w7 - final_statement.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import time\n",
    "import datetime\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import KFold, cross_val_score, train_test_split, GridSearchCV\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data_train = pd.read_csv('features.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(97230, 109)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>match_id</th>\n",
       "      <th>start_time</th>\n",
       "      <th>lobby_type</th>\n",
       "      <th>r1_hero</th>\n",
       "      <th>r1_level</th>\n",
       "      <th>r1_xp</th>\n",
       "      <th>r1_gold</th>\n",
       "      <th>r1_lh</th>\n",
       "      <th>r1_kills</th>\n",
       "      <th>r1_deaths</th>\n",
       "      <th>...</th>\n",
       "      <th>dire_boots_count</th>\n",
       "      <th>dire_ward_observer_count</th>\n",
       "      <th>dire_ward_sentry_count</th>\n",
       "      <th>dire_first_ward_time</th>\n",
       "      <th>duration</th>\n",
       "      <th>radiant_win</th>\n",
       "      <th>tower_status_radiant</th>\n",
       "      <th>tower_status_dire</th>\n",
       "      <th>barracks_status_radiant</th>\n",
       "      <th>barracks_status_dire</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>97225</th>\n",
       "      <td>114402</td>\n",
       "      <td>1450265551</td>\n",
       "      <td>1</td>\n",
       "      <td>47</td>\n",
       "      <td>4</td>\n",
       "      <td>1706</td>\n",
       "      <td>1198</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-24.0</td>\n",
       "      <td>2032</td>\n",
       "      <td>0</td>\n",
       "      <td>1792</td>\n",
       "      <td>1975</td>\n",
       "      <td>48</td>\n",
       "      <td>63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97226</th>\n",
       "      <td>114403</td>\n",
       "      <td>1450277704</td>\n",
       "      <td>0</td>\n",
       "      <td>43</td>\n",
       "      <td>4</td>\n",
       "      <td>1793</td>\n",
       "      <td>1416</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>-17.0</td>\n",
       "      <td>1734</td>\n",
       "      <td>1</td>\n",
       "      <td>2038</td>\n",
       "      <td>6</td>\n",
       "      <td>63</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97227</th>\n",
       "      <td>114404</td>\n",
       "      <td>1450291848</td>\n",
       "      <td>1</td>\n",
       "      <td>98</td>\n",
       "      <td>4</td>\n",
       "      <td>1399</td>\n",
       "      <td>540</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>-15.0</td>\n",
       "      <td>2906</td>\n",
       "      <td>0</td>\n",
       "      <td>1796</td>\n",
       "      <td>1846</td>\n",
       "      <td>51</td>\n",
       "      <td>63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97228</th>\n",
       "      <td>114405</td>\n",
       "      <td>1450292986</td>\n",
       "      <td>1</td>\n",
       "      <td>100</td>\n",
       "      <td>3</td>\n",
       "      <td>1135</td>\n",
       "      <td>766</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>-42.0</td>\n",
       "      <td>951</td>\n",
       "      <td>0</td>\n",
       "      <td>2039</td>\n",
       "      <td>2047</td>\n",
       "      <td>63</td>\n",
       "      <td>63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97229</th>\n",
       "      <td>114406</td>\n",
       "      <td>1450313370</td>\n",
       "      <td>7</td>\n",
       "      <td>50</td>\n",
       "      <td>3</td>\n",
       "      <td>1053</td>\n",
       "      <td>799</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>-39.0</td>\n",
       "      <td>2086</td>\n",
       "      <td>1</td>\n",
       "      <td>1982</td>\n",
       "      <td>0</td>\n",
       "      <td>63</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       match_id  start_time  lobby_type  r1_hero  r1_level  r1_xp  r1_gold  \\\n",
       "97225    114402  1450265551           1       47         4   1706     1198   \n",
       "97226    114403  1450277704           0       43         4   1793     1416   \n",
       "97227    114404  1450291848           1       98         4   1399      540   \n",
       "97228    114405  1450292986           1      100         3   1135      766   \n",
       "97229    114406  1450313370           7       50         3   1053      799   \n",
       "\n",
       "       r1_lh  r1_kills  r1_deaths          ...           dire_boots_count  \\\n",
       "97225     17         0          1          ...                          4   \n",
       "97226     17         0          1          ...                          3   \n",
       "97227      1         0          0          ...                          1   \n",
       "97228      6         0          2          ...                          3   \n",
       "97229      7         0          0          ...                          3   \n",
       "\n",
       "       dire_ward_observer_count  dire_ward_sentry_count  dire_first_ward_time  \\\n",
       "97225                         3                       0                 -24.0   \n",
       "97226                         2                       0                 -17.0   \n",
       "97227                         3                       1                 -15.0   \n",
       "97228                         3                       1                 -42.0   \n",
       "97229                         2                       0                 -39.0   \n",
       "\n",
       "       duration  radiant_win  tower_status_radiant  tower_status_dire  \\\n",
       "97225      2032            0                  1792               1975   \n",
       "97226      1734            1                  2038                  6   \n",
       "97227      2906            0                  1796               1846   \n",
       "97228       951            0                  2039               2047   \n",
       "97229      2086            1                  1982                  0   \n",
       "\n",
       "       barracks_status_radiant  barracks_status_dire  \n",
       "97225                       48                    63  \n",
       "97226                       63                     3  \n",
       "97227                       51                    63  \n",
       "97228                       63                    63  \n",
       "97229                       63                     0  \n",
       "\n",
       "[5 rows x 109 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_train.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>match_id</th>\n",
       "      <th>start_time</th>\n",
       "      <th>lobby_type</th>\n",
       "      <th>r1_hero</th>\n",
       "      <th>r1_level</th>\n",
       "      <th>r1_xp</th>\n",
       "      <th>r1_gold</th>\n",
       "      <th>r1_lh</th>\n",
       "      <th>r1_kills</th>\n",
       "      <th>r1_deaths</th>\n",
       "      <th>...</th>\n",
       "      <th>dire_boots_count</th>\n",
       "      <th>dire_ward_observer_count</th>\n",
       "      <th>dire_ward_sentry_count</th>\n",
       "      <th>dire_first_ward_time</th>\n",
       "      <th>duration</th>\n",
       "      <th>radiant_win</th>\n",
       "      <th>tower_status_radiant</th>\n",
       "      <th>tower_status_dire</th>\n",
       "      <th>barracks_status_radiant</th>\n",
       "      <th>barracks_status_dire</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>97230.000000</td>\n",
       "      <td>9.723000e+04</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>95404.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "      <td>97230.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>57185.416744</td>\n",
       "      <td>1.444232e+09</td>\n",
       "      <td>2.630999</td>\n",
       "      <td>51.517104</td>\n",
       "      <td>3.442672</td>\n",
       "      <td>1233.405801</td>\n",
       "      <td>1147.899702</td>\n",
       "      <td>11.231996</td>\n",
