Youtube Clickbait Detector - Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "### Training set data distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "\n",
    "X_train = pickle.load(open(\"x-train\", \"rb\"))\n",
    "y_train = pickle.load(open(\"y-train\", \"rb\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(28800, 12)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "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>channel_subscribers</th>\n",
       "      <th>channel_videos</th>\n",
       "      <th>channel_views</th>\n",
       "      <th>video_comments</th>\n",
       "      <th>video_dislikes</th>\n",
       "      <th>video_likes</th>\n",
       "      <th>video_views</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2.880000e+04</td>\n",
       "      <td>28800.000000</td>\n",
       "      <td>2.880000e+04</td>\n",
       "      <td>28800.000000</td>\n",
       "      <td>2.880000e+04</td>\n",
       "      <td>2.880000e+04</td>\n",
       "      <td>2.880000e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>3.902016e+06</td>\n",
       "      <td>4432.620417</td>\n",
       "      <td>1.544025e+09</td>\n",
       "      <td>2508.441736</td>\n",
       "      <td>1.277371e+03</td>\n",
       "      <td>2.255264e+04</td>\n",
       "      <td>2.590318e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6.059595e+06</td>\n",
       "      <td>14052.457741</td>\n",
       "      <td>3.155282e+09</td>\n",
       "      <td>9599.632181</td>\n",
       "      <td>8.233652e+03</td>\n",
       "      <td>1.405234e+05</td>\n",
       "      <td>2.617725e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>1.863240e+05</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>3.315150e+05</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>1.174793e+08</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>2.700000e+01</td>\n",
       "      <td>3.440000e+02</td>\n",
       "      <td>3.560450e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.435299e+06</td>\n",
       "      <td>692.000000</td>\n",
       "      <td>3.962555e+08</td>\n",
       "      <td>385.000000</td>\n",
       "      <td>1.690000e+02</td>\n",
       "      <td>2.791000e+03</td>\n",
       "      <td>3.451205e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>4.713043e+06</td>\n",
       "      <td>2065.000000</td>\n",
       "      <td>1.469754e+09</td>\n",
       "      <td>1880.250000</td>\n",
       "      <td>8.480000e+02</td>\n",
       "      <td>1.515450e+04</td>\n",
       "      <td>1.718346e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.452421e+07</td>\n",
       "      <td>142439.000000</td>\n",
       "      <td>2.631521e+10</td>\n",
       "      <td>793927.000000</td>\n",
       "      <td>1.004742e+06</td>\n",
       "      <td>1.759136e+07</td>\n",
       "      <td>3.773968e+09</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       channel_subscribers  channel_videos  channel_views  video_comments  \\\n",
       "count         2.880000e+04    28800.000000   2.880000e+04    28800.000000   \n",
       "mean          3.902016e+06     4432.620417   1.544025e+09     2508.441736   \n",
       "std           6.059595e+06    14052.457741   3.155282e+09     9599.632181   \n",
       "min           0.000000e+00       12.000000   1.863240e+05        0.000000   \n",
       "25%           3.315150e+05      303.000000   1.174793e+08       50.000000   \n",
       "50%           1.435299e+06      692.000000   3.962555e+08      385.000000   \n",
       "75%           4.713043e+06     2065.000000   1.469754e+09     1880.250000   \n",
       "max           3.452421e+07   142439.000000   2.631521e+10   793927.000000   \n",
       "\n",
       "       video_dislikes   video_likes   video_views  \n",
       "count    2.880000e+04  2.880000e+04  2.880000e+04  \n",
       "mean     1.277371e+03  2.255264e+04  2.590318e+06  \n",
       "std      8.233652e+03  1.405234e+05  2.617725e+07  \n",
       "min      0.000000e+00  0.000000e+00  0.000000e+00  \n",
       "25%      2.700000e+01  3.440000e+02  3.560450e+04  \n",
       "50%      1.690000e+02  2.791000e+03  3.451205e+05  \n",
       "75%      8.480000e+02  1.515450e+04  1.718346e+06  \n",
