqr09ySLP8S3QvPD/RnDZ6IcnpzV1KFwC3NN02AKua5VU9dUnSgOzzyCHJ9cAwcEySUeByuncnHQ5sbO5Ivae5M+kdwMeS7AR2AR+sqt0Xsz9E986n1wBfax4AVwI3JrkI+AHwnhl5Z5KkKdtnOFTVeX3Knxun7U3ATeOsuw84qU/9WeCMfY1DknTg+BvSkqQWw0GS1GI4SJJaDAdJUovhIElqMRwkSS2GgySpxXCQJLUYDpKkFsNBktRiOEiSWgwHSVKL4SBJajEcJEkthoMkqcVwkCS1GA6SpBbDQZLUYjhIklomFA5J1ifZnuShntrRSTYmeax5PqqpJ8lVSUaSPJDkrT19VjXtH0uyqqd+SpIHmz5XJclMvklJ0uRM9MjhGmD5HrU1wO1VtQS4vXkNcDawpHmsBq6GbpgAlwOnAacCl+8OlKbN6p5+e+5LknQATSgcqupOYMce5RXAtc3ytcC5PfXrquse4MgkC4CzgI1VtaOqngM2Asubda+rqrurqoDrerYlSRqAQ6fRd6iqtgFU1bYkxzb144CnetqNNrW91Uf71FuSrKZ7hMHQ0BCdTmfSgx4bG+PSk3dNut9MmMp4D5SxsbFZPb5BcV76c176m0vzMp1wGE+/6wU1hXq7WLUOWAewbNmyGh4envTgOp0Oa+96cdL9ZsKW84cHst+J6HQ6TGU+5zrnpT/npb+5NC/TuVvpmeaUEM3z9qY+Chzf024hsHUf9YV96pKkAZlOOGwAdt9xtAq4pad+QXPX0unAj5vTT7cBZyY5qrkQfSZwW7PuhSSnN3cpXdCzLUnSAEzotFKS64Fh4Jgko3TvOroSuDHJRcAPgPc0zW8FzgFGgJeACwGqakeSjwP3Nu0+VlW7L3J/iO4dUa8BvtY8JEkDMqFwqKrzxll1Rp+2BVw8znbWA+v71O8DTprIWCRJ+5+/IS1JajEcJEkthoMkqcVwkCS1GA6SpBbDQZLUYjhIkloMB0lSi+EgSWoxHCRJLYaDJKnFcJAktRgOkqQWw0GS1GI4SJJaDAdJUovhIElqMRwkSS1TDockv5pkU8/j+SQfTvLRJE/31M/p6XNZkpEkjyY5q6e+vKmNJFkz3TclSZqeCf0N6X6q6lFgKUCSQ4CngZuBC4FPVtWf9LZPcgKwEjgReAPwV0ne1Kz+FPAuYBS4N8mGqnp4qmOTJE3PlMNhD2cAj1fV95OM12YFcENVvQw8mWQEOLVZN1JVTwAkuaFpazhI0oDM1DWHlcD1Pa8vSfJAkvVJjmpqxwFP9bQZbWrj1SVJAzLtI4ckhwG/DVzWlK4GPg5U87wWeD/Q75Ci6B9QNc6+VgOrAYaGhuh0OpMe79jYGJeevGvS/WbCVMZ7oIyNjc3q8Q2K89Kf89LfXJqXmTitdDbwnap6BmD3M0CSzwBfbV6OAsf39FsIbG2Wx6u/SlWtA9YBLFu2rIaHhyc92E6nw9q7Xpx0v5mw5fzhgex3IjqdDlOZz7nOeenPeelvLs3LTJxWOo+eU0pJFvSsezfwULO8AViZ5PAki4ElwLeBe4ElSRY3RyErm7aSpAGZ1pFDkp+je5fRB3rK/yXJUrqnhrbsXldVm5PcSPdC807g4qra1WznEuA24BBgfVVtns64JEnTM61wqKqXgF/Yo/bevbS/AriiT/1W4NbpjEWSNHP8DWlJUovhIElqMRwkSS2GgySpxXCQJLUYDpKkFsNBktRiOEiSWgwHSVKL4SBJajEcJEkthoMkqcVwkCS1GA6SpBbDQZLUYjhIkloMB0lSi+EgSWoxHCRJLdMOhyRbkjyYZFOS+5ra0Uk2JnmseT6qqSfJVUlGkjyQ5K0921nVtH8syarpjkuSNHUzdeTw61W1tKqWNa/XALdX1RLg9uY1wNnAkuaxGrgaumECXA6cBpwKXL47UCRJB97+Oq20Ari2Wb4WOLenfl113QMcmWQBcBawsap2VNVzwEZg+X4amyRpHw6dgW0U8PUkBfzXqloHDFXVNoCq2pbk2KbtccBTPX1Hm9p49VdJspruEQdDQ0N0Op1JD3ZsbIxLT9416X4zYSrjPVDGxsZm9fgGxXnpz3npby7Ny0yEw9uqamsTABuTfG8vbdOnVnupv7rQDZ51AMuWLavh4eFJD7bT6bD2rhcn3W8mbDl/eCD7nYhOp8NU5nOuc176c176m0vzMu3TSlW1tXneDtxM95rBM83pIprn7U3zUeD4nu4Lga17qUuSBmBa4ZDkiCSv3b0MnAk8BGwAdt9xtAq4pVneAFzQ3LV0OvDj5vTTbcCZSY5qLkSf2dQkSQMw3dNKQ8DNSXZv6wtV9ZdJ7gVuTHIR8APgPU37W4FzgBHgJeBCgKrakeTjwL1Nu49V1Y5pjk2SNEXTCoeqegL4h33qzwJn9KkXcPE421oPrJ/OeCRJM8PfkJYktRgOkqQWw0GS1GI4SJJaDAdJUovhIElqMRwkSS2GgySpxXCQJLUYDpKkFsNBktRiOEiSWgwHSVKL4SBJajEcJEkthoMkqcVwkCS1GA6SpBbDQZLUMuVwSHJ8km8keSTJ5iS/39Q/muTpJJuaxzk9fS5LMpLk0SRn9dSXN7WRJGum95YkSdN16DT67gQurarvJHktcH+Sjc26T1bVn/Q2TnICsBI4EXgD8FdJ3tSs/hTwLmAUuDfJhqp6eBpjkyRNw5TDoaq2Adua5ReSPAIct5cuK4Abqupl4MkkI8CpzbqRqnoCIMkNTVvDQZIGZDpHDn8nySLgLcC3gLcBlyS5ALiP7tHFc3SD456ebqO8EiZP7VE/bZz9rAZWAwwNDdHpdCY91rGxMS49edek+82EqYz3QBkbG5vV4xsU56U/56W/uTQv0w6HJD8P3AR8uKqeT3I18HGgmue1wPuB9Ole9L/uUf32VVXrgHUAy5Ytq+Hh4UmPt9PpsPauFyfdbyZsOX94IPudiE6nw1Tmc65zXvpzXvqbS/MyrXBI8rN0g+HzVfVlgKp6pmf9Z4CvNi9HgeN7ui8EtjbL49UlSQMwnbuVAnwOeKSq/rSnvqCn2buBh5rlDcDKJIcnWQwsAb4N3AssSbI4yWF0L1pvmOq4JEnTN50jh7cB7wUeTLKpqf0BcF6SpXRPDW0BPgBQVZuT3Ej3QvNO4OKq2gWQ5BLgNuAQYH1VbZ7GuCRJ0zSdu5Xuov91hFv30ucK4Io+9Vv31k+SdGDNyN1KmphFa/5iYPvecuVvDWzfkg4+fn2GJKnFcJAktRgOkqQWw0GS1GI4SJJaDAdJUovhIElqMRwkSS2GgySpxXCQJLUYDpKkFsNBktRiOEiSWgwHSVKL4SBJajEcJEkthoMkqcVwkCS1zJpwSLI8yaNJRpKsGfR4JGk+mxV/QzrJIcCngHcBo8C9STZU1cODHdncsa+/X33pyTt53374G9f+7Wrp4DQrwgE4FRipqicAktwArAAMh4PcvkJpfzKYpKmbLeFwHPBUz+tR4LQ9GyVZDaxuXo4leXQK+zoG+OEU+s1p/2YOzks+MSObmXPzMkOcl/4Ohnn5xYk0mi3hkD61ahWq1gHrprWj5L6qWjadbcxFzkt/zkt/zkt/c2leZssF6VHg+J7XC4GtAxqLJM17syUc7gWWJFmc5DBgJbBhwGOSpHlrVpxWqqqdSS4BbgMOAdZX1eb9tLtpnZaaw5yX/pyX/pyX/ubMvKSqdWpfkjTPzZbTSpKkWcRwkCS1zJtw8Os5XpFkS5IHk2xKcl9TOzrJxiSPNc9HDXqcB0KS9Um2J3mop9Z3LtJ1VfMZeiDJWwc38v1rnHn5aJKnm8/NpiTn9Ky7rJmXR5OcNZhR719Jjk/yjSSPJNmc5Peb+pz8vMyLcOj5eo6zgROA85KcMNhRDdyvV9XSnnuy1wC3V9US4Pbm9XxwDbB8j9p4c3E2sKR5rAauPkBjHIRraM8LwCebz83SqroVoPm3tBI4senzZ82/ublmJ3BpVb0ZOB24uHnvc/LzMi/CgZ6v56iqvwV2fz2HXrECuLZZvhY4d4BjOWCq6k5gxx7l8eZiBXBddd0DHJlkwYEZ6YE1zryMZwVwQ1W9XFVPAiN0/83NKVW1raq+0yy/ADxC99sd5uTnZb6EQ7+v5zhuQGOZDQr4epL7m68kARiqqm3Q/UcAHDuw0Q3eeHPh5wguaU6RrO859Tjv5iXJIuAtwLeYo5+X+RIOE/p6jnnkbVX1VrqHvRcnecegB3SQmO+fo6uBXwaWAtuAtU19Xs1Lkp8HbgI+XFXP761pn9pBMy/zJRz8eo4eVbW1ed4O3Ez3FMAzuw95m+ftgxvhwI03F/P6c1RVz1TVrqr6KfAZXjl1NG/mJcnP0g2Gz1fVl5vynPy8zJdw8Os5GkmOSPLa3cvAmcBDdOdjVdNsFXDLYEY4K4w3FxuAC5q7UE4Hfrz7dMJ8sMf58nfT/dxAd15WJjk8yWK6F2C/faDHt78lCfA54JGq+tOeVXPy8zIrvj5jfzvAX88x2w0BN3c/5xwKfKGq/jLJvcCNSS4CfgC8Z4BjPGCSXA8MA8ckGQUuB66k/1zcCpxD94LrS8CFB3zAB8g48zKcZCndUyNbgA8AVNXmJDfS/fsrO4GLq2rXIMa9n70NeC/wYJJNTe0PmKOfF78+Q5LUMl9OK0mSJsFwkCS1GA6SpBbDQZLUYjhIkloMB0lSi+EgSWr5/48lKP0UvqWoAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# people couldn't write and using comma all the time\n",
    "\n",
    "df.loc[:,'avgword']=df.loc[:,'reviewwordlen']/df.loc[:,'reviewsenlen']\n",
    "# review length stat\n",
    "df.hist('avgword')\n",
    "df.loc[:,'avgword'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:08.924037Z",
     "start_time": "2018-11-12T15:32:08.917194Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(df['avgword']>200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:08.933541Z",
     "start_time": "2018-11-12T15:32:08.926555Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14844    The manager Johnny is an idiot I do not know i...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print(df[df['avgword']>200]['text'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## People tend to star higher"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:09.090167Z",
     "start_time": "2018-11-12T15:32:08.936679Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# star distributions\n",
    "#df['stars'].hist(bins=5)\n",
    "#df['stars'].value_counts().plot(kind='bar') this code doesn't sort ordinal\n",
    "ax=df.groupby('stars').size().plot(kind='bar',title=\"Frequency Count of Stars\",\n",
    "                                   rot=0,color='#3d87ff')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:09.098430Z",
     "start_time": "2018-11-12T15:32:09.092876Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 834 restaurants\n"
     ]
    }
   ],
   "source": [
    "print('There are',len(df.address.unique()),'restaurants')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:09.133485Z",
     "start_time": "2018-11-12T15:32:09.101047Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "business_id          0\n",
       "name                 0\n",
       "biz_stars            0\n",
       "state                0\n",
       "city                 0\n",
       "neighborhood     12179\n",
       "address              0\n",
       "postal_code          0\n",
       "review_id            0\n",
       "text                 0\n",
       "date                 0\n",
       "cool                 0\n",
       "funny                0\n",
       "useful               0\n",
       "stars                0\n",
       "user_id              0\n",
       "type                 0\n",
       "puncount             0\n",
       "lan                  0\n",
       "reviewsen            0\n",
       "reviewsenlen         0\n",
       "reviewlen            0\n",
       "reviewword           0\n",
       "reviewwordlen        0\n",
       "avgword              0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check na values for each column\n",
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## sentiment analysis (dr healey model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:22.635874Z",
     "start_time": "2018-11-12T15:32:09.136199Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 13.1 s, sys: 269 ms, total: 13.4 s\n",
      "Wall time: 13.5 s\n"
     ]
    }
   ],
   "source": [
    "# df['sentiment']=df['text'].apply(str.split).apply(sentiment.sentiment)\n",
    "# df=pd.concat([df['sentiment'].apply(pd.Series),df.drop('sentiment',axis=1)],axis=1)\n",
    "df=pd.concat([df['text'].apply(str.split).apply(sentiment.sentiment).apply(pd.Series),df],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:22.686179Z",
     "start_time": "2018-11-12T15:32:22.639508Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 37.5 ms, sys: 5.54 ms, total: 43 ms\n",
      "Wall time: 41.3 ms\n"
     ]
    }
   ],
   "source": [
    "df=df.drop(df[df['valence']==0].index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:24.716466Z",
     "start_time": "2018-11-12T15:32:22.690077Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "df.plot.scatter(x='valence',y='arousal',c='stars',fontsize=10,\n",
    "               colormap='RdBu_r',ax=ax,alpha=0.5,\n",
    "               figsize=(8, 6))\n",
    "ax.set_title('Sentiment vs. Star',fontsize=20)\n",
    "ax.set_xlabel('Valence',fontsize=15)\n",
    "ax.set_ylabel('Arousal',fontsize=15)\n",
    "plt.xlim(0,10)\n",
    "plt.ylim(0,10)\n",
    "plt.savefig('sentiment-all.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:24.722968Z",
     "start_time": "2018-11-12T15:32:24.719201Z"
    }
   },
   "outputs": [],
   "source": [
    "colors=['#6c0521','#7d14a5','#c84678','#f5914b','#0d3a67']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:24.729204Z",
     "start_time": "2018-11-12T15:32:24.725855Z"
    },
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# for i in range(1,6):\n",
    "#     fig, ax = plt.subplots()\n",
    "#     df[df['stars']==i].plot.scatter(x='valence',y='arousal', color=colors[i-1],\n",
    "#                                     ax=ax, alpha=0.5,title='Sentiment vs. Star '+str(i),\n",
    "#                                     figsize=(8, 6))\n",
    "#     plt.xlim(0,10)\n",
    "#     plt.ylim(0,10)\n",
    "#     plt.savefig('sentiment'+str(i)+'.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:25.018264Z",
     "start_time": "2018-11-12T15:32:24.732310Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "df[df['stars']==1].plot.scatter(x='valence',y='arousal', color='#0d3a67',fontsize=10,\n",
    "                                ax=ax, alpha=0.5,\n",
    "                                figsize=(6, 6))\n",
    "ax.set_title('Sentiment vs. Star',fontsize=20)\n",
    "ax.set_xlabel('Valence',fontsize=15)\n",
    "ax.set_ylabel('Arousal',fontsize=15)\n",
    "plt.xlim(0,10)\n",
    "plt.ylim(0,10)\n",
    "plt.savefig('sentiment-1.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:25.417951Z",
     "start_time": "2018-11-12T15:32:25.021201Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "df[df['stars']==1].plot.scatter(x='valence',y='arousal', color='#0d3a67',fontsize=10,\n",
    "                                ax=ax, alpha=0.5,\n",
    "                                figsize=(6, 6))\n",
    "df[df['stars']==5].plot.scatter(x='valence',y='arousal', color='#6c0521',fontsize=10,\n",
    "                                ax=ax, alpha=0.5,\n",
    "                                figsize=(6, 6))\n",
    "\n",
    "ax.set_title('Sentiment vs. Star',fontsize=20)\n",
    "ax.set_xlabel('Valence',fontsize=15)\n",
    "ax.set_ylabel('Arousal',fontsize=15)\n",
    "plt.xlim(0,10)\n",
    "plt.ylim(0,10)\n",
    "plt.savefig('sentiment15.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:25.705029Z",
     "start_time": "2018-11-12T15:32:25.420723Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfQAAAGDCAYAAADd8eLzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzt3Xl8nHd9J/DPV9LoPi3Jl3wGO3YunENJHBKH2BwJwQ0JhEKALKRsQ5cF0pZuF+hrl8IuS+mLQtmFUtKWFEoIWSgp2RByQBLI6cR24sRHfMSn5EuydVqWZqT57R+f+fUZyyNpJM1opGc+79dLL2uuZ37zTKLP87vNOQcRERGZ2QpyXQARERGZPAW6iIhICCjQRUREQkCBLiIiEgIKdBERkRBQoIuIiISAAl1kGjKzvzez/5brcojIzKFAF0mTmV1jZs+ZWZeZnTSzZ83s8gwc92Nm9kzyfc65P3LO/Y/JHnsCZflLM/tRDt73AjN7zMw6zKzTzDaZ2Y2Jx64zs5YsvOfdZrbTzOJm9rFMH19kqinQRdJgZtUAHgLwfwDMAtAE4EsABnJZrhD5fwAeBzAHwGwAnwHQnYkDm1nRCA9tAfBJAJsz8T4iuaZAF0nPuQDgnLvPOTfknDvtnHvMOfeqf4KZ/YGZ7UjUMh81s8VJjzkz+yMz2514/DtG5wH4ewBXmVmvmXUmnv/PZvY/E79fZ2YtZvbnZnbczI6Y2c1mdqOZ7Uq0Fnwh6b0KzOxzZvaGmZ0ws/9rZrMSjy1JlOWjZnbQzNrN7C8Sj90A4AsAPpAoy5bhJyFx3J8Nu+9bZva/E79/zMz2mlmPme0zsw+PdWLNrAHAUgD/4JyLJn6edc49Y2YVAH4FYH6iTL1mNt/MrjCz5xO1+SNm9m0zKx52vv+zme0GsDvV+zrnvuOc+w2A/rHKKDITKNBF0rMLwJCZ/cDM3mVmdckPmtnNYBi+F0AjgKcB3DfsGOsBXA5gFYDfB3C9c24HgD8C8LxzrtI5VzvC+88FUAq2DPx3AP8A4CMALgOwBsB/N7NzEs/9DICbAbwVwHwAHQC+M+x41wBYAeBtidee55x7BMD/AnB/oiyrUpTjPgA3JlosYGaFic/y40T4/m8A73LOVQF4C4BXRvg8yU4A2APgR4kLlTn+AefcKQDvAnA4UaZK59xhAEMA/gRAA4CrEp/jk8OOezOAKwGcn0YZRGY8BbpIGpxz3WAIOjBM28zswaTw+QSArzrndjjnBsFgvDi5lg7gr5xznc65gwCeBHDxOIoQA/AV51wMwE/AIPuWc67HObcNwDYAb04qy18451qccwMA/hLArcOanr+UaGXYAjY9pwrvVOfhANhEfXPirnUA+pxzLyRuxwFcaGZlzrkjibKNdUwHYC2A/QD+BsARM/udmS0f5TWbnHMvOOcGnXP7AXwPvIBJ9lXn3Enn3Ol0PpvITKdAF0lTIqw/5pxbAOBCsPb7t4mHFwP4VqIJuBPASQAG1qi9o0m/9wGoHMfbn3DODSV+9wF1LOnx00nHWwzggaSy7ABrtHOSnj+ZsvwYwG2J3z+UuO1r0x8AWxyOmNkvzWxlOgdMXHx8yjn3pkT5TwH44UjPN7NzzewhMztqZt3gBVTDsKcdGsdnEpnxFOgiE+Ccex3AP4PBDjA8PuGcq036KXPOPZfO4TJcvENgs3dyWUqdc60ZKstPAVxnZgsA3IJEoAOAc+5R59w7AMwD8DrYmjEuzrlDYBeBP7epyvTdxPGXO+eqwe4OG36o8b63yEymQBdJg5mtNLPPJkIMZrYQrKX6pua/B/B5M7sg8XiNmb0/zcMfA7AgeVDXJP09gK/45n4zazSz94yjLEvMbMS/Dc65NgBPAbgHwL7EOACY2RwzuynRlz4AoBdsGRiVmdWZ2ZfMbFliQF8DgD9AcG6PAag3s5qkl1WBo+B7E60A/ynNz5f8vsVmVgpeCETMrHS0zy0y3ek/XpH09IADrDaY2SkwbLYC+CwAOOceAPA1AD9JNAFvBQdzpeMJsA/8qJm1Z6Cs3wLwIIDHzKwnUdYr03ztTxP/njCz0aZz/RjA25FUOwf/nnwWwGGwy+GtSAxUM7M1ZtY7wrGiAJYA+DUY0lvBC4KPAf/eGnIfgL2JboT5AP4MbO7vAVsB7k/z8yV7DOyqeAuAuxO/XzuB44hMC8bxKCIiIjKTqYYuIiISAlkLdDP7fmIRjK1J980ys8eNi2s8Pnwur4iIiExMNmvo/wzghmH3fQ7Ab5xzywH8JnFbREREJimrfehmtgTAQ865CxO3dwK4zjl3xMzmAXjKObciawUQERHJE1Pdhz7HOXcEABL/zp7i9xcREQmlkXYhyjkzuxPAnQBQUVFx2cqVaS04JSIiMuNt2rSp3TnXOJ7XTHWgHzOzeUlN7sdHeqJz7m5wbiiam5vdxo0bp6qMIiIiOWVmB8b7mqlucn8QwEcTv38UwC+m+P1FRERCKZvT1u4D8DyAFca9nD8O4K8AvCOxR/E7ErdFRERkkrLW5O6cu22Eh96WrfcUERHJV1opTkREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhEBOAt3M/sTMtpnZVjO7z8xKc1EOERGRsJjyQDezJgCfAdDsnLsQQCGAD051OURERMIkV03uRQDKzKwIQDmAwzkqh4iISChMeaA751oBfB3AQQBHAHQ55x6b6nKIiIiESS6a3OsAvAfAUgDzAVSY2UdSPO9OM9toZhvb2tqmupgiIiIzSi6a3N8OYJ9zrs05FwPwcwBvGf4k59zdzrlm51xzY2PjlBdSRERkJslFoB8EsNrMys3MALwNwI4clENERCQ0ctGHvgHAzwBsBvBaogx3T3U5REREwqQoF2/qnPsigC/m4r1FRETCSCvFiYiIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIqlFo0BnJ/8VkWmvKNcFEJFpqLUVeOQRYHAQKCoCbrgBaGrKdalEZBSqoYvImaJRhnllJTB/Pv995BHV1EWmOQW6iJypr4818/Jy3i4v5+2+vpFfo+Z5kZxTk7vkt2iUQVVeDhQX57o000N5OZvZ/Xnp6+NtH/DDqXleZFpQDV3yV2sr8C//Atx/P/9tbc11iaaH4mKGcm8vcPgw/73hhtQXPNlunlfNXyRtqqFLfkoOIl8LfeQR4PbbVVMHWMO+/faglg4wWIe3ZKRqnu/s5P2TPY+q+YuMiwJd8lM2gyiVmdC0P7yM/me0YB1v8/x4yqILLpFxUaBLfspWEKUyE2qaI5VxrGD1zfOPPAK0twPxOLB+/eRDd6ovuERCQH3okp/G0088GTNhCthoZUxnxHtTE7BuHcO8oAB44onJj0dIvuACsnvBJRISqqFL/hreT5yNmt9MqGmOVsZ0WjKiUYb47NmZax5Prvl3dgatBtPlnIlMQzkJdDOrBfCPAC4E4AD8gXPu+VyURfKcbzbOlkw17WezD360MqYTrKkuCNrbgSNHgHnzJl7eqbjgEgmRXNXQvwXgEefcrWZWDEDtaBJOmahpZrsPfqwyjhWswy8IDh0CXnqJj5WWTq682b7gEgkRc85N7RuaVQPYAuAcl+abNzc3u40bN2a3YCLZNNEadjTKOfLJg9J6e7Mz2jvdMqZ6nr/o6O9nmF95JbBgQXbLKxJiZrbJOdc8ntfkooZ+DoA2APeY2SoAmwDc5Zw7lfwkM7sTwJ0AsGjRoikvpMiEpQq8idY0p7IPPp0yjtRa4GvxR47weQsWjK+8M2FaX5jofIdSLgK9CMClAD7tnNtgZt8C8DkA/y35Sc65uwHcDbCGPuWlFJmITDePT+X0urGkM4Vt3jw2s4+nvMnnDADWrAGWLVPQZMtMmEYpE5KLaWstAFqccxsSt38GBrzIzJaNKWpTNb1uJMlLr6YzhW285U0+ZyUlwObNwFe/Ctxzz8Smvmmp2NHNhGmUMmFTXkN3zh01s0NmtsI5txPA2wBsn+pyiGRctprHczXae3hNbt269FoLxlNef84iEeCZZ4CaGqCwkLfHO/VtPDXPfG1yngnTKGXCcjXK/dMA7k2McN8L4I4clUMkczLdPJ5qKdapkqp5/YknGOpPPDH2iP10y+vPWUcHF6YBeLu2FmhrSz9oxrNUbD43OU+nLhzJuJwEunPuFQDjGr0nMu1lcjGUqQid0WqpI9Xk6uoy21rgz9lDDwHd3XzP1auBWGx8QZNuzTPf14jXgj2hppXiRDIpE83jUxE6o10wRKNBn+pU1OSamoA77uBguKef5nvH4+MLmnRrnmpy1oI9IaZAF8m0yTaPZzt0RrtgaGsLgr6nhz9VVUHoJz+eyZaD4mLg/PM5un0iQZNuzVNNzqQFe0JJgS4y3WQ7dEa6YOjsPDvoOzuBW25hnzZw9iI3o7UcTGTgWaqgST6OL3+qY6ZT81STs4SYAl1kusl26Ix0wQCkDnofsp2dXAmuvJx93KO1HGRqDEDycXp6eF9yi0FjY+o93EejJmcJKQW6yHSUzdBJHoh2+DCPv349a+GjtQx0dHBZV3/fypWcOz685SBTYwCSjxOJcI760BDw1rdym9b77mO4A+O/aJhMk3O+TnmTaU/7oYuMJVeLlRQXM2SnKjRGWxTGb5F62WUM095e4MUXOY0NOPP8JM8t9/f39wcL0KR7PpO7BgYGeIw9e4CnngKefRZ47jk+r7qaq9NNdoGUdMrV2spuh/vv57+T3fddJINUQxcZTRjnLPuab20tVwtLrkGnahmIRrlG+5EjwLFjDPSiImDxYuD0aQZb8vmpqQGOH2fwHj3Kx+bOBW68ETh1Kr3zGY3yoqGvj9PZioqA/ft5kTB7NlsL9u9nrb2wkI/7zzKRNePT+Z57e4Gf/xyYNYsXEfk25U2mPQW6yEjCOmd5rFH0yc3RPuh6e4Ff/QpYsYJB190NvP468OSTQENDcH7uuw8oKwM2bQJ27ADmzAEqKni8hx8OFo0Z6XxGo6yFP/ggsG0bn1NQwBHwDQ38LtraADM+3wyorw/KUzTKn7SRQjud77m1lWH+wgt8v0suYXnybcqbTGtqchcZSTprl89EyYPigJFH0Uej7GcHWCtdtAjYu5f7ncdiwDnn8Dn+dZEIsGUL4BxDb+FC9rGvWMG+7q6uM0erDz+f+/YB3/kO8OUv8+KhshJ405t4AVBVBVx6KcN9YIAtA42NLEdbG+etr1wZbPAy/HMcP87PkmoN87G+Zx/4dXU8D2bAyy8HLQf5NuVNpi3V0EVGMp3nLE9m7/J0R9Hv2cP11aureXtoiEE3OMjjlpUF56W8nM3gAIM2EmGfd3s7w9eMQQowCH2TdVERw/iZZ4DvfpfN5y0tDO5Dh4CLLmK5nGNZ9+7lMU6f5vHnzuVrzj2XZRj+3fhaeXc3m+fXrQue52vXY33PPvBnz+YYgk2bgJMn+fPe96p2LtOGAl1kJNN1znK6/fqjPW+sUfTRKFdtKy9nk/nQEHDgAGvclZU8ZiwGXHcdw7izk6+74AIGcW8vn++nulVXcxDbNdcAjz/O2nRdHX8+9Slg506G7tVX8z19qPvlYIeG2Gd+ySW8f+tWXkD4i4xNm4DPf56fJxoNPrNvSq+qYuvBs8+yLz95admxvufkwK+vB668ku99xx08tsg0Yc5N/63Gm5ub3caNG3NdDMlX02maUjR69uIuvb1n9+un+7yRdHZyJHdxMZuXT51iiH7oQwy3bdvYhL56Nae8VVRw0NxDD7H/e3AQOHGCfejd3axhDw0Bn/40m6337ePFwU9/yn+d4zzzeJxN608/zd/nzuV71NYycKurgQULgF27OEDviisY1hUVfG4sBuzezRXnIhHeXrCAgd/dzde9853sPli3jhcUyYP/Ui1gA7C14umng/OjPdsly8xsk3NuXHuejFpDN7MeAKkS3wA451z1eN5MZEaaTstkpurvbW9nmM6bF5RzssvH+lppeTlr4e3tDLyaGjZdRyIM5ro64F//lc3vL7zA2nhBAW/39TGkfbO6c3ztu97FgHWO/d9DQzx2fT1fs3Urm+ff9z6+9tlngTe/maPqX32VtfmODl6stLcztH/zGwb/kSNs8n/ySfbh79wJLFnC3yMRYPlyvu6aazgNz/e5r1nDkAfYsuEfS17MBmDT/q5dDPfnnw/HrAcJjVED3TlXNVUFEZE0DO/vbWnhYi8A52L7gJls/39yM7Rvvr7rLobYiRMM88su4/E2bmQTemEhA7azkyFaVcXArqnh+1dWstn82WdZ6583j8/1K78NDLAWXVPD2q9zDPe9e9lcf+oUw7+zk8+NxxnwkQgvLHxNf/t29nc3NDCQ9+zhZzp8mLX1zZt5jPPO48XHCy8Ajz7KWv7KlazhX3EFa/wvvcTPf+ONvOj4wQ+Ad7xD09ZkWhpXH7qZzQZQ6m875w5mvEQiuTKdmtZHc9VVrCG2tzNwrriCNdDhATPR/n9/HhobuY57SwsDsrSUI9uBoKn6+ee5wMwbbwAHD/L4sRiP4Vywktz8+RwY19XFpvxolOFcWMjadkkJa/XNzbxYGBxk0/355zM8e3t54VBZyddWVDCMKyo4172/nxcK8ThfO28e33/uXIb/nj0sx4EDPNbu3UGrw6xZrHU3NvLxggJOgRsc5OeKx4Ff/xpYtYqfrSAxOSgfd2qTaS2tQDezmwD8DYD5AI4DWAxgB4ALslc0kSk0ExaQSS4jwBoywDAHzg6Y8Swf60O8oyNobj50iE3WfmrYNdewhrtqFfDb37J2/OKLDDg/YGzfPtawy8o4Qt1fTMydy5r9vn0M1KIijlTv6WHQV1fzc5w8ybDt6WHg19YC114L/OIXfF08HpQ5Huf7Dg0FzeUdHXxtdTU/08qVDHM/Kv3ECR4zEmHz/MmTvGiIxVhuPxWtp4eD6AoLebFRVQW88go/azwenJPkdfBFcizd/xL/B4DVAH7tnLvEzNYCuC17xRKZQpNdQGYqavapyrhly9jN6un0//sLhf7+oMZfXw/8+McMtNJSHvfBB4ELL+T9lZUMQ4Dh6AfAlZUxmIuKGHqDgwzEnh42nUciLI/v+y8pYaDPng289hpfX1jIx0+cYPN4by9rzENDQaj39vI1JSV8r/p6Pu67HZYsYe3dtxT4EfMAz5Pv8x8YYMtCYyPfq6mJAb95M8vlL5Z6e/n6j3+cFzLPPccLgvnz+b633Tb9LgAl76Qb6DHn3AkzKzCzAufck2b2tayWTGSqTGYA2VTV7Ecq45o1bPb208ZWr+bvI60BP/ziI/lCwTeXb94MXHwxn+/7tvv6GKQnTvD+N97g5+zq4rFOn2ZYV1cHAb9zJ4/b2MhAP3QoCNehIR6vvJwLzyxZEtSYy8qCgXTHj7NJPR7nT2FhcAy/1n1DAx8bGGDZFi0CPvlJnq/t24FvfjNYQtY59sVXVAQ7tUUifL/ubtbwBwc5CC8eZxdDXx+b8wsL2XrQ38/a/JIlfN3u3Rzdf8cdZ654N5GLvExcHM6UriPJuHQDvdPMKgH8DsC9ZnYcQIolmURmoPEMIOvtZVjV1/OPZTaXhk3+w5yqjADD7AMfYC3zwQeBr3+d969adXatMbkmHo9zulldXbAq2lNPcfT54CDDrb+fIWbGQC4s5H1dXby9ezeD9dQplnX3bgbdtm28zyy40Ono4HMiEQavH9TmR7e3tDDI/Wf2jw83NMTjlpTw4uOyy3hB09PD++rq+B35lonnnuPzh4b43r6P/dQpvt/ixazVz5rF1w8NcdBbcTG7Bw4f5gC+qiqO9h8aYktCbS0vCvwc+e7u4AJwohd5mbg4nAldR5I16Qb6ewCcBvAnAD4MoAbAl7NVKJEple4Aso0bgW9/m0ESiQAf/ejkpoaNJtUf5uQy+ulUDzzAfzs6GEDz5vH28Fqjr4kPDLDm3NfH8P7CF/j8557jvO6FC1lTbm1l0AIM2xMnWKM9doz96LFYUHOurmYZ581jKPb3B+E9NMSm9f5+BigQ1KQLCxm+v/41jzl3LmvxqYI8ma9lO8em+FOn+P4LFnBa2oEDLHNvL8sYjbIFwSsvZ1kuvTQYGFdYyPetqWFQd3fzwmT5cn6G5maej1iMzz19mueopYW/DwzwOygvH/siL1UNOhP7BoR17wFJ27hGczjnBs3seQArAXRnp0giOTDWALLeXoZ5TQ0DrLsb+P732cSd6aVhR/vDfPvtDPSf/pShVFvL26+8wt9LSniMwkK+zl9c9PUxVHfuZI20poZh9+ijbLZ/8kkGU1UVsHYtn79sGT9Pfz+brt94gxcSAwNsbj56lOWrq2MY9vcH5R8a4o9fUMY51pT9pioA7y8p4fF6eznIzA86S0dfH1epKy5mGU6c4GtLSoD/+l/5HW3YELRIFBTwImPBAn42/z5dXfzc7e3A9dfzAuGZZ4LugNZWXsxdf32whv3WrexLLyzkhcFVV3Ew4e/93sgXecCZC9Qk16D991NezveYyMXhZNcekBkv3UD/HYA1ZlYH4DcANgL4AFhbFwmH0QaQnTjBP7R+XfPqatZWzz2Xg70yuTTsaH+Ya2sZPC+8EMzfvuiioDnc136HhoKaud9AxY8Kr6nh84qLGaSRCAPp5ZfZ9FxUxNeUljIEDx5kzX3RIuCxx3js6mq+fyzG4/pQjMf5OBCEue8zj8V4n1dYGDy3tTV43nhFowzhgQF+F0VFbE04eTI4JwUFwR7t8+ZxTv2SJRw97+ew79nDC6c5c9hKceutPG+XX85WhDfe4OeureW89OpqnsPCQrYutLWxPKm6Rg4c4EXCCy/w/quu4gWZv1Dr6OCAxEiE5/3884PBg+maznsPyJRIN9DNOddnZh8H8H+cc39tZi9ns2Ai00p9Pf/Y+o1FDh9m8/D27fwDnKmlQH0AA6n/MCevse7XEd+4kaEOsJkYCPYqf+CB4EJj/Xo2sx89yuDs6OAAuNde465mCxaw2X5ggDXTNWv4XqdO8TP6cDp5kiE2MBC0Ihw/zuOZMbQHB4P+9/p6PhaLnflZh4f3RMLc6+1lIPs+9tbWoIxeQUHQVP5P/8TvK/lc7drF2vLVVwfbv9bUMIR92Vav5u2CAh7HjBcNS5fyIqK29uyukcFBjm0oKuLza2p4/LVr+VhnJ2v3V1zBLoS+PrYsfPGL4/vvabruPSBTJu1AN7OrwBr5x8f5WpGZr7KSm4h8+9usmR86BLz73cEo6OefDwJioqOMk/vNe3r442vh/g+zH81+1VUMBV/rvukmvn9nJ8PtF7/gtK7hK5p98YvAv/0bA+PkSY5mr6piyDc0MJR27mSNdPt23vfKKwyijg4GZE0NAz8e5wVAVxcDz9cE/a5rvnbe1nZmzTxbklsG/EWRl9yM39vLz/Hoo/x382Z+n62tQb/84sX8nvfv54XI4sXBBUNVFV8zfz5bNfr7eb6/8Y0z5/93dgYXVNXVbAXYsoXnOB7nefJz2AcH2QriF8I5eZItB+M1nrUHJHTSDeU/BvB5AA8457aZ2TkAnsxesUSmoeZmBvrevexz9qumJTeJt7WlP8p4+GYgw/vNOzu5UlvyFDTfrFpWxhqen2blLyba2xlUGzeyZnzJJQxlX76mJtbUOzpY666tZe10yxa2OhQV8b3mzGGwb9nC5506Fczl9lPMurr4mB901tt75ufzTe7TTW8vz3VZGX8OHOD9JSV8bNcuXuDMn8+Wh7VrGcgDA2wBufNODiJsaWE4L1oU9OdfdlnQdeO/s7o6nruiIp7/EyeCXeLWr+d3kNxUHovxvE60qXw67T0gUyqtQHfO/RbAb5Nu7wXwmWwVSmTaqqzk6mMvvXR2k3jydp1jjTIePor9qqtS95sP/+Ocao319et5IfHQQwyV0tJgwZaXX+Z2nwBrmxs28LmPP85QOXyY77N/PwPYD247fTpoZvdT1i64ICg7wOf46W0ziZ+P7leoKyjgjxk/k3N8Tn09b/uV+fyWsU1NvIDau5fPffxxnquDB3nRt3Ytn+8vvpKn1znHPvm1a8/solFTuWRAuku/PokUu64559ZlvEQi091IfZWDg+mNMk41it2PfE5nQNPwZlWAW6VGImwO9ruQ+f7ZXbv4nL/+62Bv8/POYw38+HFOcfPNs36Am1+vvKSEYRWPs6/dOdZiCwtZvhmw/fJZfPP7qVMM7p6eYECeD/qiIl7szJ3L5/T28qKnoQH41a+CJWf37w9q0xUVnPlw+eXBQj3JF1+XXjryWAs1lUsGpNvk/mdJv5cCeB+0sIzks1R/gKPR9EYZp7Pq21i1tOSae2cnj9fYGPTJlpVxcZnWVvb9FhaA5jl2AAAgAElEQVQyoJYuZQitXs0Q6+xk83tVFW/v38+g8rVufzHiR6RHIsEI+IGB6dmkni7nghq5r4X71eh8zbqxkd0WL77IGQ1veQubzJ94IugyiUT4+nPP5cXBiRPBgMXxBLWaymWS0m1y3zTsrmfN7LcpnyySL0ZrDk8Vyj4AfPP88OBftow/qf74jzTQLnlUfCzG8HnhBdYou7pY2ywvZ83y6FEOYvMj3P067XPnMpT86Hcfar6W7o/vy+yXX43FgoFv6c4dn278uS8tPbNlo6uLjx89ygFxZvwXYLdFdzdvd3UFXR9mPI/19Wd/X5NdElYkDek2uc9KulkA4DIAc7NSIpGZbKQa2fA+81WrOOAsVfAPD+yRFiMZaVT80qXBFLHdu9ls3tnJ0G1pYdPw8eOcqrZ1KwO9o4P319XxOKP1i/uuBW+mhjkQXJj4+fTJn6uggLXtH/6Qzejt7Wz58DvLmXHg3L59fH5XF/Cnf8p/hw+MbGwc+XsUyRBzafSBmdk+sA/dwKb2fQC+7Jx7JrvFo+bmZrdx48apeCuRzItG2ced3Gfe28s12H3Te6raWmtrMNCtvJzN5OXlwWvvvz84Znc3Q/o97+EgLb9F6KOP8rGjRxlA8Tjw1rdy6pnvK961i2G1cyfXMf+7vwsWY/Fzt/2+5n5KWhj5OfTD76utDVoxliwJlnmtquLz6+q4mctHPsKpgsnfS18fA7+wMFgTPvl7HO+yrKrh5w0z2+Scax7Pa9Jtcl86sSKJyIh95oODDItU/MA5P4e5ooKBcN11fN2JE8Ex29v5mF/NLh5nzTEWY9A8+SRDx+9i5sN7xQq2FDgXDAzbvZv3b9sWbD0KMDz8yPmZOBAuHak+lx/xXlLC7+DoUZ7LxYs5dc+fo3icF1LRKJ9/zjk8/93dnKPu57XX1Jz5PY60LGuq4NbGKzKGdJvcIwD+E4BrE3c9BeB7zrnYiC8SyXdj9ZmPNs/YXwQ0NgZ92X7Uut//22/5+dJLDJTaWtYQn3qK4bNjB4M6EmHz+tGjXNrVl+211zjVyi+eMjDARWSOHmX419QEzfVmvD+fDQzwPPjpaTU1XITm8GE+9t73MnwffpgXRaWlvHgqKmKgnzjB1pClS4PvMdV/A6mCu7FRG6/ImNId5f5dABEAf5e4fXvivv+YjUKJzHjj6TNPJdUc5r4+3l6/nn/Yb7iB06Q2beLzlyYa0pYt45rr7e18j4ULgz3Ni4t53CuuCHZfGxzkymRDQ6xZ+gDv7w8WsfGDxPKN31DGL3fb2xu0jPT2BnP1y8q4fe3Klaw1x2K8mNq/n+uy+7nvBw7w92XLUoexb5nx6+jH47w92qYvCnRJSDfQL3fOrUq6/YSZbclGgURmvFTzzLdsGbvPPFk6c5hraviHf8UK/h6NAr/7HZvbjx7l+5eU8PmHDvHCoKaG4XP6NMvxrnexnx1gTfPgQb6fH7Hd0MAwP3WKx4lE+Hu+8Ju6eH4b1l27GLZNTfz38GFeAPmd7QCGejTK52zYwEBfuJDfa20ta93D+dUGDx8OLgbnz+dj2nhFxpBuoA+Z2Zucc28AQGLp1xk8AVWmRL4O4JlIn3kqI42Y9yPfH32UTeRdXVyK1Iy17KuvZmiXlHCg3MGDHIx1002suftdz3wYXXgh51UfO8bABljWeJy10MpKPpZqjfR8kGpanq8Zt7cH3RKxGLs55s7l82MxXvxs2sQLrzlzOI99zhwGdqradVERm/Fratit0t3N2+XlWk1OxpRuoP8XAE+a2V5wpPtiAHdkrVQy8+XzAJ5MbmM5fK578sj3khKGzRtvMMAXLmQAv/46a/KvvRYsNDNvHgdnveMdQV/5G28wWFpa+LqCgqCf2K89XlHBXdjq67lZi98iNF/43ds8/7uf7haPB4MWlyzhcrAdHbyvtJTN9Dt28EKuu5sXT379+6IUf34HB9ls39rKZv6CAt4eHNRqcjKmMQPdzAoAnAawHMAKMNBfd84NjPpCyV+pmpzzaQBPtrax9Oc1eYnXvXuD0BkY4HrrmzezP72lJQiesjLW6v1WoI2NPE5NDfDLX7Jsvt8c4LH90qZ+cZWGBg7smsnzziciefS7/905fq9VVbzYqajguTt1iqPgm5r4nFde4bm84gpeQP3gB+wiOe884N57OR4i+UK3vJzfzYIFwep8AwPBxaBWk5NRjBnozrm4mf2Nc+4qAK9OQZlkphupyTmfBvBksjbluy6i0bOXeK2rY2D49dV37+bjJSVB83hhIW9XVzNk9uzh7wBr3X7vcr8Eql8prqqKTe3xOGuLhw7lX5iPxnddtLTwfDY1cZrg4cOspfulchcu5AWVH/VeXMzfjx7lwLnPfObMRYWGb76jpnVJU7pN7o+Z2fsA/NylsxKN5LdMNjnPZJmoTSV3XQAMgNraYInXvj72yUYirMkBXNWsoQF4//vZ/P766wzj1avZ137wIMP++utZg6yo4Ov9uuaRCI+5d2+wuIzfplUCJSWslVdX8/zU17PZ/dxz+d96YSHHOTQ1cXEZP6OgoYHPb2gIZj7Mnh0cV03rMkHpBvqfAqgAMGhm/WCzu3POVWetZDJzZavJOV8kz18f3nXR08NzCnDk+6pVHHTV0BAs/Xr4MI+xYgVrh75vvK2NTed+FPaDD7LGffXVrNn7gXR+k5GiIgaWX69c1/Jn8ruz+a1WL7uMXRK/+x1w6638zj73Oa4cd+wYz21TE89zf//oF0hqWpcJSHeluKrEeu7Lwd3WREanWsbEJNfIo1H2yfraW3k5A/uWW3g+fYvHtm38d86cM7dU9b9ffDEH0Z0+zVBfuDBoEo5Geawrr+Tzzz+fNfevfCXY73toKNhVrbR05u1/ni2xGM+dX2nv6acZ3KWlwKuvAtdcwyb2b3yDfecNDfweDh3i+Tx9mhdk45n5IDKKdFeK+48A7gKwAMArAFYDeA7A27JXNJnxVMsYn+GDCbu6WPteuJDNur7WXlt75nkd3hqyfj3v9/eVlABf+AIXodm9mz/xON/jnHN48bBiBW+vXQssWsTa/4svcqDc6dNsPo7FglXrhPxYg5ISjmOor+fvbW3c+Ob88xnwH/kIv48VK3i+zzmH88vXr9f/I5Ix6Ta53wXgcgAvOOfWmtlKAF/KXrFE8tDwwYQ1NZyydPIkw2KkrouRWkP8fR0dnH++Zw9bAHytu7eXzcWXXsqlYdvaWMv0K9tFo0EfelFRMFiuoYE1z3zn56fX1/PclZbyZ/ZsnqeenuDCqbaW30dnZ9AqMvzCLBNOnuQgvQULgmV+JW+kG+j9zrl+M4OZlTjnXjezFVktmUi+STWYsLExvRXmfGtINMrQSH7u/fezdl1fz+Devj0IaYCD5r72NdYim5oY1r/8JR/v7Aw2H/GroMVi4d6kZSw+yP3udb6lo72da7wfO8YLpVmzzqyBt7Vld22Ghx9mV4m/YPiLvwBuvDFzx5dpL91AbzGzWgD/BuBxM+sAcDh7xRLJQyMNJqysTO/1qRbz6eoCfvtbHnv/fobRrFlsRvdLlTY1scb+3HP898ABLm1aXs73Hhzkc2tq+Lj60Hl+h4Z4Pg8c4MyCujqem1mzOCbhlluCwB5tbQZg8mNNTp5kmFdV8Xvq6uLt1atVU88j6Q6KuyXx61+a2ZMAagA8krVSieSriQ4mTBUYDz3EWmNrK/t1neMgOz/grqCAARCP8+eZZxhIAwP8t6CAz/PP8SEGBJuW5Nu89KKiM+fq19fz/Jw8Cdx2G3dcO3GCUwo3bOAYiBtu4NTAVGsz7NnDjXcmW2tvaeF3VVPD2zU1LEdLiwI9j4x7hItz7rfOuQedc3m4qLPIFJhI/+rw/vdIhE2/r7zCgVl+P/N4nKvI+elsbW18Xm8vfwYG+Bwf/oWFwSYvAwNsbgeCzVvyjV/61Y8l8F0ksRhr5bNnAxs38vubP5/nzu9r77e79WMXAI5ZqKw887kTWS9/wQJ+x35XvK4u3l6wIDOfW2YEDVkVCYPk/vf2di5o4vc7Hxri3PJzzglWLrv+emD5ctYwGxs5N7qqiscqKGAYJDe1v/vdHOTlR7nH48FCNvmkrIw/RUXBmvdz53JswvLlwJEjPGfJNfHBQf6sWgU8/jhbTh5/nN+Jf47/t7+fxxhvqM+axT7znh5ON+zp4W3VzvNKun3oIjKWXO4u5/vf/cYt5eUM7YcfBp56inPU43GGd20t+9BnzWJT8Lx5rIG/9BKbaX0tdNYsTpdbu5aPlZQwzDo6gqb2wkJeJPjQCrtTp3jhs2gRWzec4/n60IeABx5g7XvDBl5ELV0aTDUsKuKqcG9/e7BG+65dPKb/b+bQIZ5ngKPlx9v8fuON7DPXKPe8pUAXyYRc7S6XfBHR1MSBWH19XBEOYO26sZFLkhYVMXBWruTI9u5u1jCvvZZNxseOsXYfizHIZ83isbZu5cYiRUUc+DU0FKz/XlQU9MuHnVnQCtLTw89fW8vvevdufgdvvMGLnXvvZcAuWMDH/QWPXyQoFuMFwdq17Gdvb2eYX3klXzPRDY1mzVKQ5zEFushk5Wp3uVQXEY2NDGO/tWdREVeKW72atXFfKzz/fDYVRyJ8XkkJ8PWvAz/+MQdq+Y1eIhEes7iYtXrfpFxQwJr9oUP5Neq9rIwXNF1dvFj6wz8M1myfPTu4aCooYJfHBz/ICy2/XWpfH3/8Ovzl5cA73xl0Zfg+73zc0EgmTYEuMlm52F1utIsIP/Wtv5/BftFFDJPTp8+e1+7L70dvf+xjXIRm924O7urt5YpxJ08GS536ec7d3cGod9+MHGZ+pkBZGS+Ompt5Dn79a9bMCwsZ3nV1HDBXXc1z2dTEc7tuHW/7LpG3vY3HeuIJfielpZnd0CiXXUCSEwp0kcnK5u5yI/1RHu0iInnq26WXAt/7HsM4EgE+9amz57UPX/DkTW8CfvYz4Phxvr8fme0cLwqAYI/12lq+PuxhDvACKR4PtkMtK2OYl5QAy5ax73rXLp7zJUuCUef33MPfi4rYUuK7RPwsgc5OnvtMbmiUqy4gySkFushkZWt3udH+KI91EeHfe/v2MwdibdnCPbv948Nr+vv3A5//PEM6EmFw9/czzJ0LmoZ9s3NFRX6sGuc/oxnPZWMjuyy2b+e5W7GCof6rXwUXX8uXs3/87W9n87xvai8vDy6wkr83vzzsZGvVueoCkpxToItkQqZ3lxvrj3JxMZtwH3qIIetHRaeqxSfvtX348JldAck1/ViM/b5mrH1GowzukpJgIJwP9YKCMxeoyYcaekFBMJ5g1Srg93+fU8Rqa3ku9u7lNMDCwuCcr1wZLPbiW1HWrOFiMqku/jKxoVEuuoBkWshZoJtZIYCNAFqdc+tzVQ6RjMnk7nJj/VFubWXfq19XfN26s5tUk2vxkQinm/n7Uz0nHud7VlQEzcVDQwz4ykq+f0EBw97PUfc19tJSvj55/nRZWdBEHwY1Newfb28HfvITfv6+Pg4i7Opi2P+H/8Aw7+gAPvxhTmUb3oqybBl/stW/nc0uIJnWcrmwzF0AduTw/UWmr+Q/ysCZf5STa++LFjFAnnji7MVIfFfAoUMMlqee4nSrtrazn9Pby4Fv8TjwjnfwuJEI36+ujqOvKyp4n1/ytbSUzc7z5wcD5iorg/7igoJgTvtMVVLCGvi8eRzoFovxPDc2sjl96VLOIrjsMuCSS9h3XlPDc1BQEJzbw4f5r6+NZ2u3NeDM73T4+0qo5aSGbmYLALwbwFcA/GkuyiAyrY3WL+8HUaXTpOpXgbvySv4OnN2fmtxdcOONvDgoKeFI96uv5iCwkydZhqYmNi0fOMD39VPeKiqCPdujUb6nHxHuL0pmGjMuIlNWxnDu7w+2SPXT13xrRSTC23553NH6xdMZfT7ZEeqZ7gKSGSFXTe5/C+DPAVSN9AQzuxPAnQCwaNGiKSqWyDQy0h/l8TSp7tnDgVhVVXzOJZcwgIaHf3Kt8fbbOd/89GmOdO/pYXDdcAP7iIuKGFyLFzPEW1vZxDx/PkfE+5Hvfo/1+vpgwZrBwaCbYLryAwh9n/nAAFfQq6kJlnv1U/dOn+btJUuAbdt44TN8PENyV0w6o88zNUI9k11AMiNMeaCb2XoAx51zm8zsupGe55y7G8DdANDc3BzyIbQiI0j1RzndUfXRKDf/8NugAgz3Sy8dvT+1uJhNzPPmsUnZB3B/P5eTLSwE3vIWLqbS0cGQ+73fCxajicXYpzwwwMcWLOBr6+rY/F9QwNXlpqOCgjNH7Le0BPvGv/nNbL4+fZqBf8klbKnwn/eii7jy27JlqYM0ndHnGqEuk5CLGvrVAG4ysxsBlAKoNrMfOec+koOyiMxM6TSp+qbuq65if288zvvWrBk7HJIvGlKtRFdZCdx5J2vwvj+9spKhtmEDA7u8nKO8X32VxzNjqB8/HvStT6fpbskbz/hlXufO5UVTNMqLlSVLuLTq6tUM729+k5/b70P+/PMMdC+56Tyd0ecaoS6TMOWB7pz7PIDPA0Cihv5nCnORCRirSdU3zZeVMXw6OhgWyYEzmpEuGpKDvryctwHgvvs4z92MA+vq6oJ52wcOsGY7fz5Dzo8DyAVfC/ctD/6iwm8RW1DAz+UccPQoa+MrV7Lcp06xC+KGGzjIsKyMLRkDA5yTvmwZuyvmzQsW7PEL0rzznWN3lWiEukyC5qGLhFWqWvb69eOr6aW6aEgV9NEoa6nXXccgj8UY2u9+N0MvHmd/fjTKAWR+GdWpmr+evPhNYWEwqM0Hpu8b9xvOVFXxc5w+zXPnP9Py5Rwk6FywD3p3N1stDh1iX7vfXa2nh+dn504+d9Mm4BOfYPCP1FWSrUWKJC/kNNCdc08BeCqXZRAJtWyNdh4e9L553y+oEokwkCorgZtvBjZvZkiWljIUX3yR4ecXsOnrC+a8Fxby31gsvXL4mrafRz+cP6afcudHog8NBVPx/Mh0H+ZVVcFqbtXVbFm48kq+x6ZNwA9+APziFyz/wABr5l1d3Ixl0SJ2K7z8MgcIVlSwOf7oUa6Pf8cdQetGqu9DI9RlglRDFwm7dEY7D58mNd5pU6maiv1xm5qAL38Z+NGPuG0rwIF50ShHhe/bx/D0teh4PFjn3A9ISxXUfpMUv/HMSHuyRyIM/eLi4Li1taxZFxXxIqOqigFfU8NjxeP8/eab+bvfknTTJg6O++UvOZ7g0CG+fts24L3vZR87EGwz293N1w4M8Hl+hb3a2tHPp0aoywQo0EXy3fBpUqtWsS98PNOmhjcV9/Tw/gce4DHWrWNzs5/udeoUcP/9wE9/yttVVQxS39/sF8mprAymuw0Xj3Oxl6EhHs+5s3d98/3lfmOZnh7ed/Ikm/2rq4Pa+PHjLOvChaxNDwzw/W+9leF95AiP6fvaBwYY1osWsbXh0KFgW9VYjAvObNvGY5WWcsva4mL1h0vWKNBF8tnwaVLd3cC3v80NRWbPHt+0Kd9U3NnJIK+t5TFbWoAvfQm4/PJgjnZTE6e6dXUFK9idOMGQ9E3fpaUMxro6BnIsxjCuqODzkud5NzQEA9iAoE/cLHi+b8ovLuaFg6/V+xaJgQFeGBw9GrQW+GZ/P5WvtJRl9vPzS0t5nPJyjvDv6OAxiop4LqLRYL19f9GjmrdkiQJdJJ8NnyZVUMAQKyzk7fJyLgrjR26nM90teQGcWIyDwIqKWJstKOBo+P5+4JVX2Nw+OMha8vHjbEafNQu49loG5htvcIGaV1/l87q6eFw/sK63l7/7FoHkGvrQEN93aIjPKS4O1ln3S+gODAQ16LIy/uun2M2axRp88gXNunW8ODn3XDa/m3F51Q99iP3syXvN+/PwiU+oP1ymhAJdJJ8N7/v2/dc+FFtagJde4u9FRZzDPtLCKamO6RekKS9nAAPAs88yvMvLWXtuaWGQLlvGMKypYY07EuH9b30ra/ff/z6PW1HBtdP7+znIzM8XP3WK4VpXx7Du7WXAX3QRm8OPHQua9P2a82Vlwec9fZrHisd54bFgAcvR1hbMA6+rY1lmzeIiOxs38rHaWta+h+81D6g/XKaMAl0kn6WaJvWpT7EP/eBBhvkVVzB8n38eeOYZ4JprOP1tpH715GP297OWfsklDM4jRxjgNTX86e9naM6dy+DduZOBevw4+/Kvu45lmTUL+NjH+Pr2dgbuxRezDC+/zGA+eJAXAIODQfN6aWnQhF9Rwdf19vJn2TI+7j+3X/imqIifYe3aM9dlB/ivH9zmF5k5eZIj11OFucgUUqCL5LtU06Te/OZgENjcucCTTzKA/fSvsfrVU2344vdib2piiO7bx/Du7WVt99OfPnOEvN+N7M1vDsoG8LUAb997L2vWfue3Bx7g7yUlwTS28nIG8KpVHJjW1cV932fPZpO+398dYK384ouDeerDdypLdQH03vcqzGVaUKCLyNnNwsmDwDo6gs1UioqCzVfGWo50+IYvfhGXgQEGr98rvb6etWcf4MPDcXjZ/Fx3gC0FPT2sxff3s5bf2spArqwMxgLU1HAzmVmzWIYLLmCfd18f+8mjUX7eVauAj36UZRmp31vzxGWaUqCLSGq+NvrQQxz9PjjINcyHN0OneywffDfdxBpyJMLAXb06WGd+tO1Efc08eR/xpibgM5/hwLvHH+c68s4FW7729HDQ2wUXsDm/s5Ov/8M/5Gv6+liGyko25ZeWBsdP52JFZBpRoIvMRJPdLztdTU3sH16zhju3RaMM38lMv1q2jKPYIxGGZyzGpu2RLhBaW4F/+Rf2lRcWclGa2247sw9/0ybOB6+rA374Qz538WL2i/f3c1DcueeySf2zn2WA793LNdq3buX7O3fm0rhTdY5FMkSBLjLTZGq/7HQVF7PvedmyzARccTGD85FH2HQ/2nrl0SjD/Fe/Yr94QQHDuqoK+PCHeQ6i0aCvvLwceN/7GOr19QzypUt5IVBdzYFzfu75qlWcc9/fz+N++tN8LjD151gkAxToIjNJLvfLzmQzc7r90MePc933SCTYIOXwYQb1PfcEr+vpCY5VX88LkFWr2NweibD27Zd9BXget2zhrnB+Lfjt2zkdDtCe5DIjFeS6ACIyDqn2yx4cDEaGzyR+wNxIIdnayqVhDx1iMHd38/5YjHuT19VxYJtfF72zk2E/MADcdRdr8Q0NfF19Peepr1oVDHgbHAz2dq+uDs5jmM6x5BXV0EVmknzZL9u3RMyeDVx4IbBrF6e51ddzXfWLLuLIdYCfvaoKuOWWYK10P91tzRpOuYtGeX9yH/lo5zEfzrGEjgJdZCbJl/2yk2vQ117LaWhHjwLNzdws5emnzw7c4bX90fr+xzqP+XCOJXTMOZfrMoypubnZbdy4MdfFEJk+Mj0Ce7qN6PaD4ZI3jUlekS1Tg9ZG+9zT7ZxIXjGzTc655vG8RjV0kZkokwPUpuOI7rFWZMvU4i6jnUfNNZcZRoEuks9yOWp+LCOFdnLN2Q+IExEFukheSzWiu7Nz7GVdp8rwWvJ0bE0QmSY0bU0knyWPmgeyM6LbL9sajU7+OL41Yf58/vvII5M/rkhIqIYuks+yPWo+kzXq6d6aIJJjCnSRfJet3cMm0j8/2sjyfJmDLzJBCnQRyc6Ibl+jjkS4+UlJSbDiWqr3Gqs2P5HWBE09kzyiQBeR7Cgv5xrrmzdzc5ShIWD58tQ16nRr8+NpTdAAOskzGhQnItlnNvrj41k/faw14AFeIDz0EH9vbNQAOskLqqGLSHb09XGN9euv54YpJSXcLjVVk3um+8f37AGeeYZLxxYUcBe10Zr7RUJANXQRmbxUU9N8SMdirCHHYiOHtO8f7+3ljmm9vRMfbR+Ncq338nKgooLrwD//fFAmkZBSDV1EJmekvurxDmLL1Gh730y/ejXw8stBzXzNGtXOJdQU6CIycWMNZhtvSGditL1vGSgvB667jhcTsRh3XRMJMTW5i8jEpTOYLZ1BbJmU3Hzf1gY4d+Y+6CIhpRq6iEzcdF3sJVuL5YhMY6qhi8jEZXIwWzbKNpUtAyI5phq6iEyOasMi04ICXUQmLxtLx4rIuKjJXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAlMe6Ga20MyeNLMdZrbNzO6a6jKIiIiETVEO3nMQwGedc5vNrArAJjN73Dm3PQdlERERCYUpr6E754445zYnfu8BsANA01SXQ0REJExy2oduZksAXAJgQy7LISIiMtPlLNDNrBLAvwL4Y+dcd4rH7zSzjWa2sa2tbeoLKCIiMoPkJNDNLAKG+b3OuZ+neo5z7m7nXLNzrrmxsXFqCygiIjLD5GKUuwH4JwA7nHPfmOr3FxERCaNc1NCvBnA7gHVm9kri58YclENERCQ0pnzamnPuGQA21e8rIiISZlopTkREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIkrMJ2IAAAYQSURBVCISAgp0ERGREFCgi4iIhIACXUREJAQU6CIiIiGgQBcREQkBBbqIiEgIKNBFRERCQIEuIiISAgp0ERGREFCgi4iIhEBOAt3MbjCznWa2x8w+l4syiIiIhMmUB7qZFQL4DoB3ATgfwG1mdv5Ul0NERCRMclFDvwLAHufcXudcFMBPALwnB+UQEREJjVwEehOAQ0m3WxL3iYiIyAQV5eA9LcV97qwnmd0J4M7EzQEz25rVUkkDgPZcFyLkdI6nhs5z9ukcZ9+K8b4gF4HeAmBh0u0FAA4Pf5Jz7m4AdwOAmW10zjVPTfHyk85x9ukcTw2d5+zTOc4+M9s43tfkosn9JQDLzWypmRUD+CCAB3NQDhERkdCY8hq6c27QzD4F4FEAhQC+75zbNtXlEBERCZNcNLnDOfcwgIfH8ZK7s1UW+Xc6x9mnczw1dJ6zT+c4+8Z9js25s8ajiYiIyAyjpV9FRERCYFoHupaIzT4zW2hmT5rZDjPbZmZ35bpMYWVmhWb2spk9lOuyhJGZ1ZrZz8zs9cR/z1flukxhY2Z/kvg7sdXM7jOz0lyXKQzM7Ptmdjx5eraZzTKzx81sd+LfurGOM20DXUvETplBAJ91zp0HYDWA/6zznDV3AdiR60KE2LcAPOKcWwlgFXSuM8rMmgB8BkCzc+5CcFDzB3NbqtD4ZwA3DLvvcwB+45xbDuA3idujmraBDi0ROyWcc0ecc5sTv/eAfwS1cl+GmdkCAO8G8I+5LksYmVk1gGsB/BMAOOeizrnO3JYqlIoAlJlZEYBypFhDRMbPOfc7ACeH3f0eAD9I/P4DADePdZzpHOhaInaKmdkSAJcA2JDbkoTS3wL4cwDxXBckpM4B0AbgnkS3xj+aWUWuCxUmzrlWAF8HcBDAEQBdzrnHcluqUJvjnDsCsOIFYPZYL5jOgZ7WErGSGWZWCeBfAfyxc6471+UJEzNbD+C4c25TrssSYkUALgXwXefcJQBOIY0mSklfog/3PQCWApgPoMLMPpLbUkmy6RzoaS0RK5NnZhEwzO91zv081+UJoasB3GRm+8Guo3Vm9qPcFil0WgC0OOd869LPwICXzHk7gH3OuTbnXAzAzwG8JcdlCrNjZjYPABL/Hh/rBdM50LVE7BQwMwP7HXc4576R6/KEkXPu8865Bc65JeB/x08451SzySDn3FEAh8zMb2jxNgDbc1ikMDoIYLWZlSf+brwNGniYTQ8C+Gji948C+MVYL8jJSnHp0BKxU+ZqALcDeM3MXknc94XEan4iM8mnAdybqADsBXBHjssTKs65DWb2MwCbwdkxL0MrxmWEmd0H4DoADWbWAuCLAP4KwP81s4+DF1PvH/M4WilORERk5pvOTe4iIiKSJgW6iIhICCjQRUREQkCBLiIiEgIKdBERkRBQoIvkOTPrzXUZRGTyFOgiIiIhoEAXCRkz+5qZfTLp9l+a2RfN7DdmttnMXjOzlDsXmtl/MbOXzOxVM/tS4r4lif3F/yGxF/ZjZlaWeGyZmf3azLYkjv2mkY4jItmlQBcJn58A+EDS7d8HcA+AW5xzlwJYC+BvEst3/jszeyeA5eDWxRcDuMzMrk08vBzAd5xzFwDoBPC+xP33Ju5fBa7rfWSM44hIlkzbpV9FZGKccy+b2Wwzmw+gEUAHuN3lNxPBGge3Ip4D4GjSS9+Z+Hk5cbsSDOaD4KYcfmngTQCWmFkVgCbn3AOJ9+0H/v3CINVxfpeFjysiCQp0kXD6GYBbAcwFa+wfBsP9MudcLLHzW+mw1xiArzrnvnfGnWZLAAwk3TUEoAyptzge8Tgikl1qchcJp5+AO7vdCoZ7Dbgne8zM1gJYnOI1jwL4AzOrBAAzazKz2SO9gXOuG0CLmd2ceH6JmZWP9zgikhmqoYuEkHNuW6JJvNU5d8TM7gXw/8xsI4BXALye4jWPmdl5AJ5PdK/3AvgIWCMfye0AvmdmXwYQA/D+UY4z5n7OIjJx2m1NREQkBNTkLiIiEgIKdBERkRBQoIuIiISAAl1ERCQEFOgiIiIhoEAXEREJAQW6iIhICCjQRUREQuD/A8cvkyXq624RAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "i=1\n",