       "      <td>0.357009</td>\n",
       "      <td>0.362285</td>\n",
       "      <td>...</td>\n",
       "      <td>3.349553</td>\n",
       "      <td>2.448339</td>\n",
       "      <td>0.689119</td>\n",
       "      <td>-6.901922</td>\n",
       "      <td>2332.247886</td>\n",
       "      <td>0.518503</td>\n",
       "      <td>1309.227790</td>\n",
       "      <td>1286.310820</td>\n",
       "      <td>40.599095</td>\n",
       "      <td>41.337036</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>33007.123878</td>\n",
       "      <td>5.515393e+06</td>\n",
       "      <td>2.835761</td>\n",
       "      <td>32.564211</td>\n",
       "      <td>1.111741</td>\n",
       "      <td>566.588895</td>\n",
       "      <td>464.111662</td>\n",
       "      <td>9.041620</td>\n",
       "      <td>0.663889</td>\n",
       "      <td>0.626704</td>\n",
       "      <td>...</td>\n",
       "      <td>1.155609</td>\n",
       "      <td>0.813459</td>\n",
       "      <td>0.710122</td>\n",
       "      <td>40.701397</td>\n",
       "      <td>715.806850</td>\n",
       "      <td>0.499660</td>\n",
       "      <td>853.921365</td>\n",
       "      <td>851.009148</td>\n",
       "      <td>27.871645</td>\n",
       "      <td>27.064873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.430199e+09</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-84.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>28545.250000</td>\n",
       "      <td>1.440815e+09</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>767.000000</td>\n",
       "      <td>746.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-31.000000</td>\n",
       "      <td>1818.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>256.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>57160.500000</td>\n",
       "      <td>1.446338e+09</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1175.000000</td>\n",
       "      <td>1113.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-16.000000</td>\n",
       "      <td>2268.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1824.000000</td>\n",
       "      <td>1798.000000</td>\n",
       "      <td>63.000000</td>\n",
       "      <td>60.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>85755.750000</td>\n",
       "      <td>1.448829e+09</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1704.000000</td>\n",
       "      <td>1479.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>2778.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1974.000000</td>\n",
       "      <td>1974.000000</td>\n",
       "      <td>63.000000</td>\n",
       "      <td>63.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>114406.000000</td>\n",
       "      <td>1.450313e+09</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>112.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>3319.000000</td>\n",
       "      <td>4332.000000</td>\n",
       "      <td>47.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>300.000000</td>\n",
       "      <td>8452.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2047.000000</td>\n",
       "      <td>2047.000000</td>\n",
       "      <td>63.000000</td>\n",
       "      <td>63.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            match_id    start_time    lobby_type       r1_hero      r1_level  \\\n",
       "count   97230.000000  9.723000e+04  97230.000000  97230.000000  97230.000000   \n",
       "mean    57185.416744  1.444232e+09      2.630999     51.517104      3.442672   \n",
       "std     33007.123878  5.515393e+06      2.835761     32.564211      1.111741   \n",
       "min         0.000000  1.430199e+09      0.000000      1.000000      0.000000   \n",
       "25%     28545.250000  1.440815e+09      1.000000     22.000000      3.000000   \n",
       "50%     57160.500000  1.446338e+09      1.000000     50.000000      3.000000   \n",
       "75%     85755.750000  1.448829e+09      7.000000     75.000000      4.000000   \n",
       "max    114406.000000  1.450313e+09      7.000000    112.000000      6.000000   \n",
       "\n",
       "              r1_xp       r1_gold         r1_lh      r1_kills     r1_deaths  \\\n",
       "count  97230.000000  97230.000000  97230.000000  97230.000000  97230.000000   \n",
       "mean    1233.405801   1147.899702     11.231996      0.357009      0.362285   \n",
       "std      566.588895    464.111662      9.041620      0.663889      0.626704   \n",
       "min        0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%      767.000000    746.000000      2.000000      0.000000      0.000000   \n",
       "50%     1175.000000   1113.000000     11.000000      0.000000      0.000000   \n",
       "75%     1704.000000   1479.000000     19.000000      1.000000      1.000000   \n",
       "max     3319.000000   4332.000000     47.000000      8.000000      5.000000   \n",
       "\n",
       "               ...           dire_boots_count  dire_ward_observer_count  \\\n",
       "count          ...               97230.000000              97230.000000   \n",
       "mean           ...                   3.349553                  2.448339   \n",
       "std            ...                   1.155609                  0.813459   \n",
       "min            ...                   0.000000                  0.000000   \n",
       "25%            ...                   3.000000                  2.000000   \n",
       "50%            ...                   3.000000                  2.000000   \n",
       "75%            ...                   4.000000                  3.000000   \n",
       "max            ...                   9.000000                  9.000000   \n",
       "\n",
       "       dire_ward_sentry_count  dire_first_ward_time      duration  \\\n",
       "count            97230.000000          95404.000000  97230.000000   \n",
       "mean                 0.689119             -6.901922   2332.247886   \n",
       "std                  0.710122             40.701397    715.806850   \n",
       "min                  0.000000            -84.000000    900.000000   \n",
       "25%                  0.000000            -31.000000   1818.000000   \n",
       "50%                  1.000000            -16.000000   2268.000000   \n",
       "75%                  1.000000              8.000000   2778.000000   \n",
       "max                 13.000000            300.000000   8452.000000   \n",
       "\n",
       "        radiant_win  tower_status_radiant  tower_status_dire  \\\n",
       "count  97230.000000          97230.000000       97230.000000   \n",
       "mean       0.518503           1309.227790        1286.310820   \n",
       "std        0.499660            853.921365         851.009148   \n",
       "min        0.000000              0.000000           0.000000   \n",
       "25%        0.000000             36.000000         256.000000   \n",
       "50%        1.000000           1824.000000        1798.000000   \n",
       "75%        1.000000           1974.000000        1974.000000   \n",
       "max        1.000000           2047.000000        2047.000000   \n",
       "\n",
       "       barracks_status_radiant  barracks_status_dire  \n",
       "count             97230.000000          97230.000000  \n",
       "mean                 40.599095             41.337036  \n",
       "std                  27.871645             27.064873  \n",
       "min                   0.000000              0.000000  \n",
       "25%                   3.000000              3.000000  \n",
       "50%                  63.000000             60.000000  \n",
       "75%                  63.000000             63.000000  \n",
       "max                  63.000000             63.000000  \n",
       "\n",
       "[8 rows x 109 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_train.describe()"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "Описание признаков в таблице:\n",
    "match_id: идентификатор матча в наборе данных\n",
    "start_time: время начала матча (unixtime)\n",