       "max      1.004742e+06  1.759136e+07  3.773968e+09  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "#### Plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "train_set = pd.concat([ X_train, y_train ], axis=1)\n",
    "clickbait_set = train_set[ train_set[\"label\"] == 1 ]\n",
    "non_clickbait_set = train_set[ train_set[\"label\"] == 0 ]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "Views distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py:52: FutureWarning: reshape is deprecated and will raise in a subsequent release. Please use .values.reshape(...) instead\n",
      "  return getattr(obj, method)(*args, **kwds)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAikAAAGBCAYAAACjNCEAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAPYQAAD2EBqD+naQAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzs3XmYFOW5/vHvLbIpMOBCUAEhCu4G1wgqGPeI8bjkxD0iiUZRzzFu+UkSJRq3RHE3LtFoODEekhjjEhENwagsKoJxwSUeEJVNRYGoLMLz+6OqoenpGWZ6mumamftzXX311FtPVb1dPT3z9LtUKSIwMzMzy5r1Kl0BMzMzs2KcpJiZmVkmOUkxMzOzTHKSYmZmZpnkJMXMzMwyyUmKmZmZZZKTFDMzM8skJylmZmaWSU5SzMzMLJOcpFiTJunPkr6Q1LmWmN9JWi7pK+lySBrRaJUsM0kjJEVB2UxJ99ZzPwPSfdV47mrYbo1jSdovPaffrs9+1nKMDdK67Vdk3ZD0eL3Kdbz6kNRa0huS/l9W6lQfRd6/zdNz3a9I7L2S/t1I9RovaXw9t2kt6R1J566jalmFrV/pCpg10N3AkcAJwG2FKyVVAUcBj0bEvLS4P/B+o9WwcRwFLKrnNgOAS4F7gU/X8bHqawOSugGML1j3GMl7OGcd16Emw4AuwM0VOn5DFb5/m5Oc65nAtEpUKDWsvhtExHJJlwHXSxoVER+vg3pZBbklxZq6x4HZwNAa1h8PtCdJZgCIiEkR0aySlIiYGhHvrMtjSGrfWMeqTUR8mL6HSxv72JLWBy4E7omIzxr7+A2RlfevJhHxekS8XsKmvwcC+EGZq2QZ4CTFmrSIWAHcB+wmaaciIaeSfON+PFdQrLtHUjdJd0h6X9IySTMkXZr+U8rFvCDpsYLtXkn3t0de2dFp2U7p8qaS7pT0nqSlkj6U9JykA9f2+iQNljQt3W6GpAtqiCtswl9P0k8kvZl2h30q6Z+S/jtdPwL4ZRo+I61v5LpX0v09mr6WqZKWkLZs1NK11E7SSElz02M+LWmXgnoWbdJPuxVmpj/3Aj5MV12aV7d70/VFu1YkDZX0sqQlkhakXYHbFTnOvyVtLemv6c/vSbpOUtti57bAEcAWwKg6xNapTmncaZLeSt/n1yWdkH9O8uIulTQ53dciSS9J+p4kFcTV6f1L3+8X0s1+k3euRxTsr9bzJalXut2Fkn6UHuOL9P3uq6Rb5mpJsyUtTM9D14JjVPvdkNRW0iWSpqfn8GNJf5c0IBcTEcuA/wVOLzwP1vS5u8eag3uA/0fSmvLDXKGk7YE9gavTZKYoSd2A54GVwGXAOyTdCT8BepEkOgBPAWdLap02M38F2BH4AjiI1X/sDwTmRcQr6fIoYFfgx8BbQOd0eePaXpSkA4C/ABOB44BWwEXAV2o9G4mLgBHAz4F/AK2BbdNjA/wa2Ag4Bzia1V0n+d9kdwW2S/cxA1hby8GVwEvA94Gq9PjjJe0SEf9XhzrnzAEOBcaQtID9Oi3/sKYNJF2cHv/3wMUk53YEMFHSHhHxdl54a+DhdN/XAQOBnwILSd7/2gwG5tflG39d6yTpdOAO4E8kv79VJAlFsaSpVxo7K13ei6TbaYsida/L+/cSye/3b9K4XBKe39JYn/N1FvDP9LlzGv8IMBlYTvIZ3RK4luR9PaJInYBVrVaPA/sCNwDjSP5n7QX0BCbkhY8HziT5PL6CNR8R4YcfTf5B8kfqQ6B1Xtm1JM3AfQpiAxiRt3w7sBjoWRB3fhq7fbp8QLq8b7p8Iknf/q3AuLzt3gJ+l7e8GLi+hNc0CfgAaJdX1hH4OPnorhE7E7g3b/kRYOpa9n9B+np6FVk3E/gS6FvDuvxj7ZfuZwqgvPItgWXAXQXv0/gi+7wXmJm3vEnh+5S3bkh+vUn+GX4OPFYQ1wNYUvBe3Jtu+58FsY8Bb9ThPXkdeLxcdSJpzZ4DTCqI65meu5m11GU9kn/aPwU+Kjj39Xn/dk/rPqSG92Wt54skeQqSMS3r5ZX/d1r+l4Ltr0/LO9X0uwGcnMZ8vw7vy9Zp7Bn1/Zz5ke2Hu3usubib5B/bEbDqW9hJwDOx5rfoYg4H/g7MlrR+7sHqLqJB6fNzJP9gct00B5H8YR0DDFAyI6UH0Iek1SXneWCIku6XvSS1XtuLkbQhsAfwYEQsyZVHxGKSBGRtnge+Juk2SYdI6lSHbQr9MyLeqkf8/ZH+xwCIiHdJvu1+o4Rj10d/knFH9+YXRsR7JN++DyiID6qfw3+SJFVrszkwv4x12gboBowuiJtF8vu2Bkn7S3pK0kJgBUnrxGUkrTRdC8Lr+/7VpD7n668RsTJveXr6/FhBXK68Zy3H/SbJ5+2eOtQx955sUYdYa0KcpFhz8UeS5udc18xhJN0id9e4xWpfAb5F8gc///Faun4TgDRZeI7VScoBwJMkiUorkmbpg9J1+UnKsSTjZr5P0nWzQNJv026mmnQh+XzOLbKuWFmhq0haSvYiSbY+lvQ3SbvXYduc+s6eqamutXZrlUFu/8XqO7vI8T/PT/xSS4F2dThWe5J/nOWqU+55XpG4Ncok7QmMTRdPA/YmSWSvyKtbvnLNfqrP+VpQsLxsLeW1nfNNgdkFSU9NcvUrPAfWxHlMijULEfGFpN8Dp0najKTvezHwhzps/hHJN8Mf17B+dt7PfwMuS/9hdAeejIjFkl4gSVA2B95KvzHn6vYRcC5wrqSeJK09V5N88z20hmN+QvINtlgiU1tykzvml8BIYKSS66AcSDI+4glJPSLi87XtIz1+fdRU1/xpoUtIxlwU2qSex8qX2/9mRdZtTvL+lstHJGN51qaudcrFFRtnVHg+jyNJng/PTxokHVlDHer7/mXNh8A+ktarQ6KSe0/K+V5bBrglxZqTu0laNC4kaUl5oI7/jB8lGXD3TkS8WOSRn6Q8RZLcXw68HxFv5JUfCOzPmq0oa4iIWRFxC0kLzK61xH1G0mVztKRV3zYldSRp9amziPg0Iv5IMnZmI5LxA5B8G4byffs8Pn92haQtSa7FMj4vZibQt2BmyMZpXL761G0iyeDlk/ILJXUneT/+Vrfq18kbwFZlrNObJK1N3ymI60n1cxIk40xW5MW1Jxm70RDl/j0ol8dJWlqG1CH2q+lzKVOYLcOcpFizEREvkrSInEsyI6EuXT0Al5B8Q50g6cy03/8wScPSaZzd82KnkLRyHEySaOQ8BXyN5BvxqiRFUlU6TfQCSYdLGqRkGvGhBdsX81OSb9NPSjpS0jEk/9zWen0OSY9IukrSMZIGSjqZ5Ly8C+TG6ORmQfy3pP6Sdk+ToFJ1Bf6sZNr0CSTnYQlJ11POKJJE6X8kHSzp+DRujYvDpWNv3gX+I43bXTVczTUiPiVJGo9Iu9G+KekkknFGS4CfNeA1FRoP7CRpg9qC6lqntIXgUuDrkv6Y/t6dQPK7MYdkxlnOY0AH4H5JB0k6DniG1UlGqd4hSahOVHL14N0lbd7AfZbD70nO1+2SrpF0aHp+fpa+9nx7kSRv/2j0Wto65STFmpu7AQGvR8TkumwQEXNIZjiMJWmFGUPyz3QoyWyFT/JiV7K6ZSC/xWQiSfKwkuQPa84SkumXJwO/I/l2+H3gGpJxBbXV60mSq+l2IrkOxEiSaap1GUj4d5KporeT/MP7OUmCMygilqf7H0+SQHwLeJZkCvVuddh3TYaTJBa/Ses4B/hG5F04LCKeA04BdiCZXv2TtA7ji+zveyQzZB5O6zaipgNHxFUk5/VrwEPALSRjigbUYeB0fdxP0lo3eG2Bda1TRNwJnJ7G/ZkkabkamErelYAjYhzJ7+ROJANZryAZi3V1Q15Q2to4lGR8zFiSc316Q/ZZDmmX5WEkvx9Hkfy+/BbYh+T3LN+RJIN263PlZGsClDcY38zM1kLSI8D6EfHNdXiMziRT2R+KiIonDFkmaSuS1sFD0sTemhEnKWZm9SBpR5JWjgER8cLa4uuwv24kg7b/TjKQdkuSi7ptC+weEa/VsnmLJ+k3QPeIOGitwdbkeHaPmVk9RMSrkk6lDrOs6mgpyWDm20jG63xOciG/M5yg1C69ntE7rDnuyZoRt6SYmZlZJnngrJmZmWWSkxQzMzPLJCcpZmZmlkkeOFtH6ZU0Nye51LqZmZnVT0eS+zHVeTCsk5S62xx4v9KVMDMza8K6Ax/UNdhJSt0tBnjvvffo1KmUu96bmZm1TIsWLaJHjx5Qz94IJyn11KlTJycpZmZmjcADZ83MzCyTnKSYmZlZJjlJMTMzs0zymBQzM6tRRPDll1+yYsWKSlfFMqxVq1asv/76JFfrKB8nKWZmVtSyZcuYM2cOn3/+eaWrYk3ABhtswGabbUabNm3Ktk8nKWZmVs3KlSuZMWMGrVq1YvPNN6dNmzZl/5ZszUNEsGzZMj788ENmzJhBnz59WG+98owmcZJiZmbVLFu2jJUrV9KjRw822GCDSlfHMq59+/a0bt2ad999l2XLltGuXbuy7NcDZ83MrEbl+kZszd+6+F3xb5+ZmZllkpMUMzMzyyQnKWZm1uJI4qGHHgJg5syZSGLatGl12nbIkCEceeSRNa4fMWIE/fr1K2sdWyoPnDUza6ZGjKh9uWw7XpdKPNbcuXO54ooreOyxx/jggw/o2rUr/fr149xzz+WAAw5YI7ZHjx7MmTOHTTbZpAwVLp85c+bQpUsXIEmkevfuzdSpU8uSADUVTlLMzKxZmTlzJnvvvTedO3fmF7/4BTvvvDPLly/niSee4KyzzuKNN95YI75Vq1Z069atQrWtWRbr1Njc3WNmZs3KsGHDkMTzzz/Pt7/9bfr27csOO+zAeeedx6RJk6rFF+vuee211xg8eDCdOnWiY8eO7LvvvrzzzjtFjzdlyhS6du3KFVdcsUb5HXfcsWoK93/+53/y6aefrlr3wgsvcNBBB7HJJptQVVXFoEGDeOmll9bYPr+7p3fv3gDssssuSGK//fYr6dw0NU5SzMys2ViwYAFjxozhrLPOYsMNN6y2vnPnzmvdxwcffMDAgQNp164d48aNY8qUKQwdOpQvv/yyWuz48eM54IAD+NnPfsaPf/zjVeX/+te/GD16NI888ghjxoxh2rRpnHXWWavWL168mFNOOYVnnnmGSZMm0adPHw477DAWL15ctE7PP/88AE899RRz5szhwQcfXOvraA7c3WNmZs3Gv/71LyKCbbfdtuR93HrrrVRVVfHAAw/QunVrAPr27Vst7i9/+Qsnn3wyd9xxB8cff/wa65YsWcJ9991H9+7dAbj55psZPHgw1113Hd26dWP//fdfI/6OO+6gS5cuPP300xx++OHVjrXpppsCsPHGG7eobiC3pJiZWbMREQANuoT/tGnT2HfffVclKMVMnjyZY445hvvuu69aggLQs2fPVQkKQP/+/Vm5ciVvvvkmAPPnz+eMM86gb9++VFVVUVVVxb///W9mzZpVcr2bo4omKZIGSnpE0mxJIananC5J20l6WNJCSYslTZLUM299W0k3S/pI0mdpbPeCffRMj/NZGneTpPLdAcnMzDKhT58+SGL69Okl76N9+/Zrjdlqq63Ydtttueeee1i2bNla43NJU+55yJAhTJkyhRtuuIEJEyYwbdo0Nt544zrtqyWpdEvKhsDLwNnFVkraCngWeAPYD/gacDmwJC/sBuAo4DhgH6AD8KikVuk+WgGPpcfaJ407Briu7K/GzMwqaqONNuKQQw7h1ltv5bPPPqu2Pn/wak123nlnnnnmGZYvX15jzCabbMK4ceN45513OPbYY6vFzpo1i9mzZ69anjhxIuutt96qbqNnnnmG//qv/+Kwww5jhx12oG3btnz00Uc1Hi93Z+EVK1astf7NSUWTlIh4PCJ+EhE1jQC6AvhrRFwUEVMj4v8i4rGImA8gqQr4HnB+RDwVEVOBk4CdgAPTfRwMbA+clO7jKeB84DRJndbl6zMzs8Z32223sWLFCvbcc0/+9Kc/8fbbbzN9+nRuuukm+vfvv9btzz77bBYtWsRxxx3Hiy++yNtvv82oUaNWddXkdO3alXHjxvHGG29w/PHHrzGwtl27dpxyyim8/PLLqxKS73znO6vGk2y99daMGjWK6dOnM3nyZE488cRaW3C6du1K+/btGTNmDPPmzWPhwoUlnp2mJbMDZyWtBwwGfiHpCWAXYAZwVUTkLsG3G9AaGJvbLiJmS3oVGAA8AfQHXo2I2Xm7fwJom27/9xqO3zaNyelYjtdlZtbkNebF3ErQu3dvXnrpJa644grOP/985syZw6abbspuu+3Gr371q7Vuv/HGGzNu3DguvPBCBg0aRKtWrejXrx977713tdhu3boxbtw49ttvP0488UTuv/9+IElCjj76aA477DAWLFjAYYcdxm233bZqu3vuuYfTTz+dXXbZhZ49e3LllVdywQUX1Fin9ddfn5tuuonLLruMSy65hH333Zfx48fX/+Q0McoNMqo0SQEclUtAJHUD5gCfAz8hSSYOBa4EvhERT0s6AfhNRLQt2NdYYEZE/EDSnUCviDi4IGYpMCQifl9DfUYAlxaWL1y4kE6d3ABjZuW1Lq4O25B9LlmyhBkzZtC7d2/atWvX8MpYs1fb78yiRYuoqqoCqIqIRXXdZ6XHpNQmV7e/RMT1ETEtIq4GHgXOWMu2AvKzr2KZWGFMoauAqrxH91pizczMrMyynKR8BHwJvF5QPh3Ize6ZC7SR1KUgpiswLy9mjUnlaXzrvJhqImJpRCzKPYDiV9gxMzOzdSKzSUpELANeALYpWNUXeDf9eQqwHDgot1LSZsCOwIS0aCKwY1qeczCwNN3ezMzMMqiiA2cldQC2zivqLakfsCAiZgG/BP5X0j9YPSblWyTTkYmIhZLuBq6T9DGwALgWeAV4Kt3nWJLWmFGSLgQ2SmPuqk+/mJmZmTWuSs/u2Z01Z9eMTJ/vIxnU+mdJZwAXAzcBbwLHRMSzedv8kKRbaDTQHvhbuu0KgIhYIWkwcBvwHPAFcD9Q8zBqMzMzq7iKJikRMZ5kAGttMfcA99SyfglwTvqoKWYWUP1mCGZmZpZZlW5JMTOzeig2jTjjly0xK1lmB86amZlZy+YkxczMzDLJ3T1mZlYvjdm91Ny6soYMGcKnn37KQw8ld3fZb7/96NevHzfccMNatx0/fjzf+MY3+OSTT+jcuXO19TNnzqR3795MnTqVfv36la2OleSWFDMza1aGDBmCJK6++uo1yh966CGkWudqNLoHH3yQyy+/vNLVWMONN97Ivffeu2p5v/3249xzz61IXZykmJlZs9OuXTuuueYaPvnkk0pXpVYbbbQRHTtm6/61VVVVRVtqKsFJipmZNTsHHngg3bp146qrrqox5k9/+hM77LADbdu2pVevXlx33XVrrO/VqxdXXnklQ4cOpWPHjvTs2ZM777xzrcd+7bXXGDx4MJ06daJjx47su+++vPPOO0VjC1spli5dykUXXUSPHj1o27Ytffr04e677y667RdffMHgwYPZa6+9WLBgwaryN954gwEDBtCuXTt22GGHNe6WvGLFCr73ve/Ru3dv2rdvzzbbbMONN964xn6HDBnCkUceuernp59+mhtvvBFJSGLmzJlrPQfl4iTFzMyanVatWnHllVdy88038/7771dbP2XKFL7zne9w3HHH8corrzBixAh++tOfrtHNAXDdddex++67M3XqVIYNG8aZZ57JG2+8UeNxP/jgAwYOHEi7du0YN24cU6ZMYejQoXz55Zd1qvd3v/tdHnjgAW666SamT5/O7bffTocOHarFLVy4kIMPPphly5bxt7/9jY022mjVugsvvJDzzz+fqVOnMmDAAI444gg+/vhjAFauXEn37t0ZPXo0r7/+OpdccgnDhw9n9OjRRetz44030r9/f0477TTmzJnDnDlz6NGjR51eSzl44KyZmTVLRx11FP369ePSSy+t1hoxcuRIDjjgAH76058C0LdvX15//XV++ctfMmTIkFVxhx12GMOGDQPgRz/6Eddffz3jx49n2223LXrMW2+9laqqKh544AFat269at918dZbbzF69GiefPJJDjzwQAC++tWvVoubN28exx57LFtttRW///3vadOmzRrrzz77bI455hgAfvWrXzFmzBjuvvtuLrroIlq3bs3PfvazVbG9e/dmwoQJjB49mu985zvVjlVVVUWbNm3YYIMN6NatW7X165pbUszMrNm65ppruO+++3j99dfXKJ8+fTp77733GmV77703b7/9NitWrFhVtvPOO6/6WRLdunVj/vz5AHzzm9+kQ4cOdOjQgR122AGAadOmse+++65KUOpj2rRptGrVikGDBtUad+CBB/LVr36V0aNHV0tQAPr377/q5/XXX5/dd9+d6dOnryq7/fbb2X333dl0003p0KEDd911F7Nmzap3fRuDkxQzM2u2Bg4cyCGHHMLw4cPXKI+IajN9IqLa9oXJhiRWrlwJwK9//WumTZvGtGnT+Otf/wpA+/btS65rXbcdPHgwzzzzTLXEqza51zp69Gh++MMfMnToUMaOHcu0adM49dRTWbZsWUl1Xtfc3WNmZs3a1VdfTb9+/dbodtl+++159tln14ibMGECffv2pVWrVnXa7xZbbFGtbOedd+a+++5j+fLl9W5N2WmnnVi5ciVPP/30qu6eYq6++mo6dOjAAQccwPjx49l+++3XWD9p0iQGDhwIwJdffsmUKVM4++yzAXjmmWcYMGDAqi4soMZBvTlt2rRZo3WpMTlJMTNbRwovRNbcLkzWVOy0006ceOKJ3HzzzavKzj//fPbYYw8uv/xyjj32WCZOnMgtt9zCbbfd1qBjnX322dx8880cd9xxXHzxxVRVVTFp0iT23HNPttlmm1q37dWrF6eccgpDhw7lpptu4mtf+xrvvvsu8+fPrzZe5Nprr2XFihXsv//+1cbI3HrrrfTp04ftttuO66+/nk8++YShQ4cCsPXWW/Pb3/6WJ554gt69ezNq1CheeOEFevfuXWu9Jk+ezMyZM+nQoQMbbbQR663XOB0xTlLMzKxemmKydfnll68xg2XXXXdl9OjRXHLJJVx++eVsttlmXHbZZWsMmi3FxhtvzLhx47jwwgsZNGgQrVq1ol+/ftXGv9TkV7/6FcOHD2fYsGF8/PHH9OzZs1pXVc7111+/RqKSG59y9dVXc8011zB16lS22mor/vKXv7DJJpsAcMYZZzBt2jSOPfZYJHH88cczbNgwHn/88RrrdMEFF3DKKaew/fbb88UXXzBjxgx69epVvxNTIhXrg7PqJHUCFi5cuJBOnTpVujpm1gTUpyWlrrH1uQtyQ1pylixZwowZM+jduzft2rWr+4bWYtX2O7No0SKqqqoAqiJiUV336YGzZmZmlklOUszMzCyTnKSYmZlZJjlJMTMzs0xykmJmZjXy5Aqrq3Xxu+IkxczMqsldiOzzzz+vcE2sqcj9rpRyS4Ca+DopZmZWTatWrejcufOq+9RssMEG1S4jbwZJC8rnn3/O/Pnz6dy5c52v2FsXTlLMzKyo3F1vc4mKWW06d+5c9jslO0kxM7OiJLHZZpvRtWtXli9fXunqWIa1bt26rC0oORVNUiQNBC4EdgM2A46KiIdqiL0DOB34YUTckFfeBbgJOCItehg4JyI+zYvZCbgF2BNYANwBXB4eEWZmtlatWrVaJ/+AzNam0gNnNwReBs6uLUjSkcDXgdlFVt8P9AMOTR/9gFF523YCnky33QM4B7gAOK/h1TczM7N1paItKRHxOPA4UOOALElbkLSCHAI8VrBuO5LEZK+ImJyWnQZMlLRNRLwJnAi0A4ZExFLgVUl9gfMkjXRripmZWTZVuiWlVpLWI2kV+WVEvFYkpD+wMJegAETEJGAhMCAv5uk0Qcl5Atgc6FXLsdtK6pR7AB0b9GLMzMysXjKdpAA/Ar4kGXNSTDeg2LDz+em6XMy8gvXz8tbV5GKSZCf3eL8O9TUzM7MyyWySImk34L9Jumlq65Iptk4F5YUxqqE831VAVd6je60VNjMzs7LKbJIC7At0BWZJ+lLSl8CWwHWSZqYxc4GvFNl2U1a3lsyleotJ1/S5sIVllYhYGhGLcg9gcWkvw8zMzEqR5SRlFLAzyWyd3GM28EuSQbQAE4EqSXvmNpL0dZKWjwl5MQMltcnb98Hpvmauw/qbmZlZA1T6OikdgK3zinpL6gcsiIhZwMcF8cuBuemsHSJiuqQxwF2SfpCG3Qk8moshmaJ8KXCvpCuBPsBw4DLP7DEzM8uuSrek7A5MTR8AI9OfL6vHPk4EXgHGpo9/AifnVkbEQuAgkjElLwK3pccZ2cC6m5mZ2TpU6eukjGf1INa6xPcqUrYAOGkt270CDKxn9czMzKyCKt2SYmZmZlaUkxQzMzPLJCcpZmZmlklOUszMzCyTnKSYmZlZJjlJMTMzs0yq6BRkMzPLhhEjal82qwS3pJiZmVkmOUkxMzOzTHKSYmZmZpnkJMXMzMwyyUmKmZmZZZKTFDMzM8skJylmZmaWSU5SzMzMLJOcpJiZmVkmOUkxMzOzTHKSYmZmZpnkJMXMzMwyyUmKmZmZZZKTFDMzM8skJylmZmaWSU5SzMzMLJOcpJiZmVkmVTRJkTRQ0iOSZksKSUfmrWst6RpJr0j6LI35raTNC/bRRdIoSQvTxyhJnQtidpL0tKQvJH0g6RJJaqzXaWZmZvVX6ZaUDYGXgbOLrNsA2BW4PH0+GugLPFwQdz/QDzg0ffQDRuVWSuoEPAnMBvYAzgEuAM4r4+swMzOzMlu/kgePiMeBxwEKGzYiYiFwUH6ZpHOA5yX1jIhZkrYjSUz2iojJacxpwERJ20TEm8CJQDtgSEQsBV6V1Bc4T9LIiIh1+yrNzMysFJVuSamvKiCAT9Pl/sDCXIICEBGTgIXAgLyYp9MEJecJYHOgV00HktRWUqfcA+hYtldhZmZma9VkkhRJ7YCrgfsjYlFa3A2YXyR8frouFzOvYP28vHU1uZgk2ck93i+h2mZmZlaiJpGkSGoNPEBS32EFq4t116igvDBGNZRHRh4yAAAfJUlEQVTnu4qk5Sb36F7X+pqZmVnDVXRMSl2kCcpooDewf14rCsBc4CtFNtuU1a0lc6neYtI1fS5sYVkl7R5a1UXkyUBmZmaNK9NJSl6C0gf4RkR8XBAyEaiStGdEPJ9u83WSlo8JeTFXSmoTEcvSsoNJZvvMXMcvwcyaoREjal82s/Ko9HVSOkjqJ6lfWtQ7Xe4paX3gj8DuJDN0Wknqlj7aAETEdGAMcJekvSTtBdwFPJrO7IFkivJS4F5JO0o6ChgOeGaPmZlZhlW6JWV34O95yyPT5/uAEcAR6fK0gu2+AYxPfz4RuAkYmy4/TN51VyJioaSDgFuBF4FP0uPkjmVmZmYZVOnrpIxn9SDWYtY6ECQiFgAnrSXmFWBgvSpnZmZmFdUkZveYmZlZy+MkxczMzDLJSYqZmZllkpMUMzMzyyQnKWZmZpZJTlLMzMwsk5ykmJmZWSY5STEzM7NMcpJiZmZmmeQkxczMzDLJSYqZmZllkpMUMzMzyyQnKWZmZpZJTlLMzMwsk5ykmJmZWSY5STEzM7NMcpJiZmZmmeQkxczMzDLJSYqZmZll0vqVroCZWdmMGFH7spk1KW5JMTMzs0xykmJmZmaZ5O4eMysfd7eYWRlVtCVF0kBJj0iaLSkkHVmwXpJGpOu/kDRe0g4FMV0kjZK0MH2MktS5IGYnSU+n+/hA0iWS1Biv0czMzEpTUpIi6RRJg/OWfyHpU0kTJG1Zj11tCLwMnF3D+ouA89L1ewBzgScldcyLuR/oBxyaPvoBo/Lq1gl4Epid7uMc4IJ0v2ZmZpZRpbakDAe+AJDUnySJuAj4CLi+rjuJiMcj4icR8WDhurSl41zgioh4MCJeBU4BNgBOSGO2I0lMvh8REyNiInAacLikbdJdnQi0A4ZExKvpsa4EznNripmZWXaVOialB/Cv9OcjgT9GxJ2SngPGl6NiQG+gGzA2VxARSyU9DQwA7gD6AwsjYnJezCRJC9OYN9OYpyNiad6+nwCuAnoBM4odXFJboG1eUcdicWZWoqYyfqVovYqVmVm5lZqk/BvYGJgFHMzq1pMlQPsy1AuSBAVgXkH5PGDLvJj5Rbadn7d9N2BmkX3k1hVNUoCLgUvrWFczW1eKJQlZTWjMrKxK7e55Evi1pF8DfYHH0vIdqJ4QNFQULKugrHB9XWJUQ3m+q4CqvEf3tdbUzMzMyqbUlpSzgJ+TdPscExEfp+W7Ab8vR8VIBslC0toxJ6+8K6tbQuYCXymy7aYFMd0K1ndNnwtbaVZJu4dWdRF5+IpZM9NUupvMWrCSkpSI+JQiM3IiopzdIzNIEoyDgKkAktoAg4AfpTETgSpJe0bE82nM10laPibkxVwpqU1ELEvLDiaZ7TOzjPU1MzOzMir5OimS9pX0P+m04y3SspMl7VOPfXSQ1E9Sv7Sod7rcMyICuAEYLukoSTsC9wKfk0w7JiKmA2OAuyTtJWkv4C7g0Yh4M93n/SQtIvdK2lHSUSSzk0amxzAzM7MMKvU6KceQzJD5AtiV1bNgOpIkAHW1O0krydR0eWT682Xp8i9IEpXbgBeBLYCDI2Jx3j5OBF4hmQU0FvgncHJuZUQsJGmN6Z7u47b0OCPrUU8zMzNrZKWOSfkJcEZE/FbScXnlE4BL6rqTiBjP6kGsxdYHyVy/EbXELABOWstxXgEG1rVeZmZmVnmldvdsA/yjSPkioHORcjMzM7N6KTVJmQNsXaR8H+D/Sq+OmZmZWaLU7p47gBslDSW51sjm6eXxr2X1eBIzM2tmfG09a0ylTkH+haQq4O8k98X5B8kMmmsj4pYy1s/MzMxaqFJbUoiIH0u6AtiepNvo9Yj4d9lqZmbZ4QuflV+1c1i4bGYlJykAEfE5ybReMzMzs7Kqc5Ii6cG6xkbE0aVVx8ysCXPriFlZ1aclZeE6q4WZmZlZgTonKRFx6rqsiJmZmVm+Bo1JkdSV5MJuAbwVEfPLUiszMzNr8Uq9d08nSaOAD4CnSaYgf5DecLCqnBU0MzOzlqnUK87+Gvg6cDjJZfCr0p93J7kLsZmZmVmDlNrdMxg4JCKezSt7QtJpwJiGV8vMzMxaulJbUj6m+GyfhcAnpVfHzMzMLFFqkvJzYKSkzXIFkroBvwQuL0fFzMzMrGUrtbvnTJK7IL8raVZa1pPk/j2bSvpBLjAidm1YFc3MzKwlKjVJeaistTAzMzMrUOpdkH9W7oqYmZmZ5WvQxdwAJHWgYGxLRCxq6H7NzMysZSspSZHUG7gF2A9ol7+K5OqzrRpcMzMz800LrUUrtSXld+nzUGAeSWJiZmZmVjalJik7A7tFxJvlrIyZmZlZTqnXSXkB6FHOipiZmZnlK7Ul5fvA7ZK2AF4FluevjIh/NrRiZmZm1rKV2pKyKbAV8BuSVpVpwNS857KQtL6kn0uaIekLSf8n6RJJ6+XFSNIISbPTmPGSdijYTxdJoyQtTB+jJHUuVz3NzMys/EptSbmHJBk5nnU7cPZHwBnAKcBrJHdZ/g3JPYJuTGMuAs4DhgBvAT8BnpS0TUQsTmPuB7oDh6bLdwKjgG+to3qbmZlZA5WapGwJHBER/ypnZYroD/wlIh5Ll2dKOp4kWUGSgHOBKyLiwbTsFJLE6QTgDknbkSQne0XE5DTmNGBimsh48K+ZmVkGlZqkjAO+BqzrJOVZ4AxJfSPiLUlfA/YhSUwAegPdgLG5DSJiqaSngQHAHSSJzsJcgpLGTJK0MI0pmqRIagu0zSvqWL6XZZYBhdffqHY9DjOzyio1SXkEuF7STsArVB84+3BDK5a6BqgC3pC0guQicT+OiN+n67ulz/MKtptH0tqTi5lfZN/z87Yv5mLg0lIqbWZmZg1XapJye/p8SZF15bzi7LHASSRdN68B/YAbJM2OiPsKjplPBWXFxswUxhS6ChiZt9wReL+O9TYzM7MGKvUGg6XOCqqvXwJXR8QD6fIrkrYkaeW4D5iblncD5uRt15XVrStzga8U2femVG+BWSUilgJLc8vJ8BczMzNrLI2VbJRqA2BlQdkKVtd7BkkSclBupaQ2wCBgQlo0EaiStGdezNdJupFyMWZmZpYxJd8FWdKGJMlAT6BN/rqIuKmB9cp5BPixpFkk3T27kEw3vic9Tki6ARgu6W3gbWA48DnJtGMiYrqkMcBdkn6Q7vdO4FHP7DEzM8uuUu+CvAvwV5KWjg2BBcAmJMnBfKBcSco5wOXAbSRdOLNJZuxclhfzC6B9GtMFmAwcnHeNFIAT0zrlZgE9DJxdpjqamZnZOlBqS8r1JK0cZwKfAnuRzPD5H1ZfZK3B0kTjXFZPOS4WEyT3Lh9RS8wCkgG4ZmZm1kSUOialH3BdRKwgGSPSNiLeI7n665XlqpyZmZm1XKW2pCxn9fTdeSTjUqaTXK6+ZxnqZWZm9VXtgnyFy2ZNS6lJylSSS9O/BfwduEzSJsDJJBd3MzNrUnwBXrPsKbW7Zzirr0vyU+Bj4Fck1x75QU0bmZmZmdVVqS0pr5FcsZWI+FDSMOAo4PWImFauypmZmVnLVWpLyl+A7wJI6gxMIrl+yUOSzixT3czMzKwFKzVJ2RV4Jv3526y+od93gf8qQ73MzMyshSs1SdkAyF0s7WDgwYhYSdKismWNW5mZmZnVUalJyr+AIyX1AA5h9ZVcuwKLylExMzMza9lKTVIuA64FZgKTI2JiWn4wyfRkMzMzswYpaXZPRPxR0rPAZsDLeav+Bvy5HBUzMzOzlq3kuyBHxFxgbkHZ8w2ukZmZmRmld/eYmZmZrVNOUszMzCyTnKSYmZlZJjlJMTMzs0xykmJmZmaZ5CTFzMzMMslJipmZmWWSkxQzMzPLJCcpZmZmlklOUszMzCyTnKSYmZlZJjlJMTMzs0wq+QaDjUXSFsA1wDeB9sBbwPciYkq6XsClwOlAF2AycFZEvJa3jy7ATcARadHDwDkR8WljvQ6zRjFiRN3KzMyagEwnKWly8Rzwd5IkZT6wFZCfXFwEnAcMIUlgfgI8KWmbiFicxtwPdAcOTZfvBEYB31rHL8HMLHuKJq7FyswqK9NJCvAj4L2IODWvbGbuh7QV5Vzgioh4MC07BZgHnADcIWk7kuRkr4iYnMacBkxME5k3G+WVmJmZWb1kfUzKEcCLkv4gab6kqWmCkdMb6AaMzRVExFLgaWBAWtQfWJhLUNKYScDCvJhqJLWV1Cn3ADqW7VWZmZnZWmW9JeWrwJnASOBKYE/gJklLI+K3JAkKJC0n+eYBW6Y/dyPpJio0P2/7Yi4mGetiZmYlKOxV8vAoq6+st6SsB7wUEcMjYmpE3AHcRZK45IuCZRWUFa4vFlPoKqAq79G9PhU3MzOzhsl6kjIHeL2gbDrQM/15bvpc2CLSldWtK3OBrxTZ96ZUb4FZJSKWRsSi3ANYXFOsmZmZlV/Wk5TngG0KyvoC76Y/zyBJQg7KrZTUBhgETEiLJgJVkvbMi/k6SetILsbMzMwyJutjUq4HJkgaDowmGZNyevogIkLSDcBwSW8DbwPDgc9Jph0TEdMljQHukvSDdL93Ao96Zo+ZmVl2ZTpJiYgXJB1FMj7kEpKWk3Mj4nd5Yb8gucjbbay+mNvBeddIATiR5GJuuVlADwNnr+Pqm5mZWQNkOkkBiIhHgUdrWR8kVyEaUUvMAuCkctfNzMzM1p2sj0kxMzOzFspJipmZmWWSkxQzMzPLJCcpZmZmlklOUszMzCyTMj+7x8ysVMXuFeP7x5g1HW5JMTMzs0xykmJmZmaZ5CTFzMzMMslJipmZmWWSkxQzMzPLJCcpZmZmlklOUszMzCyTnKSYmZlZJjlJMTMzs0xykmJmZmaZ5CTFzMzMMslJipmZmWWSkxQzMzPLJCcpZmZmlklOUszMzCyTnKSYmZlZJq1f6QqYWR2MGFH7cgvj02HWMjSplhRJF0sKSTfklbWVdLOkjyR9JulhSd0Ltusp6ZF0/UeSbpLUpvFfgZnVZMSINR9mZk0mSZG0B3A68M+CVTcARwHHAfsAHYBHJbVKt2sFPAZsmK4/DjgGuK5xam5mZmalaBJJiqQOwO+A04BP8sqrgO8B50fEUxExFTgJ2Ak4MA07GNgeOCkipkbEU8D5wGmSOjXiyzAzM7N6aBJJCnAr8FiaYOTbDWgNjM0VRMRs4FVgQFrUH3g1Lc95Amibbl9U2o3UKfcAOjb8ZZiZmVldZX7grKTjgF2BPYqs7gYsi4hPCsrnpetyMfPyV0bEJ5KW5cUUczFwaUmVNjMzswbLdJIiqQdwI3BwRCypz6ZA5C1HHWIKXQWMzFvuCLxfjzqYtXjFBsB6UKyZ1VWmkxSS7piuwBRJubJWwEBJZwOHAG0kdSloTekKTEh/ngt8PX+nkrqQdBOt0cKSLyKWAkvztmnYKzEzsxp5WrkVk/UxKX8jGQTbL+/xIskg2tzPy4GDchtI2gzYkdVJykRgx7Q852CSBGTKOq6/mZmZlSjTLSkRsZhkEOwqkj4DPo6IV9Plu4HrJH0MLACuBV4BcoNsxwKvA6MkXQhslMbcFRGLGuWFmJmZWb1lOkmpox8CXwKjgfYkrS9DImIFQESskDQYuA14DvgCuB+4oDLVNTMzs7pocklKROxXsLwEOCd91LTNLODwdVszM7NmqNrgkMJls3Un62NSzMzMrIVykmJmZmaZ5CTFzMzMMslJipmZmWWSkxQzMzPLJCcpZmZmlklOUszMzCyTnKSYmZlZJjlJMTMzs0xykmJmZmaZ5CTFzMzMMslJipmZmWWSkxQzMzPLJCcpZmZmlklOUszMzCyTnKSYmZlZJq1f6QqYtVgjRtS+bGbWwrklxczMzDLJSYqZmZllkpMUMzMzyySPSTGzevNwGjNrDG5JMTMzs0xykmJmZmaZ5CTFzMzMMinzY1IkXQwcDWwLfAFMAH4UEW/mxbQFrgWOB9oDfwOGRcT7eTE9gVuB/dP93A9cEBHLGumlmJk1b9UGJxUum9VPU2hJGUSSXOwFHESSWI2VtGFezA3AUcBxwD5AB+BRSa0A0ufHgA3T9ccBxwDXNdJrMDMzs3rKfEtKRByavyzpVGA+sBvwD0lVwPeAkyPiqTTmJOA94EDgCeBgYHugR0TMTmPOB+6V9OOIWNRYr8fMzMzqpim0pBSqSp8XpM+7Aa2BsbmANBF5FRiQFvUHXs0lKKkngLbp9tVIaiupU+4BdCzfSzAzM7O1aVJJiiQBI4FnI+LVtLgbsCwiPikIn5euy8XMy1+Zxi/Liyl0MbAw7/F+DXFmZma2DjSpJAW4BdiZZIDs2giIvOWoQ0y+q0habXKP7nWvppmZmTVUk0lSJN0MHAF8I3/WDjAXaCOpS8EmXVndejKXghaTNL41BS0sORGxNCIW5R7A4jK8DDMzM6ujzCcpStxCMg15/4iYURAyBVhOMvMnt81mwI4k05UBJgI7puU5BwNL0+3NzMwsYzI/u4dk+vEJwH8AiyXlWkQWRsQXEbFQ0t3AdZI+JhlQey3wCvBUGjsWeB0YJelCYKM05i7P7DEzM8umzLekAGeSjAkZD8zJexybF/ND4CFgNPAc8DnwrYhYAZA+DwaWpOtHp/EXNMorMDMzs3rLfEtKRKgOMUuAc9JHTTGzgMPLWDUzMzNbhzKfpJg1KdUuC15DmZmVpPDj5I9X89YUunvMzMysBXJLilkz52+eZtZUuSXFzMzMMslJipmZmWWSu3vM6sJ9JmZmjc4tKWZmZpZJbkkxa4LcsGNmLYFbUszMzCyT3JJiZoBbZ8wse9ySYmZmZpnklhQzM2tcRZvpipVZS+ckxczMmiV3YTZ97u4xMzOzTHKSYmZmZpnk7h6zjCjWFO3maTNryZykWMvlrMDMLNPc3WNmZmaZ5CTFzMzMMsndPdb8ZGzeYcaqY2bWZLglxczMzDLJSYqZmZllkrt7rGlwn4mZrSOe6JddTlLMSuCcyayRVPtwFS5bc9aiunskDZM0Q9ISSVMk7VvpOpmZmVlxLaYlRdKxwA3AMOA54AfA45K2j4hZFa2cZYZbSMzMsqPFJCnAecDdEfHrdPlcSYcAZwIXV65aLZg7gs2snBqha8hfZBpXi0hSJLUBdgOuLlg1FhhQwzZtgbZ5RR0BFi1aVN7KXXXVmssXZzRfqk896xq7dGn1sprOb0HsVRdXj1t1mLXE1hQHsJQ1Y1dVpyC2rnEN2Welj7/GW1Hi8bP2mip9fJ/T8h+/pHNa8DdqaZHvqXX8U1RjXJHD1PinsDCuttimqtT/nYqIMlcleyRtDnwA7B0RE/LKhwOnRMQ2RbYZAVzaaJU0MzNr/rpHxAd1DW4RLSl5CjMyFSnLuQoYWVC2EbCgTHXpCLwPdAcWl2mfVnc+/5Xl8195fg8qqyWe/47A7Pps0FKSlI+AFUC3gvKuwLxiG0TEUqCwTbJsfT2Scj8ujogy9yHZ2vj8V5bPf+X5PaisFnr+6/06W8QU5IhYBkwBDipYdRAwofoWZmZmVmktpSUFkq6bUZJeBCYCpwM9gdsrWiszMzMrqsUkKRHxv5I2Bi4BNgNeBQ6LiHcrVKWlwM+o3qVkjcPnv7J8/ivP70Fl+fzXQYuY3WNmZmZNT4sYk2JmZmZNj5MUMzMzyyQnKWZmZpZJTlLMzMwsk5ykVICkYZJmSFoiaYqkfStdp5ZC0ghJUfCYW+l6NVeSBkp6RNLs9FwfWbBe6XsyW9IXksZL2qFS9W1u6nD+7y3yeZhUqfo2N5IulvSCpMWS5kt6SNI2BTFtJd0s6SNJn0l6WFL3StU5a5ykNDJJxwI3AFcAuwDPAI9L6lnRirUsr5FMQ889dqpsdZq1DYGXgbNrWH8RyR3Kzwb2AOYCT0rq2DjVa/bWdv4BxrDm5+GwRqhXSzEIuBXYi+TioesDYyVtmBdzA3AUcBywD9ABeFRSq0auayZ5CnIjkzQZeCkizswrmw48FBHN7L6X2ZPeOPLIiOhX6bq0NJICOCoiHkqXRXIfjxsi4pq0rC3JrSp+FBF3VKyyzVDh+U/L7gU6R8SRNW5oZSNpU2A+MCgi/iGpCvgQODki/jeN2Rx4j+Q6Xk9UrrbZ4JaURiSpDbAbMLZg1VhgQOPXqMXqkzZ/z5D0gKSvVrpCLVRvkvtprfo8pPfMehp/HhrTfmlXxFuS7pLUtdIVasaq0ufcjWp3A1qz5mdgNsnFRv0ZwElKY9sEaEX1mxrOo/rND23dmAx8FzgEOI3kvE9Ir0ZsjSv3O+/PQ+U8DpwI7A+cT9LlNi5t0bIySlsORwLPRsSraXE3YFlEfFIQ7s9AqsVcFj9jCvvYVKTM1oGIeDxv8RVJE4F3gFNI/oBY4/PnoUJyXQypV9N7m70LDAYerEytmq1bgJ1Jxp2sjT8DKbekNK6PgBVUz5C7Uv3bpDWCiPgMeAXoU+m6tEC5WVX+PGRERMwhSVL8eSgjSTcDRwDfiIj381bNBdpI6lKwiT8DKScpjSgilgFTSEZ55zsImND4NbK0WXs7YE6l69ICzSD5I73q85CO2xqEPw8VkXZ79sCfh7JIp9jfAhwN7B8RMwpCpgDLWfMzsBmwI/4MAO7uqYSRwKi0WXUicDrQE7i9orVqISRdCzwCzCL5tvIToBNwXyXr1VxJ6gBsnVfUW1I/YEFEzJJ0AzBc0tvA28Bw4HPg/savbfNT2/lPHyOAP5EkJb2AK0lafP/cqBVtvm4FTgD+A1gsKddquDAivoiIhZLuBq6T9DHJe3ItSevuUxWpccZ4CnIFSBpGcn2IzUhGcf8wIv5R2Vq1DJIeAAaSDGL+EJgE/DQiXq9oxZopSfsBfy+y6r6IGJIOJrwU+AHQhWRg81l5AwutAWo7/8CZwEMk12vqTJKo/J3k8/BeY9WxOUunfRdzakTcm8a0A35Jksy0B/4GDPN7kHCSYmZmZpnkMSlmZmaWSU5SzMzMLJOcpJiZmVkmOUkxMzOzTHKSYmZmZpnkJMXMzMwyyUmKmZmZZZKTFDMzM8skJylm1iCSxqeXty/3fv8h6YRK16OOx35B0tGVOLZZc+Z795hZ5kg6nOTuyA/UY7OjSW7WVgmXA9dKeigiVlaoDmbNjltSzCyL/gv4TX3+4UfEgohYvA7rVJvHgCrgkAod36xZcpJiZmUjqYuk30r6RNLnkh6X1Kcg5jRJ76Xr/yzpPEmf5q3fBDgQeDiv7PfpzSHz99Na0keSTk2X1+jukdRG0i8kfSDpM0mT0xvuocSHko7Ji58maX7ecn9Jy9M7CSNphKRZkpZKmi3pplxsRKwA/goc38BTaGZ5nKSYWTndC+wOHAH0BwT8VVJrAEl7A7cDNwL9gCeBHxfsYx/gc2B6XtnvgCNyCUPqEGBD4E811OU3wN7AccDOwB+AMZL6RHJn1X8A+6X16gJsD7SWtH26/X7AlIj4t6RvAz8kuVtzH+BI4JWC4z0P7FtDXcysBB6TYmZlkbaYHAHsHRET0rITgfdI/qn/ATgHeDwirk03e0vSAODwvF31AuYVdPU8AXwGHAWMSstOAB6JiEVF6rIVSatG94iYnRZfK+lQ4FRgODAeOD1dNxB4GZhFkpy8nj6PT9f3BOYCT0XE8jTu+YLDfgD0lLSex6WYlYdbUsysXLYDvgQm5woi4mPgzXQdwDZU/+deuNweWJJfkCYGfwBOBJC0IfAfJC0sxexK0orzlqR/5x7AIGCrNGY8sEPavTQoXR4PDJK0PjAAeDqN/UNar/+TdJeko9KYfF+Q/E1tW0OdzKye3JJiZuWiWsqjyM81bfcR0KXIfn4H/7+du3eNKgijOPw7iqWIClGwUvCjiFjEaGNIoQEN2NhYiKgEP7AUxULwH1CxCAYhoBDUQqsEQbAxa4xWJhBUREEIwUJFFD9BkNdiJmQILpvAktziPM3dHebeO7vFcnbmnUtNUgvQRQoyD+rccwnwF2jLx9KPfHwBfCYFlE7gImnW5wLQTgolTwAiYkrS5nzfPUAfcE5SZw5QAKuAXxHxu86YzGyeHFLMrFlekX5TdgLTyz2rgU3M1Je8BnbMOm/7rPfjwFpJKyPiS9H+lBQiDgL7gHsR8afOWMaBpUBLRIz8r0NEhKTHpBmZVmAE+A4sA04BY+VuoRw+hoAhSdfyZ9kKjOUurcVrM2sCL/eYWVNExFtgEOiXtEvSNuAWqVZjMHfrBbrzjp6Nkk6SAkc5uzIOfCIVvZbXD+AOKUB05WvXG8sb0szLgKQDktZLapd0XlJ30XWYVNsyERHfioLaQ8zUoyDpqKQeSa2SNgCHScs7k8W1OoCHDb8oM5szhxQza6ZjwHPgPvCMtJTTPb0kEhGjpJBxhlSouhe4SlGDkrfz3iDXn8xym7QL5z0wOoexDABXSHUxQ6RZnqmizyPSjMtw0VbLbbWi7StwPN9zAtgN7M81N0haR6phudlgTGY2D0p/HMzMFoekfmBLRHQUbWuAl0BbREzWPbkiJF0CVkTEiYadzWzOXJNiZgtK0lnS81F+kpZ6jgCnyz4R8UFSD2nrb+VDCvARuNywl5nNi2dSzGxBSbpLegbJcuAd0BsR1xd1UGZWSQ4pZmZmVkkunDUzM7NKckgxMzOzSnJIMTMzs0pySDEzM7NKckgxMzOzSnJIMTMzs0pySDEzM7NKckgxMzOzSvoHU3DJnkuQ11AAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10a95ef28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4), dpi=100)\n",
    "clickbait_views_log = clickbait_set[\"video_views\"].apply(np.log).replace([-np.inf], 0)\n",
    "non_clickbait_views_log = non_clickbait_set[\"video_views\"].apply(np.log).replace([-np.inf], 0)\n",
    "plt.hist(\n",
    "    [clickbait_views_log, non_clickbait_views_log], \n",
    "    bins=40, \n",
    "    color=[\"red\", \"blue\"], \n",
    "    label=[\"Clickbait\", \"Non-clickbait\"], \n",
    "    alpha=0.5\n",
    ")\n",
    "plt.title(\"Views distribution (logarithmic)\")\n",
    "plt.xlabel(\"log(views)\")\n",
    "plt.ylabel(\"samples\")\n",
    "plt.legend(loc='upper right')\n",
    "plt.savefig(\"views-distribution-log\", bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "Likes distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py:52: FutureWarning: reshape is deprecated and will raise in a subsequent release. Please use .values.reshape(...) instead\n",
      "  return getattr(obj, method)(*args, **kwds)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1079dd1d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4), dpi=100)\n",
    "clickbait_likes_log = clickbait_set[\"video_likes\"].apply(np.log).replace([-np.inf], 0)\n",
    "non_clickbait_likes_log = non_clickbait_set[\"video_likes\"].apply(np.log).replace([-np.inf], 0)\n",
    "plt.hist(\n",
    "    [clickbait_likes_log, non_clickbait_likes_log], \n",
    "    bins=40, \n",
    "    color=[\"red\", \"blue\"], \n",
    "    label=[\"Clickbait\", \"Non-clickbait\"], \n",
    "    alpha=0.5\n",
    ")\n",
    "plt.title(\"Likes distribution (logarithmic)\")\n",
    "plt.xlabel(\"log(likes)\")\n",
    "plt.ylabel(\"samples\")\n",
    "plt.legend(loc='upper right')\n",
    "plt.savefig(\"likes-distribution-log\", bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "Dislikes distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py:52: FutureWarning: reshape is deprecated and will raise in a subsequent release. Please use .values.reshape(...) instead\n",
      "  return getattr(obj, method)(*args, **kwds)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x109624470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4), dpi=100)\n",
    "clickbait_dislikes_log = clickbait_set[\"video_dislikes\"].apply(np.log).replace([-np.inf], 0)\n",
    "non_clickbait_dislikes_log = non_clickbait_set[\"video_dislikes\"].apply(np.log).replace([-np.inf], 0)\n",
    "plt.hist(\n",
    "    [clickbait_dislikes_log, non_clickbait_dislikes_log], \n",
    "    bins=40, \n",
    "    color=[\"red\", \"blue\"], \n",
    "    label=[\"Clickbait\", \"Non-clickbait\"], \n",
    "    alpha=0.5\n",
    ")\n",
    "plt.title(\"Dislikes distribution (logarithmic)\")\n",
    "plt.xlabel(\"log(dislikes)\")\n",
    "plt.ylabel(\"samples\")\n",
    "plt.legend(loc='upper right')\n",