    "fig, ax = plt.subplots()\n",
    "df[df['stars']==i].plot.scatter(x='valence',y='arousal', color='Red',\n",
    "                                ax=ax, alpha=0.3,title='Sentiment vs. Star '+str(i),\n",
    "                                figsize=(8, 6))\n",
    "plt.xlim(0,10)\n",
    "plt.ylim(0,10)\n",
    "plt.savefig('sentiment'+str(i)+'.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:30.019798Z",
     "start_time": "2018-11-12T15:32:25.713417Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "\n",
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(111, projection='3d')\n",
    "\n",
    "for c, m, zlow, zhigh in [('r', 'o', -50, -25), ('b', '^', -30, -5)]:\n",
    "    xs = df['valence']\n",
    "    ys = df['arousal']\n",
    "    zs = df['stars']\n",
    "    ax.scatter(xs, ys, zs, c=df['stars'], marker=m)\n",
    "\n",
    "ax.set_xlabel('Valence')\n",
    "ax.set_ylabel('Arousal')\n",
    "ax.set_zlabel('Stars')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:30.027896Z",
     "start_time": "2018-11-12T15:32:30.022315Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['valence', 'arousal', 'business_id', 'name', 'biz_stars', 'state',\n",
       "       'city', 'neighborhood', 'address', 'postal_code', 'review_id', 'text',\n",
       "       'date', 'cool', 'funny', 'useful', 'stars', 'user_id', 'type',\n",
       "       'puncount', 'lan', 'reviewsen', 'reviewsenlen', 'reviewlen',\n",
       "       'reviewword', 'reviewwordlen', 'avgword'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:30.072785Z",
     "start_time": "2018-11-12T15:32:30.030406Z"
    }
   },
   "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>valence</th>\n",
       "      <th>arousal</th>\n",
       "      <th>business_id</th>\n",
       "      <th>name</th>\n",
       "      <th>biz_stars</th>\n",
       "      <th>state</th>\n",
       "      <th>city</th>\n",
       "      <th>neighborhood</th>\n",
       "      <th>address</th>\n",
       "      <th>postal_code</th>\n",
       "      <th>...</th>\n",
       "      <th>user_id</th>\n",
       "      <th>type</th>\n",
       "      <th>puncount</th>\n",
       "      <th>lan</th>\n",
       "      <th>reviewsen</th>\n",
       "      <th>reviewsenlen</th>\n",
       "      <th>reviewlen</th>\n",
       "      <th>reviewword</th>\n",
       "      <th>reviewwordlen</th>\n",
       "      <th>avgword</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11717</th>\n",
       "      <td>2.08</td>\n",
       "      <td>7.49</td>\n",
       "      <td>YQ--LJ7pvjiDSqNv0TuKTQ</td>\n",
       "      <td>Grimaldi's Pizzeria</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NV</td>\n",
       "      <td>Las Vegas</td>\n",
       "      <td>The Strip</td>\n",
       "      <td>3327 Las Vegas Blvd S, Ste 2710</td>\n",
       "      <td>89109</td>\n",
       "      <td>...</td>\n",
       "      <td>njOy2Wo3v1ioDgvu6nR9oQ</td>\n",
       "      <td>Dine-in</td>\n",
       "      <td>0.059322</td>\n",
       "      <td>de</td>\n",
       "      <td>[Sehr gute Pizza zum selbst zusammenstellen., ...</td>\n",
       "      <td>2</td>\n",
       "      <td>118</td>\n",
       "      <td>[Sehr, gute, Pizza, zum, selbst, zusammenstell...</td>\n",
       "      <td>24</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       valence  arousal             business_id                 name  \\\n",
       "11717     2.08     7.49  YQ--LJ7pvjiDSqNv0TuKTQ  Grimaldi's Pizzeria   \n",
       "\n",
       "       biz_stars state       city neighborhood  \\\n",
       "11717        4.0    NV  Las Vegas    The Strip   \n",
       "\n",
       "                               address postal_code   ...     \\\n",
       "11717  3327 Las Vegas Blvd S, Ste 2710       89109   ...      \n",
       "\n",
       "                      user_id     type  puncount  lan  \\\n",
       "11717  njOy2Wo3v1ioDgvu6nR9oQ  Dine-in  0.059322   de   \n",
       "\n",
       "                                               reviewsen  reviewsenlen  \\\n",
       "11717  [Sehr gute Pizza zum selbst zusammenstellen., ...             2   \n",
       "\n",
       "       reviewlen                                         reviewword  \\\n",
       "11717        118  [Sehr, gute, Pizza, zum, selbst, zusammenstell...   \n",
       "\n",
       "      reviewwordlen  avgword  \n",
       "11717            24     12.0  \n",
       "\n",
       "[1 rows x 27 columns]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[(df['valence']<3) & (df['stars']>4) & (df['arousal']>7)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### NLTK Sentiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:30.078928Z",
     "start_time": "2018-11-12T15:32:30.075585Z"
    }
   },
   "outputs": [],
   "source": [
    "from nltk.corpus import sentiwordnet as swn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:30.085758Z",
     "start_time": "2018-11-12T15:32:30.081794Z"
    }
   },
   "outputs": [],
   "source": [
    "def normalize_accented_characters(text):\n",
    "    '''\n",
    "    removes accents from characters and replaces with non-accented equivalent\n",
    "    '''\n",
    "    if type(text) != str:\n",
    "        text = (unicodedata.normalize('NFKD', text.decode('utf-8'))\n",
    "                           .encode('ascii', 'ignore'))\n",
    "    return text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:32:30.101840Z",
     "start_time": "2018-11-12T15:32:30.088328Z"
    }
   },
   "outputs": [],
   "source": [
    "def analyze_sentiment_sentiwordnet_lexicon(review, verbose=False):\n",
    "    '''\n",
    "    analyzes sentiment of text using \n",
    "    '''\n",
    "    # pre-process text\n",
    "    review = normalize_accented_characters(review)\n",
    "    \n",
    "    # tokenize and POS tag text tokens\n",
    "    text_tokens = nltk.word_tokenize(review)\n",
    "    tagged_text = nltk.pos_tag(text_tokens)\n",
    "    pos_score = neg_score = token_count = obj_score = 0\n",
    "    \n",
    "    # get wordnet synsets based on POS tags for NN, VB, JJ, RB\n",
    "    # get sentiment scores if synsets are found\n",
    "    for word, tag in tagged_text:  \n",
    "        ss_set = None\n",
    "        if 'NN' in tag and list(swn.senti_synsets(word, 'n')): # singular noun\n",
    "            ss_set = list(swn.senti_synsets(word, 'n'))[0]\n",
    "        elif 'VB' in tag and list(swn.senti_synsets(word, 'v')): # Verb\n",
    "            ss_set = list(swn.senti_synsets(word, 'v'))[0]\n",
    "        elif 'JJ' in tag and list(swn.senti_synsets(word, 'a')): # Adjective\n",
    "            ss_set = list(swn.senti_synsets(word, 'a'))[0]\n",
    "        elif 'RB' in tag and list(swn.senti_synsets(word, 'r')): # adverb\n",
    "            ss_set = list(swn.senti_synsets(word, 'r'))[0]\n",
    "        \n",
    "        # if senti-synset is found\n",
    "        if ss_set:\n",
    "            # add scores for all found synsets\n",
    "            pos_score += ss_set.pos_score()\n",
    "            neg_score += ss_set.neg_score()\n",
    "            obj_score += ss_set.obj_score()\n",
    "            token_count += 1\n",
    "        \n",
    "        # if no words have synset, return None \n",
    "        if token_count == 0:\n",
    "            return None\n",
    "        \n",
    "    # aggregate final scores\n",
    "    final_score = pos_score - neg_score\n",
    "    norm_final_score = round(float(final_score) / token_count, 2)\n",
    "    final_sentiment = 'positive' if norm_final_score >= 0 else 'negative'\n",
    "    if verbose:\n",
    "        norm_obj_score = round(float(obj_score) / token_count, 2)\n",
    "        norm_pos_score = round(float(pos_score) / token_count, 2)\n",
    "        norm_neg_score = round(float(neg_score) / token_count, 2)\n",
    "        # to display results in a nice table\n",
    "        sentiment_frame = pd.DataFrame([[final_sentiment, norm_obj_score,\n",
    "                                         norm_pos_score, norm_neg_score,\n",
    "                                         norm_final_score]],\n",
    "                                         columns=pd.MultiIndex(\n",
    "                                             levels=[['SENTIMENT STATS:'],\n",
    "                                                ['Predicted Sentiment',\n",
    "                                                 'Objectivity',\n",
    "                                                 'Positive', 'Negative',\n",
    "                                                 'Overall']],\n",
    "                                              labels=[[0,0,0,0,0],\n",
    "                                                      [0,1,2,3,4]]))\n",
    "        print(sentiment_frame)\n",
    "    return final_sentiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:10.899747Z",
     "start_time": "2018-11-12T15:32:30.104756Z"
    }
   },
   "outputs": [],
   "source": [
    "# takes a while to run\n",
    "df['pred_sentiment'] = df['text'].apply(analyze_sentiment_sentiwordnet_lexicon)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:10.911112Z",
     "start_time": "2018-11-12T15:36:10.902560Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    negative\n",
       "1    negative\n",
       "2    positive\n",
       "3        None\n",
       "4        None\n",
       "Name: pred_sentiment, dtype: object"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['pred_sentiment'].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:10.960902Z",
     "start_time": "2018-11-12T15:36:10.914582Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total non matching: 2\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "9996     Love the environment. Great waiter Nick. Super...\n",
       "11717    Sehr gute Pizza zum selbst zusammenstellen.\\n\\...\n",
       "Name: text, dtype: object"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# how many not matching\n",
    "highstarlowsent = (df.stars == 5) & (df.valence < 3)\n",
    "print('total non matching:',highstarlowsent.sum())\n",
    "df[highstarlowsent].text.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:10.971808Z",
     "start_time": "2018-11-12T15:36:10.963304Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total non matching: 14\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'This restaurant is very dirty. Staff is never pleasant to customers, and the food is over priced.'"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# how many not matching\n",
    "lowstarhighsent = (df.stars <2) & (df.valence > 7)\n",
    "print('total non matching:',lowstarhighsent.sum())\n",
    "df[lowstarhighsent].text.iloc[3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Clean"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Clean Function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:10.984786Z",
     "start_time": "2018-11-12T15:36:10.974084Z"
    }
   },
   "outputs": [],
   "source": [
    "stopwords = nltk.corpus.stopwords.words('english')\n",
    "stopwords.extend(('pizza','pizzas','would','dont','didnt'))\n",
    "string.punctuation\n",
    "nolem = []\n",
    "wn = nltk.WordNetLemmatizer()\n",
    "ps = nltk.PorterStemmer()\n",
    "\n",
    "def clean_text(text, join=False, stem='lemmatize'):\n",
    "    # lower case and remove punctuation, here word is char so need '' to join these char back\n",
    "    text = ''.join([word.lower() for word in text if word not in string.punctuation])\n",
    "    \n",
    "    text=re.sub(r'\\d+', '', text)\n",
    "    \n",
    "    # break down to tokens (single word)\n",
    "    tokens = re.split('\\W+', text)\n",
    "    \n",
    "    # special words no lemmatize, and remove stop words\n",
    "    # elif list comprehension: [if-1 condition_1 else_2 for_statement <condition_2>]\n",
    "    \n",
    "    if stem == 'lemmatize':\n",
    "        if join == True:\n",
    "            text = ' '.join([word if word in nolem else wn.lemmatize(word,'v') for word in tokens if word not in stopwords ])\n",
    "        else:\n",
    "            text = [word if word in nolem else wn.lemmatize(word,'v') for word in tokens if word not in stopwords ]\n",
    "    elif stem == 'stem':\n",
    "        if join == True:\n",
    "            text = ' '.join([word if word in nolem else ps.stem(word) for word in tokens if word not in stopwords ])\n",
    "        else:\n",
    "            text = [word if word in nolem else ps.stem(word) for word in tokens if word not in stopwords ]\n",
    "    #nostop = [word for word in tokens if word not in stopwords]\n",
    "    #text = [wn.lemmatize(word) if word not in nolem else word for word in nostop]\n",
    "    #text=[]\n",
    "    #for word in tokens:\n",
    "    #    if word in nolem:\n",
    "    #        text.append(word)\n",
    "    #    elif word not in stopwords:\n",
    "    #        text.append(wn.lemmatize(word))\n",
    "    return text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:25.007533Z",
     "start_time": "2018-11-12T15:36:10.987240Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 13.9 s, sys: 86.1 ms, total: 14 s\n",
      "Wall time: 14 s\n"
     ]
    }
   ],
   "source": [
    "df['cleaned_text']=df['text'].apply(clean_text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:25.017543Z",
     "start_time": "2018-11-12T15:36:25.010383Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    [displease, come, later, schedule, cold, box, ...\n",
       "1    [order, location, literally, front, apartment,...\n",
       "2    [busy, place, take, good, care, right, away, w...\n",
       "3    [stop, pick, order, wing, way, home, work, go,...\n",
       "4    [night, children, every, week, last, two, time...\n",
       "Name: cleaned_text, dtype: object"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['cleaned_text'].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Supervised Learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:25.666603Z",
     "start_time": "2018-11-12T15:36:25.019878Z"
    }
   },
   "outputs": [],
   "source": [
    "from collections import defaultdict\n",
    "import gensim\n",
    "import polyglot\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.cluster import KMeans\n",
    "ner_categories = defaultdict(int)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:25.675014Z",
     "start_time": "2018-11-12T15:36:25.669041Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "$(document).ready(\n",
       "    function() {\n",
       "        function appendUniqueDiv(){\n",
       "            // append a div with our uuid so we can check that it's already\n",
       "            // been sent and avoid duplicates on page reload\n",
       "            var notifiedDiv = document.createElement(\"div\")\n",
       "            notifiedDiv.id = \"bc50cd6d-a6a0-4efc-9909-534592c84e2d\"\n",
       "            element.append(notifiedDiv)\n",
       "        }\n",
       "\n",
       "        // only send notifications if the pageload is complete; this will\n",
       "        // help stop extra notifications when a saved notebook is loaded,\n",
       "        // which during testing gives us state \"interactive\", not \"complete\"\n",
       "        if (document.readyState === 'complete') {\n",
       "            // check for the div that signifies that the notification\n",
       "            // was already sent\n",
       "            if (document.getElementById(\"bc50cd6d-a6a0-4efc-9909-534592c84e2d\") === null) {\n",
       "                var notificationPayload = {\"requireInteraction\": false, \"icon\": \"/static/base/images/favicon.ico\", \"body\": \"Cell execution has finished!\"};\n",
       "                if (Notification.permission !== 'denied') {\n",
       "                    if (Notification.permission !== 'granted') { \n",
       "                        Notification.requestPermission(function (permission) {\n",
       "                            if(!('permission' in Notification)) {\n",
       "                                Notification.permission = permission\n",
       "                            }\n",
       "                        })\n",
       "                    }\n",
       "                    if (Notification.permission === 'granted') {\n",
       "                    var notification = new Notification(\"Jupyter Notebook\", notificationPayload)\n",
       "                    appendUniqueDiv()\n",
       "                    notification.onclick = function () {\n",
       "                        window.focus();\n",
       "                        this.close();\n",
       "                        };\n",
       "                    } \n",
       "                }     \n",
       "            }\n",
       "        }\n",
       "    }\n",
       ")\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%notify\n",
    "y = df['stars']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:25.687301Z",
     "start_time": "2018-11-12T15:36:25.677429Z"
    }
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(df['text'], y,\n",
    "                                                    test_size=0.3,\n",
    "                                                    random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Count Vectorizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:40.843747Z",
     "start_time": "2018-11-12T15:36:25.689942Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 10.4 s, sys: 75 ms, total: 10.4 s\n",
      "Wall time: 10.5 s\n",
      "CPU times: user 4.57 s, sys: 39.9 ms, total: 4.61 s\n",
      "Wall time: 4.64 s\n"
     ]
    }
   ],
   "source": [
    "count_vectorizer = CountVectorizer(analyzer=clean_text)\n",
    "count_train = count_vectorizer.fit_transform(X_train)\n",
    "count_test = count_vectorizer.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:40.872606Z",
     "start_time": "2018-11-12T15:36:40.846787Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['', 'aa', 'aaaaaaaaye', 'aaaaaahhhh', 'aaaaand', 'aaaahhh', 'aaawwwwesucks', 'aaron', 'aback', 'abandon']\n"
     ]
    }
   ],
   "source": [
    "print(count_vectorizer.get_feature_names()[:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TFIDF Vectorizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.237079Z",
     "start_time": "2018-11-12T15:36:40.876002Z"
    }
   },
   "outputs": [],
   "source": [
    "# Initialize a TfidfVectorizer object: tfidf_vectorizer\n",
    "tfidf_vectorizer = TfidfVectorizer(analyzer=clean_text)\n",
    "\n",
    "# Transform the training data: tfidf_train \n",
    "tfidf_train = tfidf_vectorizer.fit_transform(X_train)\n",
    "\n",
    "# Transform the test data: tfidf_test \n",
    "tfidf_test = tfidf_vectorizer.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.279783Z",
     "start_time": "2018-11-12T15:36:56.239759Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['', 'aa', 'aaaaaaaaye', 'aaaaaahhhh', 'aaaaand', 'aaaahhh', 'aaawwwwesucks', 'aaron', 'aback', 'abandon']\n",
      "['zusammenstellen', 'zzzzz', '¼', 'à', 'ça', 'è', 'écopé', 'était', 'été', 'évidemment']\n"
     ]
    }
   ],
   "source": [
    "print(tfidf_vectorizer.get_feature_names()[:10])\n",
    "print(tfidf_vectorizer.get_feature_names()[-10:])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Naive Bayes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Count Vect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.289075Z",
     "start_time": "2018-11-12T15:36:56.282615Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.330716Z",
     "start_time": "2018-11-12T15:36:56.291515Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.618324272465622\n",
      "confusion matrix:\n",
      " [[1638   79   38   53   45]\n",
      " [ 268   70   73   84   44]\n",
      " [ 151   37  101  256  108]\n",
      " [  94   18   35  508  594]\n",
      " [ 137    4    6  263 1550]]\n"
     ]
    }
   ],
   "source": [
    "nb_classifier = MultinomialNB()\n",
    "nb_classifier.fit(count_train, y_train)\n",
    "pred = nb_classifier.predict(count_test)\n",
    "print(metrics.accuracy_score(y_test, pred))\n",
    "print('confusion matrix:\\n',metrics.confusion_matrix(y_test, pred, labels=[1,2,3,4,5]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### TFIDF vect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.361925Z",
     "start_time": "2018-11-12T15:36:56.333911Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5738727214582667\n",
      "[[1736    0    0    1  116]\n",
      " [ 365    0    0   11  163]\n",
      " [ 216    0    0   25  412]\n",
      " [ 115    1    0   15 1118]\n",
      " [ 119    0    0    3 1838]]\n"
     ]
    }
   ],
   "source": [
    "nb_tfidf_classifier = MultinomialNB()\n",
    "nb_tfidf_classifier.fit(tfidf_train, y_train)\n",
    "pred_tfidf=nb_tfidf_classifier.predict(tfidf_test)\n",
    "print(metrics.accuracy_score(y_test,pred_tfidf))\n",
    "print(metrics.confusion_matrix(y_test,pred_tfidf,labels=[1,2,3,4,5]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.368155Z",
     "start_time": "2018-11-12T15:36:56.364891Z"
    }
   },
   "outputs": [],
   "source": [
    "alphas = np.arange(0,1,0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.565790Z",
     "start_time": "2018-11-12T15:36:56.371337Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Alpha:  0.0\n",
      "Score:  0.5567636712503997\n",
      "\n",
      "Alpha:  0.1\n",
      "Score:  0.6044131755676367\n",
      "\n",
      "Alpha:  0.2\n",
      "Score:  0.5989766549408378\n",
      "\n",
      "Alpha:  0.30000000000000004\n",
      "Score:  0.5929005436520627\n",
      "\n",
      "Alpha:  0.4\n",
      "Score:  0.5889031020147106\n",
      "\n",
      "Alpha:  0.5\n",
      "Score:  0.5842660697153822\n",
      "\n",
      "Alpha:  0.6000000000000001\n",
      "Score:  0.5810681164055005\n",
      "\n",
      "Alpha:  0.7000000000000001\n",
      "Score:  0.5786696514230892\n",
      "\n",
      "Alpha:  0.8\n",
      "Score:  0.5769107771026543\n",
      "\n",
      "Alpha:  0.9\n",
      "Score:  0.5749920051167253\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.6/site-packages/sklearn/naive_bayes.py:472: UserWarning: alpha too small will result in numeric errors, setting alpha = 1.0e-10\n",
      "  'setting alpha = %.1e' % _ALPHA_MIN)\n"
     ]
    }
   ],
   "source": [
    "# Define train_and_predict()\n",
    "def train_and_predict(alpha):\n",
    "    # Instantiate the classifier: nb_classifier\n",
    "    nb_classifier = MultinomialNB(alpha=alpha)\n",
    "    # Fit to the training data\n",
    "    nb_classifier.fit(tfidf_train,y_train)\n",
    "    # Predict the labels: pred\n",
    "    pred = nb_classifier.predict(tfidf_test)\n",
    "    # Compute accuracy: score\n",
    "    score = metrics.accuracy_score(y_test, pred)\n",
    "    return score\n",
    "\n",
    "# Iterate over the alphas and print the corresponding score\n",
    "for alpha in alphas:\n",
    "    print('Alpha: ', alpha)\n",
    "    print('Score: ', train_and_predict(alpha))\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.585464Z",
     "start_time": "2018-11-12T15:36:56.569123Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 2, 3, 4, 5])"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nb_classifier.classes_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.646401Z",
     "start_time": "2018-11-12T15:36:56.588533Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 [(-12.368395555817438, 'aa'), (-12.368395555817438, 'aaaaaaaaye'), (-12.368395555817438, 'aaaaaahhhh'), (-12.368395555817438, 'aaaaand'), (-12.368395555817438, 'aback'), (-12.368395555817438, 'abatt'), (-12.368395555817438, 'abbiamo'), (-12.368395555817438, 'abby'), (-12.368395555817438, 'abide'), (-12.368395555817438, 'abigail'), (-12.368395555817438, 'abita'), (-12.368395555817438, 'abject'), (-12.368395555817438, 'abnormally'), (-12.368395555817438, 'abomination'), (-12.368395555817438, 'abominations'), (-12.368395555817438, 'aboutb'), (-12.368395555817438, 'abouthenceforth'), (-12.368395555817438, 'aboutits'), (-12.368395555817438, 'abovementioned'), (-12.368395555817438, 'abram')]\n",
      "2 [(-5.117760043918758, 'manager'), (-5.095302959817916, 'make'), (-5.086321897723973, 'location'), (-5.07130455065702, 'delivery'), (-5.054508724183976, 'wait'), (-5.041272263558145, 'take'), (-5.030157405751849, 'back'), (-5.025616366485592, 'give'), (-5.014673225417806, 'never'), (-4.985027408825054, 'one'), (-4.976363988302847, 'service'), (-4.88010204065801, 'minutes'), (-4.817734312712102, 'tell'), (-4.789227588421362, 'place'), (-4.613914008347055, 'go'), (-4.469613198847129, 'say'), (-4.407025354097927, 'time'), (-4.369724194801662, 'call'), (-4.29468091470758, 'get'), (-3.4793629839426963, 'order')]\n"
     ]
    }
   ],
   "source": [
    "# Get the class labels: class_labels\n",
    "class_labels = nb_classifier.classes_\n",
    "\n",
    "# Extract the features: feature_names\n",
    "feature_names = tfidf_vectorizer.get_feature_names()\n",
    "\n",
    "# Zip the feature names together with the coefficient array and sort by weights: feat_with_weights\n",
    "feat_with_weights = sorted(zip(nb_classifier.coef_[0], feature_names))\n",
    "\n",
    "# Print the first class label and the top 20 feat_with_weights entries\n",
    "print(class_labels[0], feat_with_weights[:20])\n",
    "\n",
    "# Print the second class label and the bottom 20 feat_with_weights entries\n",
    "print(class_labels[1], feat_with_weights[-20:])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:36:56.653584Z",
     "start_time": "2018-11-12T15:36:56.649549Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.linear_model import SGDClassifier\n",
    "svm=SGDClassifier(loss='hinge',max_iter=500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:37:04.515397Z",
     "start_time": "2018-11-12T15:36:56.657127Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6499840102334505\n"
     ]
    },
    {
     "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>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1743</td>\n",
       "      <td>25</td>\n",
       "      <td>18</td>\n",
       "      <td>25</td>\n",
       "      <td>42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>336</td>\n",
       "      <td>60</td>\n",
       "      <td>49</td>\n",
       "      <td>46</td>\n",
       "      <td>48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>157</td>\n",
       "      <td>35</td>\n",
       "      <td>131</td>\n",
       "      <td>196</td>\n",
       "      <td>134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>75</td>\n",
       "      <td>19</td>\n",
       "      <td>61</td>\n",
       "      <td>385</td>\n",
       "      <td>709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>58</td>\n",
       "      <td>3</td>\n",
       "      <td>14</td>\n",
       "      <td>139</td>\n",
       "      <td>1746</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      1   2    3    4     5\n",
       "1  1743  25   18   25    42\n",
       "2   336  60   49   46    48\n",
       "3   157  35  131  196   134\n",
       "4    75  19   61  385   709\n",
       "5    58   3   14  139  1746"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "svm.fit(tfidf_train, y_train)\n",
    "pred_svm=svm.predict(tfidf_test)\n",
    "print(metrics.accuracy_score(y_test,pred_svm))\n",
    "pd.DataFrame(metrics.confusion_matrix(y_test,pred_svm,labels=[1,2,3,4,5]),\n",
    "            index=range(1,6),\n",
    "            columns=range(1,6))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:37:04.523695Z",
     "start_time": "2018-11-12T15:37:04.518238Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['valence', 'arousal', 'business_id', 'name', 'biz_stars', 'state',\n",
       "       'city', 'neighborhood', 'address', 'postal_code', 'review_id', 'text',\n",
       "       'date', 'cool', 'funny', 'useful', 'stars', 'user_id', 'type',\n",
       "       'puncount', 'lan', 'reviewsen', 'reviewsenlen', 'reviewlen',\n",
       "       'reviewword', 'reviewwordlen', 'avgword', 'pred_sentiment',\n",
       "       'cleaned_text'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:37:48.118295Z",
     "start_time": "2018-11-12T15:37:48.108616Z"
    }
   },
   "outputs": [],
   "source": [
    "X_strain, X_stest, y_strain, y_stest = train_test_split(df[['valence','arousal']], y,\n",
    "                                                    test_size=0.3,\n",
    "                                                    random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:37:50.258515Z",
     "start_time": "2018-11-12T15:37:50.255376Z"
    }
   },
   "outputs": [],
   "source": [
    "s_svm=SGDClassifier(loss='hinge',max_iter=500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:37:54.249320Z",