    "lobby_type: тип комнаты, в которой собираются игроки (расшифровка в dictionaries/lobbies.csv)\n",
    "\n",
    "Наборы признаков для каждого игрока (игроки команды Radiant — префикс rN, Dire — dN):\n",
    "r1_hero: герой игрока (расшифровка в dictionaries/heroes.csv)\n",
    "r1_level: максимальный достигнутый уровень героя (за первые 5 игровых минут)\n",
    "r1_xp: максимальный полученный опыт\n",
    "r1_gold: достигнутая ценность героя\n",
    "r1_lh: число убитых юнитов\n",
    "r1_kills: число убитых игроков\n",
    "r1_deaths: число смертей героя\n",
    "r1_items: число купленных предметов\n",
    "\n",
    "Признаки события \"первая кровь\" (first blood). Если событие \"первая кровь\" не успело произойти за первые 5 минут, то признаки принимают пропущенное значение\n",
    "first_blood_time: игровое время первой крови\n",
    "first_blood_team: команда, совершившая первую кровь (0 — Radiant, 1 — Dire)\n",
    "first_blood_player1: игрок, причастный к событию\n",
    "first_blood_player2: второй игрок, причастный к событию\n",
    "\n",
    "Признаки для каждой команды (префиксы radiant_ и dire_)\n",
    "radiant_bottle_time: время первого приобретения командой предмета \"bottle\"\n",
    "radiant_courier_time: время приобретения предмета \"courier\"\n",
    "radiant_flying_courier_time: время приобретения предмета \"flying_courier\"\n",
    "radiant_tpscroll_count: число предметов \"tpscroll\" за первые 5 минут\n",
    "radiant_boots_count: число предметов \"boots\"\n",
    "radiant_ward_observer_count: число предметов \"ward_observer\"\n",
    "radiant_ward_sentry_count: число предметов \"ward_sentry\"\n",
    "radiant_first_ward_time: время установки командой первого \"наблюдателя\", т.е. предмета, который позволяет видеть часть игрового поля\n",
    "\n",
    "Итог матча (данные поля отсутствуют в тестовой выборке, поскольку содержат информацию, выходящую за пределы первых 5 минут матча)\n",
    "duration: длительность\n",
    "radiant_win: 1, если победила команда Radiant, 0 — иначе\n",
    "Состояние башен и барраков к концу матча (см. описание полей набора данных)\n",
    "tower_status_radiant\n",
    "tower_status_dire\n",
    "barracks_status_radiant\n",
    "barracks_status_dire"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Удаляем признаки - итоги матча\n",
    "data_train.drop(['duration', \n",
    "         'tower_status_radiant', \n",
    "         'tower_status_dire', \n",
    "         'barracks_status_radiant', \n",
    "         'barracks_status_dire'\n",
    "        ], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Формируем массив с целевой переменной\n",
    "y_train = data_train['radiant_win']\n",
    "del data_train['radiant_win']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((97230,), (97230, 103))"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape, data_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Подход 1: градиентный бустинг \"в лоб\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Один из самых универсальных алгоритмов, изученных в нашем курсе, является градиентный бустинг. Он не очень требователен к данным, восстанавливает нелинейные зависимости, и хорошо работает на многих наборах данных, что и обуславливает его популярность. Вполне разумной мыслью будет попробовать именно его в первую очередь.\n",
    "\n",
    "1.Считайте таблицу с признаками из файла features.csv с помощью кода, приведенного выше. Удалите признаки, связанные с итогами матча (они помечены в описании данных как отсутствующие в тестовой выборке).\n",
    "\n",
    "2.Проверьте выборку на наличие пропусков с помощью функции count(), которая для каждого столбца показывает число заполненных значений. Много ли пропусков в данных? Запишите названия признаков, имеющих пропуски, и попробуйте для любых двух из них дать обоснование, почему их значения могут быть пропущены.\n",
    "\n",
    "3.Замените пропуски на нули с помощью функции fillna(). На самом деле этот способ является предпочтительным для логистической регрессии, поскольку он позволит пропущенному значению не вносить никакого вклада в предсказание. Для деревьев часто лучшим вариантом оказывается замена пропуска на очень большое или очень маленькое значение — в этом случае при построении разбиения вершины можно будет отправить объекты с пропусками в отдельную ветвь дерева. Также есть и другие подходы — например, замена пропуска на среднее значение признака. Мы не требуем этого в задании, но при желании попробуйте разные подходы к обработке пропусков и сравните их между собой.\n",
    "\n",
    "4.Какой столбец содержит целевую переменную? Запишите его название.\n",
    "\n",
    "5.Забудем, что в выборке есть категориальные признаки, и попробуем обучить градиентный бустинг над деревьями на имеющейся матрице \"объекты-признаки\". Зафиксируйте генератор разбиений для кросс-валидации по 5 блокам (KFold), не забудьте перемешать при этом выборку (shuffle=True), поскольку данные в таблице отсортированы по времени, и без перемешивания можно столкнуться с нежелательными эффектами при оценивании качества. Оцените качество градиентного бустинга (GradientBoostingClassifier) с помощью данной кросс-валидации, попробуйте при этом разное количество деревьев (как минимум протестируйте следующие значения для количества деревьев: 10, 20, 30). Долго ли настраивались классификаторы? Достигнут ли оптимум на испытанных значениях параметра n_estimators, или же качество, скорее всего, продолжит расти при дальнейшем его увеличении?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Data preprocessing   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(97230, 103)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_gb = data_train\n",
    "X_train_gb.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "match_id                           0\n",
       "start_time                         0\n",
       "lobby_type                         0\n",
       "r1_hero                            0\n",
       "r1_level                           0\n",
       "r1_xp                              0\n",
       "r1_gold                            0\n",
       "r1_lh                              0\n",
       "r1_kills                           0\n",
       "r1_deaths                          0\n",
       "r1_items                           0\n",
       "r2_hero                            0\n",
       "r2_level                           0\n",
       "r2_xp                              0\n",
       "r2_gold                            0\n",
       "r2_lh                              0\n",
       "r2_kills                           0\n",
       "r2_deaths                          0\n",
       "r2_items                           0\n",
       "r3_hero                            0\n",
       "r3_level                           0\n",
       "r3_xp                              0\n",
       "r3_gold                            0\n",
       "r3_lh                              0\n",
       "r3_kills                           0\n",
       "r3_deaths                          0\n",
       "r3_items                           0\n",
       "r4_hero                            0\n",
       "r4_level                           0\n",
       "r4_xp                              0\n",
       "                               ...  \n",
       "d4_deaths                          0\n",
       "d4_items                           0\n",
       "d5_hero                            0\n",
       "d5_level                           0\n",
       "d5_xp                              0\n",
       "d5_gold                            0\n",
       "d5_lh                              0\n",
       "d5_kills                           0\n",
       "d5_deaths                          0\n",
       "d5_items                           0\n",
       "first_blood_time               19553\n",
       "first_blood_team               19553\n",
       "first_blood_player1            19553\n",
       "first_blood_player2            43987\n",
       "radiant_bottle_time            15691\n",
       "radiant_courier_time             692\n",
       "radiant_flying_courier_time    27479\n",
       "radiant_tpscroll_count             0\n",
       "radiant_boots_count                0\n",
       "radiant_ward_observer_count        0\n",
       "radiant_ward_sentry_count          0\n",
       "radiant_first_ward_time         1836\n",
       "dire_bottle_time               16143\n",
       "dire_courier_time                676\n",
       "dire_flying_courier_time       26098\n",
       "dire_tpscroll_count                0\n",
       "dire_boots_count                   0\n",
       "dire_ward_observer_count           0\n",
       "dire_ward_sentry_count             0\n",
       "dire_first_ward_time            1826\n",
       "Length: 103, dtype: int64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_gb.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Заменяем пропуски на средние значения по столбцу\n",
    "X_train_gb = X_train_gb.fillna(X_train_gb.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- N_estimators =  10\n",
      "Time elapsed: 0:00:28.202836\n",
      "[0.6628393  0.6690325  0.65778594 0.66141503 0.66504144]\n",
      "--- N_estimators =  30\n",
      "Time elapsed: 0:01:21.018898\n",
      "[0.68679836 0.69049064 0.68466605 0.68999469 0.69272758]\n",
      "--- N_estimators =  50\n",
      "Time elapsed: 0:02:03.592194\n",
      "[0.69548905 0.69712892 0.69344154 0.69845041 0.70048886]\n",
      "--- N_estimators =  100\n",
      "Time elapsed: 0:04:24.031007\n",
      "[0.70398016 0.70602234 0.70366821 0.70758957 0.70955201]\n",
      "CPU times: user 4.84 s, sys: 826 ms, total: 5.67 s\n",
      "Wall time: 8min 16s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# Подбираем оптимальное число деревьев\n",
    "\n",
    "kf = KFold(n_splits=5, shuffle=True, random_state=42)\n",
    "scores = []\n",
    "N = [10, 30, 50, 100]\n",
    "for n in N:\n",
    "    print ('--- N_estimators = ', str(n))\n",
    "    model = GradientBoostingClassifier(n_estimators=n, random_state=42)\n",
    "    start_time = datetime.datetime.now()\n",
    "    model_scores = cross_val_score(model, X_train_gb, y_train, cv=kf, scoring='roc_auc', n_jobs=-1)\n",
    "    print ('Time elapsed:', datetime.datetime.now() - start_time)\n",
    "    print (model_scores)\n",
    "    scores.append(max(model_scores))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.6690324994558224, 0.6927275801527567, 0.700488856402555, 0.7095520064860807]"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x112197978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(N, scores)\n",
    "plt.xlabel('n_estimators')\n",
    "plt.ylabel('score')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 656,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "gb = GradientBoostingClassifier(n_estimators=100, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 657,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 24s, sys: 693 ms, total: 1min 24s\n",
      "Wall time: 1min 25s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=3,\n",
       "              max_features=None, max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=1, min_samples_split=2,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=100,\n",
       "              presort='auto', random_state=42, subsample=1.0, verbose=0,\n",
       "              warm_start=False)"
      ]
     },
     "execution_count": 657,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "gb.fit(X_train_gb, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Что указать в отчете**:\n",
    "\n",
    "В отчете по данному этапу вы должны ответить на следующие вопросы:\n",
    "\n",
    "1.Какие признаки имеют пропуски среди своих значений? Что могут означать пропуски в этих признаках (ответьте на этот вопрос для двух любых признаков)?\n",
    "\n",
    "2.Как называется столбец, содержащий целевую переменную?\n",
    "\n",
    "3.Как долго проводилась кросс-валидация для градиентного бустинга с 30 деревьями? Инструкцию по измерению времени можно найти ниже по тексту. Какое качество при этом получилось? Напомним, что в данном задании мы используем метрику качества AUC-ROC.\n",
    "\n",
    "4.Имеет ли смысл использовать больше 30 деревьев в градиентном бустинге? Что бы вы предложили делать, чтобы ускорить его обучение при увеличении количества деревьев?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Ответы**:\n",
    "\n",
    "1.Пропуски:\n",
    "first_blood_time              - 19553\n",
    "first_blood_team              - 19553\n",
    "first_blood_player1            - 19553\n",
    "first_blood_player2            - 43987\n",
    "radiant_bottle_time            - 15691\n",
    "radiant_courier_time           -  692\n",
    "radiant_flying_courier_time   - 27479\n",
    "radiant_first_ward_time       -  1836\n",
    "dire_bottle_time              - 16143\n",
    "dire_courier_time             -  676\n",
    "dire_flying_courier_time      - 26098\n",
    "dire_first_ward_time          -  1826\n",
    "\n",
    "Например,\n",
    "first_blood_time (игровое время первой крови) - первой крови может не быть в первые 5 минут игры, \n",
    "first_blood_team (команда, совершившая первую кровь) - соответственно, этот признак так же будет пустым\n",
    "\n",
    "2.radiant_win\n",
    "\n",
    "3.Кросс-валидация для градиентного бустинга с 30 деревьями заняла 0:01:19, метрика качества AUC-ROC равна 0.69272.\n",
    "\n",
    "4.Как видно из графика, качество модели продолжает расти при росте количества деревьев, но незначительно: на 100 деревьях метрика качества AUC-ROC равна 0.70616 - что ненамного превышает результат с 30 дереьвями. В принципе, можно остановиться на количестве дереьвев 30. Для увеличения качества можно также поиграться с такими параметрами GradientBoostingClassifier, как: learning_rate, max_depth,min_samples_leaf. В целом, для уменьшения времени обучения модели можно сократить глубину деревьев (по умолчанию, стоит 3) или уменьшить размер обучающей выборки."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Подход 2: логистическая регрессия\n",
    "\n",
    "Линейные методы работают гораздо быстрее композиций деревьев, поэтому кажется разумным воспользоваться именно ими для ускорения анализа данных. Одним из наиболее распространенных методов для классификации является логистическая регрессия.\n",
    "\n",
    "Важно: не забывайте, что линейные алгоритмы чувствительны к масштабу признаков! Может пригодиться sklearn.preprocessing.StandartScaler.\n",
    "\n",
    "1.Оцените качество логистической регрессии (sklearn.linear_model.LogisticRegression с L2-регуляризацией) с помощью кросс-валидации по той же схеме, которая использовалась для градиентного бустинга. Подберите при этом лучший параметр регуляризации (C). Какое наилучшее качество у вас получилось? Как оно соотносится с качеством градиентного бустинга? Чем вы можете объяснить эту разницу? Быстрее ли работает логистическая регрессия по сравнению с градиентным бустингом?\n",
    "\n",
    "2.Среди признаков в выборке есть категориальные, которые мы использовали как числовые, что вряд ли является хорошей идеей. Категориальных признаков в этой задаче одиннадцать: lobby_type и r1_hero, r2_hero, ..., r5_hero, d1_hero, d2_hero, ..., d5_hero. Уберите их из выборки, и проведите кросс-валидацию для логистической регрессии на новой выборке с подбором лучшего параметра регуляризации. Изменилось ли качество? Чем вы можете это объяснить?\n",
    "\n",
    "3.На предыдущем шаге мы исключили из выборки признаки rM_hero и dM_hero, которые показывают, какие именно герои играли за каждую команду. Это важные признаки — герои имеют разные характеристики, и некоторые из них выигрывают чаще, чем другие. Выясните из данных, сколько различных идентификаторов героев существует в данной игре (вам может пригодиться фукнция unique или value_counts).\n",
    "\n",