    "plt.savefig(\"dislikes-distribution-log\", bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "Comments distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py:52: FutureWarning: reshape is deprecated and will raise in a subsequent release. Please use .values.reshape(...) instead\n",
      "  return getattr(obj, method)(*args, **kwds)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113a04c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4), dpi=100)\n",
    "clickbait_comments_log = clickbait_set[\"video_comments\"].apply(np.log).replace([-np.inf], 0)\n",
    "non_clickbait_comments_log = non_clickbait_set[\"video_comments\"].apply(np.log).replace([-np.inf], 0)\n",
    "plt.hist(\n",
    "    [clickbait_comments_log, non_clickbait_comments_log], \n",
    "    bins=40, \n",
    "    color=[\"red\", \"blue\"], \n",
    "    label=[\"Clickbait\", \"Non-clickbait\"], \n",
    "    alpha=0.5\n",
    ")\n",
    "plt.title(\"Comments distribution (logarithmic)\")\n",
    "plt.xlabel(\"log(comments)\")\n",
    "plt.ylabel(\"samples\")\n",
    "plt.legend(loc='upper right')\n",
    "plt.savefig(\"comments-distribution-log\", bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "### Feature extraction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "import re\n",
    "import emoji\n",
    "from gensim.parsing.preprocessing import *\n",
    "\n",
    "\n",
    "def tokenize(string):\n",
    "\n",
    "    \"\"\" Tokenizes a string.\n",
    "\n",
    "    Adds a space between numbers and letters, removes punctuation, repeated whitespaces, words\n",
    "    shorter than 2 characters, and stop-words. Returns a list of stems and, eventually, emojis.\n",
    "\n",
    "    @param string: String to tokenize.\n",
    "    @return: A list of stems and emojis.\n",
    "    \"\"\"\n",
    "\n",
    "    # Based on the Ranks NL (Google) stopwords list, but \"how\" and \"will\" are not stripped, and\n",
    "    # words shorter than 2 characters are not checked (since they are stripped):\n",
    "    stop_words = [\n",
    "        \"about\", \"an\", \"are\", \"as\", \"at\", \"be\", \"by\", \"com\", \"for\", \"from\", \"in\", \"is\", \"it\", \"of\",\n",
    "        \"on\", \"or\", \"that\", \"the\", \"this\", \"to\", \"was\", \"what\", \"when\", \"where\", \"who\", \"with\",\n",
    "        \"the\", \"www\"\n",
    "    ]\n",
    "\n",
    "    string = strip_short(\n",
    "        strip_multiple_whitespaces(\n",
    "            strip_punctuation(\n",
    "                split_alphanum(string))),\n",
    "        minsize=2)\n",
    "    # Parse emojis:\n",
    "    emojis = [c for c in string if c in emoji.UNICODE_EMOJI]\n",
    "    # Remove every non-word character and stem each word:\n",
    "    string = stem_text(re.sub(r\"[^\\w\\s,]\", \"\", string))\n",
    "    # List of stems and emojis:\n",
    "    tokens = string.split() + emojis\n",
    "\n",
    "    for stop_word in stop_words:\n",
    "        try:\n",
    "            tokens.remove(stop_word)\n",
    "        except:\n",
    "            pass\n",
    "\n",
    "    return tokens"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "Frequency analysis of clickbait titles tokens (stems):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a160adda0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "\n",
    "\n",
    "countvect = CountVectorizer(\n",
    "    lowercase=False,\n",
    "    tokenizer=tokenize\n",
    ").fit(clickbait_set[\"video_title\"])\n",
    "bag_of_words = countvect.transform(clickbait_set[\"video_title\"])\n",
    "sum_words = bag_of_words.sum(axis=0) \n",
    "words_freq = [(word, sum_words[0, i]) for word, i in countvect.vocabulary_.items()]\n",
    "words_freq = sorted(words_freq, key = lambda x: x[1], reverse=True)\n",
    "\n",
    "labels = list()\n",
    "freqs = list()\n",
    "\n",
    "for word in words_freq[:30]:\n",
    "    labels.append(word[0])\n",
    "    freqs.append(word[1])\n",
    "\n",
    "plt.figure(figsize=(6, 4), dpi=100)\n",
    "plt.bar(range(30), freqs, alpha=0.5)\n",
    "plt.xticks(range(30), labels, rotation=\"vertical\")\n",
    "plt.title(\"Top 30 tokens in clickbait titles\")\n",
    "plt.ylabel(\"occurrences\")\n",
    "plt.savefig(\"top-30-clickbait-tokens\", bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}