WeightsAndResiduals.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import glob\n",
    "import datetime\n",
    "import itertools\n",
    "from time import sleep\n",
    "import os\n",
    "import seaborn as sns\n",
    "from itertools import chain"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2017-06-20 21:51:22.869660\n",
      "Using this many files 4088\n"
     ]
    }
   ],
   "source": [
    "print datetime.datetime.now()\n",
    "validFilePaths = []\n",
    "for f in os.listdir(\"data/anomaly_data\"):\n",
    "    filePath = os.path.join(\"data/anomaly_data\", f)\n",
    "    if os.path.isdir(filePath):\n",
    "        continue\n",
    "    if os.stat(filePath).st_size <= 3:\n",
    "        continue\n",
    "    validFilePaths.append(filePath)\n",
    "    \n",
    "numF = int(1 * len(validFilePaths))\n",
    "print 'Using this many files {0}'.format(numF)\n",
    "validFilePaths = np.random.choice(validFilePaths, numF, replace=False)\n",
    "df_list = (pd.read_csv(f) for f in validFilePaths)\n",
    "df = pd.concat(df_list, ignore_index=True)\n",
    "df = df[df['radiant_win'].notnull()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df2 = pd.DataFrame(np.loadtxt('data/goodAnomaliesResidual.csv', delimiter = ','), \n",
    "                 columns = [\"match_id\", \"residual\"])\n",
    "\n",
    "df = df.merge(df2, left_on='match_id', right_on='match_id')\n",
    "\n",
    "matchDictionary = {}\n",
    "heroColumns = [filter(lambda x: z in x, df.columns) for z in ['hero_id']][0]\n",
    "\n",
    "# for each row, this function will return all the heroes that the match belongs in\n",
    "# for example, if the match had an AM and Spectre, we should place this match in each of those bags\n",
    "def heroesForMatch(row):\n",
    "    pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df = df[~df['match_id'].duplicated()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df = df.sort_values(by='residual', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "columns = df.columns\n",
    "numericalFeatures = ['match_id', 'positive_votes', 'negative_votes', 'first_blood_time', 'radiant_win',\n",
    "                    'duration', 'kills', 'deaths', 'assists', 'kpm', 'kda', 'hero_dmg',\n",
    "                    'gpm', 'hero_heal', 'xpm', 'totalgold', 'totalxp', 'lasthits', 'denies',\n",
    "                    'tower_kills', 'courier_kills', 'observer_uses', 'sentry_uses',\n",
    "                    'ancient_kills', 'camps_stacked', 'abandons', 'residual']\n",
    "numFeatures = [filter(lambda x: z in x, columns) for z in numericalFeatures]\n",
    "numFeatures = list(chain(*numFeatures))\n",
    "df = df[numFeatures]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[3215497967,\n",
       " 3215097220,\n",
       " 3215312675,\n",
       " 3215162338,\n",
       " 3215181105,\n",
       " 3215377994,\n",
       " 3215271631,\n",
       " 3215289100,\n",
       " 3215279863,\n",
       " 3215253055,\n",
       " 3215448302,\n",
       " 3215416370,\n",
       " 3215261563,\n",
       " 3215294948,\n",
       " 3215155134,\n",
       " 3215350255,\n",
       " 3215341221,\n",
       " 3215306187,\n",
       " 3215302636,\n",
       " 3215316820,\n",
       " 3215190109,\n",
       " 3215355847,\n",
       " 3215130854,\n",
       " 3215259897,\n",
       " 3215394671,\n",
       " 3215152225,\n",
       " 3215274711,\n",
       " 3215366609,\n",
       " 3215449992,\n",
       " 3215376405,\n",
       " 3215268540,\n",
       " 3215420902,\n",
       " 3215332931,\n",
       " 3215327547,\n",