    "4.Воспользуемся подходом \"мешок слов\" для кодирования информации о героях. Пусть всего в игре имеет N различных героев. Сформируем N признаков, при этом i-й будет равен нулю, если i-й герой не участвовал в матче; единице, если i-й герой играл за команду Radiant; минус единице, если i-й герой играл за команду Dire. Ниже вы можете найти код, который выполняет данной преобразование. Добавьте полученные признаки к числовым, которые вы использовали во втором пункте данного этапа.\n",
    "\n",
    "5.Проведите кросс-валидацию для логистической регрессии на новой выборке с подбором лучшего параметра регуляризации. Какое получилось качество? Улучшилось ли оно? Чем вы можете это объяснить?\n",
    "\n",
    "6.Постройте предсказания вероятностей победы команды Radiant для тестовой выборки с помощью лучшей из изученных моделей (лучшей с точки зрения AUC-ROC на кросс-валидации). Убедитесь, что предсказанные вероятности адекватные — находятся на отрезке [0, 1], не совпадают между собой (т.е. что модель не получилась константной)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(97230, 103)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_lr = data_train\n",
    "X_train_lr.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Approach #1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Заменяем пропуски в данных на 0\n",
    "X_train_lr = X_train_lr.fillna(0) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Важно: не забывайте, что линейные алгоритмы чувствительны к масштабу признаков! \n",
    "# Может пригодиться sklearn.preprocessing.StandartScaler.\n",
    "scaler = StandardScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train_lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_scores(C, score):\n",
    "    plt.plot(C, score)\n",
    "    plt.xlabel('10^n')\n",
    "    plt.ylabel('score')\n",
    "    plt.show() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def tuning_model(X, y):\n",
    "    scores = []\n",
    "    C_range = range(-4,3)\n",
    "    C = [10.0 ** i for i in C_range]\n",
    "    for c in C:\n",
    "        print ('--- C = ', str(c))\n",
    "        model = LogisticRegression(C = c, random_state=42)\n",
    "        model_scores = cross_val_score(model, X, y, cv=cv, scoring='roc_auc', n_jobs=-1)\n",
    "        start_time = datetime.datetime.now()\n",
    "        print ('Time elapsed:', datetime.datetime.now() - start_time)\n",
    "        print (model_scores)\n",
    "        scores.append(np.mean(model_scores))\n",
    "    plot_scores(C_range,scores)\n",
    "    max_score = max(scores)\n",
    "    max_score_ind = scores.index(max_score)\n",
    "    return C[max_score_ind], max_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cv = KFold(n_splits=5, shuffle=True, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- C =  0.0001\n",
      "Time elapsed: 0:00:00.000072\n",
      "[0.70966753 0.71051714 0.71180715 0.71178691 0.71340715]\n",
      "--- C =  0.001\n",
      "Time elapsed: 0:00:00.000084\n",
      "[0.7146791  0.71587433 0.71631789 0.71700923 0.71838037]\n",
      "--- C =  0.01\n",
      "Time elapsed: 0:00:00.000044\n",
      "[0.71489759 0.71627903 0.71624905 0.71736076 0.71840857]\n",
      "--- C =  0.1\n",
      "Time elapsed: 0:00:00.000013\n",
      "[0.71488151 0.71626834 0.71617246 0.71738508 0.7183599 ]\n",
      "--- C =  1.0\n",
      "Time elapsed: 0:00:00.000013\n",
      "[0.71487632 0.71626572 0.7161657  0.71738524 0.71835425]\n",
      "--- C =  10.0\n",
      "Time elapsed: 0:00:00.000015\n",
      "[0.71487588 0.71626501 0.716164   0.71738562 0.71835375]\n",
      "--- C =  100.0\n",
      "Time elapsed: 0:00:00.000012\n",
      "[0.71487564 0.7162651  0.71616387 0.71738572 0.71835372]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1121ac630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 7.88 s, sys: 2.02 s, total: 9.9 s\n",
      "Wall time: 1min 17s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# Подберем параметр регуляризации С на исходном датасете\n",
    "C1, score1 = tuning_model(X_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.01, 0.7166390000880288)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C1, score1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Approach #2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Удалим категориальные признаки из данных\n",
    "del X_train_lr['lobby_type']\n",
    "for n in range(1, 6):\n",
    "    del X_train_lr['r{}_hero'.format(n)]\n",
    "    del X_train_lr['d{}_hero'.format(n)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train_scaled2 = scaler.fit_transform(X_train_lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(97230, 92)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_scaled2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- C =  0.0001\n",
      "Time elapsed: 0:00:00.000065\n",
      "[0.70947131 0.71054883 0.71177751 0.71169704 0.71359801]\n",
      "--- C =  0.001\n",
      "Time elapsed: 0:00:00.000018\n",
      "[0.71452753 0.71590583 0.7163856  0.71691469 0.71860973]\n",
      "--- C =  0.01\n",
      "Time elapsed: 0:00:00.000019\n",
      "[0.7147459  0.7163128  0.71633048 0.71724976 0.71864826]\n",
      "--- C =  0.1\n",
      "Time elapsed: 0:00:00.000037\n",
      "[0.71473702 0.71630672 0.71626466 0.71727287 0.71859946]\n",
      "--- C =  1.0\n",
      "Time elapsed: 0:00:00.000014\n",
      "[0.7147351  0.71630497 0.71625722 0.71727556 0.71859379]\n",
      "--- C =  10.0\n",
      "Time elapsed: 0:00:00.000015\n",
      "[0.71473466 0.71630423 0.71625645 0.71727649 0.71859226]\n",
      "--- C =  100.0\n",
      "Time elapsed: 0:00:00.000023\n",
      "[0.71473475 0.7163042  0.71625606 0.71727679 0.71859226]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10bf6c438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 6.98 s, sys: 1.23 s, total: 8.21 s\n",
      "Wall time: 56.8 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# Подберем параметр регуляризации С на датасете без категориальных признаков\n",
    "C2, score2 = tuning_model(X_train_scaled2, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.01, 0.7166574405534297)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C2, score2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Как видно из графика, удаление категориальных признаков никак не повлияло на качество модели - значение коэффициента регуляризации и значение метрики ROC-AUC почти не изменились (разница получилась небольшой - в стотысячные доли). "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Approach #3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Количество идентификаторов героев в игре: 112\n"
     ]
    }
   ],
   "source": [
    "heroes = pd.read_csv('dictionaries/heroes.csv')\n",
    "print ('Количество идентификаторов героев в игре:', len(heroes))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Формируем \"мешок слов\" по героям\n",
    "\n",
    "def bag_of_heroes(X,heroes):\n",
    "    X_heroes = np.zeros((X.shape[0], len(heroes)))\n",
    "    for i, match_id in enumerate(X.index):\n",
    "        for p in range(5):\n",
    "            X_heroes[i, X.loc[match_id, 'r%d_hero' % (p+1)]-1] = 1\n",
    "            X_heroes[i, X.loc[match_id, 'd%d_hero' % (p+1)]-1] = -1\n",
    "    return X_heroes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_heroes = bag_of_heroes(X_train_lr,heroes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(97230, 112)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_heroes.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>102</th>\n",
       "      <th>103</th>\n",
       "      <th>104</th>\n",
       "      <th>105</th>\n",
       "      <th>106</th>\n",
       "      <th>107</th>\n",
       "      <th>108</th>\n",
       "      <th>109</th>\n",
       "      <th>110</th>\n",
       "      <th>111</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.732526</td>\n",
       "      <td>-2.544364</td>\n",
       "      <td>1.540688</td>\n",
       "      <td>-1.244228</td>\n",
       "      <td>1.400808</td>\n",
       "      <td>1.525972</td>\n",
       "      <td>0.734957</td>\n",
       "      <td>0.969743</td>\n",