     "start_time": "2018-11-12T15:37:52.430225Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.4774544291653342\n",
      "      1  2  3   4     5\n",
      "1  1457  0  0  57   339\n",
      "2   340  0  0  19   180\n",
      "3   294  0  0  28   331\n",
      "4   374  0  0  58   817\n",
      "5   431  0  0  58  1471\n",
      "0.4798503166378814\n"
     ]
    }
   ],
   "source": [
    "s_svm.fit(X_strain, y_strain)\n",
    "s_pred_svm=s_svm.predict(X_stest)\n",
    "print(metrics.accuracy_score(y_stest,s_pred_svm))\n",
    "print(pd.DataFrame(metrics.confusion_matrix(y_stest,s_pred_svm,labels=[1,2,3,4,5]),\n",
    "            index=range(1,6),\n",
    "            columns=range(1,6)))\n",
    "print(metrics.accuracy_score(y,s_svm.predict(df[['valence','arousal']])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## save scoring result to 'model_star'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:38:11.509528Z",
     "start_time": "2018-11-12T15:37:56.485941Z"
    }
   },
   "outputs": [],
   "source": [
    "df['model_star']=svm.predict(tfidf_vectorizer.transform(df['text']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:37:05.171868Z",
     "start_time": "2018-11-12T15:37:04.526717Z"
    }
   },
   "outputs": [
    {
     "ename": "KeyError",
     "evalue": "'model_star'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/indexes/base.py\u001b[0m in \u001b[0;36mget_loc\u001b[0;34m(self, key, method, tolerance)\u001b[0m\n\u001b[1;32m   3077\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3078\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3079\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mKeyError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mKeyError\u001b[0m: 'model_star'",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-70-f1a7e132d612>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'missalign'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'model_star'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m<\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'stars'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36m__getitem__\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m   2686\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_getitem_multilevel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2687\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2688\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_getitem_column\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2689\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2690\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_getitem_column\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36m_getitem_column\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m   2693\u001b[0m         \u001b[0;31m# get column\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2694\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mis_unique\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2695\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_get_item_cache\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2696\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2697\u001b[0m         \u001b[0;31m# duplicate columns & possible reduce dimensionality\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/generic.py\u001b[0m in \u001b[0;36m_get_item_cache\u001b[0;34m(self, item)\u001b[0m\n\u001b[1;32m   2487\u001b[0m         \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcache\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2488\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mres\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2489\u001b[0;31m             \u001b[0mvalues\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_data\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2490\u001b[0m             \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_box_item_values\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalues\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2491\u001b[0m             \u001b[0mcache\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mres\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/internals.py\u001b[0m in \u001b[0;36mget\u001b[0;34m(self, item, fastpath)\u001b[0m\n\u001b[1;32m   4113\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4114\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0misna\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 4115\u001b[0;31m                 \u001b[0mloc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   4116\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4117\u001b[0m                 \u001b[0mindexer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0misna\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/indexes/base.py\u001b[0m in \u001b[0;36mget_loc\u001b[0;34m(self, key, method, tolerance)\u001b[0m\n\u001b[1;32m   3078\u001b[0m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3079\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mKeyError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3080\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_maybe_cast_indexer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3081\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3082\u001b[0m         \u001b[0mindexer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_indexer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmethod\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mmethod\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtolerance\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtolerance\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mKeyError\u001b[0m: 'model_star'"
     ]
    }
   ],
   "source": [
    "df['missalign']=df['model_star']<df['stars']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:37:42.651901Z",
     "start_time": "2018-11-12T15:37:42.554421Z"
    }
   },
   "outputs": [
    {
     "ename": "KeyError",
     "evalue": "'missalign'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/indexes/base.py\u001b[0m in \u001b[0;36mget_loc\u001b[0;34m(self, key, method, tolerance)\u001b[0m\n\u001b[1;32m   3077\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3078\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3079\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mKeyError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mKeyError\u001b[0m: 'missalign'",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-71-100f08f3c7b2>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'missalign'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36m__getitem__\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m   2686\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_getitem_multilevel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2687\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2688\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_getitem_column\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2689\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2690\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_getitem_column\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36m_getitem_column\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m   2693\u001b[0m         \u001b[0;31m# get column\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2694\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mis_unique\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2695\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_get_item_cache\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2696\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2697\u001b[0m         \u001b[0;31m# duplicate columns & possible reduce dimensionality\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/generic.py\u001b[0m in \u001b[0;36m_get_item_cache\u001b[0;34m(self, item)\u001b[0m\n\u001b[1;32m   2487\u001b[0m         \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcache\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2488\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mres\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2489\u001b[0;31m             \u001b[0mvalues\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_data\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2490\u001b[0m             \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_box_item_values\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalues\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2491\u001b[0m             \u001b[0mcache\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mres\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/internals.py\u001b[0m in \u001b[0;36mget\u001b[0;34m(self, item, fastpath)\u001b[0m\n\u001b[1;32m   4113\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4114\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0misna\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 4115\u001b[0;31m                 \u001b[0mloc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   4116\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4117\u001b[0m                 \u001b[0mindexer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0misna\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/pandas/core/indexes/base.py\u001b[0m in \u001b[0;36mget_loc\u001b[0;34m(self, key, method, tolerance)\u001b[0m\n\u001b[1;32m   3078\u001b[0m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3079\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mKeyError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3080\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_maybe_cast_indexer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3081\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3082\u001b[0m         \u001b[0mindexer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_indexer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmethod\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mmethod\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtolerance\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtolerance\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mKeyError\u001b[0m: 'missalign'"
     ]
    }
   ],
   "source": [
    "df['missalign'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:38:13.798101Z",
     "start_time": "2018-11-12T15:38:13.791163Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7513433122241412"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics.accuracy_score(df['stars'],df['model_star'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:38:16.236907Z",
     "start_time": "2018-11-12T15:38:16.204000Z"
    }
   },
   "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>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5850</td>\n",
       "      <td>39</td>\n",
       "      <td>29</td>\n",
       "      <td>39</td>\n",
       "      <td>84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1003</td>\n",
       "      <td>617</td>\n",
       "      <td>73</td>\n",
       "      <td>91</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>377</td>\n",
       "      <td>49</td>\n",
       "      <td>1053</td>\n",
       "      <td>351</td>\n",
       "      <td>345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>166</td>\n",
       "      <td>27</td>\n",
       "      <td>89</td>\n",
       "      <td>2004</td>\n",
       "      <td>1848</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>131</td>\n",
       "      <td>7</td>\n",
       "      <td>33</td>\n",
       "      <td>284</td>\n",
       "      <td>6137</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      1    2     3     4     5\n",
       "1  5850   39    29    39    84\n",
       "2  1003  617    73    91   118\n",
       "3   377   49  1053   351   345\n",
       "4   166   27    89  2004  1848\n",
       "5   131    7    33   284  6137"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(metrics.confusion_matrix(df['stars'],df['model_star'],labels=[1,2,3,4,5]),\n",
    "            index=range(1,6),\n",
    "            columns=range(1,6))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cluster K-means"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:42:54.326564Z",
     "start_time": "2018-11-12T15:38:18.554036Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# run kmeans\n",
    "num_clusters = 5\n",
    "km = KMeans(n_clusters=num_clusters, n_jobs=-1) # instantiate kmeans object, with n clusters, and tell it to use all processors\n",
    "km.fit(tfidf_vectorizer.transform(df['text']))\n",
    "clusters = km.labels_.tolist()\n",
    "df.loc[:,'cluster'] = clusters  # assign cluster results to dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:42:56.659483Z",
     "start_time": "2018-11-12T15:42:56.628318Z"
    }
   },
   "outputs": [],
   "source": [
    "# get details of clusters\n",
    "cluster_details = {}\n",
    "\n",
    "# get cluster centroids\n",
    "ordered_centroids = km.cluster_centers_.argsort()[:, ::-1]\n",
    "\n",
    "# get key features for each cluster\n",
    "# and also list restaurants and reviews associated with that cluster\n",
    "for cluster_num in range(num_clusters):\n",
    "    cluster_details[cluster_num] = {}  # assign cluster number as dictionary\n",
    "    cluster_details[cluster_num]['cluster_num'] = cluster_num  # assign cluster num\n",
    "    key_features = [feature_names[index] for index in ordered_centroids[cluster_num, :20]] # top 20 features of cluster\n",
    "    cluster_details[cluster_num]['key_features'] = key_features\n",
    "\n",
    "    businesses = df.loc[df['cluster'] == cluster_num, 'name'].unique().tolist()  # list of biz in cluster\n",
    "    review_ids = df.loc[df['cluster'] == cluster_num, 'review_id'].unique().tolist() # list of reviews in cluster\n",
    "    cluster_details[cluster_num]['businesses'] = businesses  # assign businesses\n",
    "    cluster_details[cluster_num]['review_ids'] = review_ids  # assign review_ids"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:42:59.005491Z",
     "start_time": "2018-11-12T15:42:58.982226Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cluster_num': 0,\n",
       " 'key_features': ['great',\n",
       "  'service',\n",
       "  'place',\n",
       "  'good',\n",
       "  'food',\n",
       "  'friendly',\n",
       "  'love',\n",
       "  'staff',\n",
       "  'amaze',\n",
       "  'beer',\n",
       "  'price',\n",
       "  'go',\n",
       "  'best',\n",
       "  'awesome',\n",
       "  'back',\n",
       "  'come',\n",
       "  'nice',\n",
       "  'delicious',\n",
       "  'time',\n",
       "  'atmosphere'],\n",
       " 'businesses': ['Pizza Hut',\n",
       "  \"Domino's Pizza\",\n",
       "  \"Grimaldi's Pizzeria\",\n",
       "  \"Papa John's Pizza\",\n",
       "  'Little Caesars',\n",
       "  \"Papa Murphy's\",\n",
       "  \"Blaze Fast-Fire'd Pizza\",\n",
       "  'Mellow Mushroom'],\n",
       " 'review_ids': ['2UV-a8pXUdbK0fdurFwXfQ',\n",
       "  'AJPx2FNBtrFom1Ke7hiA1A',\n",
       "  'hF8Neuh3r2x-1W09hl6fvQ',\n",
       "  'AWqyHf84pWFro3eTTUkjow',\n",
       "  'y4PRnR0PLg1wTIOl-ygD1Q',\n",
       "  'dhTHNsSgKKgiscCg1HyKPQ',\n",
       "  'gA_uXxTArxFqRsZtmORVlg',\n",
       "  '1NJn3kyc_gwnKobVTqyK6Q',\n",
       "  '9y11DeOFELzwIh4TK8pJVw',\n",
       "  'XK55eCQ2_p3w7ELXqXPT_Q',\n",
       "  'f3ME5RRO8-mYOgNibDop3g',\n",
       "  'dCDLRnZXmPG3OQ4uc8t73g',\n",
       "  'eiDeX2o4yeBiqBe1KFx3xg',\n",
       "  't0k1CtMqqZ1xBVhWqRMA-A',\n",
       "  '1UP9lZUoRQ008LlMcTl4cA',\n",
       "  'WisSCWZFuTFoy7Xq4NK8aQ',\n",
       "  'S2VxZ7krYsFzgtlPGU_tqA',\n",
       "  '0G-qjgYzZHfARPzTyLmyjQ',\n",
       "  'R8NFVz8C_uVvcmmSSYF8gA',\n",
       "  'XOhlju8gK9SECRRPubMjOA',\n",
       "  'WsNSRI_-jlN3IE_9GkfHRQ',\n",
       "  'j1YnIRmA4cNt_5REaO1xcQ',\n",
       "  'Hi5xB8y_17RxMT67tSnQ1Q',\n",
       "  'GHogO1NDhIhfs804GIUlGw',\n",
       "  'VnngXnaaIUA5H8awJaIz9Q',\n",
       "  'osRyM32_U91i0GOwgJNH-g',\n",
       "  '8n3WKuzHmFN70rX9xrrAdg',\n",
       "  'Lz5WEojXWmOhINR-britNQ',\n",
       "  '8YVKTURLhDb5o6UFgnpQDw',\n",
       "  '_ZX-GqA7Ant6wGB840qMBw',\n",
       "  'Wrh3qqpK8eGfIv4kX93Eqg',\n",
       "  'P26IN4_MrqKwwG3YFqRCuw',\n",
       "  'wLFzlPcxrvRE9zs7gv9PdQ',\n",
       "  'ubY0p5Jhm6rmIFx8Fat7Sw',\n",
       "  'n_69CqtZLvGyE4_8-PRr2Q',\n",
       "  'Qcd1EZdFy9atfsijPOaOoQ',\n",
       "  'gu1_1OltxZWjItSOD4puoQ',\n",
       "  'reyY-v72c3_8XYheqf3zhw',\n",
       "  'seq10IQTwNGUSbP7dffIRg',\n",
       "  '2fdxiLrRI-nQ01xcn3fntQ',\n",
       "  '7IxPXxPrJQCGJCGTiS4a5A',\n",
       "  'cnYya5zkmvPARKsbqwWuww',\n",
       "  'BFyhI2o96FJKjQXO87MhJQ',\n",
       "  '1Ezdx6tkdXQwWOefwOQhlQ',\n",
       "  '89hEdeeMS_NXkXNjLT-Pdw',\n",
       "  '2BM3gcZ6ZvaowN-870dovQ',\n",
       "  'dF3-cyGLEAHE4MnPJncVQA',\n",
       "  'q-xyAr7ALxOc2lmTS-f9dg',\n",
       "  'AM_HgPKUo-8lNYFJqQjVhg',\n",
       "  'oAxlOTa2m9uA3UGLtucqsQ',\n",
       "  'C_R5wX5_2Lhtyu4un5YIxw',\n",
       "  '8XlMIPpZPo10gY80XLAiQw',\n",
       "  '1_0Egh92oT1oiXn50-07gA',\n",
       "  'JMnLGFdnn4lkDZO3FhHKag',\n",
       "  '4i8qHGe58c0J3PX2dxIqpQ',\n",
       "  'tfh0Uk8_4BVK_vkKAzrpCg',\n",
       "  'BZYZmanIiwHvDETmP1v3jw',\n",
       "  'aQKl82CiKGK7yZeWE0WGKw',\n",
       "  'G0f20KhUOuSSY2C4z7wSAA',\n",
       "  'fO-jHoB8GxWSGS4C10jtIA',\n",
       "  'CQ30NPa-BJudRf2JGQRRtw',\n",
       "  'kQeoXSAflz1Wl41_5zAcbA',\n",
       "  'YvOlVlsxUdtky6PYqhLVJQ',\n",
       "  'VBrubzBGM-Go_mvMBHKvpQ',\n",
       "  'lfUl97N1JES_TV8hA5D88w',\n",
       "  'tgIwbTAK647K2ekDlQ_U4w',\n",
       "  'z64bVuX-KaUFOcgiIB_QlQ',\n",
       "  'uyoZi0ikRzE4PvIMC8vp8A',\n",
       "  'bGaJcPMZHehKxvb_0lR34A',\n",
       "  'g8z4DN4cewWHr2hFPbmnLg',\n",
       "  'VeTWQ65SytD7bohzmhWQ9g',\n",
       "  'T2NmW60ucBFTBZfZkZVEhQ',\n",
       "  'lsUyeQ9tfVviOLlj9S15qA',\n",
       "  'V6o_5Abi6OwiMs4_TuPN3g',\n",
       "  'oXxMdWPxtjqXdlHMM145lA',\n",
       "  'Z_yGAKv8HduE6zo7Px290A',\n",
       "  'uB1X5755PNO_BHp4lNPTew',\n",
       "  'pIboFTWbH7okwcy-YISBOQ',\n",
       "  'umeZqf2qJDyXVw75sflx4w',\n",
       "  '0DhPh592ES_i6JOz8j1Dug',\n",
       "  'j481lce3t8KxqXPoGVpGMQ',\n",
       "  'yWhov5XzFKhG6tvpqG1XCA',\n",
       "  'YamtAE5oALrTpdkmSeGulg',\n",
       "  '1nE3-RzrAIorHDibbQTA3g',\n",
       "  'K1tS7Wl6ymTaZlMbAHArdQ',\n",
       "  '1mdlFiMgqYpfuhdFo0yx4A',\n",
       "  'rp7luhNRP_w6b2AfhuStCQ',\n",
       "  '4kYlIx0CFnHJXLFogy8rzw',\n",
       "  'kib8GRIWFKxmoexpTVb0Kw',\n",
       "  'VtenuJ0Xy-Bd1xKGHfYZcg',\n",
       "  'N9GqrPGp38eTFzHBZSYlSw',\n",
       "  'qMTuM5r3PFv4DvUALVI1cw',\n",
       "  'v62SPH_aVlfCt2n9sT-HfQ',\n",
       "  'Xbx9Z0zeB2e-DrR3bfBUyA',\n",
       "  'D8ehIjn26-nCTTYvTEexgw',\n",
       "  'eCPtkg4axNxYPq0taCjXHQ',\n",
       "  'Ajh0UNJETN4te3cdoOrShg',\n",
       "  'iiEVPbufpun8Ab6mwrtM5g',\n",
       "  '-8JnLyDUdY2QmGIN_1u20g',\n",
       "  'EUzKbNLCH8mwLBmUAtsmMQ',\n",
       "  'NGZSKUNUKB7IAcJgx5eh9g',\n",
       "  'jf181zOJxrK5CVI3NbSwag',\n",
       "  'o7R5IFFbuLFvJ_4GA1YRUw',\n",
       "  'nb2YxPOXuYsQrqed3S8jqg',\n",
       "  '5sA0dz41LZnjz-V2eHJJqg',\n",
       "  '11CmXSC-jYoxriRlC8gGVg',\n",
       "  'qXA675qhEw7AiTla3bTeBg',\n",
       "  'wqUEtakP3xq_-GialfS32A',\n",
       "  'u8FI8XVTUt0GjwQo5Za--Q',\n",
       "  'GwIlE5IOqvL4JMPlmLmN4g',\n",
       "  'UyBxqzf84vYILwd3fCUglQ',\n",
       "  'EPue9sE-vw_h0NZCexNEHA',\n",
       "  't-awe-w12VYhmwL8mkifTQ',\n",
       "  'y00mSOW_9bhbAzJem7TGTw',\n",
       "  'OuUNmRG9E1ZmAykyc2HZEQ',\n",
       "  'yWZsMH1gNBujx1Xq3l9htw',\n",
       "  'bpQekxDR3i8kYUmdngZhtA',\n",
       "  'a3uPt4XRNKkwxkEMZUls6A',\n",
       "  'wcCi_1HU-JjRGXLFjODrSQ',\n",
       "  '0BC8qBherOorfG5kP3lr1w',\n",
       "  '3zoS9GZgQzBdGVzPSLl3AA',\n",
       "  'zmFLoDkZW-nQtMYON6LJQw',\n",
       "  'l4XaJly7YGe7zIrVq6f4Mg',\n",
       "  'GpRxLTr_pM-GWcwK8oJM2A',\n",
       "  'B3YxAiUNb5tZSJbMTXAORg',\n",
       "  '7zeiFAAB1MEgno-pfruN5A',\n",
       "  'MiwX4H9FwEovjBPotlmVdQ',\n",
       "  '1FGbYEe-umJc-0ikOCwgIA',\n",
       "  '3gELYC8rXgG-_aa8enUEnQ',\n",
       "  'PWRP6eNSc0_7nnRtcKtscA',\n",
       "  'bvamsNgSzKiVY1_cNSuRaA',\n",
       "  'htYQ3RDAmLfmN6JRiOYKUg',\n",
       "  'Pg7XUmGP__MVlLVL8rZUpQ',\n",
       "  '1mETRR4lDgiJEMCRynfWiA',\n",
       "  'RRQWnFBS88-xdcnE3wAXzQ',\n",
       "  'p31WQSDmAK2hL6zA0Qo8Zg',\n",
       "  'jESZvtu19YghUp8RhpOb3Q',\n",
       "  'OvLE6sQbrgBmSH8L1W8RVA',\n",
       "  'UQ_G3RmkSahC8tkIRwV8FA',\n",
       "  'f0sd83Usu5e1LuLzvDL5ww',\n",
       "  'd3020ZFWDz0yEyjYr4Cgwg',\n",
       "  '0Nty0Nll7pbEAMB11iRTcg',\n",
       "  'BocKGIyAe8T54IOTRuSHXQ',\n",
       "  'NIb9K1JN3tB1wyT9SSRP2g',\n",
       "  '6o12X8fYHc3WHb-7LozNOg',\n",
       "  '8ub3dhLlGtkSgc8r0gTguA',\n",
       "  '-nsA4ZO6JZ96EWIUYnYOFA',\n",
       "  'JLLGAOLYywH3yAcuLuDrmA',\n",
       "  'x9gzgwv457L5ZGfw9Hk2jA',\n",
       "  'EBwlf3gdCAjRx8R-Il84XA',\n",
       "  '07fyN9Qos_kwZ4vThQ4YVA',\n",
       "  'ggDFzAKGIWgGCZ3aQvTbww',\n",
       "  'LWhd3iuEeoagB18UqRsUqw',\n",
       "  'RNA22f3k4jLemkEoTg-ORw',\n",
       "  'ATBiPDW_szoFFW2R4jRV-g',\n",
       "  'NXeTHe9Ru73-pcF621KlZw',\n",
       "  'zyY2m_UuYKWDI5fz4k4U2g',\n",
       "  'AI2s2qBmmMAtLWf8wcJO9w',\n",
       "  '69k3ALwHNe2O785RGpAh4g',\n",
       "  'Y2xpYo5ExALRE5xIDlgK4g',\n",
       "  '7QGdOKJgJe6jsLPcn9L8bg',\n",
       "  'bvIeyvhblcR_mD9tdYUtBQ',\n",
       "  'por0kR5sCOzQAk_f9K3FmA',\n",
       "  'vS-BnATC16wZsxsOfXtMRg',\n",
       "  'KitQIIk9T_85C8h33lNKyg',\n",
       "  'p1GW8O02GWUaa4cpjQvAng',\n",
       "  'xfyoAyDteqsU49jTY6lLQQ',\n",
       "  '_zRrWOQqMfyuDv6CRgB3Rg',\n",
       "  'ptXTXiiCBTYnycQRk3xW9Q',\n",
       "  'YArZmSHGyt2BFAMBvv9_tA',\n",
       "  'F1lIzZrvlpMmLpl57Lk0-Q',\n",
       "  'XxjsJ5ZvDvs2xXQwpyoyfQ',\n",
       "  'LbrFJPQ5irAJyDX6vTNE6Q',\n",
       "  'rBPQq8SR-mc54QFQIgZb9g',\n",
       "  '6VGkx-yLPMyyiSTrIucEUw',\n",
       "  'iIzsh_YYNhh1lPuaQi1ORg',\n",
       "  'pzVL-dadlQZ-mONVhRDQsg',\n",
       "  'cudB-FENgT51qHbCNUsIcQ',\n",
       "  '36-djnQV5M4tvuLT0jn7Ng',\n",
       "  'JPCEGDfkbmcU6FkECCrxqg',\n",
       "  'c7pJrwx5w4CC6TPHOIQaUQ',\n",
       "  'zz_ajc8Efx_UvawUfa71yQ',\n",
       "  'eHhFZUgZ_9BSr_aGtMx2xg',\n",
       "  'cmpQ2VDGPdJgk9nrdZgE9Q',\n",
       "  'HAtdBJf_24XvOD9vDFUCpA',\n",
       "  '5wLQPQ50_VbslyUeWCPmHA',\n",
       "  'mxDZQeeMVsgVKBNnMYUu-g',\n",
       "  'bgV9aPf-2lSCQnBuqNcmaQ',\n",
       "  'K3cRUyyzXnMRQrC4G-l4aw',\n",
       "  'xojrVx1mzabekakSDTYLEg',\n",
       "  'GpadnS2AwS909he4-t_iIw',\n",
       "  'VwGXL6jWVKwk5NvwW1hiJQ',\n",
       "  '7hVBAxt9Duc0vOe_RWV2pA',\n",
       "  '5vubYJuPMEVOw4YEWgSTuA',\n",
       "  'E1cFtXzF1Vzcm0ZU-4sCFg',\n",
       "  '8_2HR2lrDkdJGYhLTST6Vw',\n",
       "  'w6ReLKhxW6pGraucdEpyTQ',\n",
       "  'QcUnUemIJrVMSYdZ13zW_Q',\n",
       "  'Niq4jt_FuJYw1soMDtRaSw',\n",
       "  'MxbqZJa8eMkblS4crIrhYQ',\n",
       "  'JPVf625AdZy4jEylYQa4VA',\n",
       "  'gUsUjsRtvY7oIiS6l63ASg',\n",
       "  'nDNLG6DaqkKcWkoM-jtmRA',\n",
       "  '3CssvmMLY1JMjRw22AnPaQ',\n",
       "  'Wtt3v-MumM9-NItuDFes3w',\n",
       "  'EnwRzYHF7JCziaKj-4Gosw',\n",
       "  '4XulygL3LsRRzPxx4KgTYg',\n",
       "  '9AqaNnF4E7oWsy4bDjaQ5Q',\n",
       "  '2ILTvbIA2V0nm3DYl6Bg3g',\n",
       "  'nWx-LNgO9_qI3juoiPvChg',\n",
       "  'BFb_J1g5eDokkF_PKJvPHA',\n",
       "  '9bsCrD-cGudiF876T-gU8Q',\n",
       "  'D6_cubDDmUHK9etIo4qj5g',\n",
       "  '0BltL8IQiL_WgwoBLs0uJA',\n",
       "  'YnhWFNsSZheVgx6x4imgRQ',\n",
       "  'ZbUyZb_WG9L4TcI1g8BHgA',\n",
       "  'qiTN8TwWi8HUJn6Tc_tbUA',\n",
       "  '3LEuardhj5l8h1WNBo3B6g',\n",
       "  'E9w-q6A1YaizPobz-guPDQ',\n",
       "  '6gckmJuEcHj8VMw0_JgkYw',\n",
       "  'vVCv3Pe_Ty8SACVwzRvzTQ',\n",
       "  '-F7QtXiE5QbLdMluDfk74Q',\n",
       "  'iycsVf0NjQFJGZWVTzlA4g',\n",
       "  'muUdSfEt264xD6LN-f7Y5g',\n",
       "  'iCdQ4x-aicWh07i9xZsg4A',\n",
       "  'yxA8-JT4yj2PFePwZMK_bQ',\n",
       "  'U66i4togZdWlXMDM4btw6A',\n",
       "  'L1MMietLGJi530adslu51A',\n",
       "  'gGlDiMWBM8KAV8bq8xhSMw',\n",
       "  'u3d2muXFqSDmVdgjAYCV5g',\n",
       "  '1PKcyQE3JFzhHtqB7wbjdw',\n",
       "  'DmJd4vwl58IrTq4AR-E0nA',\n",
       "  '0UVzfXxzzZOJFa0C00DUAg',\n",
       "  'K3sshw8xPDGh2D8-yC73zw',\n",
       "  'm05TVpVBfWm8-YrdD-9ZcQ',\n",
       "  'wBYp82ppMI5iCFHeA0X2lA',\n",
       "  'nSMeRFwRO41JG1zBxz9ICA',\n",
       "  'p-FXbA6wKzOuYaruX5NiCw',\n",
       "  'oW9Ns3fxUx8syWbgdNdUyA',\n",
       "  'T0DfNVyaGielKrzZEKeZLg',\n",
       "  'EuLWE0Q7w1r81IKbyYzHrg',\n",
       "  'rMMhd8Ljrl6GoA925Qhayw',\n",
       "  'qRsJ0xzU18L2Pmp04ARBJw',\n",
       "  '8hMVFbk4b_EYYMDU09d1NQ',\n",
       "  'qAtJhHMzoXmNehhxUjxbrw',\n",
       "  'g_Pdie8Cy7UhxtOekMwz9g',\n",
       "  'c_7z5Aw2aFNWFk7socRpog',\n",
       "  '0nWolIPUlFD4J4ro3DJi-Q',\n",
       "  'o_DSKC-Wbl89XVhT0oN31g',\n",
       "  '61zldmSEXxLfTXeyjC7jjw',\n",
       "  'K_h2cga_38tcgvhAW-Q4Aw',\n",
       "  'VNhl-dD4gXXyfdhC0kDQ_g',\n",
       "  'Ziz5-0JaEN03glPMr_M8lQ',\n",
       "  '_qwMMLsJpIh0jaSow4M6CQ',\n",
       "  '5l7VkXaXykU0xjnZ_TkOdA',\n",
       "  'OogeL3Lqd6_s-jor8ue_hg',\n",
       "  'y4aqiXZl_eoD0VsBWBw2-w',\n",
       "  'zYRDwBiiN7mmCjDZaQG5Aw',\n",
       "  'Q54Ize9RPYY_uCF7JjhywA',\n",
       "  't8Ru7G0V3EUnGLkWq-31hA',\n",
       "  'A3wtCMVgF-3c_G9nzZETow',\n",
       "  'jVEZBB7M4uDodycS74Yldg',\n",
       "  'aZiFnqw9pQjpDHjSV80KDA',\n",
       "  'SFpSQn_2GrTDd6yJKtCYhw',\n",
       "  '-qLMQiDG62CyHMYGoI1UhQ',\n",
       "  '81WfoKh1tuMEK_EuxMTSeA',\n",
       "  'YEOrNQOzlDggScxOWtW0wA',\n",
       "  '5j514womAtlWJFmHPWW_0w',\n",
       "  '0d7W9EU8eGNDuwqLnTN8nw',\n",
       "  'NAld4dHyKYx8J0OW-PgEZQ',\n",
       "  'rB95ckxS4gP-Ep4_rK6uFw',\n",
       "  'e1tPcxdiujTcht6pB_buiw',\n",
       "  'orwv2-NDboB75sTg1T52uQ',\n",
       "  '6PbhDLsHehwr_Fx0xXuyIQ',\n",
       "  'Y-Ko_D3NDYkEh3VCkhVM3A',\n",
       "  's2zhR9tWkX7kBF3bz4A7pQ',\n",
       "  'DcGo21JICnEgRtjWlCogCw',\n",
       "  '_cZfOvywMMWcyqnRNKyxzw',\n",
       "  '7xWdteT4lB8nnlPgf0kClQ',\n",
       "  'BmM0MhnfgpiGeGf3cQ3I_Q',\n",
       "  '54LYYhgY9oQsbAihqt2NHw',\n",
       "  'Pp-RZPr5EuCU5KhFotlToQ',\n",
       "  'u6WbUmY2-a6wr_3aeoMjcQ',\n",
       "  '5jiVz-L17djwL2xK0K8nxQ',\n",
       "  'IVOZkb8NTO_t2m_5Hjp_Og',\n",
       "  'G6MeZOm1bG1meOZ0GlcJNQ',\n",
       "  '9xEMrZ39IGGYYgym-sgxoQ',\n",
       "  'GaPU1jZXFvhT2cBa23pNmQ',\n",
       "  '-10NllcDYLGMErpvTXYgMA',\n",
       "  'fHuLAN7vrD7kDInrQxOAEA',\n",
       "  'dDjsEcpXgJeWpnXjgTUXGw',\n",
       "  'kNSi37Z2LlgZPBIJp1wGlA',\n",
       "  'TnIw856ot3eXfinQJqEpAA',\n",
       "  'XgRTYj_LdzG4IZtVe8gMVA',\n",
       "  'rf5eWZuf-Im9jU2pftrIHg',\n",
       "  'EIqhZlSbJxGLv54rAKsOlA',\n",
       "  'eem0NsWhWWeVrMVFtxkm4w',\n",
       "  'wGRvzl7Mk8WuKkyLWLjSdA',\n",
       "  'z3CNqs6lhnhVfglzgEjJwQ',\n",
       "  'EdaEN6kIcHMIU2YV5N39mw',\n",
       "  'CuvSqZWrfbpSVsTxVqhdJw',\n",
       "  '2mEfSPj6NtZdrOcoRDn_IQ',\n",
       "  'NaQVVDQmb0pD0mdqY6etfw',\n",
       "  'm_AwYrYIOCJBAx4r4mqRww',\n",
       "  'HMJZ2z-Y2hMrQ0XNApKfEQ',\n",
       "  '_2zwjcjF6iSLbUUNgM2vmw',\n",
       "  'fdTI7GqhNXKS-H_Xo5xH6A',\n",
       "  'UocvmKdRo_fHt018_FHlwA',\n",
       "  'jNHuK_43LjbFtr3MtRtouw',\n",
       "  'xkrO2QzFQ03Dj6vbHQ3CYg',\n",
       "  'wjyMgJFPYlLpFZwWEydS2Q',\n",
       "  'sLMkw0ytCAS8BvQHxkHpZQ',\n",
       "  'ODdlL9Wlpp8wA86rkMRiiw',\n",
       "  'UuGsmQRVq5bK4YvogrEHog',\n",
       "  'A_4r9ZsYUEvl7Sxt5hKqxw',\n",
       "  'Aiv1VdM5Nyovjl36e7gKzQ',\n",
       "  'DOufDnrRP40Pe7mFvYYbHw',\n",
       "  '3vk2ytGSceal-53vOcqVTw',\n",
       "  'X4EIk1kIlpZWL82q5qfxHQ',\n",
       "  '2q9vzi6u6m-YsvLHejD8Ew',\n",
       "  'DmFJ70Q-T4ZrHpAaA6kU4A',\n",
       "  'FPfe9kbb3ar52_pBVCgyxg',\n",
       "  '7wzgUMkDIwObVqxnmphN1A',\n",
       "  'EgP1yfukpdfkzRnPedIjkg',\n",
       "  'm2DH4THeTsHTs2vKiXYJ1w',\n",
       "  '5QvlLUFmg9VVcOYc_4Esdw',\n",
       "  'm1hbTxFbMtGla95c4Q98VQ',\n",
       "  'W_p0OLIIvE9hpSu-iJScuw',\n",
       "  'zv9OL_ijlIBUemA1c69lAg',\n",
       "  'moQxPizxYY0rxDjSni40Kg',\n",
       "  'hP65ir7U6TwEor9SX8QG2A',\n",
       "  'VFHbLJsYfC9l4KfcMeEAgA',\n",
       "  'EzOyEfwgZePmsEmmRyumPw',\n",
       "  'lJkBs3K7nagj7OP_uPIzqw',\n",
       "  'OdHnRsKuNusJtxaAan8nWQ',\n",
       "  'kHSp8yA-QiV3h1O0R15V6g',\n",
       "  'NSm-LD4ehlo2R7zmrdH9LA',\n",
       "  'Q3Scv74ltIad43BRfmmf1A',\n",
       "  'H4oB-XsOqQIXlM-VqGfYgA',\n",
       "  'DRGkVsup1LvpbS2NtoJdOA',\n",
       "  'KdR6wOHvbaYdUTXeOQ7AFg',\n",
       "  'y_J73xaoqvcYWiG_G54lSg',\n",
       "  'QKtqhXyk8sg2TLmslUeZDw',\n",
       "  'IxpXa0b3o7JcLhB1w2hvXw',\n",
       "  '_QubI2pIO08pL4XvFEQS0Q',\n",
       "  '6saC9lWXnRK2fhNH39we3Q',\n",
       "  'kx4kt5a2nV1bSsvB1fPHug',\n",
       "  'iGEmWQJKpDKTypUT1bMquw',\n",
       "  'SX1nKp4lA0eLr8gboB10xg',\n",
       "  'mXVWiN5JZJyeX_k8srXjGQ',\n",
       "  'ICgNqRPNIeA9gk03ZlZ7WA',\n",
       "  'JtvWPqYM5PuRYlGmh52H7Q',\n",
       "  'vJeMvYnvMoH8OgU50_zbQQ',\n",
       "  'UAzAJpo2pds6adddQgoscA',\n",
       "  'vYgA3g_nRXllHT1905pVFw',\n",
       "  'yHijTl9SHdhQUwuQIV35Vw',\n",
       "  'Vwa6_wPGLI-PRM8ssScoCg',\n",
       "  'KorF66dRWRDNvOsbH86kGQ',\n",
       "  '7GUwHDb3hh3GeOcOiZ3Y0w',\n",
       "  'S0f7JpQ0O6bTIWI5m2-Lig',\n",
       "  'C5hBtgxUbmHPq2cihyYhFQ',\n",
       "  'VzOg7STZ_CsiEFO7YB_qtw',\n",
       "  'LPxfyEWrRi4taBVlGQOrQg',\n",
       "  'P5lnFporvOSIzSd4RQcZcQ',\n",
       "  'bJ2LIbrDYwKfhyW1YipDPw',\n",
       "  'qSgfrlh-0tb1WTvOF9MEEg',\n",
       "  '1an5z-q9RhR7f8-2Y0wrxg',\n",
       "  'nrq_mN8JEtpDdSA8Ix6tEg',\n",
       "  'A98oyIEQjRKzRKxWpY1EMw',\n",
       "  'oUXyFijjGzNLfaoWj8eFXg',\n",
       "  'En78YVXLm4RFhiWkdQDhkg',\n",
       "  'dedlzmamBKRChNpeL3Ghjw',\n",
       "  '68K4-faklnWqqF1dfIoLpw',\n",
       "  'vc8vJ6BPlltswW8meQadAQ',\n",
       "  'QxqD6BsK49QA6s5858-h7w',\n",
       "  'peT4RI1Q7a_TNqF6M2jeVg',\n",
       "  '8IEeolXfEJORggv94esEDQ',\n",
       "  '_JoliHjSiCZPg-c00rbJpw',\n",
       "  '0SaYaMofOMHAyL-pdAE9qw',\n",
       "  'F-0xB9lk0jy9DZ6mcrERZw',\n",
       "  'kPd-sFNNfsZl9CRZGExXTw',\n",
       "  'Gr388h7z3_RbOYZA1XAuSg',\n",
       "  'zGF0ltxkEfssMXOsq-iQ9A',\n",
       "  '34rQIWU399gFKSfIhUBdCQ',\n",
       "  'irKAq2YTVZ94UFELljKd4g',\n",
       "  'OyDVgpXnmaMqchIfNy1NZw',\n",
       "  '3Og0gJhxRbke38r-dr1Bug',\n",
       "  '-tJ-0N1QBuTVNmp-W36kqQ',\n",
       "  'eKyfM9Z-frpISroHYiUzTg',\n",
       "  'i5-SRHcHZYdSFItl1YIC9Q',\n",
       "  'K3nUTWHWkQXsJhAvE7wKkQ',\n",
       "  'neTCgnbm6C1O9zv1IJ4Jvw',\n",
       "  'afHvWvz0sppgPSK_K48Lzw',\n",
       "  'C8hE4DrJYbvlkbFXA8QYJw',\n",
       "  'iIEoHEVZmICifHAUNl1Wlg',\n",
       "  'Uu-G806wOQIpOp_GUGWrlQ',\n",
       "  'Rxfk7HLY3z_q8Sy3YAcqXQ',\n",
       "  '_I0R0Z953y-zOap8EsG3mw',\n",
       "  'B2sPMEa73Q6FPgd8sZJueg',\n",
       "  'NdtrulqRL9DSAs-n3P3clg',\n",
       "  'Owa8z7z_2eJfFuqOzSSFsA',\n",
       "  '-VaTo1_SJ9oahw2h3iLmDA',\n",
       "  'yU6UOsWfIAtXTIcTh4NCLg',\n",
       "  '9QkUcqDOGKsKoXKXsKsoZg',\n",
       "  'X5oP3GJOi1QuHZ2wb9yRWg',\n",
       "  'ngrRwAdywSmtFCIgfqtrRg',\n",
       "  'N5WfnCuB-rsTLsCx-o62zw',\n",
       "  'XMChlapRvZXNJ0K0VqoE1A',\n",
       "  'aLZUcKpzDPAh2EMQZ5e4zw',\n",
       "  'YJcE7rTDSexFufrhae2uwQ',\n",
       "  '7qLj5Ql5r3hK-upw_mqRHg',\n",
       "  'oDiPtjw6ucNjGeYoCp3JCg',\n",
       "  'y0YQSf6ea61SuD_T_yb-gg',\n",