       " 3215354992,\n",
       " 3215254939,\n",
       " 3215149372,\n",
       " 3215370025,\n",
       " 3215201012,\n",
       " 3215272070,\n",
       " 3215223639,\n",
       " 3215422492,\n",
       " 3215260456,\n",
       " 3215354934,\n",
       " 3215377592,\n",
       " 3215445972,\n",
       " 3215101148,\n",
       " 3215174827,\n",
       " 3215384305,\n",
       " 3215188370,\n",
       " 3215297782,\n",
       " 3215493515,\n",
       " 3215119808,\n",
       " 3215360521,\n",
       " 3215102154,\n",
       " 3215271055,\n",
       " 3215403667,\n",
       " 3215217460,\n",
       " 3215482564,\n",
       " 3215466626,\n",
       " 3215255292,\n",
       " 3215095789,\n",
       " 3215108689,\n",
       " 3215330831,\n",
       " 3215468550,\n",
       " 3215446132,\n",
       " 3215384555,\n",
       " 3215470445,\n",
       " 3215318005,\n",
       " 3215238829,\n",
       " 3215293087,\n",
       " 3215391855,\n",
       " 3215439074,\n",
       " 3215401870,\n",
       " 3215440922,\n",
       " 3215092973,\n",
       " 3215369622,\n",
       " 3215116689,\n",
       " 3215236348,\n",
       " 3215145375,\n",
       " 3215129374,\n",
       " 3215199686,\n",
       " 3215481084,\n",
       " 3215191852,\n",
       " 3215459217,\n",
       " 3215189407,\n",
       " 3215462703,\n",
       " 3215113055,\n",
       " 3215469573,\n",
       " 3215428974,\n",
       " 3215323020,\n",
       " 3215169878,\n",
       " 3215359797,\n",
       " 3215419225,\n",
       " 3215221028,\n",
       " 3215488213,\n",
       " 3215198158,\n",
       " 3215275453,\n",
       " 3215204587,\n",
       " 3215260718,\n",
       " 3215153289,\n",
       " 3215120631,\n",
       " 3215400918,\n",
       " 3215119120,\n",
       " 3215419414,\n",
       " 3215443584,\n",
       " 3215276839,\n",
       " 3215464545,\n",
       " 3215118187,\n",
       " 3215247802,\n",
       " 3215343954,\n",
       " 3215145695,\n",
       " 3215365977,\n",
       " 3215162352,\n",
       " 3215344482,\n",
       " 3215166724,\n",
       " 3215457284,\n",
       " 3215144772,\n",
       " 3215254006,\n",
       " 3215471576,\n",
       " 3215267691,\n",
       " 3215203102,\n",
       " 3215460086,\n",
       " 3215132768,\n",
       " 3215444643,\n",
       " 3215327292,\n",
       " 3215107898,\n",
       " 3215193832,\n",
       " 3215135123,\n",
       " 3215361786,\n",
       " 3215242906,\n",
       " 3215404311,\n",
       " 3215472295,\n",
       " 3215359384,\n",
       " 3215261245,\n",
       " 3215419991,\n",
       " 3215164884,\n",
       " 3215482334,\n",
       " 3215344415,\n",
       " 3215191160,\n",
       " 3215269902,\n",
       " 3215132609,\n",
       " 3215229099,\n",
       " 3215424121,\n",
       " 3215214645,\n",
       " 3215280848,\n",
       " 3215324165,\n",
       " 3215366819,\n",
       " 3215121935,\n",
       " 3215217456,\n",
       " 3215478948,\n",
       " 3215287068,\n",
       " 3215328836,\n",
       " 3215189005,\n",
       " 3215180352,\n",
       " 3215267872,\n",
       " 3215325036,\n",
       " 3215403728,\n",
       " 3215386000,\n",
       " 3215354548,\n",
       " 3215107983,\n",
       " 3215193319,\n",
       " 3215174289,\n",
       " 3215405510,\n",
       " 3215486107,\n",
       " 3215308046,\n",
       " 3215375609,\n",
       " 3215370507,\n",
       " 3215356164,\n",
       " 3215394202,\n",
       " 3215475044,\n",
       " 3215314163,\n",
       " 3215232157,\n",
       " 3215449600,\n",
       " 3215150608,\n",
       " 3215387413,\n",
       " 3215116601,\n",
       " 3215476161,\n",