       "      <td>-0.537757</td>\n",
       "      <td>-0.578083</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.732496</td>\n",
       "      <td>-2.540452</td>\n",
       "      <td>-0.927798</td>\n",
       "      <td>-0.292258</td>\n",
       "      <td>0.501314</td>\n",
       "      <td>-0.080139</td>\n",
       "      <td>-0.247570</td>\n",
       "      <td>-0.246859</td>\n",
       "      <td>-0.537757</td>\n",
       "      <td>1.017574</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.732466</td>\n",
       "      <td>-2.539231</td>\n",
       "      <td>1.540688</td>\n",
       "      <td>-0.568637</td>\n",
       "      <td>0.501314</td>\n",
       "      <td>0.151070</td>\n",
       "      <td>0.263085</td>\n",
       "      <td>1.190944</td>\n",
       "      <td>-0.537757</td>\n",
       "      <td>-0.578083</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.732435</td>\n",
       "      <td>-2.532622</td>\n",
       "      <td>-0.575157</td>\n",
       "      <td>-0.691471</td>\n",
       "      <td>0.501314</td>\n",
       "      <td>0.962950</td>\n",
       "      <td>-0.198013</td>\n",
       "      <td>0.306142</td>\n",
       "      <td>-0.537757</td>\n",
       "      <td>-0.578083</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.732405</td>\n",
       "      <td>-2.529221</td>\n",
       "      <td>1.540688</td>\n",
       "      <td>-1.182811</td>\n",
       "      <td>0.501314</td>\n",
       "      <td>0.348745</td>\n",
       "      <td>-0.124754</td>\n",
       "      <td>-0.357459</td>\n",
       "      <td>0.968527</td>\n",
       "      <td>-0.578083</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 215 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        0         1         2         3         4         5         6    \\\n",
       "0 -1.732526 -2.544364  1.540688 -1.244228  1.400808  1.525972  0.734957   \n",
       "1 -1.732496 -2.540452 -0.927798 -0.292258  0.501314 -0.080139 -0.247570   \n",
       "2 -1.732466 -2.539231  1.540688 -0.568637  0.501314  0.151070  0.263085   \n",
       "3 -1.732435 -2.532622 -0.575157 -0.691471  0.501314  0.962950 -0.198013   \n",
       "4 -1.732405 -2.529221  1.540688 -1.182811  0.501314  0.348745 -0.124754   \n",
       "\n",
       "        7         8         9   ...   102  103  104  105  106  107  108  109  \\\n",
       "0  0.969743 -0.537757 -0.578083 ...   0.0  0.0  1.0  0.0  0.0  0.0  0.0  0.0   \n",
       "1 -0.246859 -0.537757  1.017574 ...   0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2  1.190944 -0.537757 -0.578083 ...   0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "3  0.306142 -0.537757 -0.578083 ...   0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "4 -0.357459  0.968527 -0.578083 ...   0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "\n",
       "   110  111  \n",
       "0  0.0  0.0  \n",
       "1  0.0  0.0  \n",
       "2  0.0  0.0  \n",
       "3  0.0  0.0  \n",
       "4  0.0  0.0  \n",
       "\n",
       "[5 rows x 215 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_joined = pd.concat([pd.DataFrame(X_train_scaled), pd.DataFrame(X_heroes)], axis=1)\n",
    "X_joined.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(97230, 215)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_joined.shape # (97230, 103 + 112)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- C =  0.0001\n",
      "Time elapsed: 0:00:00.000084\n",
      "[0.72380823 0.72420431 0.72470572 0.72601444 0.72689278]\n",
      "--- C =  0.001\n",
      "Time elapsed: 0:00:00.000100\n",
      "[0.74488327 0.74677747 0.7440329  0.74888913 0.74737192]\n",
      "--- C =  0.01\n",
      "Time elapsed: 0:00:00.000097\n",
      "[0.74964527 0.75280632 0.7491002  0.75549532 0.75169859]\n",
      "--- C =  0.1\n",
      "Time elapsed: 0:00:00.000075\n",
      "[0.74958177 0.75316434 0.74936883 0.75610985 0.75154848]\n",
      "--- C =  1.0\n",
      "Time elapsed: 0:00:00.000096\n",
      "[0.74951317 0.75317691 0.74937053 0.75614495 0.75148098]\n",
      "--- C =  10.0\n",
      "Time elapsed: 0:00:00.000134\n",
      "[0.74950292 0.75317758 0.74936958 0.7561477  0.75147216]\n",
      "--- C =  100.0\n",
      "Time elapsed: 0:00:00.000082\n",
      "[0.74950183 0.75317828 0.7493701  0.75614732 0.75147132]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10b2f12b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 15.7 s, sys: 4.51 s, total: 20.2 s\n",
      "Wall time: 2min 47s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# Подберем параметр регуляризации С на датасете с доп. признаками  \n",
    "C3, score3 = tuning_model(X_joined, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.1, 0.7519546556305338)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C3, score3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "После составления \"мешка слов\" по героям качество модели улучшилось. Наилучшее качество достигается при  коэффициенте регуляризации C = 0.1 и равно 0.75194. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "lr = LogisticRegression(C=0.1, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 8.18 s, sys: 1.35 s, total: 9.53 s\n",
      "Wall time: 10 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=0.1, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
       "          penalty='l2', random_state=42, solver='liblinear', tol=0.0001,\n",
       "          verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "lr.fit(X_joined, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Что указать в отчете**:\n",
    "\n",
    "В отчете по данному этапу вы должны ответить на следующие вопросы:\n",
    "\n",
    "1.Какое качество получилось у логистической регрессии над всеми исходными признаками? Как оно соотносится с качеством градиентного бустинга? Чем вы можете объяснить эту разницу? Быстрее ли работает логистическая регрессия по сравнению с градиентным бустингом?\n",
    "\n",
    "2.Как влияет на качество логистической регрессии удаление категориальных признаков (укажите новое значение метрики качества)? Чем вы можете объяснить это изменение?\n",
    "\n",
    "3.Сколько различных идентификаторов героев существует в данной игре?\n",
    "\n",
    "4.Какое получилось качество при добавлении \"мешка слов\" по героям? Улучшилось ли оно по сравнению с предыдущим вариантом? Чем вы можете это объяснить?\n",
    "\n",
    "5.Какое минимальное и максимальное значение прогноза на тестовой выборке получилось у лучшего из алгоритмов?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Ответы** :\n",
    "\n",
    "1.Макс.качество град. бустинга равно 0.70616, макс.качество логистической регрессии выше и равно 0.71663. Видимо, это обусловлено тем, что град. бустинг учитывал категориальные признаки при построении модели, в то время как лог. регрессия изначально приняла их за шум в данных и эти признаки не учитывались при построении модели. Также лог. регрессия работает быстрее по сравнению с бустингом - обучение модели град.бустинга на train data заняло 1min 25s, обучение модели лог. регрессии заняло 8.12 s.\n",
    "\n",
    "2.Удаление категориальных признаков никак не повлияло на качество модели - значение показателя регуляризации и значение метрики ROC-AUC почти не изменились (разница получилась небольшой - в миллионные доли). Видимо, изначально алгоритм принял их за шум в данных и эти признаки не учитывались при построении модели, поэтому их искусственное удаление впоследствии  никак не повлияло на результат.\n",
    "\n",
    "3.112\n",
    "\n",
    "4.После составления \"мешка слов\" по героям качество модели улучшилось. Наилучшее качество достигается при значении показателя регуляризации C = 0.1 и равно 0.75195. Искусственное добавление признаков - информации о героях игры помогло модели выполнить более корректное предсказание.\n",
    "\n",
    "5.Минимальное и максимальное значение прогноза на тестовой выборке получилось (0.003431763011054745, 0.9965682369889453). Качество прогноза на kaggle на тестовой выборке у лучшего алгоритма получилось 0.75528."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>match_id</th>\n",