       "  'I2SjO6O7oZ5n08m6Uya11w',\n",
       "  'kH9VnKAWzMElhcu48s_6lw',\n",
       "  'hi0046nT-81wB5O3jPKKtw',\n",
       "  'pKnjKkLGvr7c0ebevDGR-w',\n",
       "  'iJWFrPDyJIC1R2PkQTsDwA',\n",
       "  'wKHWeRCWKyLNT5Z4aEQCaQ',\n",
       "  '0dXkJyYqGWuvJPsw9TDJ5g',\n",
       "  'bARueY6u9zjzmX8iail6mw',\n",
       "  'd69qyNypV89-kYRaKp5RCg',\n",
       "  'XETEaZBWFcDf_UziizJCuA',\n",
       "  'lNT2ajeICc1bL4oaXG5vUw',\n",
       "  'MJwgxANfy34797UNA7wIlg',\n",
       "  '5QOwWwuSwzh5Cnuq0qzD4A',\n",
       "  'nno5gbiSFJn-GkBDlvIOjA',\n",
       "  'pdxxluBE6lZxLGDuorITFQ',\n",
       "  'LDw1CoNECLRW0lwmHjpO6Q',\n",
       "  'Elb3I776-RESYyIFvLiw6Q',\n",
       "  'Izma6x3_7XnxxxHR0CL-nA',\n",
       "  'zHZeN_kVylFiRIUHrGA69Q',\n",
       "  'cAp-TBVHx2FAWIOhhE_7dQ',\n",
       "  '8p0fW-bBMFWyz6ZOPX4hYg',\n",
       "  'xJpsI1tD4FmBi5UUeXwRcw',\n",
       "  '6cK43n2zfltzRCfxfXI2jQ',\n",
       "  'RLPO2kSd3841SYs2cTxcEA',\n",
       "  'RL4qdzaaw9dTpzUPxspLJw',\n",
       "  'SazoUGSnI0-lC2Am3pNLUw',\n",
       "  'qNb0UCHCYOajhUhiRvNHUw',\n",
       "  'Lq9fUGL6QT7A9ZxeCC5ZPA',\n",
       "  'iD4qk3WamIPQicuLQVXkyQ',\n",
       "  'zViHd33Dh5FQAa3G31I3RA',\n",
       "  'F1z5dH9O4cvD553Q7A7NmA',\n",
       "  'RsPqwWsSLoim7wPTUzbEYQ',\n",
       "  'JrvKKsnt8lDczjqAdWa8AA',\n",
       "  'BD22gtWnV4MH87oMENnAng',\n",
       "  'yqN9z9Xtr4J9W7C5Re9J_A',\n",
       "  'iRYUTN-5aBftr7NsFyk5Eg',\n",
       "  'OMi_Pp7MQeVz7IDkkWYi5g',\n",
       "  'DZ1xC5w9pzXUEvyFV9mzhw',\n",
       "  'pc3n7hox8ZW_I6sVWpb6-w',\n",
       "  'ss-GLS8MgkSgDwIVjScrmg',\n",
       "  'tQgo7aoOWBVCq0pe-LkjOA',\n",
       "  'h1Riulm4p0EgBP2IrSUyZw',\n",
       "  'nfrDlQi2Ygpv3vZe39u9UA',\n",
       "  'IJVye8xz0W_CJp_y0s7xjg',\n",
       "  'VYhl5C5U_8Gxhq_BmUOyuw',\n",
       "  'uTt9IUQnSf97Xi8uV5xOww',\n",
       "  'nFaVroFc4LYpssbBK6-igA',\n",
       "  '0Nk9RjZ5h_5USLUiyO_E0Q',\n",
       "  'IZsK9-TDgAnNnDIW1Xb8_w',\n",
       "  'pgKLaPt8Lb9cbFX4dtKR7Q',\n",
       "  'OklLjZ-RxIBA8DZcyWxLhQ',\n",
       "  'aWPMTWFMchzh0hnIs9lvGg',\n",
       "  'fQjPz0Z0ruQrMlQoAw9tDg',\n",
       "  'WFtKRyrCEyFizewYjlTXJA',\n",
       "  '20fhiXdIl1zhwdPAQDKxSg',\n",
       "  'aqr7oxKnN9U-A5p_X5g-ow',\n",
       "  '5J0VrXhYBOr64nh6HI1cvg',\n",
       "  'xRZBVvCJy8LOefuwgb99Rg',\n",
       "  'C2XfwSRS1QVlDeAU6xgdaw',\n",
       "  'tw9OfiMzNqAaqXC8QGAe5w',\n",
       "  'yV1yYbVAxd8wF9uN2MkE2A',\n",
       "  'EHqY_4SZgHD44a3e1OQraw',\n",
       "  'kNs5m5bG0oy79fvMRHiWFQ',\n",
       "  'Inf1xsXSMgf3vpCS0VIalQ',\n",
       "  'IeG1iH9ek9yVfpSkhihySA',\n",
       "  'X5zBbhPcFi_8gEItjmvKag',\n",
       "  'utS81AAuzvx-U_UKOoIE0g',\n",
       "  'nbqur14i77qj_8IynaRlSQ',\n",
       "  '3dWLMGvRZmrQRkGaJE0zmw',\n",
       "  'gTslBmlwNw5-K6K2LLrdmw',\n",
       "  'Nfxj40Gdq2OhO8PmEk-6fQ',\n",
       "  'YoEuFGE5TeldSG6mpy_EFw',\n",
       "  'wbMhkWYuqx9eCgIJQ8Q4ug',\n",
       "  'k22EyxgpxJq7J9mCI2xBkw',\n",
       "  'VaWmly7z04S6zLaIwG5L2A',\n",
       "  'Zkq66j6ic1TLuCSoPMbFfg',\n",
       "  'HMcnkfJNIVUAUACTsxvhVA',\n",
       "  'qkf5LR-LE_kUk7BkK8kaLQ',\n",
       "  'NKB7IBjeZE9-6CKz9QBKng',\n",
       "  '43CW_YhnAhc9qNuKm_Lg1A',\n",
       "  'zQRDZeqe-6ZlIxmMuPUWZA',\n",
       "  'XmWi767KQqzmEAbKkmvCXw',\n",
       "  'oubybazkX1ClSHoS2vtoVg',\n",
       "  '7r4QvhKhOn43pd5ZVn0zDQ',\n",
       "  'MKZ-oPZ6GFycjo8atTYNUg',\n",
       "  'JOWzxegVo4aCeRTNQEdZPg',\n",
       "  'M12q2QEwFVjw6w2cAvBCwQ',\n",
       "  'WpuRJQ9kg-BSwKQUGPB9zg',\n",
       "  '11ujrpMxXab3vbd6j7XgBQ',\n",
       "  '8_08knWlDA6ZWTPiZq-dLg',\n",
       "  'F2GrPypqpy1Jc3Muli4Jtg',\n",
       "  'dPybNP6HtDl5kFSlawotCA',\n",
       "  'DBk_p2P_EtaqZUWTiUlQ-w',\n",
       "  '4xAEcgpc1L6kGLocu2gIBQ',\n",
       "  '3pniCWG0VIUiigDh3QroPg',\n",
       "  'Drn9MREXomjFx_Y-v5jV3Q',\n",
       "  'kovZRE6WNljXmoZ7Fza91w',\n",
       "  'm9uuw5dqFUz0A5lJGXv4Kw',\n",
       "  '4TxvNMbN0y8bBiN7uEtCiA',\n",
       "  'GWKnU98sLyazV64YOJO4OQ',\n",
       "  'OnpDiiYvd2T9KD6RlIyi4Q',\n",
       "  'g01O0JBuXHT38cI1j3wDsg',\n",
       "  '75Un9WNgyRb4My9VU0o2dg',\n",
       "  'BM_3HnLwi0nqnlEiMoEqtQ',\n",
       "  'I5knDBmh9yn1zB7MhSJTnw',\n",
       "  'NMI-Zo4OEcoy1e4QUkxkMw',\n",
       "  '_aZRrduINN6Tyoxk5BBeZg',\n",
       "  'pDLJptud9_eMt0jJFz9R_w',\n",
       "  '55p5bdHUpp_TYt7Qnleyyw',\n",
       "  'oZCBknxwEaHBKAoyAKchuQ',\n",
       "  'adRLC_e-78P-aLwqlJj5CQ',\n",
       "  '88Y-wGjKfGmJjowKVuxalQ',\n",
       "  'iNKOrGmau-5rJWWsv8kEBA',\n",
       "  'E6VLj-mqNbC6Ptp6oniqVw',\n",
       "  'TUTpkeO_uZzTBs6xLDhcQQ',\n",
       "  'E-mDcuZjW0lCTMe4elh_RA',\n",
       "  'laO1wNBuRzWFZGIuDPDhVQ',\n",
       "  'zo0MuO7p3X5XlHjbTzMHWg',\n",
       "  '3seC82yazNXJRzuuASxozg',\n",
       "  'vlDDJZZFUefPIVHU9bcRgA',\n",
       "  'kO7-cHiUrWeOrUVmAPlw3Q',\n",
       "  'xmJceZh3SJhY1zeQ8eYkrQ',\n",
       "  'pqBBElYprrFG4XHIq0G2mQ',\n",
       "  'pG0TWs9CWi1kKEQqM9_41w',\n",
       "  'r6FHrT1a3eClyLya-tcZcg',\n",
       "  'JCYitVpQDn4y3ZJkGCaBeQ',\n",
       "  '0TploMKjNCHbcjEoi7iSUQ',\n",
       "  'mRPO-eFV7x168HKV9RlSNA',\n",
       "  'nmx7in4cGalzGIZS-OJy0w',\n",
       "  'eNedMjnMUdkTNcKu23gT-w',\n",
       "  'arTH2wn7wS2Y3ncOj02JEw',\n",
       "  'HDpq4Aih6CeyT3oxTT5SLA',\n",
       "  'HAecPxTtEJPPMRyCyjDWWw',\n",
       "  'lDrMEw8FZej9LE10kkNCGQ',\n",
       "  'vjFkP7Nu_cWLInyWowy1Cw',\n",
       "  '4NUf7WTleFhUiSVY4o2E5A',\n",
       "  '6uU8AuRH_775FcKEjaLHCg',\n",
       "  '7pFwWM6aQDH9XBZhPxcBqg',\n",
       "  'AwIhEwg3NVuKXEwrSAyE2g',\n",
       "  '7bro0eSEIF9OkdAJAvDpOw',\n",
       "  'ZOj7rBusV6LIfbU0NjeU4w',\n",
       "  'ToQPBXhG81JJC71jQOjJSQ',\n",
       "  'zeyShPJ9ySFoXSG6wStXEA',\n",
       "  'IgMzaWFNLgV-xiF6-NicZA',\n",
       "  'MvuDPqDHXJjDUlMltIUrMw',\n",
       "  'fHj6_ieBdY_amxtLudxmeA',\n",
       "  'WAHq9buNzMH8ut_I3oDBvQ',\n",
       "  '7qQO6bieDdeZ7dEUARnqPw',\n",
       "  '3cB_r536Xdj8NZ5cYuJNiQ',\n",
       "  'nCDsxy40Ti8HJc83iuStfA',\n",
       "  'WKsGLSKSzsJsxhLDvBc4Gw',\n",
       "  '8zHl8NO--TXC_Ju8Z2OHow',\n",
       "  'Z3SNk_pK0J35Pe0mBlOxNg',\n",
       "  'KB6iJxTLFEPmePE2Z_VnuQ',\n",
       "  'ePw-kP4iomn5qcSqZQk8-Q',\n",
       "  'EKYG9yk9ABKT-vea4CeQOw',\n",
       "  '8mI84KrLgTbdA8tAsTz3aQ',\n",
       "  'NgkIpfkrC2MInuIdYTRfXg',\n",
       "  'ajgLvEixFBsEL-1cp8N3oA',\n",
       "  'ft_LwAlGm93eDsgsx4778w',\n",
       "  '5-R8BP1t5PLdjg0vN4MXyw',\n",
       "  'CoZExlMVklGWTK5ARj_G9g',\n",
       "  'o-xCcpJk6recQXV3vxj3_w',\n",
       "  'XBSz5tKWMwMhkT29kha8kg',\n",
       "  '6nzHh8TCzFOJVvxL3t08xQ',\n",
       "  'yd6pJkDa2Wv3nKidrIltgw',\n",
       "  'b1qwm454pb6uITdYpOPofg',\n",
       "  'Lk9r_BDf4k6y8kh-16nf5A',\n",
       "  'pot7Bhwm5lVWSrjvU_7Fjw',\n",
       "  'kpZxmKZU1zkApBcHhr3xAg',\n",
       "  '2FZycH5YJDPDfCsAmaTpaw',\n",
       "  'M5fPxv4d8_26ItHECkoZdw',\n",
       "  'A8p9FC8FA9bEeOdCwoLRjA',\n",
       "  'GBEA4sNYwl7tzoag9mpeeA',\n",
       "  'kpK3gh5svewDwxp700kQHQ',\n",
       "  'uMbHglL5KB0RfFHdrRbzfA',\n",
       "  'Lj9KT50sYHTOLMcCzD7HTQ',\n",
       "  'Xi0jH-rcaVbCr9DbPSnswg',\n",
       "  'EudBkyOgcBN1SJfUGVHmlQ',\n",
       "  '9_vAnSJ69P_lvcU2rCXu1Q',\n",
       "  'JXYzw6hIRQA9CvI6nzPptA',\n",
       "  'ixahPOFMhM2LBoH3pywUOw',\n",
       "  'y6uZ6i5sWz7odPEKEnUrqQ',\n",
       "  'lcxH_zgxCFcETg-YB-ARSA',\n",
       "  'oejrYXkHIppQ9IJ1tSRopA',\n",
       "  'ziowOxHYggm5gEN0ZShAjQ',\n",
       "  'iH4R-Z5hcoBOtxrdWHV2Vw',\n",
       "  '6u-ClRDY4Q1XrEUtyfeNCw',\n",
       "  'Vfz3RzV_-RhqSR3SISv7DA',\n",
       "  'zfBc0F96Im_ffWpYG2JIeA',\n",
       "  '48HitWZ5bGh9R4efqaTQjQ',\n",
       "  'eOFg8gg8yblafgmZcF3TpA',\n",
       "  'z7Ep2-1agcdBDT-zXy7CHA',\n",
       "  'Tj---nfzPn49zVsXlyvlyQ',\n",
       "  'HsAoVGfLm_bHHpfyPOgnTw',\n",
       "  'vGE0RMZTd_RRV07KIy0wbQ',\n",
       "  'lWgDDZgY852VdA50xKinlA',\n",
       "  'BqIKfOC7R_kGALlaGpTtzA',\n",
       "  'd96ExXVdLi8f0pxJtvHPFQ',\n",
       "  '2zw0j0H26MyjEIQGBWXMfA',\n",
       "  'NGUUoqSaXPWBdjjSsuXa8A',\n",
       "  '8yRML8b04YkKks9U0seUxw',\n",
       "  'cNGiKWi9jJD4kKHa2H8zBg',\n",
       "  'rIHFHf2TIPAfFH7NBV3a5w',\n",
       "  'ATpjfWjp2G0PdghFU6ormA',\n",
       "  'TsVXP2J27rlTFilGjbG7Zg',\n",
       "  'gkttGP-pqlF5nXarKQiX6Q',\n",
       "  'niKHeogAcc0DFfd02vX9-A',\n",
       "  'xgF88_etwB7wf167LgkueA',\n",
       "  'tN7XchkN6AjyAqUKmxtIrQ',\n",
       "  '0q3JZUpimiSWXyyNg4tR8w',\n",
       "  'RawoaPUCpwb71Pa_eLJyGQ',\n",
       "  '2F5moQsM3inTg8-8G5pQVQ',\n",
       "  '6p9isnAKRfyeL_BYwbfivQ',\n",
       "  'vQf6-GVwvkjVOo8vtthHXA',\n",
       "  'qHj7bmo_5uTfDelKsiYAmA',\n",
       "  '3-OBD8cdpNTPw2k7i5ck7A',\n",
       "  'lMdQ_13gu4jIRHlMqeacwA',\n",
       "  '1QPIsrRRWLje18XAguJj5Q',\n",
       "  'jyMe8Z9AiSCNjgZREcqkWw',\n",
       "  'sKHOGDfRzhdbbib2AKjm8A',\n",
       "  'qBvaNWD6uzA9pfMpDjVR8g',\n",
       "  '_UmSy9YuvLIIO2ExFuWnmw',\n",
       "  'HKbkPDNX8pg8tp46QuaZjA',\n",
       "  'Dn-pTXxr6WV5AaebYLhK3g',\n",
       "  'z2cw-yiB6yTYVRwxSJeTeQ',\n",
       "  'Uv7MWE1o6N2bJbqT1_fiUg',\n",
       "  'zhjAwraUThKMVJGDlLgylA',\n",
       "  'bQepL8Neo32jk9CcU7FGaA',\n",
       "  'Yk4cpZRPsC_uJmuMVFVH8g',\n",
       "  'agLeSNQzxGXL9aaGv3N_MQ',\n",
       "  '2uFt8tYRZSVTE_JlSWARzQ',\n",
       "  'Z_b3u66360Gv_hHbjXx_tw',\n",
       "  'cSPJ3L0M0h6Si5n3emkqGg',\n",
       "  'MBkbwBtn6DQSdphjdc0I1w',\n",
       "  'FsAUYISyyn0KBOAvPnriJw',\n",
       "  'w6lJh7Ru3Eno7qFd8omrFw',\n",
       "  'xZf99KrY2G_5uBV1C9DhTQ',\n",
       "  'HHtzR2By3AIydMjKYSuUDw',\n",
       "  'vHjvKAVtepT6UpnFDY29BA',\n",
       "  'oTxpq4dkEqryLvWOvO9Qdw',\n",
       "  'cl-v52VIe_R4_cAPecgerw',\n",
       "  'luElZBSwzgQT7u1bxTb6UA',\n",
       "  'RG7b7HqEHNbl-vYufWlkxw',\n",
       "  'Iqh-9jd4u2jc4zFbyzvp1A',\n",
       "  'g7DuN3z6ON7wpai6cLqCXQ',\n",
       "  'WOPfFx7DgqgVaH46TPpqaQ',\n",
       "  'KpSvH_5CZJqsHzVh1rk70g',\n",
       "  'wCaFIIiwU0xbX5u3zetkFA',\n",
       "  '653yBS71YJSLh0YwsZqFlA',\n",
       "  '_3_4cEZIe_dXD_H_IkCxLg',\n",
       "  'ccmO0PLtPASCd06VsUwz-w',\n",
       "  'di7Cy7aALgbtG4nEhBBFiQ',\n",
       "  'SvhJwMYMcqzfVobZEyUNCQ',\n",
       "  'GXMt5YoAAf4ELNhgBhlx5A',\n",
       "  'Cho7if7SKHk-0cnnlYcjcQ',\n",
       "  'wqAhMl2Hf0q5TVZw9vBxqw',\n",
       "  'T_ocXJVUaE9JHxaBmqLmJw',\n",
       "  'Bu7T68Q8UU0WsWi24LIn1A',\n",
       "  'j9ywzV8KFXHg2o0zaNtBHg',\n",
       "  '8n1t7IA2tfLHbrtCN5LYGg',\n",
       "  'V_QPto40bbd_F3NPZMkgFg',\n",
       "  '16ukfpjOECzNWdxdwpDMAQ',\n",
       "  'eq7mzYiOW8StH5q0dtxO9g',\n",
       "  'DysZUMxwD7ddu72_nLejAg',\n",
       "  'xgix3IZioIGUX93dV_diyA',\n",
       "  '3eDhFWk31GWxac3bjOo3DQ',\n",
       "  'xiq8HoLUNgcDo5dGxNiF_A',\n",
       "  '1KMSzp7-UdUmEb5NCdnTmQ',\n",
       "  's3hArPV67A-_CHe3eeC6nw',\n",
       "  'fCMV-wjD4sMMOTfslVe3pg',\n",
       "  'vBddubSbPcLNEQwkMSkDqg',\n",
       "  'gZAMB4q2pduhfU_xXSAJFA',\n",
       "  'LVSVBGYwmbVtma63r9aALw',\n",
       "  'ImGmiEh0by2jffUi82u8cw',\n",
       "  'Z9bGtaIY5T797V6mmf1u-g',\n",
       "  'wSDd-xUFtY4604fiom2q1w',\n",
       "  'n9kFHTgGBNrYTpr3Zpz_0g',\n",
       "  'BGTTYkaNamV9fs8yBLLg9w',\n",
       "  '5GQ3JFPz8peSi61WyQhj7w',\n",
       "  'UPuA6EEziIkbAf72HhYz1Q',\n",
       "  'NXhbtv2gzJfzkyJh3-16bw',\n",
       "  'g9ljR0lB9YnSOoMC_hh0Jw',\n",
       "  'oUndwddsYIgZqwsBPbW8Bg',\n",
       "  'OG-oeLM7ND4xO5Nj2Z5Swg',\n",
       "  '-WdVl-WzuiEeCPqJDiwSbw',\n",
       "  'DhsXFYCX48QIYAiOsb4BXQ',\n",
       "  'iCTxKlCXwCPr6vqBoO1yUg',\n",
       "  'y7waOFxZY4DGaMJ0RGy8QQ',\n",
       "  '46lKNMIuBbeFUC4pch4m6w',\n",
       "  'notAUlxTi-WRuq0feIeRYA',\n",
       "  'E1ODUzpIx-PjpG1lvgJs1A',\n",
       "  'ueYXDQhcOjmF2ajX4zuVzg',\n",
       "  'oyFXg8bXUK4Gxu1nAo3YEg',\n",
       "  'MvclH7KkY7ozNPhWJRUg2g',\n",
       "  'm1eZcr6YbtNqd0D62Buj0g',\n",
       "  'VBwIaYqr-Sr4udtQneyPxA',\n",
       "  '2hjyonCO70HIfnf2Sp-FXg',\n",
       "  '4fkNlMSbKgPPDYV4Ipe2Cw',\n",
       "  'GJ6fcM7z_jQje5eflGb0ng',\n",
       "  'TtiGUs4ieJCReY02U0wMDg',\n",
       "  'GrKpl576uSgbw51LQn1xZQ',\n",
       "  'cSDQ3P2EI2m107E8ryUN3w',\n",
       "  'IOM8wtyjn4tA6wt6uCa8Sw',\n",
       "  'ezznolnoFNMqZYCuvH-zuw',\n",
       "  'ZFxlK_F3NDVa9dvenye6YA',\n",
       "  'ztjZiMlEwNYpR1mzKbEuwg',\n",
       "  '2XQsOi3YSkX613mInGmWFg',\n",
       "  'VTMhG9gNpTXwuHzfCAO6zg',\n",
       "  'cKoigqSMGQqh8MJHE0RqqQ',\n",
       "  'm6sVphq_NG4174Q7djKcvA',\n",
       "  'N-0_JVIGTHd626-j2l-29w',\n",
       "  'vD390wQPJnfjo9ZcdzNQ0A',\n",
       "  'VI3E9KPa6FY5NhHW36_O4w',\n",
       "  'gfybaNXMI5LDJfq5LlR4LA',\n",
       "  'eJKzO9WXfXspXfCwR1czDA',\n",
       "  'PHKhDsNmpYuQbH6g5HkeCg',\n",
       "  'AAPBRsu0LUhWJwxIa-LgvQ',\n",
       "  'nbTfc0nPIhpHH6bWSkSLzA',\n",
       "  'POimntf5LSMVk4L1LBAHNg',\n",
       "  'JkujAXROuMckPFle-PtS0w',\n",
       "  'adRkzHfwTYi2YcdYxVhlZA',\n",
       "  '8oUysFEvvSrTT-0H4Etxvw',\n",
       "  'lEeMEQDJAClAcpPvNRnlpg',\n",
       "  'w8ein3N3u1szKigJaIdOIA',\n",
       "  'FiGJx8jiKqcJ_IX0p7mSyA',\n",
       "  'FUjYeidTaUm4WsJfyBhrdw',\n",
       "  '2XhLN4EAzbOO-7uf26iQNA',\n",
       "  'Up30iQw1nXV6GoJqdiLLUw',\n",
       "  'pvs1UqsTpCMm6Dr-FAivUw',\n",
       "  '5EI7xmLinok2ey81mnJBMg',\n",
       "  '0qDYrnqG2hEHMyKxXlv8YQ',\n",
       "  'SyhZY71k4Sa3WhSb_ObpsQ',\n",
       "  'KRNu8wMywYd-zvSJF_E6lg',\n",
       "  'mtrfSQpKtUwPrhEQ3tfRLQ',\n",
       "  'fxsblwMGP3-_uyFYjSqxFw',\n",
       "  'n1PunDAeFE29EsqaWja66Q',\n",
       "  '-po2_A8iXPNr1G58PbZjxQ',\n",
       "  'oVFLwUyHjPdCmJSDYYZ9nQ',\n",
       "  'N2cOWBWlVDzW7gGvGtr5vw',\n",
       "  'eLcgiwoR_pXkO-2Nl22w9w',\n",
       "  'OAFTGDCsILdaNm8Pwr9z6Q',\n",
       "  'M3D-FaMCJCypfzsYcdHyLQ',\n",
       "  '0R6-qkPCGUdg0CUF7TMwHA',\n",
       "  'EE6_PHWY9hArc4H4OuJA3A',\n",
       "  'Mrg-07UM4cevKWoikckCnw',\n",
       "  'kaoLTMVmBbh3tTuGHsfmiQ',\n",
       "  'v8jx8U7YlAcivEQwok9cAg',\n",
       "  'Rypu2KpD9EVJqbba_3ugGw',\n",
       "  'L0qE9EQ3AtcZwSEQY7aX_Q',\n",
       "  '2vDJ9zuHCjLCFTsGNGPjCQ',\n",
       "  'Blsjvvk93HeYQKaIJsllaA',\n",
       "  '7Aj1oTWumDpjgFR0YSmDCQ',\n",
       "  'Ko1mw3im6fH07peocC2R4A',\n",
       "  '5Jtps0nuSJB6NYtHuvwzWA',\n",
       "  'v4jfxbBbdnw3LZNaQn2lDg',\n",
       "  '4irazE-jPRhpXUHiyI8oQA',\n",
       "  'GR7XmZ-_Orqj8JRIklyEBg',\n",
       "  'UQhNfgLHdSYidaVL8bUi7A',\n",
       "  'u4mU2ILFPPrcaThW7UaDNw',\n",
       "  'PlNBqHvsKPMueUnbJa2VMw',\n",
       "  'spCh5gZ3935eGlGZZYlXCA',\n",
       "  'AVlG4hDL7xEdEKlERMa_Vw',\n",
       "  'hqTu69FTskKKlXUBuao8Hg',\n",
       "  'ecPrQIfJc2RbtoBnBuIv9w',\n",
       "  '3RY-jFrUSm5zsd35NpB_Tg',\n",
       "  'cq_4FXXtktArsceWzWz_5g',\n",
       "  'z5xNG2arhC3d0l2Bspn29w',\n",
       "  '-Df2lXKkR2dgKNFfFw6U-A',\n",
       "  'EdDojQKutFTG_rNlfJI-dA',\n",
       "  'dsYZHwXZZoJYcZefnTZRMw',\n",
       "  '4SBHWNHdpXMGP9Rc0rACgA',\n",
       "  'B_GELOAHFE_2MICWgMkD9g',\n",
       "  'sHmIhP6g-r1TBLBMBPyXew',\n",
       "  'pk_p4DesP4xR0WG1uaXsqg',\n",
       "  '4URB6vkjkWPbuTB-qLdoxg',\n",
       "  'OUKjt3vz6rMPf7vXO9xGow',\n",
       "  'K1rRfDwna09kwkDRxGy36g',\n",
       "  'D0qGz3sfAOcCefgwkiKXJQ',\n",
       "  'Brgy9HAA6Mvdtn_GefwsLg',\n",
       "  '0UIwxccg3uYe7_ZQkcWS0g',\n",
       "  'TbyTVamFGS3aaGf1W0yd5g',\n",
       "  'lBhri_Fkj1c4aA5-qjIKPA',\n",
       "  'ujV_ADdMDzcCNldhRuTO_w',\n",
       "  'Uyln7si-qVyGt8r0AkYJBw',\n",
       "  'vGtogcjwcUCt0GT3_-ihIQ',\n",
       "  'FMXqelif5F4_-cZqcwc20Q',\n",
       "  'wqmwBSPvnjvaxTHNz8HFlQ',\n",
       "  'hJbWfJE6h16D0qnpfHyb3w',\n",
       "  'hVj9o4EfB_E3euYhMSTMGA',\n",
       "  '5aT6e6K74W6UeNvANoApXw',\n",
       "  'lAKGhjw0VcRpXqBmj3IIVg',\n",
       "  'bCJ_Fj30NYG4BMEyxnumLg',\n",
       "  'W5RuSXF9cJ91kurg30WB-A',\n",
       "  'EIM_aeENmeX4fNgE2ufONw',\n",
       "  '8n0CdR7el9bw-aJAHJaB5w',\n",
       "  'RWU9fOV4b85h3mzgSl0ybw',\n",
       "  '8toBaUEYr0Yn0UDFF5GwrQ',\n",
       "  'Thg2vip6NcFmHhtoUU7Yig',\n",
       "  'RULzRwOMA6y2P-Ln_jY1Yg',\n",
       "  'lt5t4t3YNue5_Rdk5XaBZg',\n",
       "  'lizXQ1aG88dIUpXdOrU0_g',\n",
       "  'wmLRgt_V0Au3DBTVzIN8tQ',\n",
       "  'FcIj1CtRSVUY34p0C8oQ3w',\n",
       "  'gznQCRViaVC1I5oHWzlEbQ',\n",
       "  'kOQwsrF30drJ447Vb65fkg',\n",
       "  'Q4sj7AxyuGtMqH0W5A1p8A',\n",
       "  'ewMjcU_X0P0H-Qc60CXJfA',\n",
       "  'P6tIgXogsRGi6KJcQF9yzQ',\n",
       "  '9XPdol4qHOiqSI0iCs3n4w',\n",
       "  'NHmIVWYNORQHWc-GICuolg',\n",
       "  '00n3RhjbOvxnfgl-g6Pq7w',\n",
       "  'nCZ2GhKdZYIgz7-TWqLAhw',\n",
       "  'T0BUVmVKWs75xYdwaTFWbQ',\n",
       "  '1awzAe8HwDfQv-u4fNDAiQ',\n",
       "  'C7435ojin8mx-y4Nr0cKCA',\n",
       "  'Y12uo9itv2uaD1HQMS5OaA',\n",
       "  'tucfLLK-w9JKZoeyK3BzuA',\n",
       "  'zpVxJAq3dTuq_JVbuLWxSg',\n",
       "  'X_xhfWr-bgKYksJBKNG2Ng',\n",
       "  'gFKmoOO5oTy3cJdF6w35BQ',\n",
       "  'gSGldja3-Rm8OpsGzHUrwQ',\n",
       "  'yL3STu3GvUQK22bTMXWeZw',\n",
       "  'lfxTSDzd2mI5LFrwbzcJkQ',\n",
       "  'SefoujqAr__gvc4cR1d4uQ',\n",
       "  'xXVgwqiCdrcLNuU5YAzpaw',\n",
       "  'rjxCkJYWCNmkzYuQaHtpyQ',\n",
       "  'wJRBxhNi8kWJpoZEN47tvA',\n",
       "  '_I3K7GBXwXNikfaDutBI2g',\n",
       "  '6nCYPjPGiqY_91wwxYlUtw',\n",
       "  'O7icvCBBoBdqm4zBgF6Z3g',\n",
       "  'oIqsOIic2-91_gkNjPzaHA',\n",
       "  'tW1Nbby0OQZ00xdjiVP0zw',\n",
       "  'D6AwO5jIxjId4TBHEu6ZPw',\n",
       "  'mFgSvPspgh3pRDbD4U3roA',\n",
       "  's_djlPBN_nAClaf6EhnAEg',\n",
       "  'Ju5iW1lQXkkSzj5N7sICyg',\n",
       "  'JvNM27WuKFBe4dzbNCD4nw',\n",
       "  '09Ip5UIaTQ51bLc18kHN8g',\n",
       "  'oE4FcXsTN6wpD7K6gidV8Q',\n",
       "  'I5H5Q8vn0rRc0agjMYJ-Cw',\n",
       "  'YrLTyppLQjTvH1WwDPyzyw',\n",
       "  'jNDqEWfc7tOLIz_wBLToGg',\n",
       "  'DFjG-roIveWPxMs53cXf9g',\n",
       "  'uMGG33vqahYT264O2sSmgw',\n",
       "  'jGmTEohR8jvMPpDtUw_YVQ',\n",
       "  's1puMKuw3wc1_n8_tsCMbQ',\n",
       "  'O06JJqhHxTBhMxC4z9c21w',\n",
       "  'nJStPD1C0exUBxtCQTdr_g',\n",
       "  '5filPS_HCIUNA9lsCGJP8Q',\n",
       "  'tUrO3ebMN0xq0tZc_sefQQ',\n",
       "  'DBaZ_kA6phB1VDG3S9KwBA',\n",
       "  'HvfpGijkno_KQALJQCXGqQ',\n",
       "  'DkdaDcitgvDysoUUJIe1Ew',\n",
       "  'AVRjmnA44hr7wZDBNyeKVw',\n",
       "  'JddcHQiqShjRkkARvthNyg',\n",
       "  'gx8vg85GyhOYOVKRAh6eXA',\n",
       "  'yAGVISz1EUBHErd0ZvzvLQ',\n",
       "  'z2u-dsv1y23ZN3rSWdw-qw',\n",
       "  's4kF1ttLn1ysydZavP0xKQ',\n",
       "  'glIxTj5SNykz4-7NQioDnA',\n",
       "  'YGFP6YfhjCmQVSBA19Xp-Q',\n",
       "  'H5xzP94yhFgl0gWO9kHsvg',\n",
       "  'yw31BZ4qKcIdqnsNitzMSw',\n",
       "  'O61Fr2l0ddStq1zsily9BA',\n",
       "  'wKLm0DY6VzMSybkBrGlKXA',\n",
       "  'niB1T3C5hvTBB3yOHqIcEw',\n",
       "  'vC__iGBZgUBp2jK7fC88OA',\n",
       "  'Wy8e33RW-TjttAuH_gc8-g',\n",
       "  '54tom1lYpQ9OI7IsEtP1Nw',\n",
       "  'PO4hdCNWSGt3CKsRrgz51Q',\n",
       "  'WZAIBWQgZ8qMvHoFP7fgrg',\n",
       "  'zAnfCLRqzIPQToUy4CYmog',\n",
       "  'DlWFAkz8y_jFW-U8kXhfJQ',\n",
       "  'uSUkPStJa87O9QRh-I8Pqw',\n",
       "  'Jv4KMzHKs0hJ9egoyaQSOg',\n",
       "  '-WDpXVGLSYJK8I4KSRrgzg',\n",
       "  'uO4Ee3rJacHADUKpkv-nww',\n",
       "  'JtAr1WQ_Mu9RS1f6fjLHHw',\n",
       "  '_c5d4jjcUIMPWi3EVUmy3Q',\n",
       "  'aEw3LC64AQMbrMDuTJnsrQ',\n",
       "  'PXGbJGIswH6HM06F_4U8ww',\n",
       "  'b2m5HkLWSiluQqyuqg0DJg',\n",
       "  'eOIi4jWOEnM6ZFcgksZpbQ',\n",
       "  '5_nOholLhGQWNyt1C27pOQ',\n",
       "  'm4pHFDWndK7EwntYUBCvfA',\n",
       "  'Z8RF1RKG9yZdpxuroQFeZg',\n",
       "  'b3qMInKDNKEC2U366Tdgxg',\n",
       "  'uhlyxrSTxbiBJDROUGbgaw',\n",
       "  'nqXKaghFLAgWKD0W05THJA',\n",
       "  'YeAbLTzwrB4tNZeYZuZgfQ',\n",
       "  'p5le7BotxWx79qAruVdUiQ',\n",
       "  'uOuPCeIWAcm2ax7CjwTIJA',\n",
       "  'nI5Vhn36YvZwfd2WD2IT_g',\n",
       "  '0SuhGdr1VMp8BE-ezSPO_w',\n",
       "  'yF6hf0_TwjCew83WxFsJhg',\n",
       "  'FrgADMNkH_UIuzulPiznPw',\n",
       "  '8fuGIAQDH5YF_Bw9Jo8o-w',\n",
       "  'uK0G2gEEP5Gh0eV11BrMnA',\n",
       "  'wM3b5ez60c3HYml-TtyPoA',\n",
       "  's5V8DsbaJuCnkcrLE6Gmdw',\n",
       "  'wRth9wg0UREiaVJmWuywmw',\n",
       "  'H_4YIK5WgkCRZT9KTQa5Pw',\n",
       "  'jGpaTpvw2uOcCa0uQSaGUA',\n",
       "  'YhgkxFYTCoT_OOW57-Adkw',\n",
       "  'RnplTTLRxwGjBXVxLbc4xw',\n",
       "  'j8g_MYmNxPv828M3Cgkm6A',\n",
       "  'xbBbRzwVNdNZVCDz953UlQ',\n",
       "  '3gqtCj_1sZ_Y6-9AqmV1BA',\n",
       "  'mpW-CZZfYnQY9iEcx5wEwg',\n",
       "  'sAP5PSSLRBcpXdAkq8_QLA',\n",
       "  '_xtrtREZM8vrrAd9-ND_VQ',\n",
       "  'VItY14goOO7BB4rOrD9nnQ',\n",
       "  'h8SasOieU_wqO-xPVfEKXQ',\n",
       "  'GeRpzNV5tGqgwCvgz_1a7Q',\n",
       "  '4ST9rabGUGxSTcC675xvYQ',\n",
       "  'Gj4tPdDtcYbDHHtd09E_hg',\n",
       "  'OXxS5ETAI3ZiHv8Bs0Fzhw',\n",
       "  '8Q7InaT286RPsspm6YU5rw',\n",
       "  '5KxSnaCNKIuaK0WwGPWGdQ',\n",
       "  'bvqpa3YH05hpZXz8yBagWw',\n",
       "  '1VVYwfkvkityIZmpjgsyoQ',\n",
       "  'xzaxS4G1Va_LFGQ4FY6Ojw',\n",
       "  '-LTumz2OsabECAdtGx3DrQ',\n",
       "  'Ur2_t1wC6xBXc64BV3aEpw',\n",
       "  'lhs2nFPCCo9f-hCjojBvKA',\n",
       "  'LP2sJzZq3l3KqUQBgfairg',\n",
       "  'ua684v9pdMFAMei-m7V2Mg',\n",
       "  '4gjDdEOlNWUbp0i6Jpib-w',\n",
       "  'h_j4-CPsNuM05_QrDY7rRQ',\n",
       "  'pjVcK1SB17C3IuUw2BzZoQ',\n",
       "  'phGB3-B8M3GOlxuWySwpVg',\n",
       "  'cx-AJC53kOSopdXpY6TCpQ',\n",
       "  'NUADkjycQaayXWjgjz1CVQ',\n",
       "  'jAFe-fGyI92XE0CLCFoPmw',\n",
       "  'Oy6lXrYOMFOfKWdB9sZH1Q',\n",
       "  'WPDkKwCsMmGcKQKj52hEng',\n",
       "  'So9keDfd3kvdo4w6LAtr0w',\n",
       "  'SakrQmMZY4kXBh7oBKhTBw',\n",
       "  'WybYFZCa4ytmSx051itJfg',\n",
       "  'SV0DjN_Ddqz-krS6VdZz5Q',\n",
       "  'wn9Y_iWmyeg8i1w1w2ZSYA',\n",
       "  'fEzvixffTXjxNUXGg3sZMw',\n",
       "  'Q4zUX5l2wthWT-KM4q4Blw',\n",
       "  'dnBqjh2jCe8u-kk0UhQyug',\n",
       "  'KbD-ylI8ufVB6A6TqZzqqQ',\n",
       "  'fyw2HxDHT0OG1Izh4pCIsQ',\n",
       "  'JhfiVIcJTAyEtyEtk2kIsQ',\n",
       "  'OVAHuPGlAiQC0Xipe-56Tg',\n",
       "  'GWuPisrbNsNDHUMQ11ukrw',\n",
       "  'Ojk2_idBYNzTFLBCc-3Ecg',\n",
       "  'C_bxglwwSi9SN7PLU-znSA',\n",
       "  '2c1t2cRlvDcMR5r6JG4DLA',\n",
       "  'QgVPymRSSzsHLYXO9ir9DA',\n",
       "  'kQOK0A2ZiWpcfz2_oYSatA',\n",
       "  'L8Z4xIiNjiynLYDLFA6tzQ',\n",
       "  'y8dj24xEUBWwggJJt3dx4g',\n",
       "  'nRJTj67mccl_PmzHr8kqww',\n",
       "  'tW7lvIKwjz8i98uyt_XIjw',\n",
       "  'XcgTnc8jiEld7qiaB0NRwg',\n",
       "  'sovgUbP_zos-Tw7BlzORuA',\n",
       "  '4mLajcopPP4iQPPIBRvHLg',\n",
       "  '7UgaQRE4vfH4zHSj71HrsA',\n",
       "  'ZvxPSXK-7keSCXxK3Yk_8g',\n",
       "  'xdCQnecxbvof9n8s0u0UOA',\n",
       "  '90T9IyB-wdpouil319lVyg',\n",
       "  'URmOQGcc8RxMb9Ss8RrQJQ',\n",
       "  'AoVnCkdlMfsu-k-SynIgUA',\n",
       "  'LIH_bGFMcHOcJO4fwn8Teg',\n",
       "  'zkFj3twvYbBVJncd9Q-h9w',\n",
       "  'LcA1jNzuHtlBKiCt-RBqMg',\n",
       "  'XoA8ak1m-mpqYtc36HReSw',\n",
       "  'xmpuX4dJ9WqWFjUyGWrpGQ',\n",
       "  'mYJW4-M4tcRygSwdT3A6Ww',\n",
       "  'uU_2Qbq87vXYDb3PFQC8KA',\n",
       "  'Ncwtr6wiJe7nxhEd3Yx_3g',\n",
       "  'i9HVCZ6NMw1ZOS7zeGYXaQ',\n",
       "  'y8vphlB57Pd8GsqR2y6WYQ',\n",
       "  'O5NPqv-AB1zaxw_21X_rXA',\n",
       "  'yn-C0bAzuAG4HaGDaoU0QQ',\n",
       "  'Veh4zNy8hCKA2w7vNVLnzQ',\n",
       "  'pzv9bxL8e7EDQV18BZeRiQ',\n",
       "  '69wVEa6vxyOF3rguPc5udw',\n",
       "  'QuMT2HlSw218sWW_jHFGig',\n",
       "  '2Mk9Er1UXbO9XeWQNEDcww',\n",
       "  'X4VUnaklXh5JL34lVzQ--g',\n",
       "  'hxCq6htjZaRmVHeasp-rqg',\n",
       "  ...]}"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cluster_details[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LDA Topic\n",
    "https://towardsdatascience.com/topic-modelling-in-python-with-nltk-and-gensim-4ef03213cd21"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:43:01.473055Z",
     "start_time": "2018-11-12T15:43:01.334326Z"
    }
   },
   "outputs": [],
   "source": [
    "from gensim import corpora\n",
    "import pyLDAvis.gensim\n",
    "import pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:43:06.524888Z",
     "start_time": "2018-11-12T15:43:03.699943Z"
    }
   },
   "outputs": [],
   "source": [
    "dictionary = corpora.Dictionary(df['cleaned_text'])\n",
    "corpus = [dictionary.doc2bow(text) for text in df['cleaned_text']]\n",
    "\n",
    "pickle.dump(corpus, open('corpus.pkl', 'wb'))\n",
    "dictionary.save('dictionary.gensim')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:47:58.290886Z",
     "start_time": "2018-11-12T15:43:12.692215Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(0, '0.017*\"crust\" + 0.014*\"like\" + 0.014*\"get\" + 0.014*\"cheese\"')\n",
      "(1, '0.031*\"great\" + 0.023*\"good\" + 0.020*\"place\" + 0.019*\"service\"')\n",
      "(2, '0.053*\"order\" + 0.019*\"time\" + 0.019*\"call\" + 0.017*\"get\"')\n",
      "(3, '0.018*\"us\" + 0.016*\"go\" + 0.015*\"come\" + 0.015*\"get\"')\n",
      "(4, '0.033*\"blaze\" + 0.008*\"music\" + 0.006*\"mellow\" + 0.005*\"hand\"')\n"
     ]
    }
   ],
   "source": [
    "NUM_TOPICS = 5\n",
    "ldamodel = gensim.models.ldamodel.LdaModel(corpus, num_topics = NUM_TOPICS, id2word=dictionary, passes=15)\n",
    "ldamodel.save('model5.gensim')\n",
    "topics = ldamodel.print_topics(num_words=4)\n",
    "for topic in topics:\n",
    "    print(topic)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:48:04.667467Z",
     "start_time": "2018-11-12T15:48:04.374936Z"
    }
   },
   "outputs": [],
   "source": [
    "dictionary = gensim.corpora.Dictionary.load('dictionary.gensim')\n",
    "corpus = pickle.load(open('corpus.pkl', 'rb'))\n",
    "lda = gensim.models.ldamodel.LdaModel.load('model5.gensim')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:48:28.372009Z",
     "start_time": "2018-11-12T15:48:10.777738Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.6/site-packages/pyLDAvis/_prepare.py:257: FutureWarning: Sorting because non-concatenation axis is not aligned. A future version\n",
      "of pandas will change to not sort by default.\n",
      "\n",
      "To accept the future behavior, pass 'sort=False'.\n",
      "\n",
      "To retain the current behavior and silence the warning, pass 'sort=True'.\n",
      "\n",
      "  return pd.concat([default_term_info] + list(topic_dfs))\n"
     ]
    }
   ],
   "source": [
    "lda_display = pyLDAvis.gensim.prepare(lda, corpus, dictionary, sort_topics=False)\n",
    "# pyLDAvis.display(lda_display)\n",
    "pyLDAvis.save_html(lda_display, 'lda.html')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Word Cloud"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:48:34.548181Z",
     "start_time": "2018-11-12T15:48:34.527538Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import multidict as multidict\n",
    "import os\n",
    "from PIL import Image\n",
    "from os import path\n",
    "from wordcloud import WordCloud"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:48:40.513224Z",
     "start_time": "2018-11-12T15:48:40.485610Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10056"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(df['cleaned_text'].apply(lambda x: 'order' in x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:48:46.490352Z",
     "start_time": "2018-11-12T15:48:46.450307Z"
    }
   },
   "outputs": [],
   "source": [
    "text1=df[df['cleaned_text'].apply(lambda x: 'order' in x)].iloc[0]['text']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:48:52.283415Z",
     "start_time": "2018-11-12T15:48:52.278165Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['order']"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clean_text('ordered','v')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:49:00.630924Z",
     "start_time": "2018-11-12T15:48:58.182065Z"
    }
   },
   "outputs": [],
   "source": [
    "dictionary = corpora.Dictionary(df['cleaned_text'])\n",
    "corpus = [dictionary.doc2bow(text) for text in df['cleaned_text']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:49:07.958770Z",
     "start_time": "2018-11-12T15:49:06.933875Z"
    }
   },
   "outputs": [],
   "source": [
    "# corpus = [dictionary.doc2bow(sent) for sent in documents]\n",
    "vocab_tf={}\n",
    "for i in corpus:\n",
    "    for item,count in dict(i).items():\n",
    "        item = dictionary[item]\n",
    "        if item in vocab_tf:\n",
    "            vocab_tf[item]+=count\n",
    "        else:\n",
    "            vocab_tf[item] = count\n",
    "# vocab_tf.popitem()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:49:14.339862Z",
     "start_time": "2018-11-12T15:49:14.332909Z"
    }
   },
   "outputs": [],
   "source": [
    "def makeImage(text,shapefilename,outputfilename):\n",
    "    yelp_mask = np.array(Image.open(shapefilename))\n",
    "\n",
    "    wc = WordCloud(background_color=None, mode='RGBA', width=4000, height=2000,\n",
    "                   max_font_size=40,\n",
    "                   max_words=3000, mask=yelp_mask,\n",
    "                   colormap='tab20')\n",
    "#     wc = WordCloud(background_color=\"white\", max_words=2000, mask=yelp_mask, width=2160, height=1080,\n",
    "#                   contour_width=3, contour_color='steelblue')\n",
    "    # generate word cloud\n",
    "    wc.generate_from_frequencies(text)\n",
    "\n",
    "    # show\n",
    "    plt.figure(figsize=(20,10))\n",
    "    plt.axis(\"off\")\n",
    "    plt.imshow(wc, interpolation=\"bilinear\")\n",
    "    plt.savefig('wordcloud.png',dpi = 300)\n",
    "\n",
    "    wc.to_file(outputfilename)\n",
    "# get data directory (using getcwd() is needed to support running example in generated IPython notebook)\n",
    "# d = path.dirname(__file__) if \"__file__\" in locals() else os.getcwd()\n",
    "\n",
    "# text = open(path.join(d, 'alice.txt'), encoding='utf-8')\n",
    "# text = text.read()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:49:24.660698Z",