       "      <th>start_time</th>\n",
       "      <th>lobby_type</th>\n",
       "      <th>r1_hero</th>\n",
       "      <th>r1_level</th>\n",
       "      <th>r1_xp</th>\n",
       "      <th>r1_gold</th>\n",
       "      <th>r1_lh</th>\n",
       "      <th>r1_kills</th>\n",
       "      <th>r1_deaths</th>\n",
       "      <th>...</th>\n",
       "      <th>radiant_ward_sentry_count</th>\n",
       "      <th>radiant_first_ward_time</th>\n",
       "      <th>dire_bottle_time</th>\n",
       "      <th>dire_courier_time</th>\n",
       "      <th>dire_flying_courier_time</th>\n",
       "      <th>dire_tpscroll_count</th>\n",
       "      <th>dire_boots_count</th>\n",
       "      <th>dire_ward_observer_count</th>\n",
       "      <th>dire_ward_sentry_count</th>\n",
       "      <th>dire_first_ward_time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>1430287923</td>\n",
       "      <td>0</td>\n",
       "      <td>93</td>\n",
       "      <td>4</td>\n",
       "      <td>1103</td>\n",
       "      <td>1089</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>247.0</td>\n",
       "      <td>-86.0</td>\n",
       "      <td>272.0</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>118.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7</td>\n",
       "      <td>1430293357</td>\n",
       "      <td>1</td>\n",
       "      <td>20</td>\n",
       "      <td>2</td>\n",
       "      <td>556</td>\n",
       "      <td>570</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>-29.0</td>\n",
       "      <td>168.0</td>\n",
       "      <td>-54.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>1430301774</td>\n",
       "      <td>1</td>\n",
       "      <td>112</td>\n",
       "      <td>2</td>\n",
       "      <td>751</td>\n",
       "      <td>808</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>-22.0</td>\n",
       "      <td>46.0</td>\n",
       "      <td>-87.0</td>\n",
       "      <td>186.0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-34.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>13</td>\n",
       "      <td>1430323933</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>708</td>\n",
       "      <td>903</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>-49.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>-89.0</td>\n",
       "      <td>210.0</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>-26.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16</td>\n",
       "      <td>1430331112</td>\n",
       "      <td>1</td>\n",
       "      <td>39</td>\n",
       "      <td>4</td>\n",
       "      <td>1259</td>\n",
       "      <td>661</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>36.0</td>\n",
       "      <td>180.0</td>\n",
       "      <td>-86.0</td>\n",
       "      <td>180.0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>-33.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 103 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   match_id  start_time  lobby_type  r1_hero  r1_level  r1_xp  r1_gold  r1_lh  \\\n",
       "0         6  1430287923           0       93         4   1103     1089      8   \n",
       "1         7  1430293357           1       20         2    556      570      1   \n",
       "2        10  1430301774           1      112         2    751      808      1   \n",
       "3        13  1430323933           1       27         3    708      903      1   \n",
       "4        16  1430331112           1       39         4   1259      661      4   \n",
       "\n",
       "   r1_kills  r1_deaths          ...           radiant_ward_sentry_count  \\\n",
       "0         0          1          ...                                   0   \n",
       "1         0          0          ...                                   2   \n",
       "2         0          0          ...                                   1   \n",
       "3         1          1          ...                                   2   \n",
       "4         0          0          ...                                   0   \n",
       "\n",
       "   radiant_first_ward_time  dire_bottle_time  dire_courier_time  \\\n",
       "0                     12.0             247.0              -86.0   \n",
       "1                    -29.0             168.0              -54.0   \n",
       "2                    -22.0              46.0              -87.0   \n",
       "3                    -49.0              30.0              -89.0   \n",
       "4                     36.0             180.0              -86.0   \n",
       "\n",
       "   dire_flying_courier_time  dire_tpscroll_count  dire_boots_count  \\\n",
       "0                     272.0                    3                 4   \n",
       "1                       NaN                    3                 2   \n",
       "2                     186.0                    1                 3   \n",
       "3                     210.0                    3                 4   \n",
       "4                     180.0                    1                 3   \n",
       "\n",
       "   dire_ward_observer_count  dire_ward_sentry_count  dire_first_ward_time  \n",
       "0                         2                       0                 118.0  \n",
       "1                         2                       1                  16.0  \n",
       "2                         3                       0                 -34.0  \n",
       "3                         2                       1                 -26.0  \n",
       "4                         2                       1                 -33.0  \n",
       "\n",
       "[5 rows x 103 columns]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_test = pd.read_csv('features_test.csv')\n",
    "data_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(17177, 103)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_test = data_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Заменяем пропуски в данных на 0\n",
    "X_test = X_test.fillna(0) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Формируем \"мешок слов\" по героям\n",
    "X_heroes_test = bag_of_heroes(X_test, heroes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(17177, 112)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_heroes_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>102</th>\n",
       "      <th>103</th>\n",
       "      <th>104</th>\n",
       "      <th>105</th>\n",
       "      <th>106</th>\n",
       "      <th>107</th>\n",
       "      <th>108</th>\n",
       "      <th>109</th>\n",
       "      <th>110</th>\n",
       "      <th>111</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 112 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   0    1    2    3    4    5    6    7    8    9   ...   102  103  104  105  \\\n",
       "0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0  0.0   \n",
       "1  0.0  0.0  0.0  0.0 -1.0  1.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0  0.0   \n",
       "2  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 -1.0 ...   0.0  0.0  0.0  0.0   \n",
       "3  0.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0  0.0   \n",
       "4  0.0  0.0 -1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0  0.0   \n",
       "\n",
       "   106  107  108  109  110  111  \n",
       "0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "1  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "2  0.0  0.0  0.0  0.0  0.0  1.0  \n",
       "3  0.0  0.0  0.0 -1.0  0.0  0.0  \n",
       "4  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "\n",
       "[5 rows x 112 columns]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(X_heroes_test).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_test_scaled = pd.DataFrame(scaler.fit_transform(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>93</th>\n",
       "      <th>94</th>\n",
       "      <th>95</th>\n",
       "      <th>96</th>\n",
       "      <th>97</th>\n",
       "      <th>98</th>\n",
       "      <th>99</th>\n",
       "      <th>100</th>\n",
       "      <th>101</th>\n",
       "      <th>102</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.729109</td>\n",
       "      <td>-2.514875</td>\n",
       "      <td>-0.925258</td>\n",
       "      <td>1.264932</td>\n",
       "      <td>0.474746</td>\n",
       "      <td>-0.237743</td>\n",
       "      <td>-0.137094</td>\n",
       "      <td>-0.370005</td>\n",
       "      <td>-0.532732</td>\n",
       "      <td>1.003884</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.992369</td>\n",
       "      <td>0.450016</td>\n",
       "      <td>1.908352</td>\n",
       "      <td>-0.400363</td>\n",
       "      <td>1.149479</td>\n",
       "      <td>0.017743</td>\n",
       "      <td>0.554260</td>\n",
       "      <td>-0.571552</td>\n",
       "      <td>-0.983356</td>\n",
       "      <td>3.200079</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.729078</td>\n",
       "      <td>-2.513895</td>\n",
       "      <td>-0.572547</td>\n",
       "      <td>-0.966915</td>\n",
       "      <td>-1.365630</td>\n",
       "      <td>-1.207913</td>\n",
       "      <td>-1.255660</td>\n",
       "      <td>-1.145090</td>\n",
       "      <td>-0.532732</td>\n",
       "      <td>-0.593200</td>\n",
       "      <td>...</td>\n",
       "      <td>1.788142</td>\n",
       "      <td>-0.574649</td>\n",
       "      <td>0.836945</td>\n",
       "      <td>1.723080</td>\n",
       "      <td>-1.597294</td>\n",
       "      <td>0.017743</td>\n",
       "      <td>-1.177689</td>\n",
       "      <td>-0.571552</td>\n",
       "      <td>0.433564</td>\n",
       "      <td>0.600867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.728988</td>\n",
       "      <td>-2.512377</td>\n",
       "      <td>-0.572547</td>\n",
       "      <td>1.845824</td>\n",
       "      <td>-1.365630</td>\n",
       "      <td>-0.862057</td>\n",
       "      <td>-0.742715</td>\n",
       "      <td>-1.145090</td>\n",
       "      <td>-0.532732</td>\n",
       "      <td>-0.593200</td>\n",
       "      <td>...</td>\n",
       "      <td>0.397887</td>\n",
       "      <td>-0.399706</td>\n",
       "      <td>-0.817633</td>\n",
       "      <td>-0.466721</td>\n",
       "      <td>0.281014</td>\n",
       "      <td>-1.038655</td>\n",
       "      <td>-0.311714</td>\n",
       "      <td>0.668978</td>\n",
       "      <td>-0.983356</td>\n",
       "      <td>-0.673257</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.728897</td>\n",
       "      <td>-2.508381</td>\n",
       "      <td>-0.572547</td>\n",
       "      <td>-0.752903</td>\n",
       "      <td>-0.445442</td>\n",
       "      <td>-0.938323</td>\n",
       "      <td>-0.537967</td>\n",
       "      <td>-1.145090</td>\n",
       "      <td>0.968865</td>\n",
       "      <td>1.003884</td>\n",
       "      <td>...</td>\n",
       "      <td>1.788142</td>\n",
       "      <td>-1.074486</td>\n",
       "      <td>-1.034626</td>\n",
       "      <td>-0.599436</td>\n",
       "      <td>0.523376</td>\n",
       "      <td>0.017743</td>\n",
       "      <td>0.554260</td>\n",
       "      <td>-0.571552</td>\n",
       "      <td>0.433564</td>\n",
       "      <td>-0.469397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.728807</td>\n",
       "      <td>-2.507087</td>\n",
       "      <td>-0.572547</td>\n",
       "      <td>-0.386024</td>\n",
       "      <td>0.474746</td>\n",
       "      <td>0.038942</td>\n",
       "      <td>-1.059534</td>\n",
       "      <td>-0.812911</td>\n",
       "      <td>-0.532732</td>\n",
       "      <td>-0.593200</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.992369</td>\n",
       "      <td>1.049820</td>\n",
       "      <td>0.999690</td>\n",
       "      <td>-0.400363</td>\n",
       "      <td>0.220423</td>\n",
       "      <td>-1.038655</td>\n",
       "      <td>-0.311714</td>\n",
       "      <td>-0.571552</td>\n",
       "      <td>0.433564</td>\n",
       "      <td>-0.647774</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 103 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        0         1         2         3         4         5         6    \\\n",
       "0 -1.729109 -2.514875 -0.925258  1.264932  0.474746 -0.237743 -0.137094   \n",
       "1 -1.729078 -2.513895 -0.572547 -0.966915 -1.365630 -1.207913 -1.255660   \n",
       "2 -1.728988 -2.512377 -0.572547  1.845824 -1.365630 -0.862057 -0.742715   \n",
       "3 -1.728897 -2.508381 -0.572547 -0.752903 -0.445442 -0.938323 -0.537967   \n",
       "4 -1.728807 -2.507087 -0.572547 -0.386024  0.474746  0.038942 -1.059534   \n",
       "\n",
       "        7         8         9      ...          93        94        95   \\\n",
       "0 -0.370005 -0.532732  1.003884    ...    -0.992369  0.450016  1.908352   \n",
       "1 -1.145090 -0.532732 -0.593200    ...     1.788142 -0.574649  0.836945   \n",
       "2 -1.145090 -0.532732 -0.593200    ...     0.397887 -0.399706 -0.817633   \n",
       "3 -1.145090  0.968865  1.003884    ...     1.788142 -1.074486 -1.034626   \n",
       "4 -0.812911 -0.532732 -0.593200    ...    -0.992369  1.049820  0.999690   \n",
       "\n",
       "        96        97        98        99        100       101       102  \n",
       "0 -0.400363  1.149479  0.017743  0.554260 -0.571552 -0.983356  3.200079  \n",
       "1  1.723080 -1.597294  0.017743 -1.177689 -0.571552  0.433564  0.600867  \n",
       "2 -0.466721  0.281014 -1.038655 -0.311714  0.668978 -0.983356 -0.673257  \n",
       "3 -0.599436  0.523376  0.017743  0.554260 -0.571552  0.433564 -0.469397  \n",
       "4 -0.400363  0.220423 -1.038655 -0.311714 -0.571552  0.433564 -0.647774  \n",
       "\n",
       "[5 rows x 103 columns]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_scaled.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(17177, 103)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_scaled.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_joined_test = pd.concat([pd.DataFrame(X_test_scaled), pd.DataFrame(X_heroes_test)], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(17177, 215)"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_joined_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_pred = lr.predict_proba(X_joined_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.003431763011054745, 0.9965682369889453)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.min(y_pred), np.max(y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>match_id</th>\n",
       "      <th>radiant_win</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>17172</th>\n",
       "      <td>114369</td>\n",
       "      <td>0.734640</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17173</th>\n",
       "      <td>114377</td>\n",
       "      <td>0.646233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17174</th>\n",
       "      <td>114378</td>\n",
       "      <td>0.242550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17175</th>\n",
       "      <td>114393</td>\n",
       "      <td>0.635741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17176</th>\n",
       "      <td>114398</td>\n",
       "      <td>0.439654</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       match_id  radiant_win\n",
       "17172    114369     0.734640\n",
       "17173    114377     0.646233\n",
       "17174    114378     0.242550\n",
       "17175    114393     0.635741\n",
       "17176    114398     0.439654"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_kaggle = pd.concat([X_test.match_id, pd.DataFrame(y_pred)],axis=1)\n",
    "y_kaggle.drop([0], axis=1, inplace=True)\n",
    "y_kaggle.columns = ['match_id','radiant_win']\n",
    "y_kaggle.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_kaggle.to_csv('pred1_kaggle.csv',sep=',',encoding='utf-8', index = False)"
   ]
  }
 ],
 "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
}