     "start_time": "2018-11-12T15:49:20.707756Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA4YAAAJCCAYAAACcfdkbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzs3XecXVdh6Pu12+ltzvQqaTTq1ZYluWFjY9wBAwZjB0NIQgp5effmAy9ckgB5SV4Sbkh4yU1CEsJNICGhgxtgjIULtmUVq7dRmV7PnDOnn312Wev+YcsRxhDAkkbS+X0/Hz545uyz91r7r/lpl6UppQQAAAAAoHHpCz0AAAAAAMDCIgwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDjCEAAAAAAaHGEIAAAAAA2OMAQAAACABkcYAgAAAECDIwwBAAAAoMERhgAAAADQ4AhDAAAAAGhwhCEAAAAANDhzoQcAAACAcyPelG5u6eldMjs2crJn2cq19Wq1opSUXQPLVx94+vvfKc/P5175HV3X9VgsGPd96em6bjiO5yil/EDADAkhlKZpuut6TjQajNdqblVK6du2ay/A9ACcRYQhAADAJSoYjcaSre2dgXAk0jWwbLVr12vHdm5/Siql7HKl9Grf6e5uWrRsWcfq+flK9vLLF18zMjJ33LbdaiBgBItFO792bc8VuVx5NhoNJqLRUPzQofHdpmk85Xm+f77nB+Ds4VZSAACAS5BhGEYi3dwWikbjTq1anZ+eGh87dnh/rVQs+I5TD0YisVf7XrFYy3ue9Lq6Un2lkp3XNE0LBIyQbXu1jo5kT63mVEzTCHiedAcHpw6MjMwdJwqBi5+mlFroMQAAAOAc03RdU1K+6h9+bemmnlQs1jI4Orb3v9qPrmvaunV9m31fesePTx+q19362R8tgPONW0kBAAAawI+LQiGEuHbjujf3tLYuFUL8l2EopVJCiB1nc2wAFh63kgIAADQwTdO0jcuXXR8Nh5ILPRYAC4dbSQEAAC4SsUg4edOWK951+YplNyTjsRbLNAOaJnQplf/C0cHvf/aBR/5ACCF0TdNWLVm0+bZrrvzFrtaW/rrj1A6eGHr20e07/20un58SQoiu1ubFN27edM+a/iVXblw+cJ3rec7YzMyx08f69Fcf+PDhoZEdQgjtpV8pKaXSdf3MCwtKCKFJKeV5OgUAzhFuJQUAALgImIZhfuAdd31808rlN35vx+4vZgvF6avWr7n9ilUrb/rSd7f95eO7dn9JCCF0XdeuXL/m9vffdecfTszOnTp0cmh7MhZtfuPWK+7buHzgupZU8v1z+cJ03fHskcnpI8VyJbuir+fy2fn82KPP7fy308fLzOcnli9ffnlHR0dfLBZLjYyMHA2Hw/tvuummNwshhOu69UKhkG1ra+vp7u5+bmJi4tRCnRsArx1hCAAAcBGIhELxazase9OTu/d+7V+/9d0/E0KIzpbm73/it379IcPQzbHp2eNCCNGcTHTee/MbPnhseHT3J//tSx/wfN8TQoibtmy65zfuvusTV69fe6cQ4p+yhcK0EOJByzStt97wug+MTc8cf/CpZ/7pzGNu2bJlZb1er3me55TL5bznee7IyMjRZDLZUigUsq2trZ3lcjnvOA7rGAIXOcIQAADgImCZZlATQqvUasXTv6va9ZJSUoaCgejp3y3v6728Ld3UO5ObH3vz9df86ttuuE4JIcTSnu4eQ9fNZb09G3/aY+7evfsJTdN0XddNwzCCgUAgmclkiiMjIydc17VPnTp1WErpmaYZjkajrdVqdU7xnBJwUSIMAQAALgLlWjU/OjN7bMPyges2LB94uliuZN+4ddON4WAwdvjU8PbT26UT8fZwMBDta29bnorFWs7cx/HR8T0TmczJn/aYXV1dW3Rd1z3Ps03TjAghZDqdXl4qlSY8z7OVUn69Xi+dDsdarZYTQrCmIXARIgwBAADOga6upt7m5nj7iRPTh2o1p6brmpZMRpp0XTOy2XKmv79t+cxMYbJSqZd/mv05rlf/8mPbPvUbd7/1f/73e+/+q0K5khVKqAefevYfdx85tu30dkopqZQQjzyz/V/2Dp544pVX8Co1u/ije391HR0dGx3HKVer1YxpmkHP82xN00QwGEw0NTX1K6V8IYSWzWYHLcuKaJqm/Zc7BXBBIgwBAADOgVgsmNi6tf/6K65Ycu2mTUt2NDXFTtxww+o7jhyZ2NvcHBPvfe9198/MFCaXLet40jQNa/Pm/uuGhzPHd+8e/kGlYv9ILGpCiFAwGJPS9//90cf/fGhi6pDjunauWJqp2v+5/XQuN1Kp1fJN8Vjr+GzmhO/7L1/B0zRNe2UoSqWkUkrqhvEjy5gdOnToy0opT0rpn44+TdMMKaVnGEbgpd9pnufVNE0zpZRcLQQuUoQhAADAOTI7W5x84YXh5269dcPdQ0OZwYmJ3HBTU7TlyJGJvZlMcfqZZwYfq1Ts0q23bri7UKjmli5tXzk+njslhBh85b40TdOWdHau1jVdr9n1smkYAcPQzTbTsJKx6GyhXMkJIcThU8M7Dp4ceu6mrVfcu+/4yadbm1InhBLKsszg2v4lHbFI5HC5Wi2c3q/v+/4nfuvXJ7pbW5b2trcNVO16Wdc1o1iuZOuuO3ceTxeABUQYAgAAnCPBoBXq7Gzq9X3pRSKBaFtbolvThAiFApFy2S62tye7p6fVmON49VDICh87NnVgbq4082r7ClhWqFSt5pPxWMtvvuOuP/el8oUQQkrf33Vk8HstqeRfzOUL01W7Xu7taPuz97/lzj/8v9/19k9Nz2WHfan8aCSUiIbCyT/+7OfvF0IUztz3Y8/v/Pd333bz//gfv/gLn53L5yeUEupLjz/5V339KyvT40MnTCsQSKZb2tx63Y6n0i2OXau5nuuYhmnZtUo5NzczdR5OJ4BziDAEAAA4B+bnK3NTU/nxpqZI865dp56WUknbdiuO49UDATO4d+/o8319zf2eJ/1du049vXhx24pisZb3POm+2v6uXLf69jduveLef//OY38+Oj17zPN91zQMa1Fnx6q733D9b41MTR8VQnxWCCHGpmePd7Y0f+RNr7v6ly3TDPpS+sVKJTc1lxsqVCq5WDiccH3fUUopXdN0w9C/pWuavqirc3XQskJz+cJUre6W+latX59u6ejyfNddtHTVurmZybFUurU9FI5E87nMjO953tT40AkhBGEIXOQIQwAAgLOopb27J5Vube9ctDFZ1yKhatmzg4mOjqRWCA5N+JP1mluJJjpa9GAstmfv6I5FA6vXSynlVFbNh+LtrUtW9ERWrt/sHTuwe7tS8uXnAd+wedM9xUol940nfvDpM58bjEXCO95y/TW/1tPeOnDmOPKlcmZqLjs8lc0Nl6vVvBBCWaYZDJhmcOmivs2GoRtKCZWMxVpmcrnR6WxuZHIuO5QtFKen5rLD8WQ63S2lXLl+8zVHD+x61qnbtbaOnkXjI8ePmFYguGjpqnWToycHC7nMq17hBHBxIQwBAADOomWrL9saicbjiVS6tVotlwzDNJSUKtnU0jo7NTocjSVSoXA0apqmlWpu7Uil29pN0wpI6fua0DShCTE9PvwjS0oELStkGLqlvfgempcN9PasDwUC4fniD9+CqmmalohG0wHLCkkp/blCYSoVizW7nu8s7upcPV8szghNEwHTCJqGYXW3tSydL5RmTMOwhBDDjl2rDg0e3DN84vC+fC4zM3bq6MFwNJaolArzTt2uTU8Mn7Sr1XKpOJ89x6cUwHmgsQYpAADA2ROOxuPai3Tp+94Zb/PUHKdeNwzD0HXDUEpKTdN1pZQ6vY1SSum6YbiObTtOvX7mfm+7Zut733P7LR85PDSyY3B07AVNaFp3W8vSdQP9187m5kc/+W9f+sBsbn789Pa6rmnRUDgphFBCCO3FW0910/G8esC0Qp7vu6fH5fu+F7DMoOdLVwkl7LpTFUKIeKSprSPVt2quOHlC0wxDKd+XUvqarhtS+p6hm5YvXdd2qkXHq9fO1zkGcPYRhgAAABcBQ9eNqzesveO6yze8tTWV6lZKqfliafbAyaFnt+184cuFcvmsX7nbNHDDPaZuBWYKY8f6WpZf4UvPNc1A0HXr1Xg41SaEEMOZoztHZo/usJ1q6WwfH8D5QxgCAABcJAxNMxYFAgMBTQsete395/p4i9tWbQla4VjZLmQSkaaOUi2fMXTT8ny3HjCDEamkX6jMTeQrc5PneiwAzi2eMQQAALhIWJpmbY1EbkwZRloIcVbC8PpY7LYDtr0r53mZV342PHtkxxk/HjgbxwNwYdIXegAAAACXokQi0d3V1XV5IBCIvNrn6XR6IBAIRH+WfbpKudsr1ccfL5UfOBtjDOl6+C2J5HtbTbPzbOwPwMWLW0kBAADOgba2tjXLli27TUopp6en9+Tz+eGBgYFbCoXC6Nzc3LENGza8p1QqTQ4PD39/bm5usDcQWHJlJPKGRVZgWbNptH+3VP7qzfHY2w/Z9gsPFoufNzXN+lBr25+3m2b3tnL5wS/M5/5GCCEiuh75YGvr/zxg2zuvjcZuDmpa+Pvl8kPfLhW/2G1ZS+5KJN/7pXz+78dcZ8jUNPN329v/+kvz+U+nTaPt3lTTb24Kh68d99yhipSl3dXaU5/Ozv2REEJsjURufEsi8Z6obsSfr1a2fbtU+lLW8zKWplnvSKXef3Uk+kZD08yjtr33a4XCZ8ddZ3hBTziA14RbSQEAAM6Rcrk8MzExsWPlypV3tbe3bxgcHHywqalpaVNT05JqtTo3MjLy5Pz8/JAQQoQ1LXpdNHbb05Xyd6JGKPGuVOo3vlUqfnFjOHz10kBg9X7bfv5/zWU+9q5U0290W9bi08fQNc28MhJ9Q8Iwmv49P/+3zYbR9r508+8cq9v7TE2z+gOBVSFdiwghhCaEtjwYXBfR9dgLtdozBd/P/l57+9/87Vz2Dwbr9gFHKVsqpTZHote/LZl630OFwhcK0s++NZl83zt041fDuv6pLZHoNddGY7f+r7nMx5QQKqEbTVUlywt0igGcJdxKCgAAcI7U6/VipVKZ0XXdUEp54XC4ORAIRH3fd+r1ejEcDjdblhU+vX1R+vlj9fr+PbXaM1nfm9ldqz1tS1WN6kbCV0oWfD9be5UIK0k/v61cfuCFavXZx0qlb+Z9b64vEBh45ZqHpykhRF1KO+/7OU8pJ+/72TnPmyn6fkEIIV4Xjd7qK+UHdS2UNIx0yZeFVaHQZR2m1eMoWdeE0DpNq7ciZWmvXXs253lz5+wkAjgvuGIIAABwDtRqtVw+nx/yPK8+OTm5K5/Pj3R0dFxWrVbnMpnM4Uqlkkmn08tCoVBKCFEUQghHKUcK4XtKuTWpqi+ucSg0XfvJ/5jvKuXMed70f/4sXEvTAq+2rS60n7gvQ9OMj7V3NC0OBFZcH43d4QvlCSHEnlr1mZqSlX22+/yT5fLDtyXi77pORos7q9UnE4bxnaLv53/mkwTggkEYAgAAnAOlUmlKCDH10o/ffen/j52xyZyu68eklPKM36kz/uOnfhGEUkL5Sniv/L0UQiohpKFphhBCxHQ9EdS0l69QKiGUEkKdeWVRCSEr0i/vrFaf+Ewu+2dVKUunt/WVkkIIYWjaP22vVh6/Nhq79fZ44l0znjchhHj6px0vgAsPt5ICAAAskFdE4WvxqhFpS1UVQogVweCGZtNsvSORvDesaS+/JbUiZbEmVXV5MLg+ZhjxqK7HpFJqd7X21LJgcO2aUGhTWNdjiwKBZb2BwIChaXp/MLhyUSAwMOf7M0fr9j5d0/Swpr/qm1cBXDy4YggAAHAB8IXwbSlrvlK+q4RbV7ImhZB1KWueUt5ticQ7P9HZdc/iQGCFLoT+dz09y75bKn0tpGmPVpWq+EL5p/dlK1nxlHAnXXfk8XL5gTviiXvvjCfuO1i3dx2r1/f7QnmarmuaYdS+US7/61uSyftvT6ff/WSp+LBmmn8T0vQnopVK070trf8tYZrp+Xp9+uFi8d/GPW+oy7IWvSOZ+pWkYTTXpKzsqFWfOOrU9y3kuQPw2rFcBQAAwHmg65oWDmsxK6CFfF94QgnlecqVUviGKSzpCs8UmqUMpXSh6cGAFq06shQy9YgnlePYqmYoYWqapgnx4u2jrlCOp5Qb1LSwo1TdVy/GYcQwYv6Lzyo6mlJaQNOCmtA0TyhXE0J3lXJES2u7uXT5Br1YLAQCgYTV1rHInhw/5liW0MLhmGHX68FYvN1qbeutH9r/jOM4pZiujFSt7PiO61qGGVJCybqUtmFZZq1eL1dqtWLVrpc83/+R21oBXNi4YggAAHAeBINaZP368HWxmN6kG8IsFGRmZtobcRxld3ebA5OT3knLEsFkymiJRfVUNKon8wU/owmhzeflzPCQc7BS/7HLQlTP/CEUCrZEg4EWz/ftqULpoBA/+vyh2bs4piqVon3q+F5xw833OELUVUtbq+46dT3d0qmFQuHy/heetlJNMT+eCGnBYFOvUzHWtCxZXrXrJaWkVEoooQmRisVax2Yyx6u2Xdxx6Mh3X3ksABc+whAAAOA8MAzNjET0ZLks821tZt9Q1pkan3AGOzqsJelmo3Niwj2+ZElwbaHgzzWljY56XdUiYT2ezfqToyPOkXpd2j/1sXTd6kolNtQ9rySEOPhq28hSIaccuyY813EP7X9WC4fjqlYt6fFkWhbyc8L3fU3XDe/k4D4tGAwrz3DHTxzfXcskxgxdN3Rd1yu1WlEIITRN0wOWFQoHg1GuFgIXJ8IQAADgPKjXZfX48fpuwxSB6Wl3OJv1J6UvvGpFFk6ddPbNz/szw8PuoXzen52c9E6Ypha0bVn2feE5jvqporAvne69c/26O3tSTXvrnlfWfsLSFDI/f+bag8d+3HZaMBgy2jr7VLmUn52fnxVCjL/adulEos3x3PpPM04AFx7CEAAA4DxwXeWKVw+w6Zf+J4QQB17LMVpisZbb16y5fWhubnLn8PCuoGXGX8v+hBBCuK7jT08MKdd1f9JmuWJx9jUfC8CCIQwBAAAuMbbrFnOV6tDpF9W8FurFJTXO1rIaAC5QhCEAAMAFIhYKRjviiQ7LNKyIFYjMlEozqUg4ZWi6cXx29rjtunYqEkn1pFI90WAwKpWSM8XizGguNyrPeNW8UkJKpWQiHIqv7+lePFsqzU4XijNCCNGRSLT3ptO9hq4Z89Xq/FhufqzqOLUzx2Fomt4e09qDpgjOVdVcPKDF40Etnq2qbNlR5faY1q6UUJmKykQCWiQRFImyI8rZqspKXnkPXJQIQwAAgAvE8rb25R++5eYP5yqV3LK2tmXHZ2ePm7puDrS1Dnz0wYc+unN4ZOdlvb2XvXvrlneHLCsUMIxA1XGqv/fAg78nhBgTQgj10mL38VAo9p4rt95/9dL+qz/zg2c+Y+h6pjUWa/3gG9/wwd6mdK8QQri+7z6wf/8D4UDgkZrjvPwcYyQgIvets+7L2yr/+JD3+JU95pWrW/XVDx5zHxzJi5HrFxnXt0b11qeGvaeWNOlLmiNa854puSdX83PipeMDuLj82AeSAQAAcP4tSqcXPXLg4CPfOXToO2u6Otd8YefOL4zm5kfXdHWt0TRNOzYzc+zvnnzy7377y1/57d/95gO/25VMdl3Vv+Sq09/3pfQNXTfu3XzFvVsWL97y6Sef+vRTx0885Usp33rZxrd2JJIdH3vooY998Ktf++C+8Yl992/dcn97PN5+5hh8KfxDs/JQrqZyri/c0bwc3T3p7949KXcnQyKZqapMtqqyi1L6oqApgtvH/e3bx73nfam45RS4SHHFEAAA4AJSceqVQ1NTh3Rd0zOlcmb/+MT+2VJpNhoIRLUXHxrUNvT0bLhn06Z7IoFApDOZ7OxIJDpOf9/1fffuyy+7e2VHx8oPf+MbH941PLLr9O2db1y18o2LmpsX/f7tt/2+EEJ0JVNd67q71iXD4eSZYwiaIqhpQtvUZWyaLKnJkqNKGzuNjVt7jK1tEa1tU5e+KW+LfNlR5bon6q4vfuKLaQBc+AhDAACAC4jj+Y5SSkmlpCel5/q+K5WSuqbpzdFo80fvuP2jvpT+wwcOPJyrVHNLWlqWnPmSmaZIpClgmoFEOJRY1tq2bNfwyK4zP9s9Mrr7WwcPfuvMY47ksiNn/lxxRGX/jL//WNY/NltRs54UXqHuFmxP2INZMXh4zj9se8KuuaImhBI1T/zQM4oALj6EIQAAwAVFvfycoBBCqDNe5pIIhRKb+vo2/dG3vvVHjx899ngiFEpEgy9eSTy9Td3z6v/w9NP/4PvS/9DNN32oYNcKlmE86vq+d2hq6lDQNINPHT/xVM11a6aum4auGyW7XjpzBI6vXCHExCsGNnrGf+fO8qQBLDDCEAAA4CJRdZzqcDY7/IaVK94QCwZjqzo7VwUMM6B+6I2kSlXrTvXZU6ee7UgmOt5/7TXvL9frZcswnlnd2fmZP7nrzX/yoTfe9KHhbHY4GghEy3Wn/NUXXviqECK/gFMDsMAIQwAAgAvEXLk09/CBgw+Xbbs8ks2NfPfw4e+6UrrPnRp6rly3y7Ol0uwnH/veJ29fu+b2DT09G3aNjOzaNTyyK1et5l78fnnukYMHH5ks5Ccdz3ObopEHdU3T05Fo2jQM88DExIGPPfTwx25fu/b21Z2dq4u1WvHI9PQR2/NefiNpMhpPdre098yXCrnZfG4mFAiEDMMwi5Vy8dXGPNDVN1ColguZfC5zvs4TgLNPY6kZAACAC8fijtSinpZYz7OHJp79SWsCRkNWxPOlV3d9RwghgpYRMA3drNhu9ec9djgYCl2/fssN16y57HXbj+x79tnDe565ctXGq8czU2OZQm6mv7NvIB1Ppo+Onjri+K6zZtHA2lV9S9c8sW/HtheOH9r98x4XwMJjuQoAAIALSFdzrOuygY7Lznxu8JWCATNw1eruq3rbEr1CCGEaurGqr2XVZQMdl72WY3u+79ccuzYyOzlycmr0hOO6jq5reltTc3tPa2ffko7u/mK1XLzzyte/+fJla65wPNd1fc/VNY2/KYGLHLeSAgAAXGAuG2i77OP3X/Px/+stm2a/8tTRr1w20H7ZFcs7rxBCia//YPDrt2xaMnD/G9feP50rT9+6uf+Rzcs7C++/Y8P7LdOw7nn9qu7H94w8/parlr1lUXti0US2PPGNHwx+47Yt/belYqGUoWvGM4fGn3n+yOSOVx7X9Vx33ZLl2clQeGImNzft+Z5bqlaKyWg8ZTtOba6Yn9t/6ti+27Zcd8d8uTi/7+SRPV3NbV2CRe2Bix5hCAAAcIGZyVdn/uaBF/7m1s1Lbr3vxtX3RUJW5Pmjk89HQ1b09i1Lb//iE4e/2N+Z6n/uyMRze47P7AkHzfC2vaPbXM93v/nM4DdX9jWvXNqVWvrVp499dcvKzi1Xr+m+ekVvesUzB8efeWLf2BN116v/uGPbTt0uVssFX0q/I93aefnA6k2+kr4vfS9XLORcz3OHpsZPDU2Nn3rTlTfeJZWU5Vq1fD7PD4CzjzAEAAC4wJSqTqlUc0q249ttqWhbUzzUFLLMUKZQzew+Pr3b9aTrSek5nnQcz3ejIctyPd89/czhdet7k0u7m5Yswzw3AAAgAElEQVS+fkPf62t1r3ZyMn/y8oH2y8fnyuMV2/mJzyAenxg5LoQ4fnooQoi/fJXN/l4IIXRdf0pKydVC4BJAGAIAAFxguppjXW+6cuBNnc2xzm17R7b1d6b6a3W3NleozQ1NF4aqdbcqlBDXrum5dlVfi+1LNVSte9UrlndcsXlF5+hcoTa2/9Ts/lNT+VPFar04m6/MnotxEoXApYMHhQEAAC4gQ1OFoe/tmHy8asva/hO5g88fnt7xzIHp56KiJRXUw+Elqd7lugwY+48XDtrlgKsJXXc96b5wfPqF549MPu940pmer0w/sv3kI1XbrdYdv+5L5T+0/cRDE3PFVy5aDwBCCJarAAAAuKBs6Fx5hVBCtcdbuj3puxWnWpouz00uTfeuCBqBYDqSaq04tfK8Xcg6nlvfP310V8WpVV7rcYPBYOiqq666Tdd1fdu2bV87G3MBcPHgVlIAAIALSMgMhqOBSCwZiqWE0LS2aLpTE0JLh1MtQghh6IZZc2uVbCU/m44kWw3dOCt/z+m6rieTyWbtVZaeuOqqq25tbm7ueOSRRz6nXnFVwQgZMavZ6nPz7qRf8fNnYywAzj+uGAIAAFxAdE3TNE3XlVBKEy8uZXhmjGmapiulpFRSGbqh+9KXQgiRTCbT8Xg8NTU1NZJKpVqam5s7Tpw4sT8WiyVjsViqVCrNp1KptmQymVZKqVwuN5PJZCY9z3ODwWBo6dKlay3LCmUymfHJyclhIYSwLCuwfPnyjTfffPN90Wg0/vDDD/+zlFIeOXJkl+u6jhBC6JYejG+I36EZmlXYXfi68pR7/s8agNeKK4YAAAAXEPliBPo/YZOXPzsdhUIIsWbNmi1bt2699XOf+9yf3Hzzzfdec801d37kIx95+/r1669dt27d1U899dQ3rrrqqjs6Ojr6AoFAqFwu57/5zW/+oxBibzQaTdxyyy3vXrdu3dXbt2//thDi40IIEQqFIjfccMPbN2zYcK2u6/qdd96p+b7vDQ0NHRZCOEIIIXRhGAmjVdqydC7OB4DzgzAEAAC4SBi6/vJtnr6U8szPCoVCznEcO5lMNre1tfUODw8f6e3tXZZIJJrq9XptbGzsRKVS+WIul5tuampqu//++z+8Zs2arUKIvblcbjYWi33svvvu+9CZ+yyVSvloNPrHjuPUlVLys5/97P8rpVR/+qd/+p8bKaH8op8x4kaLpmnaOT4FAM4RwhAAAOACF7QCgWQ0lmxLNZutqXRbxa6Wu5rbapPZ2cnT2xSLxVy9Xq91dnYuVkqpkZGRo8uXL7/MNE1rfn5+tl6v17q7u/tvueWWX0gkEun+/v61Q0NDR17z4F6631UP64lgd3CNEGLPa94ngPOO5SoAAAAucLquGyt6+1cJTYh0IpW2TCuQiMQSZ25TKBSytm1Xly9fftn8/PzsqVOnDq1cufKKYDAYzmQy47fccssv3HzzzfcdPHhw+7Zt274yOjo6qL3ktYxN0zVdKeULX3he3pt6bTMFsFAIQwAAgAtcNBSOdDW3dbc3tbS3JJpa25Lp1kgoFDlzm3K5XKhUKsWlS5eum5mZGZ2cnBzq6OjosywrODs7O7Fu3bqrh4aGDu3Zs+eJ2dnZccuyAj/NsaWU0vM8x7KsoGVZQcMwjDM/V77yNUsLC10YVqvVfzbnDeD84VZSAACAC9xcYT4rhPiPl37c+2rbSCnlTTfdlGtubu7I5XKztVqtXK/Xa7FYLDk1NTU8MjJydOnSpWuvvvrq2zs6Ohb39PQMHDlyZKcQQvT09CzduHHjqu7u7n6llNq8efONk5OTQ9PT0yNCCHtkZOTYDTfccPfNN998Xz6fzwQCge86jlMX4sUrhmbUTHt5b6o2VNt5Ps4HgLOPMAQAALhEnDhx4uBTTz314MzMzGitVqs88cQTX/d93yuXy4VHH330C69//evftnr16q0nT5488M///M9/VKlUikopddVVVy1bv379NcViMaeUUlu3br15x44d35udnR13Xddtamp6IhQKRfr7+9eWy+WuXbt2bRNC1IUQQknle0Vv2ogaaSEEL58BLlKsYwgAAICfmx7UI/F18VusVqu/fKD8qD1uH1zoMQH42RGGAAAA+LlpuqabKbNTD+oxWZN5N+/OLPSYAPzsCEMAAAAAaHA8YwgAAHCJiKebW6xAIKgbphmOxRKObVfzszNT9Vq1KoQQ0VSqKdnS1qHpul7Jz+dKuWzGME0z3dnVmxkbHfI9zxNCiFiqKR2KRuPzM9MTmqZpyda2jnA8kfQ9zytkZqeqxUJBNww93pRuEUIIwzStSDLV5Np2LTM+Oix931/I8wDgZ0cYAgAAXCK23Pamd66+5nU3FTKZ6Xi6uVU3DGPntx/6imGYX44kk0033PveX+1evmKtYRhGIZOZfu7Br32hlMvO/uIff/Izn//Yh39dCHFUCCGuvuvu97Qv7l/2wF9/8g/a+5cuv/Zt9/xiKBqNa5qmjR09vC+SSPxtMByR173zvven2jq6SvPZTOeSgZX5zMzU1z/1id8XQlQW9kwA+FmxjiEAAMAlpLV30dLDz/3g8c9/7MO/fvjZpx+//I23vTWSTDZtuPGNdzZ3dfd986/+/GP/8f99/Lcdu1bbdPPtb6uVSsXJk4OHV1559Q1CCBFNppp6lq9cO3Rg704lhLrqzW9/9+iRg3v/9eMf+cBDf/v///Gaa6+/pW/V2o1CCGEFQ+HW3r4lR579weOf++jv/Nqj//sf/qJerRKFwEWIK4YAAACXkPFjR/aPHNy/uzSfyy7btHnP+utvvC2WTKWXrr/syp4VK9ff9isf+B0pfa+zf2BVpVCYj6ZS6QNPPfHolW+66z7DND+zZP3GlbphGBODxw6aViCw9nXX35Jobm7rXrZijRBCpDu7entWrFo/cujAC7qua2NHD+8fOXxwj10pl4UQhQWePoCfE2EIAABwCamVSgXXqZ9eY1AqJYRhmVYwEomOHzu6/+DT33/U9zx3/xOPf6uUy2UKmcy0XamUX/+ud/9a19Jlq7qXrVxbyM7N5qYmRq1gMCw93zv87A8ez01NjAohxP4nHn9k6uSJY6ePZ1fKRbdetxdqvgDODsIQAADgEqKUlEKJH3rtvPSln52aHLWCwfCxHdufrJWKBd00TaWU9BzHtYJBd2j/nue33PGWd2m6rk8MHjlQymXnYk3plrmJseFCZnbq4NNPPqqUlIZlBby6Uw9GItEXj6eUELzmHrjYEYYAAACXOKWU2vHIA19682/+949e/da73zMzfGownm5pmxk+NWgY5vNSSefo8889cd9H//Cvjzz7g8fHjhzeJ6WUVjBY2PXoI1/dcvub3hkIhcJ2pVxKd3b1Pv/wA1/0Pc9d6HkBOHsIQwAAgEvE1KkTR2vlUsGp12pCCFHKZTNHnnv6e7VyqZCfnZlevHb9n11+821v7Vq6bFUhk5meGDx6QCkplZQq2dp2dN+27z08PXTy2OzoyEkhhHDr9XokkfhSvVqp9G+8/EpdN4zZ0aET0vc833PdkUMHXnCdel2xMDZw0WOBewAAAABocCxXAQAAAAANjjAEAADAzyW2NvbGxGWJNy30OAC8djxjCAAAcAkzQkZMs7SgX/Hn9ZAe10N6zM25E3pIj+qmHpSOrBlhI6EFtIgQQsmaLPpVP6+kklaT1SUdWTXCRkJowpC2LPoVf17owgi0BBY1Xdf0BiGECnYEjyipfHfOHVFSyYWeM4CfHWEIAABwCYuti90a7g9vnntk7hNN1zf9SnR19A16QL8rviF+a6gntK50oPTt5ObkOwLtgQFN16z6dP1Yblvu00KI4z2/3vMF+5S902qxFhlRo8nJOENzj8x9Qrqy1vqm1t+LDESuVlL5wZ7gOlmXlcnPTv6SEKKy0HM+2/RgJK4Hoym/nJtS0vde+blmWgEj1tzl5aeHF2B4wFlBGAIAAFzC/LKfEb7wjKjRZDVbfc6UczTQHhgwYkazrMuKM+uczD+T/1c3545Zaau35Y6WD4f7w1uEEMd1Uw+aSbN9+svTv6NZWrjz3s6/DC0OXVHcXfza1L9P/XbbnW2/65W9zNy35j6plFLibxd6tj8fTdM1I97cbSbaFvvl7IRfK2Wt5t5VynfrXm5i0Ii3dIZ6Vl9bGXz2a0YkaVnp7hW+Xc75xcyo2dS1LNSzOm219W8QQnxqoecC/LwIQwAAgEuYV/TmlK88q9nqE0qo+nR9MLwofJke0CNeyctIW5aNhNEWWxO7SY/qTVaLtcgIGXEhhJCurJUPlr/rZJwRIYTo/qXuCStldWq6ZgglLp1X2xtmINC+9HIz1bG0dqo4p2mabsbT3UaspdsWSslaKXv6SqGmm5aZ6ui3rNBav5yd1KxQVHmOLTTdWOhpAK8FL58BAAC4hHllL6N85YZ6Qxu9vDflTDuD4SXhzZqhWV7em0puSb4zdWXqXnfenbDH7P1+yc8KXeiapmlCCumVvLnT+1JSeUIXhhBCW8ApnX3S99y50UOymp8x4+kuq3XxOqEZhnSqRSOSbD29maYbAauld7XyXFsIIQKti9f5xblxZ3Zor6Zpl9Y5QcPhiiEA4KwydF1f1JxelK/W8oauG5lSaU4IITRN141kW58Riqf9cnZSKKWkUysJ3TA10wr6lXxmoccOXIr8sp+VjqyG+kLrywfKjzpZZzTQFuh3591xr+jNpK5Jvac2XNtd3FN8wEyaHckrkm8/4+tKqR/7MhmllPKFLkyhC0PTNP9iXeheM8yAEUt3Gcn2xUIIoTzXtlr71ki7kpe1Ui7QsWxzoGNgk1+ZnzZiTR1msn2xrNfKzuzwgUDnsissf/E6Wa+VFngawGtCGAIAzqqAYQTeftllb58qFKdOzWVOCSHmhBBCKSnNRKuvh6JNmmEGjETrIiF910i09imnWtJM61+V5zoLPHzgkqN85Te9rmneSlt9bs4dk7YsCyWUHtAjbtYdcXPuWLA9OBBeHL4ivDh8ebAnuK5ytPLEf7VfWZdVZ9Y5FVkReV1sXexmv+znNEPbqXzln4dpnV2+63jTg7vd2ZN7le85uqYZ9vALj2lSKkMThhBC1Mb2PeO4TlHTNEPohqEpqTShRH1k3zappC9+fEADFwXCEABwVrm+737vyNHvVV2nmi1Xsqd/r+m6bqW7o0pJXwihecXZYc0MhP1aMetX5qeIQuDcscft/dUT1WfdvDspbVkoHSw9Jmsy71f8fO77ub9vel3T+5qua/rl6qnq83Pfnvukl/PGlFKq7S1tT3tFb+bl/YzYL3h5b0op5StfeVba+pYe1GPJzcl3egVvyh6x9wghLqowNA090J1KrNc0odmOV1DCEJomtIBhRHRds4QSypPSiYWCbbZj5QOmEbVdr6hpmmboWmC6UD5ke+4l9yZWNB7tIr3iDwC4QFmGYd64csWNVcepjs/nx4fm5oYXekwA8OOEA1Zyy5LeX6p7XsnzZd009GC+ao+3xqPLKnVnrlCzJz1f1pLhUE8qEuo1dD1Qc9z5+WptRNOEPpYr7KzUndxCzwN4rbhiCAA4qzRN05KhUFITQjN03tIH4MLmen718OTsw1JJz9B0S2hCk1J52XL1pOv7VUPXAkoJWfe8ynSxdNiXyhFCqbrrlwKmEXV9v7bQcwDOBq4YAgDOKl3TtI5koqMlFmupOW7t+OzsidOfdVhW9+Zw5PXbq5XvZbz/vD0NAAAsLK4YAgDOKtPQzS2LF29pi8fbxnLzY0KIl8MwoGnB62Kx2+5KJt/7oba2PU+Vy9867jiHKlIWHSl5xhA4j3Rd16WU0rAMy7QMq151qmd+HggHwr7ru7734vp9ZzJMw7BCVsQu2xfdmzhNw7IsIxASmqYLIYRSSrpevRa0QjFN0zXXq9eUUMoyAmEhNE0q33O8esUyAmFDNy0llKq71ZKhmwHLCAQN3Qx4vms7nl0NmKGopumG5zs1z3froUAkIZWSQijhek7Nl96PnEvgQkEYAgDOKl8qf+/Y+N6AaQZs98W1vk4bdZxTlqa9r9Oy+rZGIjfek0r9WkWq8r5abfu6cHj/pOuO5H0/6/N2P+CcijfHWlv60gkhxMnuFZ3rll6x6GoraP6jW/de/gealVcvvXHqxOxRIcTJV36/ta956da3Xf4eIcTvn8dhnxW9LcsuX9235bZoKNFStcvzlmEFf3D44X/YNHDDPfFwqn0kc2yn7VQKaxdd9aZyrZBxfcfefWLbf/S2LNvU0bRolWUGQtuPfedfutL96/pal2/qbl664dT0oWcPjjz70MYl171N1w1ztjA2eHhs53fevOWX/3Qsc/wF13eqR8Z3PSaEGF3o+QM/DmEIADirdE3Tl7a2Lo2HgvHR3PyoEGLszM9dpVwhxMmgrk/sq9WeuyWeeMfbU6lfnve9uaO2vefZavWxgKbtdJTiCiJwDiTbEh2X377+rbGmSEv3io6vR5KRaLw51j6wuf+6/k2LK5PHpg+GooFY98pOt151ylbQDLb0NffH07FWI2AEsmO5IcMyDF3X9FhTJJ1qT3ZnJ/MjtWKtuNBz++lo2lxx6pTve+6p6YPPrF181Ztst1IcnNz7RFOstbc10bU0W5oezpamh588+I2/3rT0hntak93LpvMjhwvV7OTi9lVbu9L964NWODZXnDxpO5XiSGZwV1/riis86bpjs0eeH+hc+7rh2aM7DN00nz/+2Odr9XJhoWcN/FcIQwDAWWfoulGpOz/y+nZd07Rmw2hbHQpt+pV0+opmw2yvK1X7RiH/z2Oue3JlMLjxrcnk+yZdd0QIMb4AQwcueYGQFU51JLpNywyG4qGkbmhGsjXekWpPdCZa452mZQTmJ/OjG29Z+za7XC/PnMocXf265TeH46FkKVvJ9Kzq3HDs2ZPbQrFQYvmVAzeYlhHIzxYnF3pePwvf91xPunXHq1cN3bT6O9Zc3ZLoXur7rhM0wzElhCrbhYyUvi+VlOl4+6Lu5v71udLMSDSYaM4Wp055vlOLhVp7Zgrjg1O54UNtye5lkWAsZRmB0ODkvieqdinnePUaUYiLhb7QAwAAXFqUEMqTvreqs2OVpgntzM8WWYFlH25r/9St8cQ7XSXcJyrlh/9lPvepr+Tz/7S9Uvn+NwqFf4nrRiqm68mFGj9wqcvPFCdHD0y8MLJ/bOfJXcPP2uV6KTuRH9n73YMPzA5lBpu7mxZlRrNDs0Nzx6UvPSGEcGy3NrRvbMcL397/1URLvCMQMsOLN/Ru6VvbffnJF4afKef+c83Si5FpBILJaHO3ZQbDUrx0K/sr3tAYDSWaE5GmDql839DNgGUGI9FQIp2MNHf2tiy7fCRzbGfNqRQSkXSnoRmmEkoJ3vGIiwhXDAEAZ5WuaXrYCoSfO3XqOVM3TF3XNSmlEkKIsvSLj5VLXxu06wcyvjdVkbKszvjjq6pU5YFi4fM5388s3AyAS50SQilpBsygbuh675puza7YJSWVlL6SmqEZuq7rN73/Ol3XdUPTNN1zvLpruzXpS0/XNV1JoWZOzR4tzBYnu1d2rjdMY/LVXlJzIZqaHzqUKU6cMHUzYAQtY0IOPTecPbq9plWzgXAoWJjKzNiuXamWCnnPd+sHR7c/IpRSwUAkrgvdkMr3EpF0R0eqb9XBke3fqru10pZlN93/4M7//XvlWj5j6FbA8Wrlar08/9i+L35CiL9Y6CkDPxXCEABw1kUCVuTagYFrHc9zqo5TtQzjqOv7Xk2pymC9fnDcdU65Sv3QH5HNsUj/svaWlSOul+tNxa4Pmua3655XXqg5AJcqz/Wd3FRh/Kq3b3qPJjQxP5Uft4t2XgghvLpnS1e6a16/4tbe1d0bks2JtuE94ztkXTnCF76SSpZzlTnHdqvjR6b27/ve4QfXvn7F7R39rSuEEIcWeGpCiBfXUo0kEknf8z3fc13DNE0llZJS+qZlBaQuXU9zpSudqqgJ4Vuy1r9m/frmts6eydGTg/0bNqxLNbd1fufr//J3UkkphJh9adcv/4NVNJQotCV7V6zs2XSTEEI7NrHncder20KIoVcM50de3ANcqFjHEADwcwuaZmBFR/sKTWiapgnNl8ofzmaHr+pfclVzLNZ8Ynb2xM7hkV2nt98QDm95d6rpv30yk/l/Zjz3h55JaolHl/Y2JbeELCtRqNXGJwulfflKjecMgXMkFk0kW9KtXUpJadt2NRqNJ8OhUKRYKuTsul1rSqZbS5Xi/KLuJSvncplJpZQ6OTJ4UAgh2tuTncViLd/cHG+XUvpKCTE3V5x2Xd9d6HlZwWDw+nfe9yuZsdEhoWkiEk+kHNuu+Z7rxlJNzcXs3IxuGEYpl50rZeZmBlZu3CyUUpVSYb5cys/HE03N1Wq5ePLovt2+x/ISaBxcMQQA/NyCphm8btmy65qj0ebJQn6yO9XU/bnntn8uFgzFsuVytq8p3deZTI7Plkqz/ov3k6qaUhVbyeor91X3vLInpe1JGZRK+GXb4XZS4Bxqa2nv2bB60+uK5ULO1A2zKdXcruu6fvzU0X01u1revPHqm5547rGvh0ORmGEYpqEbphBCGIau3333ljtOnpw9umpV9wbP8z0ppf/oowe+LoTILfC0hBBCOPW6Xa9VK32r113mOfW6UkqlO7t6a6ViIZJIpnLTUxMTg8cO2tVKRQgx/JP21ZZqbgsFgqGpXGbK9dwfCd/OdGtHPBJNDI4PDyYisXhPa3tvvlzKz8xnp33ps/QOLhqEIQDg51a07dKKjvbHrly85Mptx45te/OG9W8WQgjH+z/s3XmYXFd1KPq195lPzXN1Vc/zKLXmyZYn2cYjQ8AYAiQmgYT3bshNLiE3cJPcXF6cfHl870EIvEAghMCFOBibwTbYYFmyLUvWLFlq9Tx3dXfN45nP3u8PDUiyjGXL1uTz+76W1HWGPqfUVXXW2WuvZRmrGhpWzRUKc/hUE2kAgCXLmp80jOG1snxTgGVfVAipUjhZnsHvknyqYRULijqTLldPWLatX6nzcjjeCfKF7NLQ6NGXq7VKSRQkmVBCbNu2qrVykeN44cDRl5+rKdXywtLcRFWplIhN7NPbEkJJpaIW9+6deB4hQJRSqmnGq274XAnEsq25E8cPm4ahH39xxzOYYRi1UilxgihhhmE1pVYhlmWZhq69/t4AGmOJptXtvWvylVK+q77lSFVTqqvae9csF7JLx6fHX+mob473NrX1ibwwc0P/mo1bV6y75cDosb2/PLDraQBQ3+bTdTjeMk5g6HA4HI5LkqlUM9lqNXtbd9dtmWo1U9X1qsRz0lK5vDSZzU4uFItnUkZdCHtucLnuCLPeuoxlLRqUnrkwO4FgYhci+wSb8WQrytiVORuH452jWC7kAOA3VRM93YP0nJRuSik9cGDqxcnJ9OjbdnCXwLYtGwAOvKX7JLadLeUzNw9uuO3xXb98lGUYZlVH7+pCpZRXdE0hlFKb2Laqa+p8ZnF2PDUzZljWOb1YMUaIEGcOl+Pq5QSGDofD4bgkFrEthsFM0u9PapalMQgxE5nMRNzrjTMIM2evW7Ct7PcLha8xCLHovP0sEXtxhtgzfllqoOBcPDkclxtGGDGYYW1iWyxmWEIpZTHDGLZpIASIwSxLKaE8wzOzs7nJK3285xN5Qehv6VyRKxWyNV2tWbZtMQgzgAAoBcqzLOeWXB7N0LXZdGr2jex7MZ9JHRofOnDTivW3bOlbfYNLlN1AAVyS7DYt2+SYk89XvlLKpXLpVFUtV8M+IRzwiLpb5N0emffcOtiU6G0KpxbztUWvzHsV3VQIoUTkWbFc08sV1XCKbTmuKCcwdDgcDsclqfcH6nvi8Z7R5fRoSzjcwjIMaxFilVStpBjnppYVbTvPIvSUiLEkIeSqEFI0KTUQAA665ZbmYGBzwCU3FxR1GgCciySH4zIKSN7QlqY1t+2eO7TjhqY1ty2U07PdkdaBl2YOPieyvDiY6N2QruYW24KNXd899OP/D66y12hfc/vAus6+9RYhlmlbVtgbCNdUpbqYz6Qi/mCspirVhezyfL5SOmeUtLM+2OESORelQBdylYVMUcmevVzRVKVYrRRN2zaXC7klBAiFvL5wRVHKIscL/c1N/Y2xRHNbXUObZuhaWamWbuhPbuFZm2cZzBarWjFf0fJNUV9Tc8zXvOPo7I6tAw1bBY4VKopRkUVWfmrvxFNwlT2fjnceJzB0OBwOxyWpaFpl38zMPgZhhmcZnmUw2xIKtYgcLx5PpY5jhBA5VQKbRYhp5fnuG13ud7UJQu+/5nNfnDXN8Vae7wECXKpYPlJUtDnddNpUOByXG6GEcAzH90TaVkRcwfhMMTUxlp0euq/n1geGlsePYEDIJ3oCB1PH9+i2cdXNAS5WK4X5bHresAwdAcZj89MjxUqlEPL5w2WlVkkXc8s1ValKvCCd3oZlMPOFh7Y+tKW3fktnfaDzb7+/+28B4J/O3u/QzPjZbTi+BgDAYAafVVjmZQCAh3+9zlRvU7hH5Fkx7JPCNc2qlapaiRBKyopephRooaIXFvO5RY/Me1rivpaz52I7HFeKExg6HA6H45LkarVcWFHCaxsb1y4UiwslVS3tn5ndL/GcpJuWTs7qixRm2fh7ff6H3Bh7ewRhdYhhoinTnF4jSTfWuV09PzX1x03bvqiCEA6H461VM9TKeG76RL2vrnm6sDBe1qulmCvsOpwa2juamxlarGTmI+5gvKRXi5TSq67a5kRqdgLeYN9Ayya2wLH/c3VHbPXf/d7Nf3ex271etdGhmeyJ19nFJABAIuSuKyt6WTUsp0iN44pzAkOHw+FwXJKIxx3pjcd7p3K5qfZIpN3FC66WUKgl6vVEs9VqFs4qXBFh2USQYSJfzmb+x6fDkS8AABiUGnnbTm+QpPd5GdhT1vRFQHD+FESHw/E2I9QiijFWnM6ND06fVqIAACAASURBVGVVkm3wooa8srDY7ENNcRcJ61ZG0wxcwJRYTX7atDLOuKcKZKqs08qVPvZLoZuW0dccKZvW5e/BmMpVFwFg8XL/XIfjQpzA0OFwOBxvGsswzKaWlkaXILhSc3Opnni8ByOEC4pSKKlqybDtc6vyASAC1FYJPTP3kEWICzJMJKdps0uVyjGEEKYUrrrRCIfjekcoEIwAh2UIx9w4lldpfm0CrZFYJOU1yG9oYjYcXSZHu8K4K6/SfFRG0bkSmXv9PV8ZvU3hnvs3d9w/MpcfWdkaWbmiNbpiZC4/8u+/fOXfT8zmhi9mHwyDcW9juPf+ze339zdH+kWeFY9PZ45/f/vQ988eFfRIvPuBm3se2DrQsNXvFv1lRS8/vX/q6Ud2nHjEtGxLEjjx9jXNt9+/qeP+kEcKHRxfOvi9Z49/b2qxOP22PQEOxxvkBIYOh8PheNNkjpM7Y9FOjBDe0NKywbJtyybE3tzWutkmxB5Lp8cAYOT0+suWlSrZpPBQMPinSY5rGZSkzaskeXOfKK7dScmepnBgMyHULCjqLABcFT3RHI53CkIpbQtgiwAipgVm0ouTugV6SaOlggoFi4CVrtF0QEIBi4BlU7DjbhwHgMKVPvYLCbjFwP2bOu4nGyh5au/EU8/sn3rmvk3t9/3N79z4N22JwF9MpAqvW1mVUkpb63ytYa8cfv7o3PMIAXrwlt4Hm2O+5rqg+88W89UlkWeFf/jELX+7sSex8am9k0/NLM/MxALu2OkRSEngxI/fOfDxD9/a9+Hth2e27x1O7b1jTcsd/S2R/s764F+Ozued9jyOq4ITGDocDofjTStrWgVj/O2zH+Mw5hZLpcXFcnkRwbkpoUumOZ/guC/d6/V9WKWkusXlvmPaMEa/kcs9PANkTuL5YE3XM4ph5i/vmTgcDoFF/Oo6HCUUyN4Fey8FoJQCBTgZNHpc7rQoyq6Xs3CIUEqmNGZBMw3V4/b6VU2tWZZ52VMxX49b5Nw/eG7oB//4+P5/VA1LPTC2dOCf//hd/3zLysZb4NQ8v5Mu3CGHEEoxRj97Ys/EExQoxQjhVK6a+tyHNn2upc7fAgBLqzviq+9Y23LHn//Ljj9/Ys/4E2e32yGE0lXtsf7furHrtx7fNfr413528GuaYWn7R5f2f+NP3vWN29e03A4ATmDouCo4gaHD4XA4Lgkh5JwrKgZjayKTnagPBOpPjRie4WEYbz3HBX9QLHztW/ncPyAAbFFq/RUABF1yU33Au9q0iTq6nP0VAJQv53k4HO90NgF7KEOGNAs0i1D7/OWtTV0D3e19ay3LNAghBCGMlrOLc4ZpaFOz48fhKpwrV9PM2lymMlfTTAUAoC0RyGdKSiYedMXPXfO1pzX7XaK/qyHYFQ+44rLIyd0NoW6WwazIMyIAwIqWyArTss0dR2d32IS8Kg0+4pMjybAn2Rz3NT+wtfsBAIBowBV1CZwrGXIneZbhjCswv9HhOJ8TGDocDofjLXXqwujEqa9zhBk2/juB4J/kbHt5WNeOHNO0fQxC4/bJCoeIYTDPs6yLwcj5fHI4LrNTweBr3pBJ55bmCSG2ZZkGZhhWUWuVaq1SFHhB0g3tqqyqSSgl9KzKyAAn00MRQujsVjqvJR5wxT5x98rfWdEaXZEpKhnVsNREyJ1gGMScXofBiKEUqG2TVwXTAAAIIcSxmOtuCHW7RM51+vE9w6k9RybTRy71HB2Ot4rzwetwOByOtxSDMW4KhZomM5mp85dlbGvpyUr5+20837tSlDbe7HLfu2Ca0xtdrl+6Mc5SCpQieiZ9zeFwXD0WlxemAWD6Ch/GG+IWOXdd0F0n8qygGZbulXlvyCuF0kUl/XpBIQBAZ0Ow8z1bOt/z6PPDj/7HjhP/UVGMyj0b2u4ZbI8Nnl5nbKEwxnMMv7YzvhYAdp6/j3xFzc8sl2ee2jvx1CM7TjxinzUaW9PMmjNa6LhaOIGhw+FwOC5ZezTStrqhcTUFSj+4dg3XFAw2AcDfn79e1bbLAPC0iPHzAYYJx1g22StKa+73+j5aFXm0k6F7bUoN5LSrcFylEMYYACGglFBKKMIMA0ApUKAA9NRoFMaUvjql0PH2YDDCCABdKP2VUEo+eHPPBxXdVO7Z0Dbzp+9f/2BZMcp7hhb2AABwDGZ9bsG3qj2ekAVODvukcDLsSZQVo1xR9KptUxshhMJ+OZwMe5KNUW/jQ3eueEjVzTMjpC8PL748Mp8f+euP3fDXD9254t9nlsszIZ8YYjFmf/TCyI8kgR3dfnhm+93r2+5WdUs9MZs9EfJKoXjQHX/p+PxLAHD4Mj5dDsdrcgJDh8PhcFwy3bT0uWJhTjUMlWMYzi0I7tdal0OIc2Es8QgJdRzX2CMIg40c17bbsvbaGJsI/TpFy+G4mjDuQEzuvvFWLHkCVik9w3qjRz2Dd91PKbWtQmqcaNUC5qURuXPzuxDCP6KUOCPfbxEGIxxzoRjPAF81oGoRakVcKFLSoDQQxQmJQ5KbR68oJihnjwQWq3px78ji3oGW6MCDN/c+OLVUnPqb7774N8dnsscBAG4ebLr5sw9s+KzPJfjcEu9+9+aOd9++puX2ycXipCxwn8QYHfnKj/d/5YGbuh/Y/MlbNg/P5Ye//uShr2/sSW5U9ZNN6Us1rdwU8336k3cPfvLBW3ofdImcq1jVij96ceRHFICWano56BG/OrNcmrl3Q/u9v31b728Xq3px99DC7pJilFmO54htWefP13Y4Ljd0EaPoDofD4XBcNAZj3BoOt46l0+PnL/MzTPAWt+f+HlFY1cDxbUXbzu6sVZ/cXav9EgmcK+H3DYoc6x1ezPxCMYyrsgS+452L88db+GTPZn3u2E65Y+N79NTwbsRwIhdq6Ca6UgIADAgBMdSqNrn/5+fPbXO8eS4OyR9ZwX2EY4DLqzQ/nCXDbUHcJrIgLlXpUoMXN+xP2fuPLtuvnN5mS1/95oc/ftPDX/vZwa89suPEf17Mz5FdHo/b6wtwvCjmM0spzGDGH4zGlWqlRIhti7LLjRBC1XKxIEkuNy9KUqWUz+maqoRjyQZdU5VKMZ91eX0BSXJ5qpVSwbYt0x+IxKqVYoFSSiXZ7cEMZgrZ9KIkuzyxRFPL7OTwsVq17BTcclxRzoihw+FwOC5Z2O0OJf3+JADAYEM92x2LdQPAqwLDEMtG18ryjUOaeujxUunbE4YxZBBiAACE3HJAt6wKAsBOIqnjaoUF2ctHW1YCpYSPtgxiORAjeq1I9GoRMbwkNPbfWNn3ky87QeFbyyRgjhfIeMyFYnmV5nUb9EyNZjSLaqYNZlm3y3mV5i+moMxvkmzu6Glo7uixLMtsauvRTdPQ65s7epRqpWzbpmlblmnblsUwLIcwxpZpGgCUlvLZdFvPyrVqrVqeGDl6sLm9d6WuKtXU3OSYLxCOym6P17YsSxBlGWPM2MS2csup+Uq5kEs0tXcvLkxPgFOJ2XGFOYGhw+FwOC5ZyOUKdcSiHbpl6RzDcK2RSOuF1ps3zakvptOfLdhW7uzHZYxdEUmUbUIMjBHrXFQ7rkb05B8UEGa0+aFdQEydcYeSxNQVu5pbYNzBhFVYnCSGWnmtfSCMEXVSBt8ww6amyKJdfhH58yrNGzZ9Wwq2CIIklQq5zMTwkf3v+9in/yKzPD+TTs1OY8ww/lA0Pnr8wMtKtVy8/d0f/eTwK/t2DR3a/fy2+3/798Ox+sbs8sKcaWgaw7Asy/H8gZd+tZMSSprae1fs3v7Eo+tvetd7oomGlsN7djxdq5aKXQPrNk+PDx1taOnqYxjGuSZ3XHHOL6HD4XA4Ltl8sTifV2p50yYmgzGTKhZTX7jAejohOgDo5z/ewgvdW/3+97+IyMssZniMEH77j9rheINsU7Py86P60vh+otfKAAAIoRFKKWU9oQTri7XoqeHdQC0De30BalkmopQCQpjatoUk2c02t8URyw0DsS0nQHxjNIvqALB8sesPz+WG/+o7L/zVxGJx4o38nIbWrr5IvL5pZnzoqKYptbqG1o75qdEhXVNq3SvWb9aUajWdmp2K1jU0+4PhWLVcyGfTi/Pt3SvXLMyMD1fLxYKp6yohxNZVpabWKuUbbn/Pg5TYpFYtFw1D0wghtiBKcnN7z0B794q1Sq1cYllu2bJM6w0/MQ7HW8SZY+hwvINgjJDIsaJNqG3atokRwtZr9F1yON6ouM8b29zaujnsdocrml75wb59/3F6mYSxhACwQakmYew6f9v1snzLLQH/B76pVL8KgGCxWD6qW1bt8p7BmxMLhGNtdQ1t9dF4I4tZNuT1hYZmJ47XVLVWH4nVZ8uFrGGaRrFWKQ5Njx+/0sfreH1uiXcNtsUGA24hMJepzMWDrni+ouYFXpAZjpOPTaT2DzSH+45Opo96ZMGTDLmTJxZKE4PdTVsqpdLikByXbMntQW6vH4AC9oei5sjx/Vx71yAp5jPW9MQxc2zYqUR5ETDGCCGEbdt+Sz6rfJKYWN/a8PH5fOnAicX0z89f3rdq01bZ5fHOTY0OlYu5DACA5PJ4DV1TV6y78bZSLr28uDA9EQjF4rFkU+vs+IljxUJm2dR1zeMLhAxD1zSlVhFEyaXUKmVKiC3ILrfscnsNXVeBUtA1tUYoobLL7bEsy5RdHp+m1CrVSqnoFCxyXEnOiKHD8Q7S3xzpf++WzvfmK1r+qb0TT8X8rhgA7L7Sx3WtYRmWA0Bg2abTe+osG1taNobdnnBF0yp+WfKfveyhYPAzMZarf7xU+reH6+q+XbCtzNnLPZgJ7FOVnbP50l4ECNnEvmbumod9gUgkEIwhQAhjhI/PjB9LhuMNFaVaRgjjhczy/PtuvOMDX/rRd754pY/VcXFkgZNb4r6Wmm7Wgh4xKPKsuL4rsX56qTS9fzS1v1zTijG/K+aWeDellCZC7oRlE8uolpYn5rPDZphPsP1NHcBxPC3m09bk6CvY4wlQQ9dItVIESiliGIa+RcHO9UyUXZ66+pZ2XyA8VypkM6+/xW/GYMz7JLE+xysXHEXMZ5dS+exyKrM8P2tb1un3oQoAQEtn/2ipkM3kM0upQDhmG4au64ViiTeBcXPuAG9QDpnUEDjZxwMneF0hl8gK0mR+blS5cGGZ09kTpUs9L4fjreAEhg7HO8hd61rv2n54Zvu2Vc3bon452pbwt8F1HBhyHM9Hw/F6yzQMjuMFQohdU6sVluE4AABKCanUKkWPy+MHQCDwgsSyLEcppdVapVgsF3IiL7tFTvYCIARAqWHpSmOkq6emlXOy4F52ib4Iz4pyScmmgAINuCMNlm0ZuerStGFqypV+Di6nVKmUmsxmJwfr6wcRAsRgjG1yspfbj0ul7/AYCwHMREZ07eg38/lzehyuFMWNjRzfjgCBW/SFI76kq6IW0pqhvOZcrSspHvclFMVUXC7e7XKJ9u7R/bslmXcLAitMTCwP22eNxDfHk807juzdrpuGcTmOLRRyh4tFpWDbxGYYjEWRkw3D1m2b2JGIJ0Yppel0ecnjkXx+vxxcWirOmea1E4hfDggBYhjMIAAk8qxYVY3q5GJxcjFfXSyrRlkWONm0bdPnEny2TWxCKalqZrUuyNc1xryNx6enj2uZ9Dz2B6PW/MzYlT6fa5ltWUYwUpdMNrV3Ycw8Qoh9pj+kxHPeuNcz4BL4kE2Ima0q4/maMm0TYnIsIyV83gGPKMQpUFpW9cXFYvmIVxLP2b9L4IMJv3fVQrF8SNGN/OLc1JmiWQghJIqcCACgqoZayk4vlMtqEQCgkF1eAoCl/ljnqrZkQ5dX8PgtYpm6ZWgMZpjJ/NxoV6RlgGc4bra0OAUAl+X173BcCicwdDjeQUo1vdQY9TY2RL0NvU3h3kJFu67bAbAMy3W29KwqV4v5ZLyxbWFpdsLr9gUVTalQQqnf6w/vP7rn2ZV9a7fatm0JHC/6vP5QanlhihDbxgg/Ew80NbfE+jYBUAoAaGTh4K/C3kQbRgwjCW5/XaC5z7QNrSHcPlisZVP1obaVhqWr2kztJwDwjgoMS4paSgb8SZcguEqKWjodFAIApExzFgCgnufNZyqVR4+r6sGzt10pSRyHEA+AUMRX3+EWfeFsOTUBAEcv82lclBUrGtdXq3q5sTHUFgi4wi+9NPZsS0ukU5Z51/R0ZgwAzgSGuXIxO5tenHm7jiUUcke6uhIDuVwlnc9XM/feu+qBoaGFQ8GgezgW80mtrdHu4eHUkVJJyXd0xPs5juEB4Mlw2BNdu7blxl/84uij4FRDPIdh2cbkYmHStIg5vVSaXsxXl85bRQWAx876fui8vwFOjgI5I0GXCKGTzes5ThDPr1bsFoR4byJ6D4OxIHGsz7BtZdf4zD9hhMbboqFN61saHqpoehojxBRVdS5TqY6c3pYCgMxz/o1tTZ8IyGJjtlobZRhcbG6OdMRi3uT0dHbM4xErK1Y0rl1YKMy63eLie96z5gMTE+nhUMh9LJerZgAAKkatVDPUSqqcngNAgBAgoEDHczPDFb1aMm3LMKzLc1PI4bhUTmDocLyDbD88s31dV926VK6SKtX00r6RxX1X+pjeTqqm1FYPrM8RQmzD1LXZhemRmzdue186t2wDAEiSy93W3LUiGowly9ViXjM0pVKrFKfnJoZ6OgbWAQLAmGHdki8ClBKEMaMatVKhmp7lWEFGNmbKSn5pNjt6YGvv/Z8qK/llBrPccnH8kKJXi1f6/C+3pN+fXN/cvN60bbOqadULrbNomrMZy1o8//Fpwxgp2CQbcEcaXKI3xLG8VNVK2bf/qN+cEycWDrMsw5VKSoEQSubmcpOZTHkRAMAwrHMuAitK7YLPxVuBYTAeGGhoXbWqafP27UM/s21qR6O+5PHj8wcJIYRhMBOJeOKzs8JEoVDLKopR9fmkIACAphmqZRFnpPACChWtKAncPoQAKZqpvpl9sAzHtcb7bxho2nQvoYTsPPb4V/KV5VkAAJ4Vpf6mjfe2xvs2q0at/MzB7z9s2sarijI5zrjgvLuyqqX2TM7+S1Uz0iG33HZ7b8dfht2ujnxNma4P+NZYhJjPnhh/mFBq8wwjGzZRZIAQpUB4hnGta2n43aBLanlhdOpLhZo6J8u8PDjYtIlhMDs9nR1DCCGfTwoqiqGk06VUJOJNTEwsDxNCz9z0KmmVwsHU0MtFtZwjr54f+I77HHBc25zA0PGOsrazf61hmebRyZEjV/pYrgSXyLl2Hpnduf3wzHaXyLncEu++0sf0dpucGXsFIYwXFmfHS9VS/vm9239CbNsChBBQShHGTLlSzHW0dA9SSuni8sJUrpBdOnx8//OEEBr11dc0QynbxDIl3uULuKMNHYnBmy3b0POV5emqVsoSYpuEEpthWC4RalmRqy5NY/zOKz0u8Zw0k8vNzBWKc1X9woGhTSmBC1QltQFsA6hu2SaZz44d8kjBaEUtXHT1wcttbi4/fYGHL3vaq20TEg57piYmlocaGoIti4vFuUKhlh0bWx4qlZSCzycTVTUVlsXvuN/HS6XqpnYp2xNq29lyamKpMHNifeftHzso+R8BgFkAAJtYZio/+YrfFU6uar3pA88e+c//Gy7wunAAYIZlGZblCCXk/PAQIcCdscgdDUHfWpFjfcmAd9XhOfwIIdSayxf3dsUjd96zovvvj84vPTqZyb9ACLGDLhkQAtyXjN1PKbV3jEx+MVOpjQIAGIatHz48s2dwsGljXZ2/YWho4XA+X8vyPMPXanr1oYduyo6NLQ8VCrUz7XaKavm6zrxxvLM4HxSOa1bQ6w/eOrhxW8QfjC7ns0uHJ04c2LZm8526oWt7Thzd3RRLNPc0tvYdHBvaX6iW8zetXH/r7Wu2eHcPHdp1pY/9Stm6omHrj18c+/F8urSwvjuxrrXO3woAw79pm5NZPADXaqW0YrmQO++hufPXYVl2qVDKZzDGuFDMLeuGrgOAxiCEZYbLHB775SM2pRaHGUG3dWXP0E+/BZRSzTZrFrEMyzb1gxM7/rM13rf56YPff7gh3LHaIwVi8A5LIysqarE/kehvi0TalsrlJQC46BswnYIwsE523fxT4v6J3xVK+uRQ3Xxu/AiclZLpeDWGwUx3dyKZTAabKxW1xHGYm5rKjNx554r3tbVFX+zqqosNDDSs4ziGRwjhtWtbbsQY4Y6O+PTAQENHb29icGEhPwUAL7zWzxB5XlrV3XFzR0P9YL5UXkoXivOjs3MHo8FAY1t9csXQ5PSeQrmcXtHRdsOLh1/52WU8/cuKY3lB4GQPRpillNqaWSublqGffFxyY8SwlBKim2rFsHSNnEylnk2G2vb2N226/+x92cSyAGC4PbEy2t+06T7ECrLYtGIt5kQXJZaJWF60a8VlaqhVxhWIW8WlSTZQ10YMrYYQgFVcnsKiO2BkZ09cmWfj8glF6pIuj8/P86IE6Nxk0m29HZ8XWMb9wujUlziGkbb1dfwlACBCKWExfunxg8f+S1s0dMsN7c2f7oxH7pR47m8kjgOZ54MMRhyLGXfYLbfP5Uv7AUCXJM6VTAaa4nFffT5fTTc1hdvWrm29oVJRS5GI50Rvb/3Itm39725qCj87M5N9Qy0wHI5rgRMYOq5ZEi9IkiBKv9z/4i9uHtxw6wM3vetD6WI+rZum8YGb7nwwXyrmZpZTUzcMrNmaKRUyv9z/4i/6mzsGMMbXfX80jBAilFKWYRjrrKp3hFDSWR/s7G4MSStaos2vtx+WYbmgO5qwbFMHgPPn2FxxXpcroOp6jVBin5rZAZQClUXBo2p61bIvrqCGdbLy3Pz5j0dZNnmz23VvyjRn8jZJJ1jcpLBcFcBEBqX6qKkfVSyrBADAsfyMq+gLN0W712XLi5MVJb+EMcNghDCllGLMsgxmWJbhRM1QypZtXndzTvJKLf+Dfft/MJBMDFB67r39Fl7o8jDY91rbrhCl1SGGiZYq2QXT0hTNUEqUXvutVDDDYI4XRF1Vzplv6gkEQ6aha1qtVgMAECRZJoTYpq69oVEj2yY2xvjI2NjScdsm9snv0RMsy3CGYRkYo9EDB6ZepBQoIYQyDB49tR3BGA09++zxn9LX6VvlliW/LIju5XxhrlJTCrFgoHF2aXkk7PPVWZZl9LY2b6jUagXNMK/bObU8K4ibu+/+vd6G9XdxrCAZllZ77uijX2IZ7sX2uhU3bu65+/dl3u2nADA0u+/nAid9UzfVi263gnnJw8baVttqJYcFl0+d2PsEH2tfo80ffwHL/ijjjTQSXSmz/ngbUcoZIdmz2cjOHHsbT/mqsTg/Oe7yeAO8IEmU/jqFE2PEfHLrhjUvT859M12pjTQG/etljgsCnCwaI3Gsq6hoc3un5r6VrtRO3D3Q9bDEcT4AANUw8scX8j8pKtr8LT1tf1ZQtFkAeK5S0cosy/xq9+7x7bZNrFOvmeOUAhBCCMbo6dOvrSv0dDgcbysnMHRc0wzTMMpKtUQppcVqpSiLkpwu5tMvnzi6u6Wuvs3rcvteOnboxXgoXNda19Dmc3v9+Uopf6WP++0kC6LUUlef6Gls4zsbWjAAnOmb9ssD07+8f1P7/Wu74murqll98uXxJwEAXBi7WwWhe8E0p21KrZJtFwEABE72dCZX3bqYnz4OV2FguKa789alXH5G0bRK0OeNGaapGaalJ6Lh1sOj488DwPmjhW8UrdikSAFoI8e1W5SazRzXOWuaEyO6fqRoWWd+l0zL0AHgJQB4KeCONoR9iTZKKcEIY9M2dIwwI3CyJ+JLtC8VZoYYzByyyfVVqr4rFusKudwhAACfJPpYjJ+xyMng7tOR8BfcGHsNSi8Y+AQZNjKkaQct29STobbBiC/ZPp8dPwzX2IihKLtc0eaWDtuyzMJSaj7e3FrnCYYiLMftkb2+QLi+sdkydH3lLdsG1Vqt0tQ3MFVYXlro2rBpDcMwbDBed7CYXl4kZxXueT2n1jV+/T2lp78/9e8zgZ9t/3q/5y97LeVaLef3uKON8VjXM3v2fs8tS97OxobVAa87Mp/OjseCgcbOxobVP3jm2WuqHQfDMLixsbFb0zSFYRhGFEXX6RuHhBB7ampq6HTvvOZYz8bVbTc/8MS+b/9VKjd5VBa9wZpWztnENpeLc8NPH/z+w4Xq8mx73Yobb1v5wGeOTD3/GABcdGBITLWqp0ZeppapAcYs0Wslq5yexbzoRgwnWIXUGJZ9Uau0PEW0akEKr77DGtv947fpqbmqWKZpwsn31nMQQu27Brp2d8RCtwks4/bLUoPIsV4AoAxCbFc8cqdPluoNy6oFXXJLqlg+oltWTWBZz6lffDKRye1c05Ss39LW9Kmo151Kl6sjlmXbcNb7zgVeM05Q6LhuOYGh45pVUWrlY9Njr9Q0tXZwbOhAvlLKddY3d5WVSmk+szyfLRUyQa8vtFTILM6mUzM9TW19s+nUzNTS/OSVPva3E0IYdySbOxOhaLKsVEtwVmBYqGqFqeXSVEPE21DTzJrfJfoBAFZI0oZ2XuhzY+yzKDURQi+cHkXwSoFYhlkYf40fd0VhjJi6SKjFLUm+iN+XTOVyUwgQkgT+VQ3UL5ZPkrz3rRi478j8wpEaIXMLpjmtUaIWEMoSSu1Fy5or2XZeIeQ1C4rUBZt74/7GHouYOiGE5CtL07nK0pRPlhMs5gWfHEoss/MjcAXmpL2dxjOZcbcguF+amHypORxqJgC/vrsPCH8nn/9/M5aVutC2KyVpYwsvdBNqWeni/Ihb9IfPr0B4Jfndoi/gFgNLhdoSQoCSIXdyZrk8gzHClALVzZMjCL5otG7lzbfdc2LPS9vLLMfJPl8g3NDYMjc8dKSxp29loqOrz9Q0jeV5gRUEgeVYLpRINnpD4RjGDE5NjJ8AdGVPfM0a+faZGeN4Nnvy/yoSCNRrhlE7PDq2s72hfvDYxORLIP8fsAAAIABJREFUyWikfSmXn1nIZCcxxgzGmNGvsRFDhmHY/v7+DbVarRwKhepYlj3VxoaSYrGYnZmZOZOm2Z1cs20mM7JvLHVk56mHzswr88gBI+qv7+xIrLzJI/qjflekgWE4/o0cCzX1mjY/dM40B4SZIhtIdNi1/KJVXJqihdQEQgix/niLOnngafoGbh5cr3ZPzH69NxG9R+K5wHSu8NJCsXQoXa6NEKCkoKizXklMSBznz1WVycls/nnVMAuEUnsolf5ZQVFnAACOzi89ihBCPMu86c8Nh+N64QSGjmtWWalWAOB0Ks1BAAC3JGdvWrn+VptQUq5VSjzHCQAAhmUac+nFmVy5mGuta2jrb+n06IahRwOh2FI+vTiRmrtu5gpQSohNiJ2rFHNHJ0YOn73svk0d9wEFmEuX5ygAtcjJioQJlmtSCKkmWK7JBKpTRaEAAKeLqmB0dabfHh2f3IUQQh5Z8lMKtKqqJRZjVuB5WdX0cwI3jDHa3Nq6+dDc3KGarr/mBaxPknwfXrfuwxSAHk+ljh9SlTfc53ExP3W8UEnPEWpbCCFc08o51aiVDEtT0qX5McPSatZ1WIFweHFpeDKbnTRMy8hUqxlCfj0v9afl0nePaNruim1fsC3CgCSJPMICAAJZ9AQIta3XS3G8nCJ+ObK5N7n52HT2WEvc1yILrNxS52+J+uTo1FJpimXwy5ZNbIwZplos5MYP7tsNABCI1wl1bR09lFIKCCFBll1zJ44fCdYlGpamJ0fVSqXUvnrt5lIms1QtFnLL05NX9CaMIGDx038c/tBTT1W+CQBngnjdMFQAQKVqNTu7lB6ZWlg8TiildeFQU9jnTRwaHt1x9v/3tcCyLGvfvn3PAgC43W6fYRg6OTnCTVVVrZmmeSYVXRLc/mI186p0c47lxXet/sgfCZzknkkP7yXEtgixTUCXHt1TYtuIFaaBWBalJ4NASinFrJAiln5JhXGuVTLPB3yyWG/ZRIt4XB2mbe/JVZVJjsGSadsaRsCoplmsD/hWz+VLe6ezhQu9fxcB4MnT35i2rQPA9y7bSTgcVzEnMHRcVziW4/0ujz9fLuXWda/YOLO8MJUtFbMiz4v1kXijwAsiAEBzPNkq86IMCKByMsC8blBKqSyIskuUXYv5zDltAfwuwf/zfRM/n1oqTQEAnC5Vv0epPbvF5boDAKEDinKmCAWDWV7RK0WX6A1hhDGhV9cd6kyhuHDqn6+6YDufRxA8n7l922f+6D8e+SP4Df0Flyvl5c/86LHPZKvVLHmTgUnhAheQF3uc17JTaaOnS/ufE5i/UK3+3AIwT3+PEUIIAJ2qUgrjun58wTSn3aI/wmCWMyztqhp9wgjhiF+OrOmIrYkHXPGFXHUhHnDHWQaxqmGp5NR5EEJsQz157LLH6+274abNTb0Dq1Njo8c5XhDqWtu6s3OzU5hhWVPXVWLbFiWUVPK5TPPAyrXtq9baU68c3m+fnPcKAADJJN/+rrvcD7W3Cat0nSgvvKA89vzz1Ud9fib8W+/z/dfdu5Un7r7H8/tuF/b97Iny1w8f0rZ/4pPBf/jFzyv/Oj6uHwYA+OhHA385NWW8snu38kRdHdty193e32tu5vstixp79ihPPvN05d/6+sTN//0vIn/c3y9uaW3lV3zxi4ni8Ii+L50v/lV9NNwuCaLn6PjkLvtUejCDMSMKfH7/iZFnS9VqhsEY01OpqQgAU0rJm30NXQ6ni8NczLolJb8UcEcbT78PIoQQpZQKrOTqSq7e9rO93/r8+OLRnQ3hjtWm/dYFbfQCAeA7NSgEAGAw4qIeVxelQIJuuZVjGZfAMC6EALOYETBGHM8yLr8sNaSK5WsuFd3huNKcwNBxXSGEEI7luLAvEK6pSiVTLGQqarVcH2lrSISiCUXXlKV8JiXwvOgWZTfHsFxNU962HmNXgiiIUrFaKcYCoThzXqEdVTfVT9w9+InppdK0Tah9cHzpIAC81MTz7Qal+iuaui9n22daBNjUthiG5bxysC7qr++E16lg+naReJcXY4Y1LV0BQIjnRJdp6SoAAMtwPAAFwzJUgRNdAAgZplo1Tl08yTwvBWQ5cFd/341dsVhXWyTS1hGNigAAS+XyUuVUvz2R44SIxx1pDARF07ZN3bLOGdHjGIaNe73xkqqWgi5XUDNNraLrFb8k+VmM2eVKZVkzzTPbeEXRE5DlAMswrG5ZekFRCr9ppPJ6gTHGHMvwGGHm7MdthDRCCMUIoSDDRG52uzd08sJAA8//YNE0Z4MM6zMo0Qkl2LJNY7kwe4LQq6fH3kSqMPHPPzv0zwAACAGybWqfDoIsm1in5h5Bdn52qrC0OA8AoNWqlaM7fvXE8V07nxFdbk+irb3nyX/+p79r7htYM/3K4X2pifETAJTmF1Pztm2ZqYnRIcsw9LODwnCETfzhH4a+kEqZ49/61/znOA4JukYV06RGIsGJa9dJd+oGVR9/rPRljBGTzVopQqjd0sz3u1zYe3o/yXquo1giGY8H+2+7zf3bDAb2X76R+6wsYw9CgAkBMjys7f3ed+3/9YefCv0/P368/JUTJ/Q9pkl107LNVV0dekdjeFVdONg0NrdwxDBNraupYfVCOjvhkiRfyOetM0xTLVQqaVU3am31iYGJ+YWjHMtuNy3LhGvcockdP/zgDX/81c09d/9+c7T7UE/DulDM3zBJqL2sGUo55m/s1ky1Mti69bd8cqgO4GTxLrfojyRDbfUCJ8p+VzgZcEeXK2phGSPMuERvqDU+kOBYQQp64s0+OZStaqXsqYqlbwkkiCIgQFTT3lQvxqtNTTcyQ4vpJxAAagoFNmWrykRV19Onh2cpBWJTYpUUbc6yyTX/e+dwXG5OYOi4rpRqlRIA/OsFFh0+9XXayF3rt95dqJYLQY8/BADXTSqpZuja9PLCVK5cyBrWuZUvXzg2/4JiWIqimcpyQVku14wyAMDLivLcGkm+8bd8vo8f07T9APAIAACDGdYmlmlZpl5W8q9qSn45SLzLu6p16/sJpWQ2M7KPYwW5u37N7bnK0sn5hLzLT4GSdHF+tK2uf4thasrQ3L6n4VSa8cr65MqPbdz4sa0dHVubQsGmLz3w/i/plq0DAHzuxz/5HABsBwBoDAYb/3TbrX+6rql5XVcs2vVnjz3+Zxihr58e8Uj4fYkffuITP3zs8OHH3r1y5buz1Wr2udHR5za1tGzqiEY7vvDUU18AgMcAAFrC4ebP3nnH793Y0X6jmxfcZU0rP3Xs2FNxr/e7S+XyVdub71K5JNGzqqtjbXt9cqUkCp4zuXQI4Jk9+78PABMBhgl/NBD4435RWtfE8x2HVOWlvGVl7vZ6HrQp2I/W9H+zbEPzuyMNS8XZYbhKCj1YNrHhIoqJ2JZ1pnDFqdRKFQBUjufVhfGxoaa+gTXLszMTy7PT47Z1JlXx9EX7q9qbrFgh3sQwiH3qycq3lpbM6fOXEwJk547af05N6WfmErMc4l7r+CgFappUC4Vxnd/PRGdnzeFi0V4+lbarR6Js3rJAr1RIPp+3zvldnVlcPnFiembfR+6647+Pzy0ceXr3vu9tWdl/X39b66bnDhx6tDVR19den1xZqFQz8+nMRGsiMXB0bHIXnDVSfK1KF+dHn9z/nb/e0HH7x3obNtxV00rZXcNP/kuusjT1qyOP/MOGzjt/tzXet+no9K6fYoQZw1RrbtEf2dr37v8zFmjoMm1D29B55+90Nay7c/srP/qSyEmejV3v+t2QJ96qGrXyPes//vB8duzA7uFffBsx3CIQ26a/IUMDSbKLCUeT1NA1Wi5mcSAUA0DIzmUWsSS7kdcXJKVCluvu60M8L2Kvfz9YpkGU2jV9I/TU+7EGAMCzzG6LEIMQeqFRQWek0OF4E5zA0HFNE0RJaunoHyzklheXU7PTF7vdyYnm3K/csuwpVMrXVZXSkNcf2tw7uEW3TH0sNTMGAGfuPkf9cjTslcLY78LHpjLHkmFPEgCONXBcm5fBgWOatn9c//UFJiG27RK8IR1rFZF3+eAK9OWL+Oo7VKNWOja750mMGBz11XelcpNHI75kO0IIj6WO7GQwywbc0cZseXFqIT95NJWfOlPGfWhxaeiLv/zVF8fSmbFP3XTjp/7kh4/+yUKxtAAAsFgqnQl2JzKZiT9/7Md/vrK+fuVXHnzgKxc6FpbBbGs43PrVHTu/+qmbbvzUfSsG7vvmi7u+eWdvz50fXrfuwwDwmF+WfZ+/613/x2BD/eDXn3/h6/OF4vwN7W03fHDtmg+WVLUkcty/nz2yeD1Z1dV58703bPq9YqWSqWpa6Uy9SwTA4pMN1uMs21DHcY3/V3r5j/4gGPocAIBOqZa2rMV+UVxr2hUt7K1r80iByBU7kUvkkt2egC8YzeSWF3Tj5Mi1aRgGAOw99XXRPG4mUCnbedOkF0wfJASsbNZauNCys2EMGAGAopDyCy8oj91+h/tjd9/j+f1iwU7v2FH7TwDY83r7UHWjZhiWZlm2YVqm3tXUsIZlGE7V9apt26Zl26ZmGDXLtg2vyxU8MT27zzxv5P1qg73+INe/cot5/MguUiq+5mfBqcB516mvU84UYX321BcAALBNLa/gztYu8/CBnT/d+63Pn34cIYT4ePsa5HKHqgwvPjHz3KNELWWAUopYQQJACHkDCcEXqjfzC6MAkL3QsSCWZbnO3gG2rWOFNTt9AiKxeiZW10Rt22StDgsopUgQJEopoapSZcKRBGAGWxOjR+G8FO9rmWHZ18UoqMNxNXECQ8c1TXZ5vBtuuus9J468/CIATF/sdrGWcEe1qORquqqJbsEjugRCCSW6alzzqX6yIMqmbVkSL8gMYs5JJV3dHl+98+jszrvWt93ldQneWMAVAwAgAOSYpu0v2Xbeg5kzveYopaRQzcwSSux8Zfmi5uK81WpaMdMc7V7fkRi8WdWrxcZI11qeFWSEELJsyzAsrSpwsgchhBS9UkgGW1dEvIlqppyaBAAoqWoZAMofXr9uSbcsfSqbm5rO5WbO/zn2yflG5c5YbOn8NNLTCKXk+bGx558bGX1uW0/XNsUwlGeGhp5xC4L7oxvWfxQAoC9R13dje/uN33jxxW88evDQo4ZlmfUB/2RfXV3fLV2dtzz5yrEn4ayiHteT7uaGNccmp3Y/tv35ryqnUnRP+6dTf7MIcTalVt6y0qeXMQgxbsx4LAomRhgrerXglUNxdKXLc54nEoolWxs7+mz7ZKpftpBJLWcW5wK+YNQmtpWMNbTyPC82JltyfZ0rNpQqxVx3e9/o4vLCdEdr9yBQgLnFmTG/NxAOB6PJsanhQ5RS2tnaszqTW563iW1HQ/F63dDUmlItjUwMHVI1UpVk7GFYuOAoIKUAtn3e6AgFsGxqMszJbUQRS5//fNSNEGBKgc7NGaOiiB9uauJ6tm51f+C97/N9mmHQPtumNqVAKYVTXUF/LZXJTiCEsG4Y6s6Dhx8vlCvLdeFQ8/DM7P6p1NLxdKEwX1XVEiHEJpQSv9sdyZbKqTfSduNKwOFIQn7vh/5LdTk1DQBvyU1CrnflZq5/5Wbz6MEX4OyRK4SQkOjeCCfTrCklpqEv1EqYl9xcuLEPIYwpAFBDrdpKKYMwzl+o8ijieAGHo0lz+Pg+c2z4kLTtnt+2JkePElWpuh586LPGob3b9d3PPyHd9/4/sLPpeaa5vZ9qmmpn09fl+47D4XjrOIGh45pWq5SKO37+w++WS/kL3ll9LXWdsT7bJOb8idSRuo5YD8sxwuJ4egghNHk1VUJ8M8pKtbSYz6Ys27Is+9y5KrmymtvSX79loCU8oOqmOjSbHQqxbCTJcaEIw8Z9DBOygJpwqkIbxpjRTbVW1YqZK3IyAFCoZRfGFo/sRHAy8FP0Sp5leNEmlkGIbZXV/BJGDFdWcoumbWg+OZTQTOVtKyiUr9XyJrFN1TDVkqqVVMNUNdPUWMywDMb4A6tXNyT9/uTHN2/6+D39/ff86A8+CV/70Ie4vkRd33yhOC8Lgqv/hptuxwzDEtu2CCWkms9nOV4QKSXENAy9ksumi5n0Vdc38vUoml7VTVM9uwn1+Yq2nbMoWPd7fR8NsUy0gefb4hzXsMUl3/lEufy/LdvUp5aHdqeLcyM2ubrmptXXNbYnYskWmxA7GorVz6Vmxqq1Sqm5vrWHUGJ3t/eve2X48C7D0DVKCalUS4We9oH1lFI60DW4eftLT/8wGorX+32BSK6QWbxx/S3vLldK+XRueT4Zb2hrb+leOb84M+6WPT5FrVUA4NDoiL7/pptcD2zcIN/r9zM/5DgkshziMmlrLhp9zY9wWizY6b5+cbPfzxxdu066IZHk2mG/+owgILmjQ2jzenF2ft4am501TnR28mtOb6gqpKKqpNrcwg+43cwRSimp1UhlOV84u3DS6Z5y56fgn1nndHGWt+J5f/tRuFxHalfzKSMz/Qq1DA2AUqJWcoBZzlZKaSC2BUCBmIZKDbUMr/H8Ucs0SLVc4FdvuE1YuzEEhCxyfSs3gW1Z5tCR3YhlOXHbXR+iqlKhmqrYs1MnKCEW19reb06OH7vQPh0OhwPACQwd1yiMGdzZv3rjn/39v35Olt3eH33nyw8DwC9EyeXaePPd72ts6+6PxBuaXtn//LOJhrYOt9cf/OG3v/S3y6mZqZN7QOCPeRPlbGU50hRqtwzbSM/kxq+dC5nXVqiUC4Zp6G2Jxo5CpZQHgNHTyx7bNfLYYFtssKYZtWPT2WPHp7PHLUpJPce1LlvWvAnU9GDsAwBACKO2eF+fzxWqq2qlyx4YIgYxiEEcRdRaKs2MunvctyABdeZGiy+wMhsWm8XV2ox2WK/qKmKQpWm1KiXUUsxKBTBgxCCG2hece3JJLEKs0ymShFJCz2sQzrEMZ1NiH5qbOzSezpxpPfDs8PCzqVIplavVcoONzdts2zKBUrAty8zOz00jhFCys3tgaWpixBuOxADgmgsMh6Zm9n7ojtv+W8Tvr9+2Ye0R0zTPpD++Mj71Ur5cTqdMc+bRUvGb93l9H6nn+NbfCQT/dM4wJn5WKX9vV632NKGEwsm5fBfdHPxyYTDDlKulAiHEjkfqmjiOE2RJ9siy21Mo5dIv7d/5xEDPqi2LHD9VrpTyo5PDhxrqmjo4luMrtUphZn5qZGXvmhtM0zRmF6ZHt2647b0IYXzglb3PBf2hmM/tCx3JZ19gGZZlGIYBAJibM0cff6z05W3bPB+56Wb3BzSNKLte0n+8/dnq9wkFoqqkAueFNbYN9s9/Uf3WBx8IfLa/X7xhdsY8MT1tHFMUUuY4xK9eI20bHJRuZRnE1mqk9Nhj5S8TcrLnpKrSyrO/qn7vjjvdv7t5k3zfwYPqr+CsnMmLdS29lyJJdku33/MR/99+qQss21Ie+99fNo8eehG5PT5hy83vFtZvuQu53D57bmak9vgP/tFemJtECCG2ua1XfNe7H2Jb2weorqv688/+SHvu6UfEbXef3jMw4WhCuvd9nyDFQgZJ8nfVqYPPEEM9P51TBYALtnEBAOAYXqBAqWWfnDNOTdNEorTHnp8dpYauUU2t4ZmpIQBAVKmVgWV5JEkuqmk1sCzTRAgDQghs22JidQ04GIojUXKZhw/sPPkW5nA4HCc5gaHjmkSITRiW3fudL//P//b+h/7r/3C5fX6AkxUR65s7elxun+/4wV07br//I5/42SP/8qXWzv5Va7dsuwdOZbQNPTf8c7coeqqqVjbKeinRE1+BEcICzwm6cW3P/0IIIbckexayS3PiqfYcDIMxAkAYo9q+kcV9B0aXDjRGvY19zeG+XcfmXwqw7C+rtl1mEWJljN0AAJQSymB2z3R6eC+l5LJO5Mc8llxdrk1CUug1M+a0kTX+f/beOzqy677z/N37Qr1XOQIoFHJsNDqgM7rZmaFFislUsiTaWsthZFGWZ85YO+Mz6/F4d7XjtWds2R7Za0mWfCRKlpWZJHVTTXbOEd1AN3KuQuX88r13/wBAg81mEEU20GR9zqmDCrde/e5Dhfd99/f7/iaca5wPGGljSvAKYblF3iaGxQ4rZ8WQgGzONc4DVs6KKSPKaXubfTsSkayOqmdgieEQoZQgAMThV7tlvtMki8VkSddL5yYmz/3g4qUfLPaKBABgAMwihPCi+E/zN+fT9ahlmYAQio4OD1BC7lrThLqqUJso8FJHQ11Pa11k7VLNXMjm026O02WMYcIwhv4xnfriNzKZ/8Eh4A3GNJ0xjd4islcailouYowxoZTGU3PTyXRiZu/2+5/QDV3LFbKpunBDG4c5zrRMvVDKm6ZpGLlCNqnpqpLJpeIAAOPTI/2b122/96H9j3/q0rVzL6uaWnpgz8OfSGXi0Zuj/RfLSqmQK2RTeOF9SgizMEaHr1/XT9pkm8Ph8fkaundsbFqf3RodHuz/yy/R36UgMU8wVL24SseLYi6er4l99dvCf03NTI6LklPmRU5Siiin60L55ZPyD146qv+rVi4VJafHTZmL8KIlAoBOKWMch3569ap+JBQKRmKxwsQ7se84hDgOIZ4D4EzGdABACydVGHmDFeY7AResihiEkNJX/vaPxXUbdjme/L0/yf/ZFz7CLMsk05ODysRoP83n0/LjH3vK/sQn/hAA/hBX1dQ7/rfP/Jk1Ony1+Hf/7+dBEEWwTJOZhiE/+DgAYwwHQxH5g7/2O0iUZP3Ms/9My798ayRZdLh3dT/6ad1USwDwtcX7maYq8Oq2O0vr7coAkL39XENuZJPsrFTMVURhhQoVbqUiDCvctRDLsoLVtWVyixW6aehqbGZ8dHjgyrl1W3bfNzHSf1W2O5z+ULhucYwNccLW+oZNfTOzfRGwe8j13FiICXxbU9M2uyieV+abOd+VcBhzDsnu4DlOODd47RwAQCTgjHAYc20RX5vPKfkIpaQl7G3JlfQcAJxaK0lbL6nqiRIhRVhygPFO2qb/MvBuvkrwC3W5U7mnPVs8H9bj+pAyppzTo3q/NqMN2FvtspkzZ7Qp7Yqn1/NxM2NOIw4Jgl+o4z18dfZY9utW0XpVevFMNjfDY47f0dqyoykQIAghtCDiygDz7SgkQZBW1VR7eMzxDlF0uGXZLQuCrluWXu/3vaXY+2ajfecmJs59bPOmj81kczOT6fQkz3F8yOkMRfP5KIfxFKH0dk6bDFaIA+fb5fnjp77Oc9w3/W5XNc/zYjSZGscYIYQQfszl/s29fv+DDIAhQEhAIDIASgHovFgA87yiHAGAQ8s9j9djYPja+SU3XwIAwBj/YrGOjuM4jjHGKKVUFEU7AKATF15+kVJqJjJzaY7jBFmWxbNXjh8xTUuVZclXKpXmpucmohgjLp/PTxNCXpM+SymjmOP0hjVrN9R1rl4vu1xu0Z52+KpqajHH85ZpGIHaSINlGLrNbneee+GZ7zZ292wq57NpVSWljk3bdmKO4wxdU7NzsdlQfWOLWioWosOD/d279jxgKGr5yksvPgcLta+EMOJyubgnnvjE7//VX/3VH/6q+03GWG4UxAYJY7sNIVmhpIgR4hAAyhGSwgjNLGe/Q5KMzxjnT//cmhy7gQRxVLrvg0/y7as2mFcvHSOJuWmutq6Fa/R3MU0tCx1dmxBCSLr3wZ3AGFUPPfdNMhd9Tc0yFm2S/MEnfhf7/FXlp7/2RRKPva0abUKJlSrExhbb8/yqMFUtI14QcaS+DV7t1F2hQoUKFWFY4b0HJYRYpqFTSqiuqeV5MwTKOI571fvdI8ueoNMZ3NbUtE0xDIXnON4lSa5rs9Fr8Oqzr3cVlDHqsjvcNb5gDY/n09EwRhhjhNe3VK0vKHohX9bzeEmPwwDHV2232+/bYrenU4TEx3V9cPlmAEBNqjLCTHurfTsAIPY6joyMMWoVrKTgEyJ6Qh8xM+aMWCN2AHutVXl/LNZ/amzs1Gd27/rMgdWrD2imqX35yNEvA8AVG8+Lu9vbdx5Y3XWgxuOuiXg9kYfXrX241uutHU0mR5/ru/bcW409msvFeluav/LZPbs/+4X77/tCXlPz8201BPkbp05/YyabnQGAFW3I8XaxS5Jz76YND2/q6thPCDE5jvu8z+0OtdVF1rvzBb+p6rqMkaNbkjcbjOojutFvMap7eT7QLIqdI8a/OeLeLSw1VyFLVnsdDkeNLMs+m83myuVyE+FweOP4+PhLtbW1m3RdLySTyRt1dXXbh4aGngsE/B2CIDiKxWIUXqe1A8fzfHVTS8e1Yy//rKFrdU9X7859Z57/yb8kJsdHnvgP//v/PTs81J+YHB9pXLNuE6WUREcG+6ubWzsEmySVC/ns0Pkzx+598rc+5/L5Qyd+9L1/blm/YVsgUtcYnxgfRjCfIXC713W5XJ7GxsZVqqqWvV5vUJZlJ0IIXb9+/bQkSY7m5ubVgiCIQ0NDV6qrq+tTqVQsFArVptPpeCKRmNE0TZEQtrfZbN06oxplQNttYreAkCghbL+maefilhWFZWwvwAxDZ+p8P1tmGobvz/9XCTvdPi5S1yY98MhvIFGUaKmY42pqmwFjbj4tw+WlxWIWTOO2GSZcbV0rDoZqjWtXTtBCPv16r40Rxi013TvCvqZuUZDsAICypcTU4OzllzDCXG/ngQ/znGAbi/cv1nVCjbehs8bX2N03eeonlBIKANAWXrcLAGAk1nfcYw/UtNR03+NxBGsLSmZuaPby4ZKWzwAAMGKZwHM8LRVz7+AurFChwnuEijCs8J6DzV/mD3LYrRVg81AA2h+N9afLpfTxkZHjeVXLY4QwA8YKmva6tR53A4wxNp2cmw64vUGE5l0dJ+bykwAATTWeH8SzSlzVTa3G76z2OGweAIB+TbtYJwrNIkI2fgV8L5ASSSujylnBJ0SUYeWEVbAS6oR6kSgkCwBgpIwJnMMxqlOlfLN8xBa2dZISyVj74d70AAAgAElEQVQlK6kMKseoTl/jLptTlHxrVeiL25qat9lFQVYMQ0kUiwmA+fdLQdMK4+n0+Hg6PX56bPz04vMSxWLCsCyjbBjl//7zg//9RmzuRlHXi/9y/sK/6JapG5ZlnBkbP1PUtOJiXdWZsfGzzcFgfEN93Qa/w+E3CDEShULiyvTMFZOQV1ZhvS5XkMOIS+cL8dXNjVvimexUOn/39jns6WjbvaV71f3Tc4mh3RvXPY4RwjzG/D3r1z7y4tnz3742PXO63WbrtiEk/yif//qQrl+zGLO8PO9/xOX+5GJ963sBTdMy9fX1veVyOQkAyOPxNNrt9hBjjEmS5JMkyeN2u+skSfIqipJ0uVyv23sQAIASSkrZTHr19p33Sk6HKzE9OVbd2NQerI00JqYmRwmxLEKIxRijDo/X37B6zQZvqLpWV8ply9B109B1SgjJJeLR9Xvv/SDCGCemJkZFSZI6NvfuLqRTCQB4Va9SURRtmzZt2i/LsjORSEzfc889j9y4ceO8x+MJ+P3+alVVyzabTVIUpfTAAw98QpZl5/Xr109v2rRp39zc3NTLL7/8AwBQSpQUBjTtMgNGESAUs8wpF8aeGkGomzKNUcLe+VrgXwYkijZkt7sAAJAgir6//AcHLZcKQsfqTXxD06ry01/7ojU+cl164OHfkPbc/1EAAKaUi9hudwMviLfbJi0WsuqLL3zTtmPPY7Yt2x9APP8cs16bgVEXbOvp7Xzw05OJm+fc9kBtW3jd7gsjh7/DGKMUKDBgbFXdpvt5TrDBQqsT2ebw7lz98O9Np0cuAcAERhj/5v4//szNmQuHnJJnYEv7fR/zu2oac+XEdFt47a5qb/0qgRP/wiTzIhaJkh275ssMKlSoUGEpy34AWKHCclDStDIsNECHWw6GAObTMRljdDnTm94uhmkal4aunx+LTg2rCz3UFvG7ZH+moGUAQEvkyolkXkkCAGAE+LyiHNUZ0+gy1/sAACyYxowtXBZ55bqV/7dWBwCQXLgskgUA4Hmek2XZRQgxEUKYUkowxok+RRkd7x8fYIwx05w3czDm05EvLFzeiB8tuX58yfWhhQuIgmBb09q8feO67u6rwyPHfLpWHfJ5IyWBFwmHud0b1z8uiaK9b3jsxLburl3hUKB5bVvL4faG+rVr21vvuW/r5rnro2MnawL+pvqaqo5rI2MndcNUezra9pQ1rXBtZOxktlBcNpfYN6KzqWHT0OTUxSMXr/xox7ruDwIA5MvljMBzgiSKDgAAB8YuL8cH4pY1YzFmAQAohJRURpVWUexazvjfSVRVzXk8nuOmaZYty9KHhoaeK5fLCU3TchhjwTTN0tDQ0PO6rhdN01RKpVKcLqlHvRVimZbD6z3uq6qpJZZpKsViwSbLdk4QhcHzZ45hjuMMVVUSkxMj5Xwuq5XLRY7nhXI+l6WEWIaqlk8/+6OnlWIh7wmGqg1VVcqFfM7hdnszsdhMLjH3mlYGjY2NXYwx9u1vf/t/CIIg5nK55NWrV4/X1NQ0PvTQQ58aGhq6fPny5aOzs7Ojf/qnf/qtiYmJGw6Hw6WqaqmlpaX71KlTLwAAmIyZADC+dNsCQsKYYdzIE7LsK1dcqLpO3LB1PxeqHrXt2ncPAAJrdOgqt2X7A4A5HgixuFB1vbRr/xOLzzFv9p+T9h34mG3nvsexz/9dJNpsiOdFEo9NSvc/DCSTiumnjz2PeF607b7vwyQRnwGA87e+dnP16u0m0ZVzwy9+K+iubXHYXL5kbnZY0edX9ByS+3sBV03z0ueki/GJgpKdawqt2goAEyFPpM3jCISHoldervE1dtUGWtaeufmzr0+lhi6GfU3dH97x1Jf6Jk49AwB9CGEOe/0hppTfNefmChUq3L1UhGGFuxK32+d/7BO//0f/6Ytf2+8PhSOtnWs3/te//s6n1m/e+RUMgIBSRgmxyqV8jhLLMjRVMXRV9bldVTzH8RghbJiWbhMFWdWNMocRJwqCTdUNRRIFOVIVlEVekADg2nLP9ZdlwdVRhdukw27uqNmcLqhpAChSyhYWVwE6bbaeAiE5RmnWmk9lW5bawncSl8vl37Nnz68VCoWMy+XyapqmaJpWrqmpaWxpaen2er2hs2fPHsIYX6H0nTNhYIzRRCY743U5QxtXdezXDVMdmZm9WhPwNz6ya8fvXBocPjIViw9uW9P1gbl0ZtKwTH10ZravrirUFk2lxihl9AM7tn0qXyqli4qa3dWz/vHLg0NHJFG0H7/S94ymGyvOrXMR07QMnuMEmyjKCCEkCoJUE/A38Rwn6IapAABkCUkhYPBUIPjfHnZ7TqiMKk/6fO1b7Y69LxQK31nuObwdvBLyPLVVfOoDrfwHYiUW++j3lY8BAOTz+eklw0Zu89TC61y/LeVcLguvYyqyhMW0xduNG7tlDADA6wqzmZmZkQsXLhzesmXLfYODg5d0XVfpPMQwDM2yLNPr9YZM0zSSyeRsPp9PRSKR1pmZmRG/31/zRs6kC2Jx2UUhEMsyb/afB9PQ3f/xT/4REMLKv/7zX7JiPm1cv3qKb25f6/zcF/6G5rJJ/cyJF4TVa7cDAEA8OmH+5F//l+uhxz9t3/vAR0HXFHr62HPlQ89/k2eMiJRaNp7nzNPHnhND1XW+h5/4HVs4ktBjs6+qRyxp+aRD8uzy2P1hjz1QyxhjilF8w/+xapTzk8mb55uru7YLvPjM9lUP3jeTGrlSUvPJpqqubWsbex+uC7SuN4mh4/meieCSvVUAACCIIi0X84upsxUqVKiwlIowrHBX0tnStTE7MjhyI5/XLMs0+wf7zsqy3eGWHG5ZUVk6NhuPTo0Mfv1v/vQ/UEIoxng06PXUrm9v2+X3uKoxwpzA86Ikivax2ej1sdno9XAw0IQx4lojkbVHL135kSzZnMs9z7eL3Y6dhsE0y2IWzyOB55GgaVTRTaJv74ps39ReE6SM0UROScymitGMReKPuD1PlijJjxnGTQA4vNxz+FUxTVOPRqNjdrvdRQghqVQqWlVVVZfP5zOGYaixWGwikUjMvJOiEAAgEgq2blm96n5V18s2QZRVTS8ls7mZgMcd5nlO5DnMSzZRBsaYquslh2Rz2QTRzhiDRCY3bZdsLpsgyAhhxBhjl24OvUwosfLlcrqkqPl3MtZ3mvMDNw49vnfXZwJebyTgcYc/fO+ez7dGImun4vHBmYXWHdOmOfrldPr/fMTtfvIRj/tJG8JyzDSnvp3N/t1ppXxXvu9yGstzGP15X5z2PbVFfGq543knIIRYIyMjV8+ePfvi1q1b76urq2tLpVKzpmka5XK5eOPGjfOTk5M3N2/efJ8kSfbnnnvu6y6Xy9vW1rZ+bGxsQBRFSVXVksDP13YzxhiHMWeXeLthEkMUOLGoGMVFJ1WL3Fnn40WsyfGbiON+HRhjgDACYACU0gVRO4Ew/i+v3M8YVZ//4VcYpbRpzbruhsaGiDB248TszwevK4VCbt2efQ9m7n3gUVPXFUd08gr74KMfm7zWd1EavXm6M+iTprrXrOcFMWqZxit1pDemzx/c2LLnox/a8dRf58qp2b6Jkz+ZSg69YeaCaelac/XqC/XB9o31wY5NLdVrdlwcfem7hM5nQIzGrp14/vw3/iRTjE8CzHdptKhlAABwoao6VsilsNsbfBd3a4UKFe5SKsKwwl3JdHRiWOB5YWJ6pJ8yRmPxmQlBEMW6cEPbbHx6NJVOzJqm+cqPL6WUSaKYiWcyk05Z8ii6XkznCjHDNDXMYQ4hhHTTVGWb6JyIzd3geU7UDVNbyU2aZVm0O52SS5YFeyyWmzHNf3MzXN0t7RgfM64BQCwY5GvDYb4FAF7OFLTM1lXhrV2NgS7GgJ3qnzkFANF+XbtYpKQgICSkCFkR/fPsouir9brWFzQ9miqVRyn95eqQSqVSAQB+sXi7tra2aXJycjCVSr0mdfidpKSq+XgmO8VznJArxZIlRclpuqHE09mpZDY3G/J5a112h//CzaHDhmnqXqcjGA4GmuOZ7FRZVfOmZRnn+m8cwhhzHqfDr5umWlbVAkYIv/mrLy+Dk9OXm2vDX9qzcf0TV4dHj1f7/Q0Xbtz8xcmr157LFAoJAICF9OybAPB/4PkaWEQZo3+8rJH/6hDK6Ac7hNekYWOEUHcV7u4K4i4GwG4k6Y2hNB1yicjVW8/1Hhq1DplkPqW2M8h1VNlR1dlZctbGg62nhuupc6M6xQTlYpRcjBZZ9E6ltyuKUgKAv124+cNbHk4DwMDC9VtFzImFvxcxQmhDe/UGpyQ6KWO0UNYLIa89ND6XH+9uDHRPJYtTHEIcA2Buu22koOjLkt7I3qBFDFtiLrR4FwCA5HC48snE3MT1vgt7P/bkv7t46Kc/nOq/drGqsaldlGR5+MK5E0qxkOu+Z/f9g+dOH5281ne+urGpnRcEEZYYDNkE2Wm3uf19kyefUbRiBiHMuWRfDUJo5o1+ezKlxFRBzcY3tOz+MKGmEctMDhBKSHP16mlCiVnja1qdLSenESAs25yeXCk5CwBA4nNTgDkOLPO2JkcVKlR4f1MRhhXuSuYS0UkAeI1FOADMvt5ztPkWFBdCPm9UUbViWdNuPQgZut3zVioOh83V1BRsFwROTCaLcwBgCgIS6+vFzocecn2wuopv6O11xHq325sJAQsAXj52bfoYA8bSBTV9cyp90y4JdgCAbknaXCcIzRgAu03LCytgX3AYCa1VgT3pkjKeLatT8Cu6Fkaj0Yl3JrI3JpXLxzBGzy6k6i7llfSwpSccMMZRxtitDcHHFx5DS1Y0J97NuN8J2uoiayVRtP3LocP/0yLEZAwYfe2B9SssiJwVeeLlnWJdNV731BbxqaLBihwGbncjv/tbV41vzRTYzBd2iF8Yy9AxALiJEUJ/ulf8hI1DtoEkGdjbzO890MofSCks5ZGQZ2c9t/OLx/UvAsCKXjVeCkIIdUT8HQ1V7gbDIsa18eQ1j0Py2ISSrTboqq0LuetsAm9DCNBiivtyx/zLEKiNNAg2m6SWioX2zVt3yU63h9J5kRlua+/Sy+USJYR2bt2+B2GEMce/6piLwxy3sXXf3mRhdtQleauckifkdQQjVZ5Ix+nBn399Vd2mDXvW/NrOluru7YalKff3/Dodjl49Op0augiAkulCbGxX96OfvTx29PuKXkgDAMzlpm4MRa+83N2w9cGmqlVbKSMkV07NnrzxwlcBwGCWaeDqcCPnD1QDwJE7vtMqVKiwoqkIwwrvO5LZ3GtMFu5GTJMY5bJecjhsrsX7GANKCLMYAyrJyOlwIHc2S+KTE0Y/AMCDW1sedEqCc31L1XpVt9SGKncDAERreb5h0jBGWm22LjvGTowQutPGOy0h/86wx70WI8Rfno5+t6TpifX14VfaF0gC71pXF/6QyHPOZLE8NJHKnmyrDtzrtct1RVWfG4wnD+mmtSLqZpaKQsQLIufw1TBi6kCJxYilI1FGsHAQ/EaprO90muu7TXOkds2+zRs+ki0U4pduDr10ZWjkOACsiBXo5YDDCP/FfbbfSKss/ddnjL/GCPAfbBX/4IPtwgf/5qz+N9cT9Pq9Ldy9AHCzxolqVgW4VU9fM5+WBSQ/3sk//vyQ9fwLw9YLITsKfeNx+Rs9NVwPABxd7nm9VQiltC7kPm638XaTULOsmmWv0+bNl/X8yeszJ01CTZHnREoZXRCGdxyR5+SI17NBNa38XL7whu1S7KLgVQzzlbpItVgq5JOpuYlrfRdEWbJLdqfTNHS9sXvtRrVULKRnZydHLl88bXe7PZwgiJZhGKahv9Lewil7q9Y373z8eP+z/zA613ccAEFP884n2sLr9zhsbn9ZL2Ti2akbqUJ0FNh8OqhmlPMMGKOUEL+r+lBZL6RjmYkBzVSLGGFEGc07JNf347mpmx57oJZQYqYK0TFKF9yQDV0zr5x/yXK5/e/aTq1QocJdS0UYVrjrme+fjRBaSElbuA8zAEYopdxCvz7GGHtlHDAqcrzY29zcq5qGenZ84hyHMV7cxt3gRqooekmSBDkc9tYNDMxeBgCwLGZxHBo89GLxm7ksSZTLtMAYEF1nKgBAJOiMHLkydeTRHe2POiTB4XbY3AAApxTlF2VKi1U8H56zrNnlmH+qpIyUDTPdFPDtaA74dgDAM0sflwTB0xj03XN2dOorBU2LBpz21pDL2TEcT/2ixu3qbq8K7geAZ+903G8G4kSJ91Q3k3JuDtscXt5X08p0JY844RAj8+lcS4U4hzEmb7DKtpI5P3DjxZlEYrizsWHTtrXdD35w5/ZPf/bDj/edvtb/Qn2gLnhtbPgyZZTes2bTrulEbEozdM3v9gackuxUNK1sl2RHyOsLDUyO9t+cGh24b9M9B75z+Lmnl3tebxe3DdxdIdz1jxfNf0wpLEUoo0+uEwb2NvF7/TL2Hxy1Dv5Wj/BbNg59dV8zv65ssvJQmg45RHA82C482BXiuj6zRfwMAECbH7d1BHAHrEBhyGGEARDc7n07kyzcmsWx6KibuHXscoARFgJOe2tJNxIcxoNs/jeEwcJvwWIKtywKvu5IzSMYo28xBrR+1eqcUizko6MjNyx9XuwtZgJ0bu31ZmOx6eTM9MTr9YcEACDUMjnM8R5HIMxzouSUvKGG0KotuqmWNLNcSORnRgDgssDxvEdy+SmjBBiAzEtytTPosiwtPzZ7+TCHOd4l2l2tgYZVIYd/QsIcTWbHryUyY1cFjhedNoc74q6q90iuvMSJuloqFuAtmB1VqFDh/UdFGFa461lVU7NqU2PDJq9d9l6cnLpoEmLu6+zcN5FOT3TXhvt/s7d3KwJAo6nkaJ3XW+dz2H2jydToiZHREwgB8sp2r1uSXE9s6DnQFAg0nRgZPQEAZ5Z7Xm8Gz3NCc3Oog+e5V/U/I4TRllbR0nSmVFXx9a1ttp6Bfu0UAEyMRnOjD21rfWhtc2itRah1tG/qKADArGmOMwD4WbH4PboMPcVsPO9YW1fzgFuWat2yVBvLF/puHaMYRnpoLnloXX34I+Op7AkOI6Gs68l4oTTgkmzVQae9/U7H/VZgxNSZoRax5PRxDm8N7/SHCQMGCweMPIe5vesbdjfXeMZNQs2P7lm1q7M+cGkuU5ora2a5NuCsdciCY2Q2N2K38XaB5wTNsLSSuvLcSXPFUgoAUgLP9x25ePmH9dVV7Zu7Ou/73ccf/r/yWdI/nZgbRwhjHmPO43B6Dmze+eAPTxz6fl3b6jq7JNuj6cSsKIiiwPECoZTUBavrlntOvwoIAAEAUAqvCCYGwDACjBCgC1Fy4dM9wqd767jeTWG8aThNh2NFGou4UWQqT6f++LD2x4MpOrj4vJzGcl9Zrsm8AXt7GvcG3HKgqcZzrlA2CkXVKJoWuWucjWVR8LWE/Ls6a4IfiOWKfSVdj9t4wSXy3LMbGyMfMQgpBZ32ttZQYK/AYXk4UzgWy2QywBjj1vTs5KrDQ4A5HvuDGheO2DheiJFEPMYYZaLACSLPvdLr0LSo6baLbpNQUzfN/PmRZ762s+vB39675qHPUDD0uezo9bODLz1tklIx7HfWqIal2njeqveEm1ZXtfWYxNQNYhpO0e6MFhMzAieIXaHWdYlSOmYQU9ctXVtXs2pzyOGv0S1TS5bTsVgxGZV4myRwvNgfH74CAFPLuLsrVKiwgqkIwwp3PT673VdQtcLB/oGDf/TA/X8UzeWiM9nsTNDhCH5i69ZPfP/ixe8XVK3w+f37Pn92fOLsjy9f/fHHt2z5+NWZ2auLK4xrIpE11W539VAiMbSlqXEL3AXC0OGwOXXd0gEQglvqtHp65H19fdqx/fudn4jFzLHNm+UDAPCPB8+PHSSUEs2wtHODsXNnb0TPArxS6wUAoMEy4JRsVS7JVh3LFa6aFikjQMhhEwObmurqEADy2uX6oqbH02VllDJm+u1yY7xQuhFw2JvrfJ7NLslWnS4rY2/+SnceZhk6AFxavC1Wt6w309M32EJql9cpeR/c2vrgaDQ7en0ieb0+5K6PBF2RXFnLHTw/fnDnmrqdBcUomBY1961v3CfZeClf1vOyTfi+qpvL8v96PTBGyCnbvY3h6vqGmurOjoa6DbXBYMtMIjly8PTF76xvXbVBNwz9xPVLx2WbTQ64fcFYOhk9XDyTObBl50MCxwuJbCqeKmRThmkaY7GZFfk/favkNMgNpujg5gjefGoGnQrYEXx+m60jXmLxrEqzignKmRly5hPrhE+oJlMPjlgHCzor1rpwfjJHJxs8uOHMDDljUbC8EvKq5mtb0KwEjl6dOrp7XcPupx7d+NRctjx38vrsSZdd7C8qRgkAQLbZ7HVVoXa/x10zGZu7kczmZr1uV8gwTLWoKMteM0ko1YfiqV+MJtJHd3U0/yFlzEIIcQAIiTxnz6vazM1Y4mcMEJwamfwHvrF5FVdbv84avnmZc7n90v4PfJwZhg6GruJAMEwS8Rly9sQLADDXHvG1b+oIb1r8hj57M3r24d7Wh3WD6C9dmXzpnjXQzPMHj7/UN31sfWvVeq+Hl/f0ONZfGQmyzz664bPPnhp5tn88OyByghgvpWZTSjbR7K1rM6hlzuTnJixKrKxaSNs4UZIFmx0Dxnm1mBnLTA9hhHFRL+VLhlIMOvxVPOYEDnPcsu7sChUqrGgqwrDCe4Iql7NqQ0P9hkSxmMipas5vd/hHUsmRVKmcagwEGg3LMlKlUsprl7099XU9BrEMl83maguF2gAAbszFbwAA+GS77/TY2Onlnc1bIxBwVhmGpc3OZieWOpICAOgaUzZskPcn4tbkzRv62Y5O22YAgF3r6nd1RPwdGGO8obV6g2ZYGgCcW5YJLGB3yx5vyCWnNW3MJdvCJV1PaIjlgwF3K6HUxICwR7ZFKGJWz5qmD+uaWRodnzs2ncmfxxhx1W5nV0nX42PJzLHlnMdbxYiPXV16O5VX0p99dOP4zy+M/VzVLbU94m8/2jd1dM+6+j1ep817dSx5tae1qkcSeKmgGIXnz448/6n7135K5LEIyyTkX481rS07tq/tfsjvdtcAABuenr5y+Pyl703G4jcM09QB4MQt9atLa7r+/hazHQCA792x4H8FWny4eXMtt/lAK7+zzo3qfnuj+OmJLJ1wiHDu6T7z6c9tFT/3uS3i5wQOhGoHqn66z3w6o0IGAOBijFz85Drhk4dGrUNDGToEAJBRWeYHA+YPPtDGfyDsRGHNAk3kQPzyeePLsAJTAFvD3laPXfScuRE9E02XomG/I5wtaVkAGBIFwXbP+jUPPtC75cn2+rqeb71w8P85dPbCt7euXvWARYgBAN9d7vgNiyiqYWYtSnWMEMeAMR5jgcfYJnCcDABAGVCMEAcAQAuFDLKlHMDzAivkU2ZibopRSrDbG7RmJodpLptghqYBAIg8LzaE3A2EUiJwnDAwmRron0j1N1S5G2oDztqpRGGKAWOEUkIpoz89O/rTPevq9+impQ/NZId+en70p8b89/swwELaOaWEUEomcrOjhmUai48t4fJtpjn67u3BChUqvFeoCMMK7wkYA8YhzD1z5eozhmUZbVWhtkSxlLhSLF5pr6pqBwRwbHjk2OpwzWoOY+7I0NCRgqYVJjOZSZMQcyyZHNNNU3fYREeiWFwRtS9vxuxsdtI0iZnNllOW9WpheL1fO9XSIq7tv66dohTI2KhxFQCgp6W65xeXJ34xOJ0Z3L2ufveq+sAqWCZh2La5+R7RLjga1tRZNW1VXeVcOT03mryUms6Ot25p2kmiee3kjYkv13WF1/nWVNUb8YJ3kjOO69goK24+KVHZTeqdSl7gxlNTpbHI2siadfeuDuTm8rOzQ3PXiXn3pLKlC2p6a2ft1tFYdtQi1FJ0S6EMqNtuczfVeJraan1tyZyS1E1L13RrRYnBpfhcrqqyquYv3Rx6eXh65spCaumreKP61bvNbGcRjBCWeCSNZenYZI5OyjzIIgciAkAXY+TSlgj395vD3GYKQJ8dJM/2xWnf4n6ocuC+L50xvjSWpWMzBTYDAKBZTHfb0HPxMot3BnEnAoSGM3RYt0B/40iWB4sy68yN6Jm8ouc9dptneDY7bC58J7nsdt/uDet/7dLNoZd0w1QB5uu9KWNkfXvbblgBwhDg1SkXBVWf7a6tfoxGmOW1yw2zucKVsm6mPLItsiZS86hTK58rDA1cXBh+O3fsV1B1Uy2pRski1HLZRVdPa1VPJOiKFFWjyGHM5ct6/t6NjfeKHCcu/ewDABQUvbB3XcPeoMd+KZVX0gCvfH4uvdFrVqhQocLbpSIMK9z1FHWteHV29ur12dnrqvlKat3SGoo4AECj39+AEcKXpqYulU1TC4VCq09NTl3hOE7wer0ts4ZR9siyJ6GoRUmS3As2+wwhhBljlOd5yWazuQzDKJfL5eRrArnDmCYxfD5HoLbW13DhwthxWLJ6lM2SOcmGejs6bJuvX9eOqyotAQCYhJoPbGp+YG1zaG1XQ6ALY4Q/trfro8f6po/FMqU75h4Zagy2tG5qbJkeiF4WJcHOKCXlrJJu3dx0TyFZjGOEsDvorPZWe8JtW5pW5ebys0pBzQbr/c3VLaFV0wPRy+6QqybSWbPGMonR3FPfm08Uo9lobrp1c9M96ZnsBADk3iyOlcKRq1NHqn2O6mROSSZzajJX0nJHrk4d0QxLy5a07M3p9M1kTklahFllzSz/+OTQjxXdUpY77ls5P3Dj0KWbQy83hqtXtdfX9WCEDiOMMIcxZxFi3qaFx3uCkQwZhTdYkTk/Sy7Aa/v9AQBAokyTAPCNW+8v6KwIAIcXLisShyzY1zSF1nQ1BEIAAD6n5KOM0W8f7v/O4hhJFGSP0xk8dunqT6r9/gYAAMYYVTW95JAl93LFvohJiDKaTB+1CNUMi5QvTs5+K6eo0yYhquY25QwAACAASURBVEWpEc0X+3KKOqXoRubCxMw3F8a95c/edLI4/ezp4WcpZdQm8jbDIobPKflU3VILil6QRV7+0fHBHyVzShJjjBc/+7mSlnvh7OgLTllwLmR2VKhQocK7TkUYVrjrGYonhhAA0i3rDc+mR/P5aKpUSmmWpdXV1+/w+XxthBDd6XSGMcY8Qojjed6+Zs2aXxcEwR6Px/s4jrNVVVV167peKJfLCUVRUsVicRb+zVlv2cAYcY2NwTae5wSOw6+qG9mxw/FYuUzzbe22nmjUGmlpFdcDwMC18cS1Gr+zxrSIeXM6cxMYA0IZYXe4l5zLbw9pJa2Qms6M+2u9DaWMkpruj16OdIXXIYywVtILslvy2r2yn1JGUtOZMV0xyoJNkIL1vubYcLzfZhedlkGMTDQ7pRbUPKNAZ27ErtZ11/ZgHt9V323xbDkBr3VpXDy5cbsV7GXvM3k7nLLd+/EP3PeFrqaGLZphKJcHh4+EQ4HmPRt7njhxue8ZmG9sf1cjckjgMfCEAbEoWAIGgQEwi4LltoE7ZMehkQwdETn4N8MRCqbIgcgYMJOC2eJDLeuquXWHRq1DigmKyIG4q4HblVRYciJHJ0J2HBpKk2GBQzyPgbcoWCZhK24FPOi2BwNuOVBUjKJhUQMjhJ2y6Fw6hlBqGZapBX3eWg5jnuMwb5ckV3t9XU88k1l2ExRCqQWvPpF4feHv7erMb5tdwWHMMwBGKX2VcZfNhqXqGr6hpJbKhsFUs8Sydjt2EV1RLcKMbJnkSQHSlsWMW06aLMaz4tKGK1So8N7mrjp4qlDhdmjztUtvikmIBQAWAEB9fb1FKbV4npc5jhNzudw4IcSwLEsxTbNMCDFLpdIcY4xRSk2EEFYUJWlZls5xnO1dndBbxLKomcspmXDYW48WLNUX4XkQ8gWSkiXk9Pq4Kg4DBwDgc0m+sVh2bCpRmNrUHt5EKCUHL4wfJITeUSfS2cG5aw1r6jZtOLDmiUKqFC/nlaxlWkYxVUo4A85Qw9q6TZRQEh2K92MO8+vv7348Ohzvz0RzU7lEPmpz2JyZaG7KFXJVO32OYDFdSlDK6OI2qHVn51Nhnl0b1j1umKb2l9/67r/7j09+7O8RQihXKCXDgUCT3+OugQVhaHPYHKIs2JW8mr2bUn7dNuR6rJM/0F2FuydybOJSjFx6sI1/cCxLx05Ok5O9dVwvAIBBmPFEl/AEZUAFDoQXhqwXFvoVwnOD1nO6BbpTRE6RA7E7hLt76/her4S8L41bL+1r4vcBAIgcmtjbxO3ZUcftODJJjth4dEa3mLGc87+VaLoYjWfLcYwRDrrl4Gg0O2qXePvSMflSOXNlcOTokw/e/5+DXk9tpCrYuqat9R6Pw+7/+x8885+WK/Y3g8OYt/GckwFQAEAWoZosCj7KGNEMMydwnCzynINQZjaH/GsJpbok8NcsygyLEB0AgOdBDPi5MAMOmpvFNcUCTes6U4tFkpVk7GxpQetHhvVLiYS17AK5QoUKFQAqwrDC+5Tp6ekzCKGz7LX1ThNLbwiCIBQKhVGHw+EpFAppjDFHKaUcx3GEkGUVHzyPBUIosdtFZ1WVuxYABhcfGxs1+rpWS70uN+dvaRHXXb2qHgUAGJhMD3zusY2f0wyi5cta/qs/vfrV5bCV1xVDAYC/xxzGlLyq99m3Fv5eWXLfyG3GLXLhDbax4uEwzwMCRIhlvvnoeQROtBFmmXQF9jr0u13VU3NzNzP5whwsfrQQAM9hAS85edHQXbthwwfWfmjq2uzFhu7I9cREakQr66XlivutMi/mkDiRYxNjWTpmF8CeUFji+wPmDwAAmn348p5Gfo9DRA6DMGMqz6Y2hrmNdgHsfXO0rz2A29v8uG0wPd+CQuKR1B3iup8dNJ/treN6GQC7PEcu72nk9yAESLdAH8rQoWiRRhfrzlYSpkUtALC2d0V69/Y07j09MHNaN4kOS+ruVF0vO2Tpa3Pp9OSmVR373U5nIJ3LR//l57/4y9GZ2WvLF/0b45Js1avCoQdNQlQeY1tO0aa9drneJERLlcrDPrvc6JJsNbF88ZrAYclrl+pFjnPG8sVrABADANB1pqZSVrS6RmjMZsjc9Iw55PdxNYmENR0McmFZxi5KGX2vpli/28y3JZ7nNr/lFSpUeBtUhGGF9w3uQDDk8gdCicmJEdPQjTf7IQkGg+GtW7duiEajE+FwuDGVSsVEUbS5XC5fNBodxxgPLadZhmFY+sxMZtwwLF0UeXHpYxMTxvXqar7R68WhuZg5Pj1t3gQA8Dhsnsl4flISBSmZU5Meh80DANFlmQAAvI7Y+6XGvdVtrDQw5rhqb30HnW9b8ZZSQ3lOEBpCnZsSuelBAEi/uxH+8kSTqbGWutq1iWx+VhQEqb66qmN1S9O2+dqp0ivp19lYfiY9k5lo3tDQ29RTv7WQLMbX7u+6PD0QvVJIFOdWougFAMhqLFvrRrUNbtzQFyd9CAEqG6wMAOAQkX1XA9fZ6MWNU3k6ZVIwTcpMyoBuDHMbqxyoKq+xvCwgudmLm1t8uKXWhWrLJitvqOE2BO0omFJYqjOAOxu9uLHGiWrSKktvruU231OP7pkrmXMAsCLF8+qm4OrZVGG2vspdb1nUwhidXSp2DNPSB8Ynz43Nxq5jjDhg86nrIZ83kszmZpcv8tcHIcS5JFuYUmZyGIs+h73pwvj0P4s872wM+HpNQtTxVPbEXL54fVW46qGI17NhxEi9VFC12OI2LIuZMH/CbnDJpq8CADidXNrhQO5cjt4VZmfLAUYIzbcNAcZjLBHGTMYYRQhhHiPRb5dDBiFlG8+7OIwnGWOUzWvEikisUOFtUhGGFd43UEoJsSyLvcVyOo/HE4hEIi35fD7t8/mqEEI4HA43IYRQJpOJv8vhvikL6Z834TZ1W5s2yfe7XJxvYty87g/w4Z710l4A+KndJtgPX548PDSTGertivQuCMMKdxgOc1xdoK1nS/u9nyTUMjpqe342kxq5sqax92GbaHdlinMTw7G+o42hzi0hd6TNsDRlPD5wur12fcfmtv0fj2YmrkcCrS8W1Uy8o7ZnnyQ6PdPJoYsz6dGrhFrLlpp5buDmIZfT4X94Z+9vhYOBps986LE/VzStcPxK3zOzydQrPQlT05kJhNHfOn2OYHVLqCPcVtXVtK5+y+rdHQey0dxUZFX44NxI4iZZYU3SW324da7I5q7EzCtbI9zWH9+0fpzTSA5g3hl5rsTmTkxZJ2aLbHauROZKBisly2ZSMUEJ2lFQMZmS1VjWxiHbiSnrRFph6XiJxJu9uDlWYrFYicY8NuQ5MWWdsChYqsnUc7PkXFph6ZXqSAoAcH4wdn7H6sgOjBE+Nxg7t1QUOmTZ/bH79/9+e32khzJKYEk988D45DkA+LvliPnNIJQaqmFmAQAEjpOLZT3WHPLvogysvKrNChxnp5SaAACUUjNdUkZFjrP7HfbGTFl5Q6dSAIBSieTgLjLIWg5kUfD7HfZmBowKmJMAgCmGmZMF3kOBUQ5jMadoU9Ue52qnZKsqqFosr2rTAFApJahQ4W1SEYYV3hfY3R7P9kef+Lhp6FouGY92bOndhhFCQxfPnWhZv2Gb7HS5dVVRWtZv2MoosNPP/vBps1way+fzKU3T1FQqFVNVtTwxMTGg67pWKpVyK9la3x/gawf6tVMTE0Z/Z6dtc0OD2AUAMDSTGXr8nvbH60LuuguDsQt+t+xf7lgXcTpcbsPQdcM0bnsA7LA7XQf2PPxJmyjZz145eXBscrh/T+99jzdEmjr7h/rOXrp27sgdDvltwxijqlHMltRcoqTlk6libMwkhjaTHrnitgfCVd66jkwxPrm2cfsjI7Frx5P52WFFL2bLWiFdUvPJWHZiQNELmbpAW48kOr3x3NTNjroN9+aUVBQW0tiWg2Q2F/U4Hf98bXjs5MHT558mlJJULj8bTabGVF1/lZMjmxcPSVfAwapbQp2tmxvv4QRO4HiOf+gP7v2Ti89f/VeOx8+QFVQvqlqgUsZorRvXTuXZVLLMkvESSwAAKCZTYWE1aBGEOTxDHUUwDB2pQoxapo4E0QY6ITfKbAphXkA8x8f4UM6amRxit2lFghGaeKMWHyuBsmqUD10cPyTwnKCbrzYBczvsvm1rug48f/z0P81lMlOwZC75Uvk17UxWCopupK/Pxn8MAAgjxGmWVXDZxGrCmFnWjZTAYUm3SJkyRqYyubPT2fwFjBBnkrfuWFrhjbHxvLPO59lU0o2ELPA+hBASOE5GCHBB1aNovvZTdUlSDaHM4LAsZBX1TUV5hQoVXp+KMKzwvkArl4rR0eGBus6utTwvCInJiZH7P/Xpz8fGR4fCLW1dpq6pkY7ONZP91y5JDqdr20OP/vqL3/ynvwWAxR/5/MLfFZe+t5SGBrGzZ4O8v6lR6IpEhLZUypoJBvjI8Ih+GQDgsR1tj03G85M9rdU9s8nCbEvY2wIAfXc6Tskmy4Ig2BhjTNM1xWF3uPb03vfIyMRgn0203SCEmLJsdyGEkGHommVZZiTcUNvWvKrn2z/6p7/IF3MpWbI7/vC3/9NDP3jhO3+XTMdXZDra60EZZXabM51X0tFcOTWbKcanPXZ/dU/L7g8BY8xuc/mmEoPnL44e+d7axu2P+Jyh+nPDv3i6qGbjBTUbT+RmhlSjnK8LtvXUBVo3+JyhOsYoZYwtawom4nkeGMv1j42fxZJsp5p624NkySm5mtfXb129p/PA7/zdkxtjw/GBF7967H/ODs5dMzVTjawKr93/6Z3/vv/Y0EEAKN/habwukzk6mSzTJI+BV0xQ3kywcfUNHVykvo3zh8LI4w1ao4NXgRATe3whJEl2qusqQgjTUjEH0ZnbtrtY6aLQJnDih3ev6j3ZP3tS5LHY3VjTzWE8TRbSgVVdLw1NzVz2upzBqXh8SDdNdTGVVNX1cndTaHVPa1VP33iybzpRmFY0UzFu6cu6HFjzq4G3fq9kl1zPL7meefcjev+RV7XZK9PR71LGLIwQzxgwDiOeMUYJYxYCQIQyM6eo0wGHvSWvarOLaaRrN+3cF65vbqOU0uGBy+dCNXUNkfqWzuEbl88Zuq4Jgs0GCMAydX37/kc+Mj022D9w5eyxeHRyYpmnXaHCslIRhhXeF1BCaPO6niKx5tPscom52OOf/yO9Zf2GrS6fPzB04ezxrt6d+72h6tpSPpeeHrh+5c22uVy4ZdklcpxoUWqZhJiyIMgix4kZRcl0NnhdZpbPDmbYccooLag0P3izfD6VIq/UEZZUs+RzSr5wwBm+020qFnn4vid+uypQXZcrZJMnLxx9vqWhbc3Orfsfa23sWHf83EvPxBKzEwf2PPKkz+MPjUwM9l27cfnUnt57f621oX3NxrVb9w2ODlysr21sb23qXLe1554D56+eehHusoMzQolFKSU+Z6je4wiEawMtawixjPHEjbMt1d07MOK4gpqJ902ceqajtmdfxN+ydi43eZNDnBB0h1sMSytHM+P9uqmWhqNXj+qmWiprhWU9ccHVN3UijuP5xhZD7N21BgC+f7tx7Vubd+35jR1PDZ8dPfqd//LD309NZ8eXpo2KknBOyanZ5Xl3vj4WZQQAim91PM3nUlykoY1mkjE6MzUMHMcjhBAzTZ2WS3lWyGew11/FdE2BFVpX+WY0Vnsa1zaH1jZVe5ryZT1/emD2NHn1XJDP5QxtXNWxb+/mDR9ZumJ4aXD45a8/88J/o5TRh3tbH/Y4bJ7DlycP+13y1UxRzd7m5Sq8r0BUt0iBw5gjlGoIIWRZlCGEEQAApYwSSigAaAih5NLaQn+opnZiZKAPY4z3HPjQJ1OJ2em56NRYe/embeeP/fzZR3799/79tYsnXz577Gc/6d54z76LJ3/xQiGfWdEnfitUuBNUhGGF9wV2t8ez8f4PrAnVN7TUtrWvtsn2y83rel7e/tgTTw6dP3t8bnxsaHLg+uVCKjlXSKcSyenJUQ5jLPBYoIxRDiGOATCew7xFqCXynGgTOFumpGXIHTQ/4THmtjU3r414vZGyYZQVQ1ca/f5Gv8PhPzU2dmpHc8sOgxAjr6p5QikZmRwfGk8VJhaff2Fo7sK2VbXbHLLgqA+5649cnTpyp2JfCqXEGpsauT4xM3pjdm5qtKyUCqFAdeTclVMvTkyP3vC4vP7L188dCfqrI/W1je2aoZUPHn3+6YAvFH7uxR/+08Jmrvznp/7s/h///Lv/oGrKillVeitgjFEwEPTaPMBsxB1Y17nl/pKRmauqCUVCtb6HC8V8yu6W7F5n5+aG2uZuhollpouZqnCgWvQwa31w6wN1Wk3DyOTghfbQqs1tsH5vphifKGn5FCy0ZLnjc3I4XVyk3iau3bCTq61v1V4++K+vN3bk/PiJ8StTZ0uZ8qsOxERZlBECZJlE+9mXD3/R1E313Y/83YPmsikAeH7pfQgh9F4yxxiayQzX+BxfShfVNEYYOyTBsfRxSRTtXpcz9P/96Jk/7h8dP0PnxTUAzPcQDHrkYEO1uyGeLccvDs1dFHksbmir3gAAL93xyVRYMciiTWqPNDYBAqjxBcPRdGK2yhuozpYKGcMyDYHjxblMKgYLJwRv/Uwxxli5WMjJDqdL1zWVUWAOp8szeO3CadnhcqUS0WkAYLLd6VLLxXxtY1unw+XpLxfz+dvFU6HC+4WKMKzwnqa+pqq9rGr5cFVVrUsUPFYqUfC6PVW2poYNkl6Whi+eP3nz3OkjpVw2E25tfync0rbK6fMFcom5aDjgDHfU+TtMi5iEMjKTLM60R3ztgAB0w9IRQujCUOwCALymLujdgjJGp7PZ6ZKulxhjDGOETULM9TbbetMi5uGbg4cxRrik6SWew3y6VHrVgffpgdkz4YBzInBZDsxlSnOpvLosZ0hfOnnw++tXb9q1aV3vftMyjVw+m0QIY4wQEnhB2NKzo7eloa1bUZWSLNmd6P9n773D47iuu/9z77TtfRdY9A6CBSABEixilajeKMl2JFnFju04shPbyRu/9mvntf1zfsnPTmw/ceJYLokVK7JsS1axJKtRhb13AESvC2AX28vsTr/39wcIGaRISbZFgpL28zz7kDtzZ+YOZoGd75xzvgcQprCwaZLvMshsMVstXgzhcN++srKyOq+prHoscexljDGDMEKs1dCOndjxeFVV1YlNmzbdVodrFgmCYDbZdGkq0rXLbre7nB6rXWSn+nZ1b7+gCLtkJ+RwerE/UE5y2RSdHB8AjuMvNFbKyedt3F29rLyDYRm+b9/QawAwcr4x73XeT6JQ4FneZRVcHrvJXV/mrvc4TB6WwSwAPD03RtN1RSxImQ3LW2+tLytbqhNDn0slnYzGhhLJ6aFYOh/beXJip9MqOEVJFV02k2vBTqrIZQHLslxDeXVTqcdXajWZbSVuX+miqrqWk8N9x/snR/tbqusXa4auwgUyRcaHe0/lMom4VMjljux56VmEMbbanC4xk0oAAnRw54tPMwzDsizH9Rzfv8vtKwmyDMtd4tMsUuSyoygMi7yvqQwEGr1OR3B/V8/zfbte2263mN2EUEIoMUp8nvLnn3r8v+es8cPDg2c5fHY0lbavbCpZWVD0gqYTTdMNbVmtf1kyJyVHI5lRl9XkYjBmLuX5nKk3Cp15vcGSYDA1nEyONjU3L8ccJ1gNw85xHO/ieFJZWVlit9vdmqYpNpvNaXGVpLDV6oDcdMLj8QSSyeQltUvnOZ5ft3LTxoCvtNJudbgRIChIYhYhQFvWXfvhg8f3vMSxnODzBMqjicgkpe/NFLu3AmOMOI4TvF5vqc1mc05MTAy0tLSs1HVdGxoaOuX1eoO6rqsAAIlEImKz2VyCIJh7enoOrlmz5vpgMFgzODh40jAMQ1XVS/Zg4q0gqWSUxOPTxmRokBbyOSZQWvWH7iNQ42/iTKwJ3qPRos5FrZ03rtl8q9/p8SdzmeRrx/dv39dzfK+qaUp9WWXDzWuvvLW6pLyWEEKODnYffnLP9t8QQoxrVl5xvdNqdz65e/vjeblQ4FiW3bBs5abmitpFj7723P9k8rnzCumFBs0as+D2xtL2TF7J2M2C3cQzpvljVF1XRqfD3Q6r1et22ANn3JQpAEBaFON5kTXJqiFrOtE2t1VtfmrPwFNT8dyCtdApcnlAKCXZgpiZScUjsXQyynEc1xcaOR1LJ2OJbDouKUohk89dMLo3PnS6e97baQAAhDGi5zeNi2KM+y9nQ7kiRS4VRWFY5H0NBUrjmcx0RcDfwLEs73e7KtK5XIxhGI5lmLd8OpiXtPze7qm9o5H0KAACUVbFcEIMS6ouyaohmwXWLKtvdhFcCPqj0X4KAIQQghDCbrfbMzU1NdLY2NiWSCTCwWCwxm63u7u7uw/U1tYu9nq9pT6fL7hv377nL/VcdcPQe4e6j4yGhk4bhqHPxCMhTVPVZSXl+5a43CstybhZjoczsXR0ZEIqDLoR8pfIYtCg1Jh+9cm+av479U6G8cxo+tTSipoXBEMT2szmpQwgFiPAfbJ8QiTkHdeBLQSEEBKNRid37d79W0WWJVEU05OTk4OUUprL5VLT09Ojuj7b9F6W5fyTTz75oKZpSjaTTe54fcdvOJ4ziaKYNgxDnwxNDmKEEIWFjUZRWSoIK9d4DEWWSDaT4Ns6NiGGOUgN4w1hf88/3fEjq9vqu9A+Nt+3rvbg08ceuTQzfnepKimr+s6nv/R/9nUf2901MnAy6PWXsZhhKaUEEIBZMJnNgmA50Htin9Nqd37s2ts/EcukYq8e2/8yx3LctSs3XH9yuP8EAJy0m6322664+o6kmElJinzZptPKqq4wDI7uODmxIy9reTPPms9tgVOQ5OxTO3Y/iBFmLCbBzjCYLchKTtcNzWnl7B/e2Hh7mddWNp3IT6dFOU0W2ECpyNnM9Z281NdFVhX5xFDvcVlVJFXXzmdGNDb3H6fVG7x2xUe/srPn6R/MpCb6zzMWAAAuIAoBAKAoCosUmaUoDIu8rznaO/DaXENcgNkvOUKogRHCCCH0Vo20h6ZTQ0PTqSHdOMsuf77guGxsyXXDMAAA6urqRLvd7pZluZBMJiNVVVVNfr+/Qtd1XZIkMZVKRX0+XzCRSERmZmYmAoFAJbzD5urvFmTWLCD0phXZpNGTTR6a1rSxmxyOe7Lh8VQZQhX9inJqXFUHb3Q47iokwnkfywYFhEwSgwuZ6FR8CS+sdDPY62IYX4GQXAwzEYE3aQzDsJQSghBGhBIKFCjLshyllDKYYRBCGGOMdV3XVF1VGMywmq6pmqaq506NYTD2em0Bs5m3Tk+nJzRNf9ONCsIYI4Y3UUNTEMPylBgazN5ModkRlM62XqbU4nC6OYvFmlc1CXM8Yw+UBBVFkVleEApSND//BsYwDOKzuOVlJQ3LB/XxXimvFFRRlnwWlx9YhNxmh4dYy6ioFrIsZgZ0YixYeweqaQrX1NLBllXUIafLN18UAgCUNZcu3f2Lgz8h5PwtKJrX1m+ZE7cIYTTbr/q9gYU3WZxWmzOcjIUPnD6+Ny9L+YIiF5QznyezYBp45JVnf84yDMsyLLd+afvG1rrmtu1H9750bLDnyA2dG29a2bRkFc9xpxvKqssWVzcs/Ydf/PDrF7gpvmwwDEJWNgeDm1urNpt4xtQ1GuuCeQ3dKQB4HI7S2zZveKCltrqTZRgum88nX9x/6OGdx0488djOvscwQjiZk5IMxsw5f2+LLDDr6ms+oxlGwWe37igoWlLVdfGMY+tFxZj9TntHkXKO4YUyT80SgTPbLvK0ihR531MUhkXe16ia9kc3hZ5/g1LqL6tyOT3+kfGBLnWecDCbzBa/t7RiYmr0koqrCzEyMtIDAD3zFj12nmGXZZ8n0TCyCqVynpBcv6Kc0CnVRUKyAYYNlnFc9aCidCMAHNX16Xqeb3EyjGdIUXqutttv3ymKzzUKwrIsIenaxSuX+jWt0uv2lSqqKmOMcV4ScxhhXOIvrSIGMUwms9Vsslg5luXjqfh070DX4WBJeW04Oj0GAMfOnZvdbnZ94xt3/HDLlpYbb7zxO8vgPGIaCza3qbp1ixodOcHYvGVGPj2DOcFiFDJxxuoKAAAyxOQUxszMkvUbVyFASMykkmab3WGxO1ze8vIqQ9O1Hb9+5CcAZzcz51lOaPE3tJU7SqsLmiT2xoa76j1Vzb2x4VPNvtqlAsubjk+fPggIoYt0ed4RWm/3IWQyWxHLcsquV584d/3gwZFdB3977BFyAcOmVbcsNziBM7EMy3rdPn+JL8iks6m4osqXRWT+regLjfTdvv6aH3506y333dC56aYDfSf3v3R49/MMxmMIIXTVirWbt12x9XaLyWxlMcMub1jcPjwdGkIAMD4zPXaw99SB1S1ta187ceCVrR3rrplJJyKH+04dXOjzeid0NJZ2PHdw6LmpuDilndNqwm6xuD901ebPIQToJ089+/diQUpVlgaarl698qMMRsyBUyefsZg4i89p8XU0lnZgjH5DyPunDvO9zr7hsf8ocznalpaVbuNYbImkc10+m7UrVZAmjEsgEOeDEUYOiyfoc5TVcyxv0g1NmU6Odpl5K0IIM26rv7KpbLkdIYRi2enhlBgNUUqpWbA5A47yBrNgc0tqPhNJjZ9WNCkPAGA3u/x+Z3kDzwqWbCEViWenh1VdkQEAfI5grcdeWoMRZiRVTEczU4OSImYwZpiAs7zRafGVEWroiWx4NF1ITJMFfChXpMi7RVEYFinyDigPVjXccOWt9//w59/7EgBE5pb7PIGyG67c9jEA+MqlmEdHh2XrqVPSLk2jb4pqvdcZVdX5Yut1jBAilFKM0Km5Xm5zy+BsY5L/PPNvBACgurw2azJZrLquaYRSQ9UURZIl0SSYrZJcEAEAFFWRWJblVFWRgQKMTQ737nLRhQAAIABJREFUarqm/klf7JhhWGdJLVEKWUAI876qJZyvaokc6tnN+SpbMG+2y+MnXzXy6ZnwyFC/uzRYrsqyJJgt1tjUxGgiPDVeyGUzmqK86WFGQZXEvtjwKcXQFJ7heFXXlbSUTWKEcSgTGaNASVLKxBe6HpOtrGliq+ta9NGhbra+qQ1hPD0/+vnyT3b+y4VEIQDARPfUMcxgBiGEGmqaWnlOMI2GhnsAYOiSnMCfyJN7Xv5Nidu3+8oVa7beuu6q24Ief9m/PfXwd2VVVT53231/u6/n2J6fPP/Yg6lcNvng57/xn3MZC4QQurS2ac/W9rVXt1TVL97UumrL8wd3PZu/QB/Iy41wQgyvaChdUR9U6idj2UmY1xvVajbZgz5Pzfd/9cQXxqbDvQAALMMcZxmGX7ts0bVKIRIKem1BWdXljqbSjif29D8BC9RGp8ibcZhMQQvHeeJiflBgGbvDbCoLOGyLjo5PPQIAk5dyLnaLp/Tq5Xd+ySI4PJqhSJRSsqvn6R8omlywmZz+1porbtV0RbKaHF5RSseeOfSzr5h4C22r2XBrfXDpekKJwWFOGJnp2ccy7M9NvM25uumaewPOymYAShEA6p44+DsGM7+zW9wlV7fd9TcMZjmEEBLlTPzI0Ku/BIBMmad26cYlt36WUmIwmOUlRUztOv3MgwBw3l6kRYq8lygKwyIfOEr9ZVVLmltXswzHD4z2HpflQn5Jc9sanhOE7v4TB1iG4ziOE0YmhrqXNC7rnIlHQqls8sCqtjVXze1jUcOS9ury2kXtyzqt6BJFaUwmbPn6N0r+tr9fOQwA7ztheC5zYnB+g+930ux7fGq0/49MQ+x5+yEXhqqSKIe6dxJZTCPMsIQzWYx8aobI+ZQ8ngoDABhiIkxnU2nHYF6NzNuRlnNpAHj5nMVzYumyiFYDAGB/oIKqsoRsNhf2Bsrg0N4X5q/Ppwtv2ZsuG8tFMMZMsKS81mK22nVD18R8Nn1xZ/3uUOEvrQi4vCUsw0Z2nzqys6GsqrEqUFZtMZmtqqapdovVHs0kZ3TD0Ncvbd/YUlW/OJyIvmGyMjg51j84NT647Yqtd5R5A+WvHNt37vW+bDk+NHM84LYGbCbOZjPz56bzoTMtOs56IEApJZKiS4NTqcGRSHoklZNTsUwh9n5ybX0/UOaytwksa89I8tRILLeroGrp5ZVlH54rz7iULKlcdZ3XUVrz2wM//XJKjIXMgtUpytmE0+IpZTDLxjKTg3t6n/uJ1eT03rf5Sz/3OkrrWMxyDcFlG06M7npiONK9pzqwaPX17ff+/VD41K6gp3ZpwFnetKf32R8nsuGRlY1X3d1ev+nDofjA0YCrclGpu6rl13u+/5lsIRkx81aHpOQzAACdjVvvjaYnB/b1P/9fZs7q2Lbm0//cVNa2mWW4Cd24vFO/ixR5O4rCsMgHCo7j+Ruv3HZdXspnQ+HxQcPQtcVNrZ0CL5hiyVh44+qt22Zi0xMYM0xoaqy/uX5xh0EJSWWTsbl9OO0uz9Ubb1yXSEWnWfbClvzvFhgjvGat5aa/+Vvfh5ubhVVf+pL/51/7eqm6f3/+2e0v5/7H42FKrrvO/ufHjsmvbL3ado/Vgp0vb889HI8bU9dcbbvvf/4n9U1VpQoAwOo11hsCfqbyxRdzD70fo45zLERtGtHkAgAcejf3yS1uXUNSyRkjPDn6bu73YqGPDJ7Cdocbe/xBfeD0kbcyezgf9R016ziBM0dPpk/0D58+5nS4vKlMMn6x5vtu4ne6/R+/9vZPlvkC5YRQkpcL+Wf2vfpkLJWIEkrJU3u2P3HHhms/smX56q3T8ejk4NTYgKQqbxjLKJqqbmxd9fw/f+qL3+seG+gKxcITC3k+7wSLwJl9LrOvzGsLlnqspV6H2avpRAOAfXNjCrKSi6XSU7dv2fjZFc2NvxELUmbDitaGTe1tt+88dvLJ0Uh6bG5sa10gXkwjvbyYSmVPFDQtpemG5DCbSj1WS42i668quiFe6rlUB1rWDIW79oRT471nFokAAD5HEEQ5HRudOb1flDJxAIh/7ubviHazu0TgTNbmivar3DZ/5bpFN3ySZXjTmXTUhlJXVUtj2fLNLquvXDc0xWZ2BwAodVg8wZl0qC9bSEVu6Lj/G13j+57pnzz2iqwVRIww/t93/OhKVZPytSUtawAAKnwNKzKF+DSDGQ4AisKwyHuaojAs8oHCZrE5rRabo2fg5IHB0f5TPk8g6LS7fT0DJw8Mjw92b1l3zYeSqViEUEJm2+rhNz0VdTrcPoQQGhztP2kSzNaqspqmizlnQihxuZj9uk61hgZhxWOPZf4lXyDZTNqIAwCYTNi6abPtI4IJWw4eKPyOYYCLRfWJVMqI1tXzrYuXmNYBwOsch7jPf953e1+fcpBSWLCUQ8wwmGEYFiGMDV3XAAEihmFghmEAAJqWdKwZ6T91VFNV5YxRzRvwPMtzHMNjjDEABcMguqLoinEewwqGwdhs5q26bmiyrMkcx3A8zwpntqW6TgxV1WXjnNRGlmUYnmcEhsEspUB13dDMZv4to8IMgzHPsyaWxSwAAkop0TRDU5Tz17haLIIVAKBQUPLzjwcAYBjUUFVd0XVDZ/yBcr61fYPlpjvS2ulTB7DT7WOqa1v0gd6jwPMmfXS4G7vcAWy22JDbE2B8gQptsO+Y1td9+A+7Ku8OiGE57PaWYF+gAjRVBoA989ebrIIN4QtH2Nfc3lGPMWbkYTrhsLu8NqvNCQALci4XgmFYxu50+2qblraVVtQ0SHkxG5+ZCjWs3LDi4T0v/8pmsTlqG5e0TU9PjOw9fmB769orr50Y7u3q07R07NThF+12l6ev79Th3q4j+wghhjHPACuSioUJUPr03leeeC+4JGoG0cSCJtYtdtVNxnKTmk60c11Jc/l88jev7vj3bZs3PvDZj9z2HRYznKQq+dcPH39sJDRy+K4ti++kZ1JH/2xTyzIG4x7jLUzBilwarALvsfK8j2dZyrOs1W+3QY3XvYxQoveEY8/RM9cIYQbTc/5OXyw4ljepmnxeQaobmiprhTfM4SilhEEMw2KOi2Ymh3Z0P/3vOSk5MztWV9P52GRtyeI1Y9Hegwf7X/rvuW1VTc4nxeiEqsuFZw//11cay9o2t9Ws37a4svO6Mk/tdzDCA5QSur//xZ9NJ0e65o4nSpmYZrx96yCEEWJYhpuLouvqm43MihRZSIrCsMgHioJcyOmGrvu9pRXBQHmGZVkqFnKZYKC8llJKE8nYdDafTVWV1zRVltU0uZweP4Mw9ji9AavZ5vC4PIF8IX/GyKSsymK22AHBRU8lTaeNWEUlN6QqVBobU3tyOeOs9DqEAO3bW/jt4KAyZ5xCAQAdPSK9fMUV1lsZBu2qq+OXms3Y1tUl79Z1ql/sOV+I6vqW1tqmpctNJos1PDk6SCmlw30njzQsXtGpyoXC5hv+7H6Xx18yePr4QThTw4IxQjU1/sYHHth694YNzdeWlDjLCSFkaGjm9JNPHvm5y2V5OX1OmmJDQ8ni7373o4+89trp58rLPT+9++51t9x8c/vdZWXualXV5GPHxvb98Iev/CMAzD19BqfT4rrhhrbrb7ml456mptJlkqSKx46N7Xvppa4n54TbudhsJvumTYs23nFH558vXly+nOdZUzIpRg8cGHq9uTn4s8HBSM/8KAjLMuwPf/jxXwIACgZdn7n22tZ1t97acc+iRcFlGCOmvz986qGHdv0rAGwHANCH+o6TTDpuuv7Wj2unju2Sd7z8mPm6Wz9OZblgRKbHsNcXxA6nh/GXVCp7d/7WmAkviLkQYhhGuGLLCn2o77j20jM/hzN9GOfzlz++/ymMEQMXqCFzljjK9/368H9GYuGJrJhJ2ax25/nGLSQ1jUvaFrWuuqIg5jIuT6AUAAHHCQJvttqcVQ0NJrPFOp6MT+dUJd+8fM0GjBlWU1XFwBhNZzOxxFDv8Ynhvu5UMvZGrTLLsIzA8cI9W2++KpKMhQ/0nty/kOf4TkEASFZ1efuxse3ZvJJlWczaTGenkhqEEIzR8X/95eN/bTEJDosg2HIFKSUpiui28a6RMDsiSpoIAOCw8I5ieeHlQVOJ75qGEt+VWUl+I93ZaTFVTCTSB5nquhZsd0xRTVPZhuZFcB7DrotBND05UOFrWG4R7C5Fk0SW4QSD6KrL6kMAQOk5Hx4KlGalVFRW8mlNlwvTidFTgABxDG+S1EI2Kc6Me2wlVTkpHY1mQgOz9YQYq7pc4FjBXJBzqWPDOx4PxQaPX7Pi7i9X+ps6ppOjPR+76qu9Js5ij6QmThtE1zhWsGi6Ir2Vy/kcgWpfQ8OqmvUmm8khJvIxAHgUYNZYh7yHXJiLvH8pCsMiHygURZary2tfWNm25qqKYFVDd//JA32D3YdXLV97dXlpZd2O/dufTGYSMzWV9S2LGpasjCVmpgpyQWxpXLpKUWVpUf3ilQMDR49MhPp6G2qaW8V8PjM8Pnjq7Y/87nChbw3DAG16Whsi5Kw6Hlpby7/W0WG+pqyMq1+0yNQ5Nq6eTqaMmUsx1wthtTmd2VQidrzvtRe33fPZL44MdB/nON7k9gaCw32njoZDo4NH9rz8jKapbzxJra72N37967f/x+LF5cu7ukJHBgbCXSzLcDU1vsavf/22f3/00X0Put3Wf0ul8om5bTDGjMUi2Jqbg8u+8IXrvllb62+Kx3Mzk5PJMYfD7FJV4yzhYrUKto98ZPX9n/70VV/O5eTs0aNjuwsFRSwvd9d84hOb/1dlpaf2TediFawf+lDnxz796Su/HA6nQ8eOje1TVV3x+x3Bm29uv6utrXr1V7/62KcA4PTcNggBmM2c1eu1BT7+8Y2fX7GiZl0uJ2eOHRvbfybCqc6vsyKZdILKcgEIJcCyHFNaXguGoVNFLrB1jcuwzeGmmqKQQj5LFblADX1BnPGwxxfETpeP2u1ujmE5IubScE5qLcMx3HP/uv0b5AItCVqvarlJ03RVUWVpfefmm1mG5QDg+KWY/zsln8ukxWwqaRiGHhrp786LmUw6GZ0JjQ2ezovZjL+krFKSCqJcEHNOl9fP8oLgdPsCPcf27WxZvmY9AFDD+H2UQOB4vr1x8Yr1Szs2rl/aseHxnS/+KpsXL9i4+3KixGMtqQ+66kvc1pL9p6f2FxS9IKnaWX0XOZblSj2e8pQoxlLZXBQAohgj5HU6gwYhxsmR6MkSt7XEauKs3WOxbkqLyvByYDKVPTqRTB9MiIU30tgDDluzwLJ2buX6dlJaVkNlKc/4SyoRZk5ciqjhidFdT2xb/Rf/fGXrHX+TyM2MCZzZ1hs68jIhF466TSdHuyYTQyfa6zd/xO8sbwQAahBDPzz4yi/6J4+9Uu6pW7a6+Zr7I6mJXp4VLCkxOtEzceiFCm99W6WvqV3RCqLV5PTphqokc5FxAIAD/S88tHHJts8ihLCkimmryeE7NbrvaXgHJllVS8vbDd3QxGQ+Llh5G8IIYcDYa3H5aj0V9kQ+Hc0q4jtq01GkyMWgKAyLfOAYnxrth3l9tliMmGP7J56vdOBKk0QNkKkaOvrfB9IyTRc0KKQlmh4d7+uODGzvSUo0uSyAl/WOTp4+0nX4tfYgbgeKoMaFq3kG8QAUKAAdTdFRndB39QZ97oYJIXhTeiulAIYBbzpeImGEQyGtf/Vqy40uNxPo75MPF/Jkwb90PIHScoQxTkYjUwgQamlbvd7tCZQApVSW8mJDy/JVNrvrtJhLpzmO5b785Zs/s2RJ+Ypf/erATx55ZO8PIpH0NMYINTeXtX7xizd+6447Oj8+NhYfYhj8i3NTQzs76zYdPDi849/+7aVvHD48sltRdMnlsnhtNsERi+XCc+Pq60sW3XJLxz2yrMkPPvjKPz7//InHVFVXSkqc5Z/4xOb/tWxZxcpzz6O9veaKu+5a+5ejo7GBb3/72S92d08dJYRQr9ceuPfeKz57zz2zL5ZlPq/rhg4we60AAKqqfPVLlxZWPfrovh/t2tX3QipViJvNvNXvt5cmk2IMAEAPjQ/QbCZBdV1V9u14hipygfEFKtQTR3ZQVZG4+qblJJdJkpnwOAiCmUqF/MW8bm8JpYSkU3GqqjLVNRW7vSXnDjn+QtcTg4dGdl3ImbT9+mV+lmdNAJRORUIjPCeYLv7E/zAiU2MjGOOx80QHTgIAMIJgApcnQJLxMOJ5EzWIDmcMWGYOvf4qSSYiRJ11nq334Dobz6UDDlvAZ7P4ntn7wtOvHt29nUU66zQhs51H9pk8ndGMhYvwvxV5Wcvbzbx93eLydQxGTF7W8qFoNgTzxLzdavFs27Lhgef3HngIAPoAADDCTGtj/QanzeodGO3duX5JxXqMEeZYhjvYO30QimHDy4Jyt3NFa2Vwxdz7MpeDjWbFPqrIBexw+SjDsFpfz+FLlUoay0wNvnj8F/9vY7Bts99RVl9QcimdqIqsFrJHh17/da6QfCMKf2x4x2PxbHhElNJxp9X7UHN5+1UBZ0WTbmhKKD5wlFBiJHKRcb+z/N8XVbRv9TmC9ZKST+ekdIxSQnJSOgpAwe8sb1A0OX9o8JVHxqP9hwAAhsJduxDCuDqwaLVFsHuyhcS0okvvqOZyZjQ2IOXkbMPK2vWarEkIEEKAULWrrC5g8wZDfGQMLlEEtkiR81EUhkU+8HAMcKvKmFUcg7hY3oitrWDWNvlwk2aAJmlUCmVJKJSlodYS3Lpr3Njlt2L/dI5Ol1hRyZpydo1KqOoSGFdOpTlCgWAE+KHj2kMA8K5azedFkiGEGlXVfIvJhI9TCkRRyFvWNBQKJNvXrxy64Qb7Jwf6lSOhkNZnGO+uYP1j0BRFlgt58cDO3z3JcYLg8gWCyXhkOhELT508tHO7xWJzzLVeqKz01HZ21m1KJvPxOVEIAEAIpQyDux59dN+D3//+fb9eu7Zhy+7d/S8CQGz+sSRJzT/77PFH9+0bfHVOnAFA/MwLAGbrCm+8ccWipqbSpc8+e+zRPXv6t8uyNveznezsrH+ys7N+U2NjyZK5bQSBFT73ueu2VFb66n/849e/3dMzdXyuLiyRyEVra/2PfOhDqz/R1lbVGQg4ggAQApiNGAIAaJqhHjgw9NoLL5x8bO5YiOPVSeQUmMb6VVzDojgSBA5MZisYOk9lKU9i0UnQNNWYCY9jp8tnpFMxUOQ8YMxQWZbon9C380/FiEen+baVcZKIhY2Z8LiwduNNCGM034Bm3+OHH3qrdhUjxycOzKaaIlQaKK+OJWamLs3s/zDeKmWMsizHtyxdrY+N9ADGmCkJVoNhGEApBZZlSTaTgDN9KtdWsGvjBSOeDB+KvxA59GytG9dWWklFQwXbIKogaoRqKdlIAcBlKQyTWSnlc5r3swxmx2YyY7JqyJm8cla008zzlsbKyhWKuvvBuWWUUmoYhtZSU70qnRrvzUlqziDUqA+66hmMGYCFq4H+oIARRnazuyTgqmxO56OTscz0MINZDgCoQXQdI2AYhLhKj6szJythUVFiLoulStWNfHR48BTb0LwcWW1O7A+UwxnBf7ExZtsJHYLzm3z9eP6bvb2/m2tjBJl8IgwAj5xvn7HM1DCcv83EMAD8x/m2OeM8uv3M623xmLH77mXc3cMpMiypjoF8upAcPDSy09AMjRBCKpylFV6Ly2/mTJaklH5PmG0Vef9SFIZFigAABaADCWNgIEkGrqxlroyIJJKWIR3OkXBnOdO5JICWvDCov5BTaC6r0CzHANfgwQ05leamcnSqwgEVpTZUSigQBgGDLkLdYS5Hknv25J+6917X1ySJ5na8Lv4aAB5/q210nep1dXwvJUDyeZKJRPSxd3tefyiFfDYzPpzrCo309WiaOpfO2T9vSApjjOZEVm2tv9luN7uGh2d650ThHIZBSENDyWAyKcbKyz01gYA9COcIw1AoOTI+HhuaJwrfBM+zQnm5u0oQONPwcLQvnS4k5q+fmkqNx+O5mfnC0OWyeisq3LWCwAr33nvFX91yy4q7f/azv3hDBH3zmx/iXS6LJ5+Xs36/vRTOCMM50ul8oq9v+uQ8AQrIYrGzZRV1VNNUvqPzKpIXM2AYOhVzae30qQNEyufYhuYVJJtJsPVNbSQem+Ja2zfSgpilmqaQRGwKFrCVCTKZrFzz4pXY6w8yFVVN57qSynnlTU/VMcZ43m9LhBJK3U6PT1MVJeAtqbgE0/6TwBjh0lK2FgBITiRphMwECSYL4/WXIbPZin2BcpJKzGC700PksyO6dh7ZvRbG2zVjdKkGqCYWTMsCeFm9B9eHcyQcEX/fM/VyJZ6REvVB96mrO2qu3nkqtFPgGOGcIQijWeYvxAgzDMZM70Sitz+U7G+rC7SFYrkQOaetRZGLQ5W/aeUNK+//hsPiCe45/dyPMMKjjWVtG7yOYC0A/FdMzA8m8oURt9VS0zUVeTorydO1Ps96jmHMbG39EprLJvWJRJQW8guegXK5YxfAfmMTe+PeCWNviPGjmZFYf2w8MQIwK9B9Vo86kBg/7TE7fDExedn/zhd5f1MUhkWKAICoUlHRQTEIJR1B5sCNTdyN/Qmjv2uGdkVEGlnsR4vDIgkv9jOL15Qza4YtaPjItHHk+kb2eoOCMZklkwbBhmaA5jSB89wi+HcDw6CG2Yx/8Ppr+V8iBEgUSRoAIBrVJ77+tcg2WSbnuelGqKaWR1PT2tDgoHpclsmCNMw2m3lLZ2f9xmxWSodGp4+ZTJxZELDAcQxlWYZlGMwaBtFVVVdNJs5ssfCI51lFVXXVajXZWRaz2ax03n52uk40UZSzFgtvNZl4y7nr83lFlGVdOt+2czAMZq1WwU4ppYWCKqrq2cYpkqTkVVU/KzprNvMWs1mwEkKJILCCw2F2nbvf/v7pU9PTqQlV1d8UyVNVXcnl5LNvqgghwHECtlgdJJtNUk1VSWwmhG0OF1vX1GpMT45gn7+MUeQCAEI4UFJJ82KGSAWRphIzVFUXLGIIAKCePLqTra1fgp1uv/zqC7+40Di71+Zbc1v7fUu3LLrhay/9bVCVtQLDMhwAhZce3PHtwZ3jr/QMnDpgEszWSzn/P4aSEra6bbl5E0bASBLJHTuhvJ7d/doTlBgGwpgBhDE1DB0xDAMUgMrSG+Lw0W71UYwASxpIFICenDFOYgSYQcAYFAxCgUgavOVn93Jg7ZLytaKkik0V7iZCKcEYjc8ZLsmqms/m86nN7cvv8Dqdj2i6rgT9Xs8Vy5fdPDQ5dcprN3u3ttdsLXFbSxRNV4o1hhcfBrPsnRu+8BcnRnc/6bR4gxbB5gYAEOVMbOOSbZ8BgP8is6UQxtr66tSKyrI786qacJpNZaPx1B4wDB4EwYIwRvQS9fG9ENZm63rGzDi1hBYCBliqUYloRDFVmdq0qDYiTUgnF3J+50J0onXesuLuKz+2PpOKZCYJJU+5zA5lsauutcIRrOmLjXTBB6BPcZHLl6IwLPKBR9KojBF6aa55+tGwcYzF6CQAgE6ogRGa/t0g/M4glGCEDh2eNg5TCpRQShmMejACbBAwYF6J3ztpxP5HzVUieQA4K+pwxmH0TSl3DgfjaW83dyxrNW3M5Uiqt1c+cDHm9E4wDKJrmqGaTJzZ7TJ51qxpvNJmExxTU6lxv99earUK9kJByXV3Tx1bu7bhSrNZsB49OrIHAHaIopzVdaLb7ebzOlSy7KyoSybzMUXRLnAT/daXwzCIIUlqHgDAZOLMHMdwmma8YWjA86zAspibv42iaPLs8Sh87WtPfObkyYmDv68gnHdkOrv/+e/n/XvWeJLLphDGTwKcMdlDCAOhBBAAIISoYRCE8fCs5yzAbGIqpYAQAkrpH9o38N2GylIBAA6fm0J6LrXLq9ZUtVZ0vPbQnu9vvGftA6/8567vBWp9jaV1gebwwEzPkqbW1TwvmKYiE+dL8bqs0HWqSgWSAwAgBIihEYUU8u+o3igj03OjLe/JG8KMKGeqAo4ql83kmoznJue78GbEfOLF/Qd/fvd1V//vta1Lrs/Lcs5ttwdCM9HBlw8cfiQjivGxmcwYx2DuvquX3ncxsi2KnA1CCHnsJdW/PfTTL7fXbf6Iibc4AAB0Q1MEznSWq+zxialfNpT4rnKZTeVjidS+qVTmuGGLlnNL265gq2oXGdOTIwAwthDnAQDAebkqZVLp5kv5Ri2lTWEB2wSXEGTMjB0IGHCm9ncOh4DstW5ca+OQDePZen1KgfYnSH8sT+Icg9hSGyottaFSDiMuJdPURJpM5DX6xkNVu4BstS5ca+ORTTOoNi3S6RmRzsz5CvAM4sodqLzEiksIpaTUihEhs+nR412TRyf7wl2UUMKwmMUIoTpPVUmdu7LJJlgdl/BHV6TIeSkKwyJF4M1Cbr5xzJl19Nz/AwAYsy6gl2Xqk8PBeDs7Ldcnknp418784/kFNp2Z+wm3tJQvj0TSk/394VOf/ezV/3dgINz9+uu9z61aVbuxqspbr6q6Mj2dmohEMpMAAKOj0f5MppCqqvLWBQLOYDSaecMwhmEw3rSppcHnswcOHhzeEYtl/6g0nLljapqh1tb6m51OiwcA3nBvLSlxVbjdNv/8bVKpfHxyMjmm60Tr6Ki94tSpiUOaRt62Fuztnq/Ts+vX5teD0jPr6bnL4DIy66gtrai18MIMzKux5VmOC7i9JTOpeETTdd1kE+zxieRI757BV1bevPzOyHCsb/jo+N4Nd63+dPXyys5jvz30S5NgtkryOzN0WEhiMX0K3ialGwDAZTGX1/o8GxACNJFIH4yL+dlUMowZv83aVO52LKeU0oGZ+Pa8oibebn+XA2aBM9UHXfUBtzXvsZs9gAD29kzunT9GNwydZZiXwvHk2OK6mjVWk8keS6enjvb2v5oR88mgx1baXOltNvOsmVAgRWF48aGU0pSMa9EsAAAgAElEQVQYCy0q77jaanL4OIY3BVyVTctr198+lRztmj+2qcR3td9ua0IIMeUuZ7ui6WKIZRUg1ABNU4Hj+IU6DwAAhBHL2Bm/GlNHEYt4BAjUsNqHLdipRtWR+WN5BnGf7eQ/saWG2aLooDR6cWO9G9fvnjB2f3On8k2OQenV5czqP1/B/XmFA1cAABQ0Wni2X3/WaUK/ycg0a+eR7VMd/Keua2CvowAUAaDRFBn9yTH1JwxGJwAA1lQwqz63mv9cqQ2VpiSaihdo3G0GNwCAt9xdk4nlwkpByde1V6/NxHKRyWxkPJSNjDf76pa+/TdEkSIXl6IwLFLkPYpF4D1ui7kqmS+MSqqWETjWxiDMFVQ1BQAwOakOAsDfLvA0AQDA67UHmppKl0qSWgiFEsOBgLPMbjc7o9HstK4bmiQpeUqBxGLZ8JYti2/q758+efz4+H4AgMnJ5OiBA0Ov3nbbqo999KPrHggEHA9Go9kwxgg1NJS23Hnnmr9IpwupI0dGdieT+T+qcF/XDX358ur+4eGZ3nXrGq86eHB4hyCwTyqKrvh89sC9966/qbra2zB/G1nW5E2bWl7bsmXxTdu2ddw7MhLts9vNO3I5KctxLOdwmF21tf7mQkERT5+eOnGhY7MO1s/7+VotoYX0jB6hFynafKm4/5rb/vyxnc//Eua16LBZrPb7rt728Z8+/9iPACCmFNQ8ZjFnspscyen0xNLNi64fOTa+31XiCMYnU2OFWXfVhXNYfQe0VQY/1DUZeYrQtzdzQgghn81iLnXalzYEPJsxQiwAjADM9gI0cayj0uNcVe/3bo6LhSEAeE8IQwYjxmHhHesWl62bjOcmAQCaKz3NGKNTLMuZWZY1MwzDUYD4xEx0OFWQCrIspw3DUFRVLQAA+F0W/4ZlFRucVsEZiuZC13TUXVNf5u4bnk6NvPXRi/yxEEqMw4Ov/KKj4cq7Sl2VizhWMPscZXWaoUivdz35/fljgy5H24lQ+FeZgjwFQKlOiMItX/thkoiFkcPpwSaz9e0yBC4m8qTcpUwpPUQjZ0XbzzSQP2tOtW5ce18rd9+39irf2jFq7Fhfzaz/6gbhqz89qv70yLRxpMKBKh5YyT0AAPBPu5V/Skg0saWG3fKJdv4TcYnGAeCZrXXs1vvbuPv/ZZ/yLyci5ITPgnyfW81/7q87+b/+3Avy5ywcstyzjLtHYED4h53KP8QLNH5TE3vTzc3czQAGlLcEW2vaKjvz6UKCN3HmM66kEM+nZkps2SBcRg/5inwwKQrDIkXOg91ucixaVNYmikp2ZCTapyhv7/ZYU+NvEATWNDoaGzi3Ru2PASGEPFZLbZXH2ZkQC8MaIbLLbKrgWcY6nkjv5xgMJQ7bYknVUgLL6s1B/7UJMT/sspgtAYetxSbw/ul09oSi6/lqj3tNuiBNTGeyJ1XduOQ1S5KkFk6dmjikqroSDqdDPp8YNZk4c39/uAtjQLmcnDlyZHRPa2vlqh07ep8rFJR8a2vlKgD4raYZenW170f19SUtd9655tPNzcHWr33ttomvf/12pq4usGjRorLWp5468vBrr51+1rhAf7x3wvDwTO+zzx7/5Sc/ufnvPvOZrV9dt65x69///bbsN795RxXDMGwiIc6YzcJZNYxHjozseeSRvT944IGr/s8Xv3jjtwYGIt3f+taduW996894m83k8HptgRdfPPUbADivMMQmbLctsW01V5vbc125lwABcq50riEqKcgh+RRfwjeyNtZbGCrsxyZsM1WbVqhRdViZVLrOvRG6XFhUVddiN1vt85exmGHXtLSt+8WrzzwMABAZjvYhjDAllJx4qfupG/7qqr9feVPrn2UT4szxl7qfWpiZv3MEjrV97IqOT/dMzzwL8OY2Medy5gZ1yG01/8Rm4gPz1xmE6ABwsMzlkLw2a8P593B5Is6mX+//5A3LF58aiZ7iWMw1V3qbHQ5HhdPpqqKUUowx6/F4GkRRjJjNZk8wGGwXBMEhCMLjiqKIJp41DU+nh8NJMbympWxNIiclpAumhBd5N6CUUIZhdxeUXDLgrGjiOZNV1SRxKjnaFU2H5huBgU6IvKjUf30qL41RADKTzfUl+08fpXkxw69cs5XkxexCShllUukh+u//FuIzNY/ne8BW6UCVHAPcnnFjz0yeRJu8TFdKoim7gOwIAap24eoWP9Py7T3Kt3eNG7sIpbTCgVNX1jJXbqxmNpo59PIPbzDdMpQiQ4/16I+pBtUAAD7aypV/dYPw1aAdBy0cWFr8uOXxHu3xHWPGDp1Qo8XHKB9ewn0YACARSo5yAmcq5KRMZiY7TQghtZ6K8mZf7VLbmZTeIkUWkqIwLFLkHDiO4To76zuvumrJLadPTx1Lp/OJa69t7bTZBEdPz9RRQeDMS5aUt/f1TZ+cmkqNXXdd24cjkXSovj6g1db6F61b17h16dKKXd3dk39SLyITx9prvK41kYx42u+wNrks5sqkWBjNyepMvd+zuX8mvt2gs19MBqW6qhmiTRACLGYEm8AHYrn8wJKykm1ZWZlmMRJcFnelYhgiAHS/Kz+oP4B0Op8EgPk1jnEGY2belzedDCXGnA6zp2Vx+QqTibP2902fmhs8Ph4fbmgo+btbbum4Z/Pmlhva22vWGgYho6Ox/u9+9/mvvPpq9zOxWG5m/jEppVTXDU3XDe2tDC3w7DwIAOSeeOLwf2ezUvqWW9rv3rp16a35vCIeOza29/HH9/7Xli2Lb7rttpX3Ufr7p9GSpBV++9ujvwiFEqO3377qY21tVZ1Op9mtKLoci+ciBw8Mvb5378ArZ89r1jBH1w3NUEhBz+kFKSSdVMJKL2tn/ZyPq8kczPyKD/B1pnLTEqKSvHuT+5NaSpsEBIxtsW2rkTPiADD+J1+YdwkGM7jCX1JZU1JR+3/v/Yx/RePijs3LV1uBAkUI4buuvKnNajJbDWPWGTYeSo6kI5lJXTPUyd5w9ul/eeErFqfZnY2LES2l5HwuZ2mJ11PdNzZxxDCMBW+vMgfGCLdXlX/03jUr/qzS7Vr1qY2dL/zl5jXkyNjkw0fGJh9mMGbaq8ruWVJecotBqNYzPfPM6enoc7Kmva+dG18/Mf76mpayNQxGzJGByBFKgTidzupUKjVqNpvdqqrmvV5vI8aYs1qtAVmWM2RWEINF4CyqpquTsewkailDRwfCRzX97VOyi/xpCKzZOpkYPhmKD55ACCNKCRU4s1XgzXYAeOPzOpnMHHFZTFUsgwRCwWAQYoBlEFUVSTmw+3eg6zqlC1fbXOr0eZsrah3TiegUwzBsc2VdUNE11edwp2VVkVVDUw1CDMMwyGCSDKoGVbctYreV2fFT19azV1h5sI6lyRghQOw82DECHCvQ2Fx5iaRTaSpLp3xm5HMKyBm0o2BfnPYZ9PcPhEIZGgIAKLGiEoOCYWLBFJdofK4kRdKpFM3TKACAUlDF5HR6Ip8pJGvaKjsBIJSVxXRKziYs3OVvtlXk/U9RGBYpcg6aZmiVld7hkyfHDxw/Pn5AktR8Q0NJy6OP7vuR2231dXTUrs/lpPSaNQ1b9uwZ2C4IrGlwMHK6pMRZFolkQl1doSPXXLPsNvgTm9QiQAzHMBZZ09IMQhzPMNaMLE+LsjJT5rK38Qxj5hnGxjGMBSPEUKBg4liHQYhSULVEXMwPtFUGP6LoRk7StFRczA7nJPmysML22az1G5tqb1d1XRQVLaYZhmzleQ+nMOZ0f8rNM4wViyRlE4QpUVHiAABDQzN9JhP/jz/72c7vsSxmZwWWoUqSmldVXQMAQBhhxsK4DMnIsgzuv/PuH2wyDGLk84qIMMKAACGMGEopoTrV7XZ76erVq+8eHh5+ked5u9VqDRw/Hg/t2fObz5SU+JdSCkTXCclmVf13vxve/+yzvdtzOYPW1dVtNZvNh5ctW3ZzLpebHh0V048+2vPyf/zHrh+kCvkQsjs8UFbVUBgcOlEIz4zOP3ddN3SXy3IXxgiLopJjK/nlBmOwpEAy1ELdek6Pailt2tZiawYAquf0hJ7VY7yfrzNEIy7FpBEjb6QW4LJdEIQQri4pr7nvmm0fbyyvbvrLm+/6q8IZ900KAHmpID68/bcPJXKZBAAAJZRYnBZPXXvV2pJaf9OOh/f9IDoWH7S6rd6KUn99mdNTrWqaPIjRcXgHEblLBSGUCCz7ZELMD9/eseyHvzp48n6dEFXVDREAYEl5yW0NJb4rX+0d/v9MHOdcW1/5FwAADMaPG4Rob7339y6j4fRoOCGGAQEoqqFQAOjt7X2SEKLzPG/1eDyNuVxuWpKkBJxxVtK02TYtRwcjR1c1l65av7Ry/QuHR14oisKLD4NZ9s6NX/jOzu6n/x0ATs4Ju0pf44rldRs/BABfmBsrKmq0IeC9MlWQxkdjyd0sw5jZytpqA7OsEYuEYIFTH/1OTyDo9Zetbmlbqxu6ZhHM1lAsMrGkpnHZaHhieCo+M3l8uPcYAJBQhoZ+fET78ZfWC1/60BLuQzMinfnefvV7x8LGMYNSsrWOlSkAtfHoDQMeDiPObQZ3WoZ0QaOFjAIZtxncCH5fC+s2IzeDgMkoNMMzwGsGaBYOWTBCiFBKOQY4B48cAADZWG6mc1v7Ry1Os/v4i11PAgDk1EK2oEpiUkrHz/RqLFJkwSgKwyJFzoOuG5qi6Iqq6gqllGazUrpQUMW5lgYMg9kjR0b3TE+nxoeHZ3q3bl16SyyWjcTjuZlEQoyyLPMnF+Qrup6bSmdPLCoL3JDKSxOhZPqwpGpJVTcKoqzOuC3mKpvA+9xWczVFQGprS9fpmi4b8YxKeKxBGqPpdPbETCZ3usrrXo0RIELpZXHTZeE5d05WIumCFBIVNWbmObeZY52ipEfj2fywRgypoGjJOVE4hyyrEsCF7fsZC+NyrHTcLo1Lx6lG5bymKRRTgyvjq1iD6oyZcVCdqkbeSALAcC6XiyxZsmRCFMWZ8vLyGovF4ldVNed0+hZnMnLGZDK5MpncVCAQWDozM3MyEomctFgsXpvNVqrrupLL5aYKhUK8vLy8U1WJrqqA02avi2tctorqmkbrF6+gU+E3OWum04U3hJ0QFDKUUEIJ1YlMcnpSnwIAkMakI4yDCTAmxi5Pyd2GaCRNlaZlkIUZ+g7q2i4luqHrALATAHY++Pn/56c/e+mJn54a7jtBZwO31CCGQQihPz8z3lvhrtn6iY1/46v01FUsLlt+4KmjD3MCa1p/5+pPGXElc/rV/hfZMy1MFvC0zoui6/kShy1pEKLlFDWqG4YCMBtFvn9dx12xnDhgE3gfAIBBqF7jda/rj8ReBIDkgk78InImunLu7+Xc+zQAHL7QtrmCIgLA6xdpakXOA0II+exldWkxdlZv1VQ+NlnqrmqZv2xJWcmtp8PR3zUEvFusAu+zCrx/JhkP8x2rt1JZyhsz4XEAeP6SnsA8JhORSa/T5YtnUzEECOUVuWAxmSyRZDQcy6ZiJsFk1nRdB5j9nD60zbzm4ZPaw9/cqXyTUErvnLevUIaGxtNkfHMNs7nZx/TnFJpbU8F0NnqYxgePqA9mFZr7+HL+xb9azf/V6nJmdZkdj1p5ZP27dfwNo2kyOpEhEw4BOUJZGrqikrni0JRxqNyB05urmc5qF64GABBsgj0bz83k4mLU4rS4AQA4zHJZRUyHMuExjWiXZYlAkQ8ORWFYpMh5KBRUMRRKDBcKqqhphjowEO4yDGJEo9mp7u7QkWDQXVkoqHmzmbcihFBf3/SpbFZKiaKc0zRD6+mZPPqnzuFM7dERhNBRSin1lLkrwcZSi8nsGBXF3ZyJFSYdhCZFLZRTleiUWT9ocESPEWWovDbYatdk/7hW2JNIZycQmm2/cbkYm0wk00cA4Mg5i/e8G/tWE+qEucbcAQCgJbUQUUheKBMWIw4JCCEkjUpHsYDfeCKcTCaHKaWGpmn5ZDI5mE6nx4LB4IpIJHLC4XBUIoRwJBI5IYpimBCiY4w5SZKSLMua8vl8lBBiTE1NHWZZ1pxKpYaIxSboobF+0DSVKrIECGF4i6iXElbm1/SMn3mBIRk5APj1XC0eAABC6CDA5XMdz8dzB19/ZioWCSmaesEbnPJFwVZKCf3FV5/49Me+e+fDAACFrJxWCopYXxdsnT40dVRgWWEiPNMHl6nr77lghBirwAc4xmE285wb4PcPdwxC37fRwncKxgwWOLNd11VZM37fb9PEW2w2k9OflVIzqiYvSJ/VDx4UNEOVPfaSGpj3wMJrC1RrunKWwJc0Le00m4Iui7nCa7PU52Q1StKpmD4x2otYjifZ9IIaJSUy6QQAvHrucgYzuKa0vDaVy5z1QMbEINNiP1r8mVXcZx5YyRuiBuLJiHHydIycZjCMPXRCe+jj/z977x0m13Gee35VdWLnMN2TEyYBAwxyIjJBghQDBDGLskUrS76PHGRfa63VXl9pr9ZXtiytvbJsWrIVqWiKYBZIAEwAQWQMB2FyTj2d0+nTJ1TV/jEDCgQBZhIDun9AP5g5fULV6UZ3vfV99X7LxU/+z63y/8yZPFftRtX7hux9ewbsPQAAD/daDy+rwMu+ukX+6niGjXsU5HFL4P7no9Y/503I502e//VZ69d/ulb6069vk78+nefTPhl8/UnWDwBgG3YxNZUei48nhz0hdwUAAOWUBlRv2ZLylpW7z+79OVzg6FyixPtNSRiWKHEJzAI1rBnJXOJfuCKmJSP6qKBVOsPVCCFsRSTjUO/gvlpvZWMsFZ2emjo3yhjnFw3Wn3qz1xJEUSyvb2yJT02McMY5EQXRH66oSsdmpj3BsrAgyWMAYNQvrVmjuhWP4pTdsfHEUGI8NVzZXN4eqi9r7nz6zMOGZuQFWVQAABxu1deyZsGWoZOjhwDmt5B4K3hdruDaxQtvSGZzka7+wRc552zN4oU7nIriefHlM49RSnPGhHHGTtvTiCDBztgzglsoAwa0OF7sshP2GDNYHtDvP/ump6dPzf34/AWXOj+T/qp1i3O8It4A4JXI30UueK+pK/l2OS8KAa6O1/GJw8899kb7iJIgFzUzX8gWM+e3CRKRJUV0IEBIEgS5YBj5+dpZPidWL0wn45wz3bJSZyYiu4+PTvz0gp054/yqELfvJgF3eV3AVV4vCbJjNNZ7LOAq91X46xeNxXqPEyIkWyuXb2Oc2hW+Os3rDFZKmanBSn89nU6Ndl/ptn/QYZzTwcjpAzuWf/T/WNu64zGtmE24FG/ZliW3fbh34uSr1kQPzMT31wV96wqGlTJsmp9KZzvBJ0pgmgbIikNc0NoBAEeuUFcuC2WUAcCrsjW2NQhbP7tSZKYF5uIQWYwQoHInKr9toXDbF58sfjFuC+nTODDxw2m8u81lLaiodDc8PKbtPxqDE67FzhZ/lY+kdD5R5RH+/qZ29dagTAOFJBRenmFdXRHSWVVTtSCZTM6cTIun7+9WfliraFX5gp7pmmFdDtFyFCwoeMs9ldVtFR2YYCJIgoIxRjKRsEhEyaO4fQuCdW0AcOrSvSpR4r2nJAxLlLgEftUbrPaE61RRUReHW5bnTC2rmYVcNJ+MVHvL6yxqm43+mpaJbGS0aBnFd3ItQZSk1jXrNoejDU2MUhqbGB8uq65pKG9obMGEkEw8NgMABhGw4ClzlWOChWCVvx4Y8Hy6kAhU+eqIQF5VfN1T5gp7yt1VJ/ecfvAd3Yh5hmlZRUWWHbXlamv38MjRomnphmnpy1tbthw/17tPN4wCAEzNPQAAgDiIRnWapTkaZwbT3itxdTWItveL9vqm9tbaxoWyIMkXbn/iyHOP5vWClpxMjS7a1LJj5c0dd6puxVvTVtkRqi9rrmguX3RmT+/vCMGC2zGbZjUfKZhmEgBBlc+zTBaEsxw4ZZzrG5rrn1xUGb5lNJF+KVssRgJOR4NmGDGE0NR/tfdHyFvd4pDdfk3PxhdWr7p+aObsSw7J6ZNF1eVWfWG36i8fjfUcJViQHLIn0FpdXndu7OieK93uqxlJFOWw31cjioIciSdHMMbEprbFOXDT+r2zNmOUBdzlPwcAWFK//lYBi5JNLXNo5syhk0PPv6om53Qmd9YhS1MSiTtFQlSCscgy6Ti3bRP7A+Vk2eqtCGN8Uf3VeYdTQo5//7D6+b4E6/vBCfMHBgUDI8Brqsiab+2Qv1XrwbWaKTM15A29NJJ4caytIu3MOcaYJ0ARK/hkyqg74AwBwESwflHDi2nrpMPhcGGMSd2yuoUr61JBURTlnp6eExUVFXUDljWx78zI06mU+arlEMtvXPIRSrntLnNXuPyOIOccDGoWX470HB9MjvXlzcIH2qiqxPynJAxLlLgEnHPukpxup6S6NEvPW5RamWI+VeOtqCeYCGXOQEWVJ1xb46moB4CuNzzh68AYY/lMOokwIb5gqFzP5TKeslA5QghZplGEOUfNTDQ37fQ5gqZuaYpL9pTVBhoVl+ymNrX8ld6alnULttimbZgFIx8dTQxO9kfOLtrYfD0APPRu3JP3mu1rVt41NDF1+kMb1t135My5PaosuxyK4qkMBRsdsuzaf+zkr3XD6Fq7eNFYmc9bBTB77yrLgsO6YVy25h0t0AwAZC73fIl3lx2rNt74nS985a8QRjjsC4bj6VQ85AuET/SfPb735KGnAEAbOzt1Snqs85erdy6/1+l3BHf+5Y3fiAxGu4/sPvkATBnpco+/RhCIiGfTcOcdBcNKnhydeOCmjrZvWJTqR4fHfwQAu0+NTf1SEUXvzuWLvkMwlnJFY/pg/8h3NcNK7Ghv+e93r162JeBUG2v8npWf3rz6ziODw9+L5QuDm5ob//TO1YtXBR2OxjtWtX/33nVLuw/0DX47mi302QxsdhWKSgGLUtHUMmktOlETbFpuUVNnnFGRiIpb9YdzemomkZsZ8TmD1Q7ZHQi4wnWmXZzXtSvnMwIhwvK2lk0+l7PM5VC9zTXVSynjVBYFNZHJTgPA3gv3zxYS0wfPPXq/Q/YECCYiZdTSipm4Yc2+Bg5J9HEOzCGLQY8yW6anwuteYjNmZBXvsNC8cAUQQqxzXYfnuygEADApmIwDWxLCS9pDuD1vQj7sROFdbcKurAHZyRybBAUhs2Bq2Vgu4q/wVCsu2ZuN5aadPjVIREFKR3NTAAAejycAADyfz2d0Xdc0TcvJsqwqiuL0er1Bp9PpzWazSUmSFAAAASNCMBDOgTsDziPyuqatroCzrPtA316CANuMUQCYnnuUKHFFKQnDEiUuwUw+MfVU/4GH0ZyrGEYIU0ZtjDEBQMAYo6emzh62GX3FHIMQQkKhULXb7fYZhqGPjY31v5lrmUVdJ4K4G4ADxgQzRtnQ6c6jc08j25xdqzV4YuTQcOfYEeCcA0IIz7lscsoppYxOD0S7gQNns4YdCIBzhPG8HFhfClWWXSG/r1ogRFyxsOVaXTdyjHM2MRPty2hactuq5XfAOxThJd57dq6/9sNPHHnuseN9Z4997pa7//hL//K3X9y5YftHFtU1tcOcwKGWbfYfGXph8MToIVESZIQBm0Vbt03bCLjd5ZghVFsebpmvaohxzgjG/3JkePw/AAAoZQYAgG5aGYHg77w4MPI9BIAYB2ozZjDGqCiQfzjQP/xPc6acUO3G1V/ZLPzVUwN0j5Yf3Jei+OU0AsQ58OksG7+zDe2YzguLD47Rg3BBBPxqoi7Utjrsq20biXUfqfDVLiz31bYxzuhItPvIqqZt96iKy5fKxyYmEgOnhiJnXmwob1+PEN59JcsfXK0wzllOK6TKvJ7KiWhsUCCC6HGoflEQJH5RKjNGGK9tveEPJEFWZ7cghBDCAJyntdgEADwScDoaKWfWyvqaP2SMWRZlRZ9DqZ1IZo5zSvuRrKhIVV08VF5zBbr7lrEotxf48f/4zErpM3++Xv5zhwiOrMGz52Ls3D8d0f9pOs+nWT7H88l8jNqMYoJfJJ3jRxilNsIYIwSIzjnmHjly5CmEEGazgo4PDAy8DDDrzkwptdFcLUXGGBMJErbUk83VblxdsHmhYBoFMtijWf2QEWJ2riOMO+AydW5LlLgSlIRhiRKXgM0OTC5lGEEv8zMoiuLYtGnTzq1bt94mCIIAANve7PWobb2h+yKbLd7+em6Ul3puXrlXvh6xVHpicVPjNd0jo8fWtC/ckSK5qECIGA74a6PJ1Hjf6Pg7Kv9R4v3B5XB6hiITQzk9n7WobTHO2e+OvvDERzZcf7trtvB9KlgTaHSXucLj56Y6tUwhfeHx9ZUVXpeqeuPp9NTFA9r5xFz5idd8Rtj00tstmxYB4MK0856/3iT3PdJjPbwkjJesrBCWUg5UIUipdEHwx53mj5eW46UdYdIBV6EwtKlZ7B4/9tRkcug0pbYlCUQsmMODjHOGEbN7p/b8KpU30iJBAuOcMc4ZwlMzDeWeegAYudLtv9pgs1G7k/AmyiQhhFBtWcsKVXJ6AQAwxoIsOlxhb03rmbHDTwDAIxOpzCkAgPqgf/d0JnvatGkh7HG1yoLgFldvWsny2RQqFjSkKM6L1lfPW4ZSbBgAvnrhtg8DwF+/ejcK8Jrv21d9j9pzLqcXcNnPKVlAQqULV45m2GhrELdSxmmNm9d0x1j3plq86fgUu9iErUSJK0pJGJYo8S6haVoOAL63YcOG4bvuuutPrnR7rjamYvGhD21Ye9/Rs91Pa3oxO5NIjeV1PeNUFU8ik52eisUGQ35f1Y3r1670uV1lk7H4oM/t6lm9aOH6mnCoeWFj/RqPy3kkm9fmVY2//2pMxCLj9eGq+t7x4R7KGN25YftH0vls2u1wes6XPGtYXru2vqNm1cxgrAdeLZbA73aHAQFK5fKx+SwM3w3yJuTXVJM1nAPXLNAoA4okjqIaj66rIevcEriHUnzoSrfz7TCTnuijzDLOF7KvDLoqm6t8zQZat78AACAASURBVBOx3ERDpa+hULQKVUFaxTgw06Km1yl7BQEL1UFXNZSE4XvKXK28P7twmyyqrpULtt4Z9FQuuHB73jCi58scJfKFQQDAeGwY8UwmIa5Ysx2KugYIwXzkfB3B8z8DQsDY+xuNpgxod5x2KwJShlJsSLe4Hi+geExjscVhcfFgyn5NOaMSJa4kJWFYosRbQFVVx5IlS65ZtmzZJpfL5Z2amhrZt2/fr5PJ5KXcKwEAYPPmzTtzuVy6p6fn+JYtW3aJoig/8cQTP1m9evV2AIBkMhlZv379TaFQqKpQKOSPHj26t7u7+1hjY2N7R0fHNY888sgPLMuyAAC2bNnyYc45HDhw4NG6urqWjRs33lpRUVGnaVr26NGjezs7O9+Vkg9XglgqPfHUoaM/m5iJ9e87evxX2byWMG27WBMOtQgCkWxKbcYYHZiYeFkggqQXjTxjnCYymemDnacfzWla8kIHzxJXhqeOHXhSlRU1lk7OHO89ffS+63d9UhAE8Uh310sZLZ8BALBN2zAKpnYpt86Mlo8zzqhDll2A5umI802AMUZvNAjdO2Tv9cjgieR5JDPIM5QBVQRQYgUea/bj5vEMHx9Os+H3q83vJmkt9ipnXlUSVI9D9ohiQfSokse2me1zKT6CEbEps1VJUDMFo7QW+AphWHq+3Fd7aEXT1rsu3N4cLrv29GTkYQBI0tnUSUpC5RGWz6Xg9KkDSFWd/H0WW28GQghZUFO1WFXkQc44czkcSsjvrXY51HHGOC0Ui/l34zqrVq3alsvl0n19fZdMB7UZp3AJl9EGH65/vM96nJa+s0rMM0rCsESJtwDGmKiq6ozFYpNDQ0Nnr7nmmpscDocLAP72csdUV1cvAAA0NjbWt2XLlo/EYrEJt9vtvffee7dNTk4O5vP5NGPMPnPmzOGGhoZFt9122+eTyeQMpdTeuHHjrZ2dnQcA4DQhhHzjG9+47+mnn/5FIBAI33777R+TJEl++eWXD7pcLq8gCOLl2nA1YMy65j099+sxAACxrX1VGmPd7ut9iVuvFP692LnwhferjSXemON9Z44CAJi2Ze1+ce9vz44OnBGJIA5Njw/mdC0LADB2evJEsCbQsGhj8/XuoOuAoRl5mAsnrmxtrQ+43WGMED43PHoUrqJ0aEIIuW7Nyns2L1+666M3bD+mSNL3qsOhZqcqe3L2VMbnQ6GCznNuF/ZnczRpuYlhe4lLkKgvmmcpSUaKISFaXkUW6QISvSoOV8dt1enEZzWN5a50/y4EEUSIg/g45RYgINziRSxhlZlMAwYMiUhlJitwOhttGotmx7SipRmWbcTThTghmJg2NWWRyHzW1pWKBIvnRuLnrnTfPugghFFj+aL1BAvS+W0CFqWN7bd+KKeno6/em4NTkoJwQb1DUlvfCmMjPSybTkjNrcsRIS9zSueVwCEYkyULGtfXVZS3eZ3O4ODk1GlFEtUFVZVLEtlcRBDIQdumb7iE4/WQJEn+1Kc+df3IyEgPvMV1gtN5Pm3YvFTMvsS8oyQMS5R4C+i6rp06deoFRVFUzjmoqurcsWPHR1/vmKmpqeHW1tYVTqfTI4qiFI1GJ+vr6xe53W5fLBabHB4ePpdIJCKccxaNRieWLVu22efzBUdHR3tHRka6V61atR1jfKatrW25KIrSqVOnnldV1en3+8unp6eHz5w5c1jTtKz1e+E070GCKCJFcQBCCBijXC9oIMkKIoQAQoibpgGMUuwLTHHbsoBSGyGMkKI4QBQlsG2bG8UCSJKMBFECSikv6trV4I73QcewTNOpqA6fy+2jjLGjPV2HBUwEgjGBuWVIoYay5lU3L71LcSkey7CKfHZWHQAAuh/vfuzI7pM/03Q9a9n2vHlPo1kIzC7nuuT77NrVK+668Zq1H89pWnJBddUSjDFxKLLrjus2fbE79btT6UwxqijY2ddnnAiVCTUuF/b19RnHly5Vtw4MGKecLuz1ekiIcaDU5rZpcl0Skez1kjIAmFfCEBBgMSDWikGxDhgwzjlDBAnAgCIZOaSw1Jw9nv0tQqifc84LhqUDwMSVbvbVjMPj9a25aeddvcdeep5atlXV1LxooPPkS8uvvX7nRF/P6eblq67Z//Mf/0tN26Il9YuWrMjEo5Gqptb2fT/74XcZ+71wwwiT65be9ZcO2f1KSRib2VYyFxnZ//KvvwXwd69cUxYE941LWr5+z5plOgfgZ6cijwjrr2uUVqzdzvO5tD3U3zXfRCHArLN4IpON2IxaCABZlm1UBAL1Ib+v2qLMhAvWRCqKou7ateuz69atu9G2bfvo0aNP7d+//z+z2Wxy2bJlm3bu3PnpYDBYoet6fu/evb8+cODAI1VVVY1f//rX/1d7e/tawzAK999//2ej0ejk1772tT8QRZAbGqTFmsYyts0tTWNZWUYqpdx2OLBbVbGbYShNgJSYl5SEYYkSbwGv1xvYtm3bHYsXL14vy7ISCATChBARY4wvN1icmpoaXrFixdb6+vq2iYmJgXQ6Hevo6Fiv67qWzWaTixcvXnfDDTfc63A4PIqiOBobG9sJIaKmadnOzs4XNm/evCsYDFasW7fuxhMnTjyraVo2l8uln3/++Yduu+22LyxevHjtSy+99NTBgwcfg0sXZJ93CM1ty9Ubbv04y2aSwBkrPrPnV9KajTdit9vP9ULOOnf6CI1Hpxx3/MGf0dGhc/q+J3+OFMWh3nL7Z5GiOllkctg4eewZZeO2Xcjt9nPDKBgvPvsIAPRc6b79V6fM6y/7izs/+cVb1m378EQ8Mv7pf/jqx1tqGtpWNrev+tVzT/wCALKjXRPHf/JXv/kkQvCaVNFcQovmU1ryEqe+ohBCpEAg0MQYo4IgDNuXEK1bViz7yIP7n/suAgTb16y8GwAglkpPhv2B2r1d9q8Lhp2zLTDTaRoVBCSaJteDQaFqcsoacDixx+nAXgDgggBSOsWiuRxNUgq2Zc2/yALCSMAKdtMCTRMH8dtJe0qukNu4zQ0AQPqgfsTO2DNXgynJ1YJlFAtDXSePrL1p593jPd1ditPpKq9raPaVV1RFRob6GJ/9DkIYE9nldDstrzlw6tihC0UhAABltk0wuWf2t7n/hRwAgPPz51BE0QPAGcH4Z4cGR/+V8dnIPeeckZdeaGCZdAIIIfBaI5Z5gWXbFkboMQ6zkzrAOe8dGz/RVl+3un9s/KQ9ay4DAAA7duz4aHNz87Jvf/vbf+Jyubz33XffX1uWZe7Zs+eBaDQ6sXv37vsnJycHV61atf3222//QldX14vDw8PdZWVlX/zCF77w/3R1db345JNP/oxzzhhjnAjIFkWQ16xVb3K7SWBoyOiqr5MWjY9bvcUi0xgHNjGBeq/g7SlR4rKUhGGJEm+BRYsWrdm8efOHf/GLX3y7s7PzhU2bNu285557/uz1jpmZmRmTZVltbm5e2tvbe1LTtNyqVauuHR8f79c0Lfu5z33ufx05cuTpRx999AdlZWVVX/7yl/8VAIBSSuvr63tWr1593cqVK7fV1ta27N69+37btu25wdY+r9d7cvny5VtuvvnmPwqHwzUA8LX34Ta8cxAglsukCg8+8P9KK9deJ7YvXY9dLp/Vd+6EefTQ787PQCubtj8PoigBAiDhyjokK2r+P/75/wIAEFsWrSCVVY3G0UN7SE1di7CgdSmUhOEV5/bNN9zZVFXX/OALe361ccmqLQCACkVdu2H1xpuePPr84wCQ1XN6FgCuqkLOnHMmSZKrvLx8qWVZGlwi+iWJgmJYli4Lonx+m9flChomLZw8qe/TDUO7QChd0lRm6VJ1s2lxAwFAX58xb514mcl0ADhw0eZLttejuLxBhy80mZkZM+nVk9kw35BU1ekNhsojI8N9M2PD/b6ycGVlU/Oi5PTUeC4Rj0aGBnsBAGaGh3rdPn/Q6fMFLMMoXupcLsUXaixvv0aRHB6Efl/WKF/MxAHgsbqgb61NadGjKlW9kdhTumm9sv5TWrikjNu2LXas3MRzmRS8Nr1/XnBB/c8LJyeOXbgPxhh985vf3BWPx6eWL1++GQBAFEW5ra1t5fPPP787k8kkQqFQdUdHxwa32+0PBoMVkiRJAACUUpvPqUHbtl9xIhYFJDEGbHTEPKsVeNbjxsHuHuOoptG0afCiKCKZcyhNmJSYl5SEYYkSbwFZllVKqZ1MJiMul8u7Zs2a69EbGGTkcrnMl770JaO2trb14MGDj3m93qDf7w/39PSc0DQt43K5vJFIZAQhhBYvXryuoqKi/vyx0Wh0Ynp6emTLli0fmZqaGp6ZmRnjnHOn0+ny+/3lAJA6evTo09XV1Qs6Ojquea/7/67COQPOOXDOAGHMKbV5Npu8bFoSxgJc8OULBBOkOlzI5fLSyNQIHR0qpebMA1a1LFnz2EvPPjIcmRiaE4aQzmfTTkV1Cphc7d85GGMsztZ8ey1dfYMHt69eeVc0mRr3uZxlq9vbrt/QseSW/rHxUzal1puJnp09WzzkcGC3prEPjBGLgAXRKaoujK6euqrzES2dTmGM93MAzhnjCKEeQAguMH8ZAgAwi3oRAJ65nAESxgRfv+zuP6oPL1zrkFw+jAUxW0hM1Za1rDrWv//nAPAYAsBNoeDWKr9nuU9Vaq9pqs8BAIyn0sfJqk2tQlv7al7UNRwqr0GE4PmYTvpmwBgTj8cTMAxDr66ubgIAGBgY6BocHDzNOee33nrrJ6urq5ump6eHnU6nV5Ik5XL//89TLLICABx+XzpQosS7zNX+JV2ixPvK6OhoTyqVin7mM5/5WiaTSeRyuTSlswvYFy5cuHLLli0fueuuu9Y0NTUt+fKXv/wvvb29J5977rmHbrnllokFCxZ0xGKxCUEQREKIaFmWmcvl0gcOHHh0586dn9m4ceOtyWQymsvlUjA3w6nrurZ27drTW7Zs+cjhw4f3ZDKZJACA2+3279ix46N1dXVtnHOGEMJPPfXUz7/yla9cwbvz1iDllfXqzbd9GomibBw5+IQU2HwLzPUbEYJJbcNCx613bANRlHixWLCH+08DwYLjzj/8cxafmeC57Cl7dOgc9gZCHGdT9ILUoBJXDsu2LFWW1QsnTGrDlXWWbZs2e2dmD1cSxhi1LCtPX0fg7Tt64pe7tm78/PqO9puqQ6Gmj31ox1+NTUd6nj58/OeXSj29FJRyCgDpN9xxHuKUVOfq6o6NDf7qJlVUnCcnz77UHRs8vaN5440IITSSnhwAAFhasXDVmpqOjRkjlz42cfqgR3b5OspbVxKMyZ6+A7tjWjL6Rtf6r8qFSxbeaKLhcq64GCHSXLls656TP/tGTbBluUhE5dTwCw8uqllzg0vxlAEATKWznZQx0+tQajJFY9Kw7CwAgGHZOav33HFSVdNkD/R24lC4GtjVmy7MGKPZbDZx/Pjx/Y8//vgPz2/nnHO32+2/9tpr7/jpT3/6zSNHjjzV2Ni4eNOmTbdesA/MldRBV0stxxIl3oiSMCxR4i0wPj7e/6Mf/egbgUCg3LZtKxaLTfp8vjLGGPN6vSNHThzeB9VW+cPfe+hHlmmbodbAwvqV1ev1sozr8NiBEwvW1W2VFUntp6fjSTRjtFzTuG2GjBf/44fHv44AoWQyOfPUU089EI1GX0lTKxaL+fNGNJxytqN5461rgu2ObE90sn88m672lNfpBT3f/3LPyeubN9ySKKRjAiZCb2z4TNbIz9t0PTo1OWQef2kv0ws5lohPs1w2zfO52QExY5zFo5OFx3/7fUAI8Vw2xTKpmP7EQz9AqtPNTUNnyXikuO/JnyOHy8Nt2+LpZOwKd6kEAOw5duCJz9x85xe6xwbPVgRClZ+66Y7Pblu2bvvLQz2dWS2fwRhjmJ0AQJxzjhFGgACdr1nIGONvptTD+82c+YxQKBRilmUVLrVPNJWa9Lqc//jcyc6HFFFSbUatRDoznczmIux1Bo0IYcT5/Orv20EiotwSbFhkUNN4ebr72Mb6Vdf1xUfOxQrJmaUVbaskIsoOUWV/vvETdz3Ws//X2aKW4cD5iqpFa8fS08MCJsLOhdvvAYDvXum+fLBBgDEmscxkX8BVXudSfeG8no6eGz/61L1bvnQ/AHwzVzSiABCt8nmSsZzWZ1H6SkoqKQtXkZq6FpZMRMCyzKv5vcsY4zfffPPu7du339nT03MyHo9PVldXN8VisUlN07IIISxJkuzz+cq2bNmyKxwO154/tlgsatlsNtXY2NgeDodr/H5/MZVKlb6HSlzVlIRhiRJvgbl6gqNzj/NEAQAymUxSkISXEoV4un1L6w3pwUSvXMQ5ySW5u4/3HgjVlzU5/GogNZUe7zp25rmCVSzKDtmNVezo7e09yeirzWs8Hk+grq6uZcuWLTsnJib6h4ZmUyUlQZSBAdgZo2ikC5oR0bIWtUzLskzdKhZqvRUNiUImrlm69r7dmLeKaRZpLDJOI1MjIMmytHr9Dq7lM/Zgbxp7fAGQZB0YpQgAaDIR4YZeQIrDxYtFDTAhQl3DQpsxxrKZBE/EI1e6OyV+z9MnDu4pmsXifTfc9imMEf7Qms037ztx6KlfPvvEzy3G6HUbb7o7X8imVdnhGh4fONvcuHBZOFhRe66/6wgAgNvpefmaVVs2+Tz+w+lsKnGl+3MejDFxOp3hioqKFdFo9AwAvKZtZX5flaYXo/1jE69Y14uiIDXXVC8jhLxMKX1NVDsYrqpesmpDq8PpPmEaxYKsOpwYYWwUCwVBkmRCRNEyjaJR1AtXwwA8b2rZiWxktCc2dPr65o07OXAe11Izhm0VAQACDm+IMkoHEmM9ulXUW8oaFi4KNS1rDTa0Z4r59GByrGTK8R7DOWPTyZGzIW91ayI3PdxStXzb5vYP/zdVdvlyeupVBmaRTO7MxfVGxfal63k+l8ZloSqkOtwI49NXsyP0/v37fyNJkvLZz372a7Isq9FobHL3Q7vvTyQSkQcffPB7O3fu/MxNN9103+HDh/ecOnXq+fMO4MViUV+4cOGvd+3a9dmvfvWr/zE4OHgaAP7yCnenRIl3REkYlijxLqK4ZE9Zjb9BSxeSsbHEoJbRkxVN4UW5WC7SfbBvb/uWthtVh+xjWYMjzSKYMWQnCnbI5WwNe1yMc+AWpbpAsLR2+YrNH73n7r88ffbsgb17996vabM14ARBeMTr9QaLxaKu63ruwsiKW3YWWssa26ey0QnKXjsInS9Yg31diJAz4sIl64SFi9cg1eGmQwNdynU3fQwoo1bPmaNg25a4eNkG3nn8ObJw8VpSUdUAAMCSiQgIRBDbFq9mmXQcEfIiv8SAu8SVwZydPNk79wAAgBvh0/AtAHA6XO7aqvpWyzYNUZTkTC6dGJsY7gUAVFtV3xqNR8Y3r9u+SyvkszadX26HGGMhn89PJ5PJAcYunbb8Bx+6/svPn+j8LULoIOecu50O37WrVtx83ZqV93zt+z/+GAC8arKGEIEsXbtlydrNN+4CAMhn06nGlsXLMMY4m0kl3F5/UBRlOZ9NJU++9MzvYL6VrLgEjDNGGaMcgCNASBUVR72vuinsClTWeisbxzORYYQALQ63rGgta0xb1DJ6YoNdU7nY+GRmZiyaT0xf6T7MVzBGRCRENW16oYkRuBU5bDNmnDeIkQXByQG4aduXjGxTRmmZp+q7upnPGJae9zmP16xs2na3Rc3iM10Pfgfg71/Z92JRCADAMumEEAhWEFlxsmwmfjWLQgAA26L2S8+deO7gM0f2ZgvJiEPxBILuigYMgnT44PFnTh05e5RS28zqycj+PS88KouK06l4Apxz5lb92r9+9wdfyxczcTaPv3NLlHizlIRhiRLvAIKxQBl7ZQCbT2oJTPDjwH+/FgRhdIj/fg3Gv3kdamV90LfRKUlBBIByk4Xkkqry2yjnFnDOTJsWMEZCenwk+8A//+NPojmtNxaNTb1yTUKEqqqqRq/XW5bNZpNOp7NL07Q8AIBJreK56MDLulXU39cb8XYgRCR1jQvNl154XGhqWwaECEh1uvU9j/6YZVJx7HJ5SCYVR5KkIKfLa/X3nATLNKW1Gz/E9UKOVFQ3Fn7zs2+XROH8orm6vrloGMWpRHSSXRThYoyxvJZNYSIIjDFWW93Q6nF5/Ll8Ll3Q89lsPpta2bHu2ucOPf3bgq7NKxEkiqJTkiR3sVhMU0ovuV6wb3T85K6tmz5vWlYx5PdN7ty84WNLmhZseOrw8QdM236NOySlNq2qXTA42NN1/OzJQ8+t33bLHWNDvWdT8Zmp2+77k7/uO3vicOeR559ev/Wm2xwujxfmuTA0bKs4kBjrSRezSdM2jeOTp18UMBa8itufLGTi5a6yyvH09PATvc8/uKyibXXWyKc7p7uPHhnvOrC4vGWFv9wT7OKMAkApHQ8Agk5Hg0OWyjjnNJEvDIZcriqPKleOxFOH3IrsDbmdrTZjRltFaCFlzGoo848mNX2kMRRYpgiCO+B0HEvr+ji7oE7oeeLZqeELfv0txuThWWHzzTdsF50c6wdGKQiCYPeeO/5u9vlKoMpO39rWHfeNzHQf1k0t41EDFeXe2rZ4ZmqwrXrV9RhjggChWGZyIOgubzBtoxDLTg0E3eUNYW9Nq24WMi+PHHwY5rKHSpS4mikJwxIl3iaSIDgWVoa2B93Oc4mc9or1/MUpofyihfmaYSan07kuRRTcAMAN29YEjCXKuAUIkEiISik1TEoLAiaKZprxC8UnAIA9VzvK7Xb7vV5vGca4wBhjhm1ePVbwjFGeyyTFjpWbkcvlZZGpEa4X8sAZRYIgCA1NjUJjyxLABCNZVnlR1wBmDU14QcvRybEBoallGSLCDKf2B0YcBgShrEoQ6ycsc7hVVjoAOBQ51wWExDIiVOQYTQMAtzi3jhcKF5cLuOJ8bPutH+8c7DkVORKbBoBXvS6WZRrHuw7vJ0QQEUKIMcY8bm/Ati0zm8skOQCfiowPZXKZxOXqgl4pBEFQZFn2FAqFmH2ZSMz+Yyd/JYqCfO+N1//3vK5nZFFUHnn+wP2nevufu1QaKQCAZZmmrKjOlvYVa/PZVKK6rqktXFnbEI9MjHp8wVDHqg3XMsaoZRrGe9vDd07B0gsAcOSCTQ/N/fuji3ZNwGtLW5x5r9p1tVIX9K/nwDlGCIuEqDZjht+hNkySbGeF19NR7nW1U8oMAWPZYrToUZRKn6rUCIQosiC4MEIEXlsqFBDCqKli8cZ4dnoorcWnQt7q5m1Lbrt1+9I70fGBZ36ZLSRfNz1fWNCy1B7q76Lx6NTr7Xe1YFPLGI/3nwQAThARUvnoWKW/fjHGRCCYiEORMwdD3uqW+nDb6mQuOnZu/Ngeh+zyt1Qu2xbPTg8JRJQQvL47eYkSVwslYViixNuG8zKXs8WizIDL1CS7FDalBgD0vd2rzg2o7Xg8PpXNZpOpVCoKcPXVROK2bWOf/0USrqjjtm3xTDoGsqzOikPOWCoxYxx9cQ/X8hnOqM0z6QQAAuPA/od4Qctym1rY6fTCVbDu6q3gxThQK4lNEdsaX6IoqyWElEnLGskymgoJpKIBS60AHAqM5+G1deSuOA0VNQs6B7tPXqpO11x66KvKimCEEeOMu10e34ola7aOTg73xJIzk+9bg98ktm0XLcvKW5ZVuFzE0LQs06mqv1JEUd20YtmuR54/+G8ne/qftS4ss3IRmVR85uWjzz9tW5aVyyQTgVBFBGNCZqbGhhpblyyPRSZHE9HpSW22XlyJ/0IQjKTxZPaYU5bKnLIUGkumj1Z53cs4cMY5Z05JDA7HkgdcilyeLuhjWd2YbqsI3ZjQCoOZQnEioRWGLuWUiREm69tu+uT+rt98W5Wc+euW3X1fwF1eP+twjREAfOf12oUUxaFs/9BH1RtujdFYdMI8dfTZ9+wmvA/IgupUJafH7wrXZgrJiFNxB2pDrSs1I5tACIFFTZ1zzjJaYtrrDFQubdi4K56dGswX09Gwt6Ylkhrttmxj/mfplCjxJigJwxIl3iYeRamM5bS+2VnZ9w9BEKSamppmQRCk7u7uY7quXzJ6cTXA0qk4AMQv83Rk7nExFzqtXu7YqxYGQA3OdYtz0+DM6DHMrpWqY8OIafRNWdZYWOCVOucFGSHlSrf1UpzqP3e8Olhe45AVhygIxQvloXWJdYPn0031YiHfefbEAdM0dHMeRscKhUJ8YmIiAwCczUXwG6sq2//4zl1/d+F+f/OZP2KEYMHvcoXuuf7aL924fs3H/+/Pf0r72x898Mmiab5m8GgaRQMALjRcyQAA+ILhckmWlZH+c116IZ9/D7v2tvF6vTWVlZUrJycnj+Xz+YggCDLnnFFKLYyxgDEmjDGbc84JIeJcMXALYywAAKKUmiWL/8tDGTMsRnXKmCUSotYH/eur/N4V8XxhQCBYDjgdDUlNHwbOuUWZTjm3KGNmrmhEG8sCGxhwW8D4sM3YqyYmEEIo4C5vyBaSkZqylhUBV7jucO9TPy6aWuamVR//G3gDYWid7nyRToz1c84YLxTmdXrzm8GihfzA1KnnKGd20SxkM5ooJ3NTQ6al58c556ZdzI/Fzh0BAI6RKImCIOeL+WQkNdqtSA63aRsFwy7OX7O3EiXeAiVhWKLE24RyZluUFnNF45JmCaIoSQFfsNzldHs54zydTcZSmWSMc869bl9AlmQVAMDl8vg4YyyVTcWy2XTi/EDZ6/YFAv6yCoww1vR8NplOzJimYVBK7UQiMSOKomQYpVnKDxrjljU8aVmjNue2hND3KQA9qesHOABQzm0EgASEBBVjl5cQf4bSeRVJeqm789BX7v3c/2iqqmvumxjutSh9ZVDqdjh/kytceu3gXHp0EgCAEEQUBTkJQUI2S5PvU9NflzkB86pIYSavxZ89fuo/X7MrAKDzf4GDZdO3XMMxk4xHc+lE/HIpqPMBSqkprixxAwAAIABJREFUCIJCCJGcTme4o6PjXtu2jaGhoX0VFRUrXC5XeSQS6aSUmk1NTTeYppkfGxs7WFVVtRpjLIyMjDwHAKevdD/mK10TkYc453a6oI9PpDInEAD0RqJ7JEFwLq4q3/Vsz+C3Kr3uJTO5fE8kkzvNOfDjIxM/YZzTWC7fyxi3LhaFALPvZd3Ip2rLWla2VC3bmsxHx8bj/SfdijckYFF+o3bRVGIGIRQFQHA1OOW+Ebcv79iVN4x8LJePCcQn+FTVRxAmlHNq2JbhVlR33ijmOef83HTkXEPQW39wIDY+V4Jm3paEKlHi7VAShiVKvE1sygzGuR1wOhbARelxgiCKK5es2XrT9l33uRxuH+ecT0cnR37z2M/+EQCGNq/bvmvDqi23JDOJmYCvrEKRFXVotP/Mbx5/4J8AYNLvDZTdu+sTX6qvaVyIEMYFXcs9+9LeByVJfgpjxDjnLBQKVWcymQQATFyqfVcTWMFOwS2E7KwdZQYrYAkrgIGwItMEl+BnBiswixlYxAqWsGJrdhoAQPAIISQixUpY41e6D2+E4pJdLr+zLBvLRcyi9RojkvPMuQAyAACT8/NC5FXiwC8InhWKusEGbgoI7bU5nzcOntuWrd3OGGf15dWN9eXVjRc+9+SR5x+HN2GgoqrY1d4uXyNJSAGAh9+rtr5TktlsFAB++laPI05/GDjjTM/FLzewnts+b0Xhec5H/AghUiQS6QwEAs3BYLDF6XSW9ff3P5nNZicaGhqu1TQtmslkxsPhcAfnnOVyuclQKLQISsLwstDfizoKv38vWARjayKVOVHpdS9JFfTReK7Qz9gr7qHnj7lsJgnnjJ0ePfTYtR13/FlOT0Vf7H7i+7qRz7RULVs2lRp+U2s9P0iRXsY5Oz46evymJUtumslmZ8o9nvKg0xnc39O7f/2CxvWpQiGVLrC0KorqhqYFG8ZTqfHXq0taosTVTEkYlijxNrEp0w3LzuWKxmvSHb1ub+DDN9z12ZGJwe7fPfPIT0RRku+783Nf2bnjjk8DwFcBABbUt3a8vOdXBx946D/+rqmupeOOWz72xfaWjnUA8NDmddd9pLK8uvHfHvinr+rFQv6GrTv/4ObtH/lE/3B3ZzIdj2QymYSqqq4PSsSQOEnA0eLYVOgrHMASnnG0ODbaGXuGKGTYs9rzYZqncSkodSn1SrVapy5XqpXnzbg5otQojWJIXIAw+o0YEGvFMrHBSlhjVsIa5ey1NutXEpffWdaydsGW7oP9e1WXkq9dXLWikC2mk5Op0UCVr84VcIaiI/E+IhI5MZEaCVT5aouakQ/VBRcAAEQGot16vpgDmB2UKRirGUr1SlGsg7ewxvW95v/b/dPvEEQumV5tWG8uRZRSbgUCQsXIqHn23W3dOwMRggEAvVUn3PVLl9zcNZMfZ6LDzUw9J1e2NmDVHaC5xITgCp4DjAXBV9lEc7ExAEBY9YYAYWynJvuw4vQRV7DaTk8P0lxicr4MyAkhYnl5eW0gEGjWNC3qcrnKA4FAi2EYWQAA0zTzjDGbMWZRSo26urqN3d3dD2UymbHy8vIOjLEYiUQ63+g6H3TcDofP73GHk9ncDEYIq7LkTOXysYpgoH4mkRxzOR0+t6r6soVCyq2qPtO2ixxgfDKV6QSAt3X/GGdMFKRfTadGzhSMfCaZiwwDAKS1+OTh3qcuNgr6wHNmaupMrmjkeiKRHsoYjefz8Y7q6g6XIrvOTE2dWVBWtiCt6+nBWGxw59KOnc/29l7VaypLlHg9SsKwRIm3ScDlWFAf9K+LZHPn4NVrhMDnCYQqQlX1P9/9w29NRsaHAACu33zzE7ff9NH/piiq4/pNN8FkZHzweNfh/ZHo1FhleTVOZZJRn8dfBgCwbsXGG8PB8pq7dn78TwEAwmUVtQub2le5HW5fNp9ONjY2tgeDwYrp6emR97fX7w3c4ga3+axoQICISrwAAGbMHBI8QpjqNM0Zp1jECnGRIKfcBg6MFZkGHDhxk5CjzbGZqMSrNqgrUy+kfghza7XmC5ZhF6lFTYQAL1hVv6GyuXwRtak1flZy2SY1K1rCi7xhT6We1dOKS3EHq/31mGDBV+GpphY1XQFnGQA8CQDAAJiIQCwXhGqNsRzMI2Hodbp9GS2Xnqtn+CpEQRBcqkPO6wUNI4QY51wghNgXCS3OAZwu7PV6Sej9a/nrgwRRFFoWLcdOlxc7XUeZln/TKWS3Xbv1i0MvjjyRz8THiDtYTdxlNdwsakgQVcfCTXfZ2dg4N/WcVNW2HksODyvmU5xRSyqrXcQ5o0iQVCnU0JF7ec/34aJ01itJsVhMT05OHtU0LZrL5aZyudy0ZVmFQqEQEwRB0XU9IYqi0+Px1Jw9e/ZBRVF8qVRqZHx8/EVKqVUoFD5wa4TfKh6XI9DWULeqq2/goCxJatDrqUjlctHG6soleV3PtNbVrByfifbWlYdba8vDLUXT1JPZ7G4AuGzWAQAAxgTLiuqg1LapbVtEEESMCcYYEUqpDRhRDXIRZ5nXT1xivaI6ZyRZyWQLqYKiOp0IIQSc82KxULh4MiLgVv3bltVtOzMSO2NY1BidyYy9pzfpPaZrYvJ81Prg+W0LKyqSE6nUBADAVDoz1ROJ9Ibc7rK93T17UwV9XqXvlyjxblIShiVKvE3yhhGljNmqKPoufs6hOtwIAdKLv19PlddyGVGUZIwwBgDI5TNJ05yN+HE2+wdjTAAAvG5fcGRiqLvr3IkXzx+/74UnfjWTiIwLgiDquq4NDg6eQeiDZ5HNbW7ZGTtCHMTPDFYIbAkkjCmj20pZU0qNErAS1qgZNQc441ypnvVfIQ7iE5xCwIybo8CBccov6wJ5pUEIoVBdsCkTy0XmRKBXUiVHLp6POn3OQGIiNbxsx+Jdw6dGD/srfTWGZuTSkexkPqUlXjkHALI50HHbHB4wjXOvd733m8/fcs8fP3r42Ueaq+u1qUR0qlCcNUdqr29u/8FffOP/9Drdvk/ceNtvHYr6n2sXLl18w+pNZbIk7TNM85VoIiFIkCSkXqr+2pWDc+x0eUhdQ5s9OngWALKqLDsqy4KNyWx2Jp3LxxurKttVRXZdeBRGCH/p4x+rRpgIVny8WwzWtBHVEyrGRk6bsdEzatPaW5ihpQujnXudi7bcSxzeMmO67whwxpSO6z9hpaYHaWZmBDincEGxcQEhwYmJ2+TMoJzbAUEIOTB2JW07lmcs+3bTixHGyOF2e4kgisGqmvqpgd6zCBOCMMaM2jYRBFHP5bKccwsABuces8ciNHCxiJAkyWGapuZwOIKmaeY1TZvJ5XLTCCE0X6KfVxIEs2VbqsOhJoEQMeTzVUfiydHKYLA+ns5MORXFXREI1BNCBIww0fRihr2JbIjmRctWt3Ws2RCZHBk0irpWUV3fzBizJ0b6eyRJUizbMjHC2BcIlyuqw3Wu8/ALZRXVdf5AuKKoFzSEEZ4cHeiZHB3svfjc6xZWrivzqmWNFb5GVRZUgtEEnWcZGm8HjBByiOAgGAjjZBI4A4mANJWKTbkk5FRFAgVdL9R7WT0ADFzp9pYo8V5QEoYlSrxNCoaVsCgrCAS/ZrF+KpOMWrZt1lY1tIqiNIQQQnfe/LHFsXhkvFjUNYDZYt+XGxiNT432M0bpwWPPPWrbloUxxghhXDT0gtfrDVZVVTXatm3lcrmrfuYSy1h1tjiXqY3qas64jSSkONucWzlwJgbEl+WwPOhqd12v1CgSLdA0lrHL1eG6SQ7LPc7Fzo1iQKw1Z8w+K2VNShVSizFpnAUO82qQIjtkZ8u6xhV1S6pXFTUjP3pm8kT75tYbZoaivXqumCmr8TcwxmkukY/lkvkYtWyzkNFT0wPR7jU7l3+UUU4TE8mR8+fjAMxFsEekSMpROq8io+0NzR23bbrhLsMyjbMj/acX1i34u56xoZ7P3HzX50dnpkaHIxPPXLfimhtimVR8JhmflkVJwRfVADNNVsxmaUISkUIIIpTOA4GICUGqww0IY5irTVpZFmj44zt3/d3Th4//HAB++YU7Pvy/K4LBBjZbpB0AZgf+Qb+vBsuqx7Fo891MS0et5GQvs4oaMNuyMzMjTM8lXctu/BzV0jM8F5/ktqEDIGTFx85xyygIgeoWa2aoEy74vFikKCsEAMFLSIBysMsEodKBkcuFsefpXO5BhFD/2xFeVU2t7W6/PzjR33u2rKa2oXJB00JD1zVqWaYrEAh5g2Xlz/7qgfvhUhF5hADJigKMUj4XMTZNsyCK4u9EWXZRSk3LMF5Jh37Lr8EHkEgyOZrK5aJzkyCIEEwM09KffPGlHxuWXRyenD6DEEKUUvusQA5bNrUs235V1JgIgrD5jo9+cvn1N+4SJVk58vjuXyCMzyWiU+MFLZcNlVfXFvLZNGWUZpKxGVGWlfqm9o7J0f5eXzBckUknoqIkyaHymvpisZCnjNp+f7hSEEXx/BpYX7i88oZPfObPJ/p6T5cJJCMJRKoLe+oAAXwQRCEAQK0X1d7dLt4dyfOISbnpVwW/ScH0KchHOdCuGdpl2mASjAlGaLC0zrDEB5GSMCxR4m1iUWoghH5xqeeS6cTMoePPPf6hrTs/HvSHKiVRktcsv2bHLx/5ybcZZ/zW629/3XM/8vRv/u0Lf/jn//uenfd9aTIyNuBQnW7Lssznj+zbnUqlYhjj7wPMisv3oGvvK8xgOhbwc/qY3gkMKGecpg6lfgYcGLe5YcbMIX1EP8kMluecc0TQvwMA55RThNG/A8xGXBFCnYCBAAcuA5ICglCWpbPF4DkAYECYAWcEIcI4pxyAU/7uD2icEnJgBFi3QMcIMOVAAUDvO9C3v+9g334EgCzKzbHT48cBEDDGWO+hgWfA7QsAxoQjjPY+eO4BTqkNWCb7fnzou8TnD9makcNOp5tTZocRdk5b1qgDYee73f53CgIEP9zz4PcffemZh+/ZevPHPrR6880A0NNUWdv897/+9789PdJ3OuQNhDsa25YallkM+QJhgl+9JpFzYDMz9qjbhf3hsFAHAMNXpjcXwChlyUQERFHicw6jI9MzPX9z/w/vtuecV2VJcnzlX75/WzKTe2XdMUYI/8Nf/Ok+qqVntP4TT1AtFZmN/HHgnHOE8beAA4ch9DsAzmc3zw7GEcI9syfBBBijFxrVFBjLJymNtWG8NCwIVRqjWZ2h/CAzzyUonXkzwquqqqrBNE0jmUxGGJs9t20aRYQxcfv9QSIIoqEXND2XyyBCSCGbSUdHRwYMvXBJa37kdPuklWu302hkHBFyAvuD5UAI4URIQE1DA3Y4PTDY24kwSXFGr/rPrncD26Y2QijvlD0BDpwzyhnBSCHI5VREKhtWUcMIE8qoWSzol0xfrmlrX7r57o995uF//Pu/iY2PDRmFQj4dm4kgjBHnnM/NuqDZzxvKahpa2vpOHz8cnZkYHRk49zKfrSPCMcbdAABef6i8adGyVdS2bSIIhNo2JYIguANlIdU15X3p2ck9FmWWSxFdz3SOPvOD9+92vacULCh0RminScF0SsgZcqJQ1uDZM1F6hjKg/z979x0e13EeCv+dOW17A3bReyFRSLAXkRRFihQlkpIlWZYo23Isx5bsOM7jOPmS3JSbOPmSL77JvXG7luU4tmJbtlWpLkpUYxN7QSV6B3axvZ8+8/0BUoYoqlmkSJHzex4+XJw9Z3bOARY4787M+yIANJMjM6VOXHqp+8owFwsLDBnmIzj35quo1t+QjmRCsqxk3E7PT9as2HDzkiXLNtncFu/TLz76nyc6D78u2URbW/PS6fae4/vy8uxUU1nN5zr7Th6MZsIzgsRLhmae/Mmvv/+361Zcf+uilmXrs/ls6lT3sb26Pjvd7koICOfCVuy21liXYxFbz0wDRYJPqNDC2qBQIFRm2jMvmIo5O9Jg/m6K3NwEM2e+F0aJIJS7MOebZ5EWnlaUkxxCvIiQlDVJikOzIywAQKOGEboYn/o2+nBjmQuXTWXIFAJAeR3yJgFT4EAwCBgWHiyno/S0bpC5U/10y3U3XIudTg9wvAAIYZrNJJHLU6Du3f04v3rDLTgcmgBKiBGcGpbHRyaDuj6hUHrZJR+SVUU+1td1ZDoWnh4LT40ub1ywwudy+371P/5dNCkhiqbKGTmbLvIUFnMIcwghROHt3wNBRJIkImsyaUZM8zLJuIoxR5V8zpwaH6RzRv1hTvbHNzu6nktlshFFVd+WEfIvv/j5LiU0dJyo2SQl56yn/N17+R2jonMCwXdcgyFVPX3m4Wu/z+nwPM/ff//9Xw2FQmPPPffczwBABQCYGRsZQhgP09lAsf/DtEkNXaNyPgsIIa60ok6Y17wMKKUkmYjgQn8pV1RSRbKZhCEPZuAyWit5qWHEcVWB+StEXrLl1WxC4ERLhb9hSSoXmx6e6XrTZfUVp/LxIJyzlv2sQGVlbS6ZiA8cP7pfyefmBu30fI8nRwf6zrf97IcDMFs/9vl3vtLs0wtrAwtPDs6cDMWzMx/yVC9rkRyJAsArZ79u8XPNfTHSZ5yZ0n52XTTMmT7NMFcaFhgyzEeEOYwLyrxVhZW+2qVbF6469XL3U40rawsKGtz660de3Hl0YO++hhW16zp7T+8R7YKjdnHVqvGJoZEjxw6+VNFSuqhpTcMC3okG9nXufrG0sbjF7rUX5JNywizI23d373wmOZOeBqA0HcnMFNcXNbn8zlBla9nSTDQ7ExoK96p57RNb4P4szs75sITtvIsPgAm6GlYHsYitiEPCW0lpPqBaUWriAHgLQtYaUZpfLPDlHAB3IJd7uUGSFvg4vhAj4Dpk+fCErg/D22+ePjKEABXYUEGJgyvhMfCyATKPgR+Ik4G8Dvn5hXj+YJwMwjk3+8boUDcSRQkIIQAIUUXOIYvFRnO5lN7TcQhMQ0eiZIFUMlYmCDVtFuvKXlXpgMvsJqVrbKDzjvVbdiyobVy0fN7CFYUuT+HffPZr/9Pv9gXmldfMHwlODBW6fQGrxWIrdHv9VtFifcdaWQogikgSRc5y8qT8ewU+FxylBKw2B1dcWm1GwpMA8I6g/NHdr3/vfIfufH3P/83NTHeQy6gmISGEHD16dHc6nU6eqSH5Fko+em06vq6xzZgY7SOxaNCyYctd+kDvSZKIh42B0yfZNNK3I5SYwcRot01y+mQ1m3RYPf6hUNeBrJyMRFLTg6qu5Ah9Zx3Mktq6xpY1629Yse1T1/krqmq3fe1P/nrLvfdFu9/ct3uy73QXAEBt2+IVDUtWrOFFQQwOD/X1Hz20N5tMxAEAbC63e96KVetLahuaiGnoQyePHxpqP3GImLOjucU1dY3zV15znc3l8jStXhtGmMMAAF6nxXvD0pobblhWE4ok5cjJwSszu2x3xHzb+m02dZS5GrDAkGE+IpvL6ilvLl2k5tUcIIDqtorlFofkMnVTs7ms7snTwQ7TMA2gAECBSjbR4fDaCnylnvKGlbUtkbHYkK4aCi9wot1j81kdkqukvqgZc4jjBF5c1FaxIjQU7s2nleS81XUbRtsnjlS3Vaw49VLXk+RyWHt1AZhZM6ZMKO0IIUxNagAGbCSNEDWoinhkMXOzRc6RKErikpWbpGvW3wymaeQe/q9/MaPh6blt9atKp4SQZUhTeygAHdO4fh2oHtT1sTpRak6YZrRHVU6GdX36o04l5bw+v7Tu+k+rB9542oxFggAAwwkyHMnTiIhBlHgkmRRMAYMwlSFTugl6RqUZncA7kuMYo0PvlUTm2NwvFlmt1WliJu0YO97tgEvl8b0vPfIHm2/90sr5bauGgxODT+576bEKf0nliYHu4zcuv3br1pXrt+uGof/61Wd/ORyaGMIIc8qcxDMYI1RRKVRijHhBAPFSnstc2Ony8iVlNSAIIkLwoZI+9Y9NnLxY/fp9nRntfPVCtYcwh/mqmhphfssKMxqZJsGpYaGucSGtqms2pyYGaT6X5qtqmvj5LcuRIJ6iusZGDM84MzI8DgDjPMdxHqfIZZX4lKyrst1isWKs5XNyLsNzPGeTLLasks8SQqiuqko6Fg3nUqm4ripKMjwznUsm4po8m/Cpqrl10bav/smfh8dHBzVZlpffuP0zheXl1TaX6+eGrmtrbr1jR9PqdddPDfR1cbzdueUP7/+z3Q/953cBYJ+vuLT8pq987et2t8c3PdjfU7OgbVllU8uikY5TR9uHZtrrSr11pklMQq+s2SsMc7VjgSHDfESCRbCJFsE2cmr8cFG1v6Gw3FcTnYgPJ8Op6XQ4E0IYIYQRRhiQoRmqnFaSvCRYnAWOgK7o8nT/TLeaV7NWp8WtK7rCCZxQWOmr7d0/sNvitLib1zVsjk0lxniRs1gckis+lRif7AmeKmkoaklHMjMAEL7U1+CjMrJGDABi77+jqRtD/aew2+Ozbr/9PmR3uAHgbYFhxDDeUVfyLC/Hv0qB0gwhyQ8aFIpty67FHm9A3fvKE+eOdCCrzSE0LVihnTzy1g12UqEpeO9SGeedDvZhxE0zIiAkhgxj8qO29VFhjLDVih2EUFOWSW58Znrc53R/32lzONO5TCqRTScFXhAopbRnbKh7XkXN/L7Jkd7TY0PdivbOuoYIAZJlmj19Wjlks2P3pTin86GUAsmk49jp9ALHCx/m2B2bN/7pzjf2PaDq+lslBhBCaMOGDbdv2LDhDofD4Y7H4zO7d+/+7YkTJ17XNE3zer3+TZs23bVs2bKNhBDz0KFDL+3atetXqqoqgiAIn/3sZ/98YGDgVHV1ddPSpUs3hsPhiaeeeuon8Xh85r777vvHRx555HuDg4OdAABer9d/7733/s0bb7yx89SpU/taW1tXfelLX/o7i8Vie+655x56+eWXf63NCc59Pl9g8+bNdy9duvQ6nufFwcHB9ueff/6/x8bG+gRBEK699tpbN2zYcIcoilJPT8+RF1988ZcAdNIMh8aVV3f9muqaSjPphDE9MQgYczSVjFJimiSdjFFFyYNpXB7Tgy9D1cXlNaubFq0xTNM4PTHYHU+n4wtr57Upmqq47A53ZaCk6le7n3kIAGLRqclxzHGTvCRJHn+g5M2dj/1Czedz5Mz6zXV37PjDRCg4+dLPHvw/hqapC9dv3Lr203fd23vozddyqWRi2Y3b7zj49BO/Or77xZ0Yc9yWL93/rWvv/NyXAWBfTduiFYUVlTW7f/6f3x3uOHWkZkHb8pqFi1YAAGRkLVNV5Kpy2y3ul4+NvHwJLxfDMBcYCwwZ5iPKJfPxfFpOLtvWdicvclLP/v6Xa5dUrTZ0Q8tEs+GK5pJFpY1FLelIJpQKp4PViytWymkl1f1G34t2j61g2fa2uyZ7pk9xIidWtJQuEqyCdaxj4ujCTc03qzk11/V67wuFFb7albcu+RwxieEKuIp9ZZ5KJadmEZ4tfXG1OJOwYkpobO6msvyOBBjYarMjp8uHeJ4n+XyWZlLxs8XIkShJ2OeTkCBZgBAXtjviNJ/LgCCK2OUuQKIoUUIIzWaSNJ9LA8KYKy6psmzevhE5nB5jqL+dL6swzdD0GBBCkN3hwh4fn3/yNz8wIzOTH6Qf2OMtBNM0kcVqQ4IgUUXOkUw6Ts9T8++9TOn6aFDXx83zrEv7uPn9XHlrq2XNwIB2As4EvfFMKg4A8bP76MZb53eC5/gOQkzz3aZlmSYlTieXXbLEer0oIask4V5VJZd+LaWq5kksEjQjM5Mkk4oBAGCMscBz7zOqieA737h/67P73/wpzKk9V1NT07Rjx44/ffjhh/8tFouFSkpKahKJRNg0TdNms9nvuOOOL1VXVzc/8cQTP0IIoc985jPf4HleBIAfIYRweXl5/Zo1a7a9+OKLv3jooYf+WRRFKZFIROLx+AwAwHXXXXc7AHQCACxcuHBNRUVF45kkM0QUxeP/8i//8uWvf/3r3ykqKirHc36PuN3ugvvvv//bXq/X/8wzz/xXNptNWq1Wez6fTwMArFix4oZt27Z98cknn3wgn89ntmzZ8rl77rnnLx988MG/zWQySQCYmyDl3Pfo2O97+a8WGCGMMcaGrhljM8GxqkBptVWSbJqhaRzGuH9itDenym8tHyCmSVZ/6tOUUgqEEHI2KLTY7Y5v/OhnTfue+O3PsslEjBJCS2rrT3E8z3uLS8p5QZSsDqdrpKv9mJzJpAEA2jZsOtR8zbpNNqfLdc2td5Qq2WwmMjE+rOSyWX9F5WhkYnwYAGBNS/maw6eDh7Oyll3fVrkeI9TBplkyzJWBBYYM8xFpspYHgOcxxmjO4v0DmMOYmIQAQDu8fSF/15zHQ2f34wVOSARTk6ZuarzAiSd3dT6p5NQMQgiPd02dwBzmBItg1WRNTgaTk5jnBCWrfOAi21c67PYUWLfd/nmhbcl6JEgWMxwcV3Y9/XMAOIwEUbRs3HKntHr9LchqdVBNU5RXX3xY3f/603xl9Xzb7Xd/A7m9AcRxnN7fc1x+5rEHqK5rtrvv/StxweJ1gDHmyyrqwTT11P/6+z+kcj4rLlq+wXrzp+/ja+pbk3/59ZsAoPv9+uH4wn3/E3hRRBbJhj2+AE2n4/lnH/8xwngfJYRiyebEos1l5pNhAKAACCOMOSRINjOffmtE1aTUhMsgKASYHflyuzm/24398AFGQ40PMFpkGFRzublCjgO+pISvAYBLXquR5HNZmH0vv6XQ7Sq5fuWyHW9toABvm2RKZ2/0/V5PGZwz+9Q0TVNRlLzL5SqYnp4eOXjw4IvZbDYFAFBcXOxbt27dp44cOfIyx3E8AEA0Gp1ev379rQDwo7NtjI6Onn7llVceyWQybxuhXr169eP33HPPXzidTreu69p99923tr+//1QsFgsBAGiapnIcF1IU5R3rk5ubm5dVV1fPe+CBB/66s7Pz0LnPb9269QvhcHjCNE1DkiRrJBKZ2rRp010PP/wJX2FzAAAgAElEQVSwGwCSH/ByMu+if3K0H96e9CcBAG9NRcYII0Lffw0oQggIIaYgWSxnt/GCIAIA0hRZPpOsGfGC+NYHG4IkWQxd03RNVQkxTYQxfuvDR4QAnanBm8yqyWKfvVjVLWpW1rIsKGSYKwcLDBnmAiHnJGw4ExS+/3Fn9uNEXqxdXLkyl8jHCyt9tbHJxIihGZquGoppmLrDZ/cDBSJYBBtCCIlWwTZ8cvwQXA6p/C8D4qLlG/im1pX5J3/zfWNspNd6w7Z7rFtv/zL2eIf4usZqy5abvyjv/O0PtVPHXkdWm4PqmkoNXcdu72T+yd983wxNj/G1DQvsd9/7V1pZZYPWfnwvttn/zLbjD/4fIJTkfvHgtykhFP72m2df8im+urbP9cd/+f0P2g/HF+4DoWHe4vT3//WPSTQybb/zC38uLV+9xRg4fQIhnBMC1bVSceNyIz0zSgkxES9YjMT0AO8KVMLvmX3yYqNAYWxM7zaMd2aZtGFsK+KF8oRpRpOmET/f8QAAAkKCl+MKw4YRBJgtcI8AIDxjjI2Oapc8KHw3Jf7CmtvWr/1a5+DwAfouSYwQQlgSBMu528fGxvo2bNjw4Nq1a7cvWbLkuvHx8f6amponxsbG+kpLS8Xy8vK6XC63qqSkpPrsMd3d3YcwxojnZ/90B4PBUV3X33HdOzs7D+q6ri9atOjaUCg0VlRUVLFr166Hc7nZLMjvxeVy+WRZzp0NUs9VVFRUWVhYWCqK4lvn1NPTc0TTNOV8+zMX1gcJCgEA5Gw2e/uf/sWRectWXtt35OAej78ot/TGrdfmM5lkcmZmWsllM6lYZKZ59drrvUXFMxwvCFvv+/r1Ix2njuqqqi5cf/1o23WO7ZXNLYtdBYWZ+sXLFhZVVdf3HTm4p3ss0t1S7W9xWEXHsf7gsffvDcMwnxQsMGSYy4SaU3PzVtWFTZMYumrIwYGZnrplNWvzaTkpiLwk2gT74NHR/a3XzbtJVw1FV/S8JmvnrSd2tUGiJNk/+6UWEpwaMfpPnyC5bFpobH5VaG5bxZdXNfLzmpeR8MyEdvLoqySbScPckQ1D17jSinpx+TU3YqfLxxWVVCGLxXYx+gEAoHWe3G+ODnUTWc7btn/6NN8wfxGyWO1UlnNUV/PUUPNCQWXzbIMIk2x8GlscPoQwoh/wpvDjREww3B4ukM+TdwQdFoxtjZK0cFTT+iwY5xdbrWswAO5WlOMBni+tEMW6YU3rrZEkYZPDcdsCq3XXqKb1IRtgngdRUellnXGXEELaB4b2ffc3j/0Jpe8eGH7nG/c/c764cc+ePU92dHTsb2pqWr5ly5bP8jzPP/LII98lhJgzMzMTu3bt+uXevXufPrs/pZQQQqh4ZpDHNE3zfBk+ZVnOtLe377vmmmu2dnZ2vqmqqjw+Pt77Qcrc5PP5rCiKVkmSzvseSKfT8e7u7sO/+c1v/s/c7YqisN9Fl9zbfxQOPPnYQ1u+dP+f3fL1P/07YpqmIEmWY7ueezwenJ4gxDQPPPnoQ0s233RbSX1jE0YIUwB49Vc//yH86z/CaFf7sZHO9qOrP/Xpz7esWb9Zk/P5ZGQmCABQX+qt7x6NdKu6qa5pKV/DYTR6pRS5Z5irHQsMGeYyEhwKn0YI4fh0ciKXzMeHT4wdBKCAOcwRQs1cIhcd7Zg8WtlSulijlNpcFg9cAclnPjKeF5Ak2Wg+l6aGMbuezdA1MA0DJIsNW20Omssk4Tw3xtbtn76fK6uo144dfJmIFquwYPE6QOj3W7v5Pv0AAJhNxHGmH5SYgBAGjDGlhGJBmlR0JQsIcwCAACihupo3lVzqcgwKAQAMg+oYAbafJ1GMQamu0dlyIyalpk6p1ihJCzEA9vK8X0BIyBOStSFklxCyJEwzolOqUQ0QBaCigN4x0nY5mZgJ9+98Y98DOVl5z5G4r93xqYOm+ba6lVBdXT1/4cKF/lAoNDY+Pt6XyWSSTqfTy3Ecn81mk93d3YfXrFmzbWJiYiAWi4UKCgqKTdM0AODE+/XLNE3S2tq6+1vf+tb3bTabc2hoqCsUCo3P3QchhNEsPLdcyODgYGcul0tt37793rq6Oi2Xy6U8Hk9hNBqdjkajof379z+7efPmHR0dHfsnJiYGXC5Xgd1ud/X09ByBOWsomY9P/9FDeyITY0PGOaO2M2MjQ/6Kyn8rrWtsrq6sayp1Byr0yZlwU13zsoHh3pOT7R0nCiSnr6C4tKKipKph7+svPGok0unN67betbBuQT7T0z/QMTw6rBiqHHD4ih9/4pGfq7Kcu3tZ8fb1CyvXp/JqKhTPhd7tQxGGYT55WGDIMBcJxyFeEJCoaVQhZz5NtVqxXdOoYprU5HnEiyKyIASYUiCaRlXDoOcGeRPntssLXLeclpMIIZSazUrKqIpMErEZHCipxA6HG3GcKrS0eYEXRJpKRMxYNMjXz1uMJIsNYZwDhBBQAMAIe//XAzfln370AfXNPc/wVbVNoOtzMmVSCqZpIl4QAGMOYWy+Z4239+kHAAAl5F3XBhJdleE89fFgdp3RZclqxQ5KKYnHzXdkg+UA8TxCAoeAb7ZYllSKYl2WkJQVY/tJWX5zpc22oUGSWgdUtTNmmpGoYYQ0SlWiUupycQ9fNsXtAaBIEEqLeb58TNMGOYQ4nVJdQBgHx6eGPBxfYMHIUiuKTWOa1s8jJNowdoQNYzpPSFay2/63NicjKQCAzWZz3HrrrfcFAoEKQogRDIXGXnrppYczmUySUkofe+yxH9x400333H///f+vIIpSLBYLPfvMMz89czhVVTVv6Ma7lnyYnp4eDgaDowUFBSXPPvvsf6mqqgAAuN1u35133vkn3/ve964tLi6uwhShJYuXXPeFL3zhpV/84hf/Gg6Hxx966KF/2b59+73f+ta3vg8AdHBwsOPpp5/+CQCEXnnlld+Koijt2LHjW06H3S3nstlDhw7uGh3s7RY4ZHAIOJOCaRJ41wRDzIUVm56agPP8rQAAiEyMj/g8Bbkq0VsyMR2JNdc2L5sKTQwtXbByYzaXSVoUE+Sxqan2/sH+xkBVUzoYnCrwFpbs2vPcrzwuT2FNRX2znIiEUm5Zn+g73QkAsKSheF9e1fOyasjpvJpm32eGuXKwwJBhLpLWVss1W7c6v/zLXyb/CQAG7Hbs/Pa3ix7/6U/j/0OScM+GjY7PrF5tu9np5HyJhBF6+eXsLwDgfVN/G7qpI4TGzownXVV/kJEkWbjCojK+sake2exOvqKqkS+vMoDnJ7RTx1633nrXH0sbbrzLnBobkFavv8Xo7T5ijA73kER8Rlq55ibrrXf+kXbq+B5ktTpILBo0hvrbzcjMJF/XuJDEYyFx8bINXElZDZz5BJyqmmwGp4allWu3iivW3ETTqRgShKOAMOYK/WV8bUMjstocXHlVI19RpYEgTL5XPy719bvQMEZo5Upbs9WKnYIA0tznBISEFouloV4UmyUEUtQwQy6MvTlCMmkKiTJBqHZgzgXUoAohcso0Y+vs9puOyfJeAIil0+a7rkm8FKwIOZotlqX1ktRKAahOqDqua8NejivsV9WOKlGsb5SkBQU8X9StKMcWWqwrfDwX2JvNvnA6mzt57s1zX1/fye985zv3C6JocfoK/P7K6jpFtHLuQHEpx3N8Vjfkp1/c9dDzr7z6a4wQFixW60R/bwcAgKZpms1i/asqT1kdmOf/HZBIJKL/9E//9EWEEJ67/i+TySR++ctffufXv/71/wYA2Fy/5uaB6GjPUHS87xe/+AWYpkkwxid+8IMfdPOzixmRaZqGfuYDk3Q6nRQE4UfPPPPMT+t9XMOCAGr1S6bvjxbTr3bO8B1VblzVFTa7+mKkDwCCF+v7wXxwGHOcwPNiNqekegY6jwyN9XfedfMXvhkMT48OjfZ3zqtrWaLrmnqq+/he3dDUZDoZTabiUYxxrK152bVVZTXzn3zxtw+cbU/VTTWeluPLGkuWpfJqCgDevISnxzDMBcQCQ4a5SDAHvCRhG8ZwdloikizYjjHiSkv5+tWrbTfv3ZN7/Phx+WWPFwcU+XfrqSRelNwWpzchp6I85kWJF0R5djQJCCWmwPFYM96ZdOJKhz2+gG/7bV+zVNY05ePRGefGGz+nBaeG8s8+9mMYGezWX9j5M+t1mz9DGuYv0bpOHVDeePlROlsqIcRXVP+tZfO2e6zbbv0yzWZT6t5XHwegNPfIf/+bbettX7bcdMu9+qljr+d3/vYHJB4NAgBQ0zC5Av9u5HB6LNdefzvJpBPG8EAHsjvc0oYtd/GVNfPNcHBcWrfxdmFe8zL5hZ0/1ftPnxSaWn9suXbTHUL9vMV6d/ub6hsvPyZQA9BIfw+fSSYkRPgiBw6gpuWTBsYcPTOa80lDCKV+P3/SMESNUnj7NF2OgxGOGzgVi76V2ZJDaK8gCKLbYS9IZrLRo/n8HgpAvC5X0UGMXouls8EPWl/y45YmJN6tKMd5hAQnxu60aSZGNO10muOK0oQkCAU6rRtjCiEyDyBETCPYrypdU7o+er4RFXO2jEreXeh3ljQ2tWKO41xFxSUldQ3zAQAwxpyczaRMw9BdBQVFks3umB4Z7gUAReJFqaGguvHTLTd8/pXBN58rsHm7CCVmpaeklse8MJ6cHi51BsSAo6AUI4xi+WSk0O7LlDj95W3F83nV0JSh+HifU7K7Q4nwZCgRniSGadb4Kuq9FlfB8rIFwlR6ZiKWT0bmFda0WAWLDYBSt8XZ4bf7ihf4G92yoeTDqenQFK+7IxzMZDSa8duwvydi9sRkGnu3ZDzMxy+ZTkR6BjqP1FTUt5imaeiapnT1tR8yDF0dGOk7ZbXY7D5vYbGua0o2l0mFItNjALMJ1VYsumZEFEQpm/9d5tvF9UWLTZOYBW5bgWGYBofxIfMDrF9lGObyxwJDhrkEslmSjEXN6YULLddqGlW6u5UDqZT5VjkCgROE1qKGxVPpmbEiR2GZrCs5jDDnsbp8oUx0KpyNBgFg6hKewiVhzgTH68pLH6ooCjRGU6lpoAgUTctrQCRLSXGzpGSI58Du10/1D+zJ5/JvS51vTIz2A8DfnafZLgD45rkbbTbsyOdJ1oxFQgDwg7ee+L//DgCQBYD/751Nze6mn+46CgBHz24tcuDA0lpuEzn93Ehep3kSgDavlfPu7TyyN3X08Esf/kpcPtJpM5bJmPFk0oyc3YYRQtWlJY3Lmudtqi0vfSWaSE0XFXgr2+Y1eFPZbHRBfd3aSCI5MTEz0x+KxcerSormO2xWz4FTnc9UFAXqCz3usqlIdCiSSE6dL7nKpRA3jCgARM/z1Nkb5tfP2d517o7no8r5bGRibIhSSg1d10Sr1SZaLFZimqYmy3mEMY4HpyflbCal5HM5AAAec3y5u7jKby8oLnUVVUbziXClu7SmrqByflbNZ5r8dQsUQ1VaixsXD8bGTku8aAlnYsHl5QvX9UQGT3ksLp/ICVJKzSQ+3bLlnkc7X/j5YGy89/q61ds8FldBUknH24rnL983dvyV21tu+HxSycT9dm9RJJcI3922/ctjianBqfTM+HQ6PPHGqLzn97+qzMfBMHQDAI5hhI/PyWr63JxdXj2n3NIoAIDPW1i0pHW593jHoVfnJi4amIwPNFcVNr94eOjFyiJXJYXL4z3KMMxHxwJDhvm4IACem33PxWLm9Asvpn+6bJlty5q19luXLLVuqqkVfzIyrHUBAEicYCl2+stKnP4yjDA3lpweKncVV0u8aOExx4/EJ/rf+8WuXJQCtYiirbq4uEkUBIuq64phmpqm66rP5QxghDkOY+6jvIbFgm1Ll1qvxxg9Qy7AdF2vBXnrfLiOUCDxPIo7RHAoJv1EjhKeS1WpAgBvS1lPAYDnOMHrdAZM0zQIJUTkeUttWUmromk5DmMOY8S11NasklUtJytq1mGzegq9nrKFDXVrJUGw1pWXLnhmz4GfALyzDMaVRMnlcgBw6sMck9PkXKWntL2xsLr51aE3n+MxL6wob1t3aPzUnt7IcOffX/+N/+gI9R0bio317uze/evPLrr5K8VOf3lMToSf6339sTVVSzbOD9Qu+G378z+7f+WOXoOYBgCAoqvyiXjXwY5Q37Gvrrj7LwTMCSIniIqhyieDp4+Ec7HgYGysxyZYnaqpqXmdZSL9JHmvUhfFLscCqyiM2USxIK9pMZ/dVlPpLajIJ2cgkwwrVQXelQYxFbsoFsq6HuubjD+blbVkOJUPX4jfkQzDXB5YYMgwFwmZTb5AuDPBoMvJFQgikmafowQAuiwWPFxRIczbus31lQ3XOe6CM6MMeV3O7R89/ioAgGbqqm7q2kB09LRJiWkS00ir2au2sH0wGhtJZDJhDmNe4HmJEGoiBKAbhi7wvKgbppbJ/260sLRUqFm+wrbV5cTegQH1hKpSWVZodmxU61m3zv7p9nbljeUrrDd53Fzh1LQ+ONCvHt+82XH9qlW2bT4fV1xeLrzmdHLeBQst14oCkg4ezD/X0mJZjTngBB5JgAC98Xr2t0uWWjf5/Xz5xLjed/RoftfatfbbeB6JvIBEvoA/8XiP/jgAgEHAkDiQEAKU0+CKvLGmlNKiAl8snEhOTsxEBioC/oYin7cikc6Gy4v8DaFYfKxvbOL4NQtatlklyX72OI/D4bdIkiMUjY0SQkxymU4rBQDYvHLZjlePnnhkbv3S2rKSFp/bVXSsp+81nuMEr8vpz8lKOq8o2Qv9+oQSk8ccjwAhk8z+XhB5QRI5QTKIoSMAZOElK485ns7WuTCtvGTlEOZ4zPG6+c7ENZqpqbKu5g1iGhhhpBiaohiqvG/06O5wNh7Kqvn0831vPNHkr1vYUtSwOCGnY/Ahg1rm8uSz26rzmh4rdjtaQqlMV02hd20wlerU1Rwq9zpXjkTie91WSznPcVY7xkXJvDJtmMSE2dkTDMNcIVhgyDAXSSZN4kCBLlps3VhfL/E33+L6LEKz6w2Li4Wq8nK+sbxcCCMEiMPA6Tp9KxvmmfWErHD9eciqmgeAD1TfThSxtHmz4/p4zAju26s+DgCweIn1+mzWTHIc4svKhPrBQfVEQ4O05Pnn0j+Jx83pdNqMDwyoJ/wBvmLfvtyTNht21tWJbb296pFsxkxsv9n91XjcCKZTZlQ1QS4q4iubmiwr/X6+4vCh/PMtLZY1DQ3SkuZmy+o9e3KPjo3pPbJMspoGCjdbdoQQMluQ7qJeqEtMUbU8pYSsbVtwc06WUyX+whrdMDRZUbNOm9Vz/fIlO5LZbJTnOamtsW6dRZLsoWh8NJvPJyqKAg0DE5OnLrcpajaLZF/ePH+z2+HwN1SU149MB7t33HD9ppl4fPxE78AbXpeT+FyuYowQqi0vbWqsrFh87HTfq8ubm1YNTEyectqs3oDPW24RRVtFUWBe1/DwwUwuH1/R0rwllkoF2/sH98ZS6Q+UaTiSi8/opqH/wZLb//jlgX1P9UWGuzbWrd66qe6a7a8PH37RY3H5Vlctvq7Q7isajo/3z2Rj063FjUu+uvLuP8+oufQLfW88saF25Y13LLhpWaHNGzgy0bFPNlRZN3WNUkKyWj5TYPP4XZLDc1Pj+tt1ouuvDLz57DVVSzd6rS5fSskko7k4y4p8BbBLoq+torTMq2pRDmG+wudZDoAgq2phgeOshmniMo9rsWoY2bSiTEs87wQA9L4NMwzzicMCQ4a5SCYmtL5duzI/u/56x+cXL7JuPHpMfunAgfzTskwyGAPX3GK5prFBWkookO4uZf+ZrKTMBSRJyMbzSEymSCQeN2YsFmwDSgFjhHkeRI4DXlWp/Mor2V+tv85x58S41nvgQH5nLk/Tqkrz6TSJuVxcAcKIi8eMYDRmThUWcmUzIX0knSFxSoF4PDhQ6OcqGhukpQgBUhWaT6XMqEmoqaqWjNNpc0qSIdrtkrO42FOez6tZRdFljsMdpnnlBoepbDbGcdxPEQJMTGJ0DA4dIISaZyM9jBCmlBKTEIIx7gGYLRiPMWrHCHOEUpO8V2mQS6CqpLjJMIlxoL3r2bu3XP/nS+Y3bkxkMmGPw+EvKfBV5VU1CwBAKKUuu22ioijQaBFFWyydCq5bvODWeCod9LlcxaIgWKaj0eG2+vq17YND+xVVzb5x/NTjH+Z8VUNTOYy/A4DAJCYBAOAwd+TM02hT3TXb3xg+vOupnld+TSmlbSXzl58OD7b/6tQzD2qmppqEEIzw8N7RY7thdpB3NhMvUKCUUoHj//m25s2fe2lg31Ojyemha6uX38BzvPDTY4/8x2xMQIElHLky5FQtjhF68GzCIIQQAkqBUEoQQr0As+9XQimhlFKEELpc1v4yDHNhscCQYS4SXacaALx05t/5fPtj7M5VSZZJNp02Y1WVQtOiRVauqIifyeVpqqpKaOYw8A4n5xUEJGEMeKBfPV5QyJe6XLggnyMppxP7mpqklcmkGVYVkps/37JcVcmCgX71+LmvMz6m99hs2NXVqezNyyQbiRiTQAHq6oqaZJnLhMPp6ebmssWSJFh4HvO6buqDgzM9cIWvnxNtgk2wCDYlq6aJQQyKKMIYYZvH5kMAkE3kogCzAeHZY86sV7ps6hfOpWha3maRHGWBwjpKKcnLStputbinwtGhZDYbrSwqaiz0uEt9LlfA63J5i3zeinQ2H+sbGz++ZdXyzxNCzd6x8WML6uvW2K1Wd8fg0H7DMLScrGR+nyD43MDsbICIEULTmZlxMS9aHE7JLYq8RCSdRkgkXFrurkykcrGSEo+luMSNAAAEgROTyXwslfrdFGzdNIy6gsqD11YvvyHgKCgZS04ODscn+lkweGU6Z9r2Wz+LcwJA8zzbGIa5wrDAkGEuAslitbYsueY6QRDF4b7Ok5HQ5Hhjy5IVwcmRgUwq8YGKlWPMYclitcn5LFvD8XsyDKoXFPB7GhrEpW4P589mSWJiEvXXNFctLaggNYPj+a6CytLaylK5AWPEpXNSwuCiYPe7A8E4nWhd3bhpcFA/laGG3DDPuZQSjZw8Et5dXu2dL5OcItnsjlBSmYznc7EcKlWal2kb+k6e3s/zhnD4SP6FmZngkMvpcmUycrKvL9iBMcKqaiiZjJzStHcvTn4lkGyirW5Z9WqLw+LKxnMRXdFlJadmrE6Lu6DcW00p0KFjowcQRlOflHqcwUhsxGG1ui2iaD/SffrlyXBkoLqkuDmZyYYN09QVXZMnZyIDosBbgFIIRuMjiqbmTEKMA+1dzymqlhsLzvTyHCc4rFZPMpONyqqaVTVdvpD9PFMe4yQAwPz5pQtaW8uXnT49fUoujCmLqqtWq6ouSxJvMQxiiCIvjY5GBsrKfNUA8OrcdoZi4/0AcNUmumIYhrnasMCQYS4Cu8PtWbJq4437X3nqEV2brVGXz2XShmF84JEQT0GguLaxdTEAPH/ROnoViMWMEMy5hu5Cf1HXgO9wZGJ8pLimbp6zqCAwI2vhbDIRmxro7yksr6rW5Hw+mK0O2Tm36iqBUmKa5mRKmDA1TS1saGjMarpsKeL8seD0BMf5+OJGd3Nn19h+t7+o+Pjx46+ceakXz+nK4Md20pcBQRIsgarCOl7iLa5CR0BXDIUXOVFXDUW0inYlo6SsDsmVjuAQXKYjhOdSNE0GgIPnbB6b8/jcgu59cx4fmPP4GEYIzalveDELwSPTpKbLZfU4nRY3pbMZrERRsOTzuZjPZ/fb7RYnz3PCRewDwzAM8wnAAkOGucB8hcUlG7bd+cW6+W1LRwd7OuLRmen5C1dcs+W2L3zm+Uf/6wcAkGldsua62nkLFiMANDk2eHps6HTntTfc/jnJarUNdJ88EpwcGdh6x5e+WF7d0LRx247C9qN7X4mFp6+6uoUXA+I4rKuqkgyHprzFJeVO6ivMxOMRq9PpDlRW1vKCKFkdDhcvCAIxTZOYxEQYIUPTVEooBWLqhq5pqmzksol41FXoL+ZFSfL4AyWU5WN4i5xR0idf6tqJOcRRAoQQQix2yVm7pGp1IpicGOuYOJpPK0lCiPn+rV15zlf0/mIYH48OBoPJcQBAfX3BTo7DnKLosiTxVsMwtWuuadg0MhLuSybzsfdtjGEYhrmiscCQYS6wVCISPrznxZ3+4oqqA688/YimKXIiOhNcvGrDFkGULAAARaWV1aZpGPt3P/VbRc5li8uq6xwuj3f/7qd+Gw1PTcj5bKan/dC+TDIePfj6s49rqnpF1Ly7HKSjkZlsIh41dd0YOHZk33D7iUOmbuiY43hAANQkJuIwB5RSQAgBhd/l35u9lacIIUSISQghJh7o60YIY4QRIubVGeScj2mYBgCE5m5DCIWTodQkJZToV9BUWo7jOKfT6bHZbE6O4zg48xOTy+XS6XQ6ZhiGabFYrG63u4DnedE0TSOTySTz+XzmYq/Xyue1d8vimwYA8Hrtr2UyStIwTPazyzAMc5VjgSHDXGCmaZqFRWVZQ9fUOesDzc/c+6canAktCDFJIjoznUpEIwAAksU23H3y4J41mz+1Y7D75JFTR954SVXkvK5rqpzPXZG17i4VYpoEAM4mWtDP/IM5/39Yn4hpkJeDM0HQFfUhB8YYNzc3L7vlllv+MBAIVLhcLl91dXVTKpWKPfHEEz967rnnfuZ0Os3bbrvt3jVr1my3WCxWXde1rq6uQ08++eQD8PapqB+7RCLHRgoZhmEYAGCBIcNcdBzP82VV9fNuvuv+mtrGBUsKA6WZ1mVr3zZQYLM7XEAphKcnRtzeQr8gWqxyNpMqKquqbWxZsmJybKA3n81ctUXtGeZy5XA43Fu2bPmcKIrWH/7wh3+BEEL33XffP8bj8fDOnTt/nMvlMlu2bPnstm3bvvjII498b2BgoL2oqKhix44d3zQMQ8cY//2VXtOSYRiG+WTAl7oDDHMlymVTqcN7dz0FAIAAAd7tsuYAABd3SURBVMfxfNeJA69n0okYwhgP9bYfHxs63XF2f4wx5kVRTMZmgvF0JGz1uD3h4Phob8fhAxwvCAhhtniNYS5DLpfLGwgEyvv6+o6Pj4/3TU5ODvb29p6w2+0uURQtAAAbNmz4dG9v7/EXXnjhod7e3uN79ux56rXXXnt89erVNzmdTvelPgeGYRiGAWAjhgxzUci5bAYAXgcAMAzdAID2xuWrXAWlZVUVjtbFhqZpM2Ojg7woinWLlq4ygPRhj82RSyXjDqfHtXD9xq35dDrZd/TgnujkxCWdasYwzLvLZDLJaDQanD9//tKSkpIaAICGhoa2YDA4ms/nMxzHcQ8++GBZX1/fCXN2GjMAAExMTPRLkmT1er1FAPCBStgwDMMwzMXERgwZ5mNS3bpwqZLLpouqaxqqW9uW+isqa3hBEBqXr1xX1dy6JJtIxEbaTx6Rs9l0KhoJBYcHTresuXbzpe43wzDvLpPJJF566aVfeb3ewD/8wz/88pvf/OZ/TE5ODj311FMPqqqqmKZpapomW61Wx9zjznxNVVU9X2IYhmEYhvnYsRFDhvm4UArZZCJmaLrGC6JktTtc3qLScslmd/CiKCm5XObsrsmZ0HRiJjTdsHTFukvZZYZh3hshhG7evLkyn89nH3jggb8ZGxvrNQxD13X9rayrJ06ceGPJkiXXVVRU1CeTyajFYrF+5StfuWlycnIoEolMX8r+MwzDMMxZLDBkmPOQ7JK9tKGoWc4o6chYbNB8n1TuGGNUUO6tVvNaLh3NhM+3z8zoSH82EY/OjI0Mmrqu2VwuT92ixaum+nu7Jvt7O5tXr70ecxynyXI+k4hHdEXJz4wO91+cM2QY5kKx2WxOm83mbG1tXVVSUlKt67oajUanbTZbXz6fzz733HM/r6mpaf7GN77xbxMTEwM+n6/Y7/eXPfTQQ/9sGAbLasswDMNcFtDHVGOXYT5RHF67b+3dK+9Tc2r2zceP/VzNqe9ZMoIXeeGaO5bdG52Ij/Ts69/9QV4DYYyA0rMp/AEhhAAhoISwNyXDfEIEAoGybdu2fXH+/PnLDMPQCCFEFEUJY8w988wzP33zzTdfqOT5+kBpaflt69bd5/D5POOp1EBvR8fRztOnD3GmyemUak6O80zr+ljSNKOE/WFmGIZhLgE2Ysgw55FN5OLN1zYedwdcpQCzgd+iG1putTot7shYbHj45PjBmkUVKwsrC2qVjJK2OKTXzx5bs6hyhSDxlv7Dw3vf6zXODQAvdqFrhmEuLI7juE2bNq1vampavnPnzh8PDw93EUJIIBAo/9znPvfnCxYsuOb48eOvV2JcNzU9Pdr57LP7woYxJROa8/FcYIUoXRc1jVDYMKZqRbFJISSfNs0EsNqYDMMwzCXAAkPmqlISKKsSRckSjoYmTdPQOY7jeV4QCSFE1zVF0zXtfMeVzy9p85V4KgePje4vbyppcwdcxS6/o3i0ffKop8hV2rp+/k2UUlLRXLbI0A2t643eF85tY3Ext6jUiUsPThoH4zJ91yyE8wq5xpksmUkqNHUhz51hmAuLUkp5nhf8fn9pQUFBcSgUGhNF0dLS0rKyqKio6tSpU/s1TVM7KD2cozSTzmQSKqUyh5AgakgUEbJkCEkqhOSndX1cJiRHAN5z2jrDMAzDXCwsMGSuKl5PQWBebdOSVCYVz+WzKVGUrIqSzxb5Syq7+toPcpjrN4n5jmLT/uqC+shEfHi0Y+JocX2gqWZx5arpvlDX5OnpU4ZqKG2bmz8lZ5TUoi0tt77+3wd+GJ9MvK3ERIkTF//RcuF21QB1LIXH1lRyTWsr+bUng+bJvijp+3yb8PmBGBlonyHtn20VPptUabI1wO3uCpvdH9/VYRjmwyCEkMLCwpddLpdv/fr1t91yyy1fJoSQcDg8+fTTT/9k7969TxuGoQNA7MwhwfdoTv4YuswwDMMw74oFhsxVRVbyOVnJZ3VdU71uX2B8erQ/EpuZ0g1dU1U5jxDCAECsTotz0ZbWUleho8gTcJbGp5JjNYsrV5XPL1nIC5w41jF5zOG1F5bUB5q9pd6K8FhskBc56bWf7/9+Ua1/XnlTSRsAHD37usEMCd3WJJyYTJPJoQQZKnPiss4Z0tka4FojORrJqpDtCpOuiRSZmEiTiQMT5oG+KOm7dFeKYZgPIhqNBjHG33300Ue/jxBCALMjiZRSQth6YYZhGOYThAWGzFVlZHywBwB6zvPU1NkHGGNcWOkrsjmtboQQCtT4G4aPjx0sKPfW1C2rWhMeiQ0MHh3ZX7OkalVFS9liOaOkO17peaZsfvHCTCwXGTk1fri6rWIFzAkMAWZvFgkF4rch/0313E0IARIwCBmNZqYyZGp7I7/9idP6E2mVpms8uCaWpzEAiFzcK8IwzEd1JgB8xxRQu+T0GsTQVF3O8ZwgWkWHOyMn2HuaYRiGuSyxrKQMM4fVarUXFhaWhEKhcY7jOABAhBCT53khn89nP0rb8wvxvLQK6ZxOcwsC3AKbALbZqaVkrN6H6zUTtK6w2eW2IHedF9f1xUjfRIpMXqBTYxjmAnDbC0u8Dn+lxFvtU/GhDoGTrAF3eYOi59KR1PRgWUHtQpMYeiwdGmksW7SBUkrHIn1HLYLVOa9s6abeqeO7CTGNtJyYUbV8JuApb5yKDXde6vNiGIZhGDZiyFw1MIcxxpgzdEN/t30opaS2tra1qalpGaWUhsPhKQCghBDCcdwR03zveobvpffM1FCMEX5zwnxzbkp6DqMJhACZBEhcJgmM0BhLWc8wlx+/u6zBbfOV5JR0tLlixY0ZOTHjcxRVDQY79lYU1i8p8dW0msTQnFZvkSTYnLKWTemGKlsEq9Mq2V2qLmetosNT5W9clsxFp4o8lfMBgAWGDMMwzCWHL3UHGObjwPEcV7WgfNnyWxbd7Sl2l77bfg6Hw+P3+8skSbKpqir7/f4yu93urqysbMQYc2f3wxgjnuO5d2vnfBBCqLxcrF+x0naT1YacbjdX6HZzhQ4H5y4o5MpKy4T6ykphniRhCXPsQxuGuRzxmBdlNZecSYz3FjiKqymlNJENj88kJ/rdtsISRcunounpoUhqejCvpuOxTGgkLcdDOSUdS+ViwXh2ZiyYGOkqcJbUNJYu3jgW7j1yqc+JYRiGYQDYiCFzlaCUUtMghs1t9fpKPZWCRYjriq6cfd5pd3lMYhqqpoSff/75nxNCCKWUYozxmdFCTtf1t0pZrFy89sZ1KzZ+yuv2/V0iFf9Aa4YQAiSKYBF4JK5ebb/Z5caFPIf406eVw5IF28IzxnhtrdhWVi409p5WDwPAzEW4FAzDfEQV/oalxd7KpsFQ5z6D6CqC2aQzg6H2vUvrNtxlEl2LpqeHounQSFVg3nKglEYzoRGD6Fpr5aptg8GOfbFMaLS8oG5RMhdh08UZhmGYywILDJmrAsaY43jMG7qhlc8vWZgMpSYBYBIAgOM4fPenvviVYHhq9NX9ux4DgPz7tScKouSwO9wY4w886k4IJR4PF7LbTU+hny+bmjQGTJPqikpzTid4rTbkyOVIUhCQRCl9R8kMhmEuPZMYWv/UiVcnYkMnDePtdU/jmfAEAPw7xhwmZ8reYMydImR2CjpC+GcIAZJ4q4PnBLF38vjLJvn9p6czDMMwzIXEAkPmqkAIMdWcmlUySioVzgSVrJoBAPB5CgM3bfjUtuWLrtkcjYeDt9+0ozCbzyRf3vP8byySxVpTWd9SWVYzTxREKRwNTXb1tR/M5bOZuW2Lgig21DYttkpWe/dAx2FFkfONtc2L66sbF5imaQ6PD3SNTgz1EKoTQiDe26scRn0IYwwcIdRECKGRYa0LY8TZbIbDYsH2bJaw4vYMcxmKpKeHKCEmMQ3j3fYhc2qhkjmBH6WEAgD1OgJuWc2mgonRHoDZOqcWHiwjCTJ6MfvOMAzDMO+FBYbMVQEhQLxFsAiSYI1OJEbzaTkFAIAwwjar3WW12BwCL4iSZLFquqYCAEiSxVpTUd9SUlRWTQghSxes3Oh2egoA4Ddn28WY45obF67cvun2L53oPPJGz0DnkdrKhpZbb7zz/mhsZkoUJMvC5iXrXnjtqZ+rRp+COcoV+rlySoCoGlXiMSNYWirUxWLmdCZjJqam9EEAiF6aq8QwzPtJZiNTAABeK/Jsb+TXpFVId4XNrgo3rih34vKxFBlLqzTdVsy1jafIuGqAmpBpQjaovLCIW2gSMNeUooKpTGzKLVH7zfOE629u5EvCORoGgNFLe3YMwzDM1YwFhsxVQ5QEa2FlQY29Z9oLACMAALF4JOR1+x6uq2pc0HH6xP7nX9350Nn9c/lc+tCJfS/qhq6Zpml86oY7vrJs0epNcCYwNAkxKkurGm/Zcud9R08dfGXPoVeezMu57B/9wZ99fio4Prhz129/LPCi+KUdX/+HZW2rNnUPTxwp9OvlPI9Ew6Aah4EvKOBKiouFGsNUdIcDewBg8FJcG4ZhPhzdBJ0CogsCaAGlQFsCuGU4QYZ1E/TV5dzqqQydml+I5xc7cPH+MWN/VIbo1gZ+60CcDEykyMSaCn5NLE9iaQ3Ssg4yRiwZHMMwDHNpscCQuSpQCjQ5k5qc7gt1mwZ51ylgc4mCILXOa1u9YtGazR63LxAoKCqPxMNTGGO0bsVGEHjRcs9n7v/r/qGeE68d2PVoXs5lOY7jvvvtn64o8BQWXbNs/XYAAJ+nsPhE15HXX3wN/ruvL3eMUiBFxXx1ba24cGxU7zl+TH5ZlkkWY/hQWU4Zhrl0VpVzq4rsqEgxQXFbwK0T0AfiZCCtQtprRd43Rs03Kt1cpVsCt0NCDsUExSWBK6vRbPuM2b6omFtU5sJl7QNGu+5CukNEjkt9TgzDMMzVjQWGzFWB4znBU+yuQBghOS0n5z5H5/yHEEL0TP3AFYvXbrnxupvvefqlRx88PdB1dOOaLXcuXbjqegBAAADFgZLKztMn///27u5XjrIO4PgzM7vnpee09PT0hbZwWqW04EsxmoZUwRfUGAiJRS6IkGjwxhDvjPo3mBgvjDfGGGPEC0Ji0AAaSWqq0kosSoNQStPS09IWempfzuu+zowX0lhqTQxNXMjv87mcnc3+9mrn++yzs/u2bNq6fWrj5m15nv8tz/OsrurqF0/89Hv7nt/z5KXX6PW63Va7tVBVVZ1SSnmevXh8uvtyWaayqtxoBt5r6pTqyWXZ5MVWfXG2nWazrMo6/dSZ79TzB89WBx/a3nzo5Fx18pmj/WfunGrcudirF6cvVtMXWvWFbpm6M4v1zNEL1dF7tzbuXezWiy+drV4a9HsCIDZhSAjNkcboitXja7MiK4qhYujyx/r9Xq/b67SvW75ycmR4dNnQ0HC/3+91d33hgXWLiwuzx04cPZhlWb5l87bbLn/e6Tdff+3xpx79wSdv/+yu++958BuP/vIn3z1x6tjhR77yzf1bNt384X379zzZai8tjo8tv66Xsu6lKEzpX3coTSkJQniP2v1a//dFnu0pr76w86c8y56t3lpkKvJs/1XOezSllPIs+/Ol8wBgkIQhIXQWu/OnDr359+u3rL01XXEJ1mq3Fl458vLzO27b+fkv73r4W2fOnj7+9O4nfvax7bcf2v6Bj95x390PPLLUWpofHV02VtfV2y7uut1Oa/ezv31szap1G+75zBe/Ojmx5ocrxlf8/MH7vvadB+97+Nutdmuh0WgO/fG53b/K8/yFy+MQeG/7L1GYUkrp8tj7X88DgEEShoRQ9stydPnImSzP8oXzi2/7Q/qy7JcrV6x6+sLFczOrV63dMLX5xq133XXX/Y2RbO6vL+3d3ciHh65fv27qid899qNP3fHpXVNTU1s7Vat9fObIoalNN26bOTtz8tfPPP7jG9ZPbamqsjz2+tFXbtq09fs3v++WjzSbzaG5hbkL5y6cfUMUAgDwbpVZrIR/K4oi37Vr19c7nU5rdHR0/NChQ88fPnz4wI4dOz5X13W1bt26qWPHjh0siqKxevXq9SMjI2MHDhz4w/T09KuDnh0AAN4pYQhXGB8fX1GWZb8oima32233er3u8PDwSFEUjZRS6vf7vZRS2rlz591LS0vzzWZzaO/evb8Z7NTAIDRXbdgyvPHWj5dLszNDq6c+2Dl9eH8+unxV1VmaLZZPbsxSysql2ZnmxIabu2deeyEfGVu5dOQvTw96bgC4kq2kcIWFhYW5qxxuXXmg2Ww+NTExseb8+fMz/4exgHelPM+aI2P5SDnROfXqc/3ZM9Mrtn3iS51TB/eVc2dfb67Z9KG811nsnDq0L2sOj6XaXYgBeHcShvAO9Xq9Xkrp9KDnAAanPzdzvDcz/WKxfHJjXXbbVa+10D7x4p5y/typ3j9OvJw1h5dlxdBw3e+0qu7SQtVZnB30zABwNbaSAsA1mrxh1aZuq7vUmm/PFo28UfarfkopFY2iUVV1Xfb63aKRF1meF1mWsn637Jb9sj/ouQHgEt8YAsA1uv6mNbcMjw2PdxY7CyllKS+yRl7kxfLJ8bUX37h4sr3YmVu1cWJzlqU0M33uyBtHZg6mlBYGPTcAXCIMAeAadVu9paGR5ujE+pU3Di8bGisaeaO90Jnv98punVK97v1rtqWUUtmv+t12r9Vrd//jd8sAMEi2kgLANWo0i2aWZVlW5EWeZ0Vdp7qu6yrVdV1VdZUXeZHe+sAt+2Wv7FfloGcGgMsJQwAAgODyQQ8AAADAYAlDAACA4IQhAABAcMIQAAAgOGEIAAAQnDAEAAAIThgCAAAEJwwBAACCE4YAAADBCUMAAIDghCEAAEBwwhAAACA4YQgAABCcMAQAAAhOGAIAAAQnDAEAAIIThgAAAMEJQwAAgOCEIQAAQHDCEAAAIDhhCAAAEJwwBAAACE4YAgAABCcMAQAAghOGAAAAwQlDAACA4IQhAABAcMIQAAAgOGEIAAAQnDAEAAAIThgCAAAEJwwBAACCE4YAAADBCUMAAIDghCEAAEBwwhAAACA4YQgAABCcMAQAAAhOGAIAAAQnDAEAAIIThgAAAMEJQwAAgOCEIQAAQHDCEAAAIDhhCAAAEJwwBAAACE4YAgAABCcMAQAAghOGAAAAwQlDAACA4IQhAABAcMIQAAAgOGEIAAAQnDAEAAAIThgCAAAEJwwBAACCE4YAAADBCUMAAIDghCEAAEBwwhAAACA4YQgAABCcMAQAAAhOGAIAAAQnDAEAAIIThgAAAMEJQwAAgOCEIQAAQHDCEAAAIDhhCAAAEJwwBAAACE4YAgAABCcMAQAAghOGAAAAwQlDAACA4IQhAABAcMIQAAAgOGEIAAAQnDAEAAAIThgCAAAEJwwBAACCE4YAAADBCUMAAIDghCEAAEBwwhAAACA4YQgAABCcMAQAAAhOGAIAAAQnDAEAAIIThgAAAMEJQwAAgOCEIQAAQHDCEAAAIDhhCAAAEJwwBAAACE4YAgAABCcMAQAAghOGAAAAwQlDAACA4IQhAABAcMIQAAAgOGEIAAAQnDAEAAAIThgCAAAEJwwBAACCE4YAAADBCUMAAIDghCEAAEBwwhAAACA4YQgAABCcMAQAAAhOGAIAAAQnDAEAAIIThgAAAMEJQwAAgOCEIQAAQHDCEAAAIDhhCAAAEJwwBAAACE4YAgAABCcMAQAAghOGAAAAwQlDAACA4IQhAABAcMIQAAAgOGEIAAAQnDAEAAAIThgCAAAEJwwBAACCE4YAAADBCUMAAIDghCEAAEBwwhAAACA4YQgAABCcMAQAAAhOGAIAAAQnDAEAAIIThgAAAMEJQwAAgOCEIQAAQHDCEAAAIDhhCAAAEJwwBAAACE4YAgAABCcMAQAAghOGAAAAwQlDAACA4IQhAABAcMIQAAAgOGEIAAAQnDAEAAAIThgCAAAEJwwBAACCE4YAAADBCUMAAIDghCEAAEBwwhAAACA4YQgAABCcMAQAAAhOGAIAAAQnDAEAAIIThgAAAMEJQwAAgOCEIQAAQHDCEAAAILh/Ai+hGzTbXBxkAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "makeImage(vocab_tf,'Yelp.png','test2.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# play"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-11-12T15:49:31.286073Z",
     "start_time": "2018-11-12T15:49:31.279421Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[\"Papa Murphy's\", \"Papa John's Pizza\"]"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nameset=set(df['name'])\n",
    "[x for x in nameset if re.search(r'(?i)papa',x)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {
    "height": "105.50271606445312px",
    "width": "331.25px"
   },
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "204.8505401611328px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}