       " 3215498524,\n",
       " 3215223648,\n",
       " 3215347649,\n",
       " 3215288708,\n",
       " 3215250234,\n",
       " 3215123893,\n",
       " 3215102237,\n",
       " 3215294572,\n",
       " 3215169250,\n",
       " 3215362362,\n",
       " 3215401727,\n",
       " 3215453213,\n",
       " 3215312256,\n",
       " 3215332342,\n",
       " 3215312958,\n",
       " 3215223737,\n",
       " 3215285061,\n",
       " 3215413686,\n",
       " 3215324947,\n",
       " 3215327320,\n",
       " 3215267032,\n",
       " 3215454284,\n",
       " 3215357642,\n",
       " 3215488755,\n",
       " 3215224918,\n",
       " 3215115434,\n",
       " 3215486952,\n",
       " 3215459746,\n",
       " 3215330354,\n",
       " 3215345885,\n",
       " 3215129922,\n",
       " 3215300804,\n",
       " 3215475617,\n",
       " 3215272277,\n",
       " 3215423386,\n",
       " 3215130289,\n",
       " 3215483739,\n",
       " 3215153586,\n",
       " 3215135265,\n",
       " 3215299981,\n",
       " 3215316700,\n",
       " 3215448897,\n",
       " 3215206674,\n",
       " 3215475240,\n",
       " 3215411914,\n",
       " 3215494824,\n",
       " 3215208665,\n",
       " 3215390160,\n",
       " 3215251697,\n",
       " 3215466606,\n",
       " 3215391093,\n",
       " 3215162013,\n",
       " 3215299806,\n",
       " 3215321270,\n",
       " 3215287399,\n",
       " 3215375892,\n",
       " 3215195313,\n",
       " 3215144420,\n",
       " 3215379626,\n",
       " 3215137884,\n",
       " 3215375087,\n",
       " 3215303959,\n",
       " 3215437756,\n",
       " 3215336973,\n",
       " 3215416702,\n",
       " 3215205916,\n",
       " 3215220146,\n",
       " 3215440465,\n",
       " 3215270302,\n",
       " 3215094959,\n",
       " 3215298099,\n",
       " 3215165027,\n",
       " 3215433320,\n",
       " 3215298273,\n",
       " 3215455368,\n",
       " 3215363503,\n",
       " 3215100697,\n",
       " 3215283180,\n",
       " 3215410568,\n",
       " 3215291565,\n",
       " 3215385605,\n",
       " 3215288289,\n",
       " 3215148792,\n",
       " 3215094660,\n",
       " 3215347239,\n",
       " 3215352617,\n",
       " 3215107489,\n",
       " 3215474947,\n",
       " 3215404711,\n",
       " 3215147040,\n",
       " 3215180855,\n",
       " 3215286517,\n",
       " 3215308199,\n",
       " 3215399746,\n",
       " 3215142113,\n",
       " 3215138351,\n",
       " 3215276184,\n",
       " 3215233902,\n",
       " 3215292517,\n",
       " 3215166848,\n",
       " 3215477347,\n",
       " 3215246062,\n",
       " 3215136191,\n",
       " 3215281167,\n",
       " 3215459906,\n",
       " 3215480662,\n",
       " 3215287547,\n",
       " 3215337213,\n",
       " 3215276888,\n",
       " 3215228456,\n",
       " 3215282906,\n",
       " 3215354545,\n",
       " 3215243693,\n",
       " 3215131987,\n",
       " 3215220488,\n",
       " 3215496947,\n",
       " 3215129208,\n",
       " 3215335248,\n",
       " 3215477407,\n",
       " 3215386429,\n",
       " 3215189721,\n",
       " 3215355241,\n",
       " 3215114144,\n",
       " 3215446878,\n",
       " 3215138380,\n",
       " 3215391557,\n",
       " 3215452559,\n",
       " 3215388921,\n",
       " 3215457606,\n",
       " 3215245376,\n",
       " 3215182714,\n",
       " 3215425334,\n",
       " 3215146405,\n",
       " 3215303729,\n",
       " 3215297974,\n",
       " 3215488250,\n",
       " 3215291673,\n",
       " 3215165758,\n",
       " 3215215978,\n",
       " 3215226575,\n",
       " 3215426458,\n",
       " 3215298264,\n",
       " 3215375401,\n",
       " 3215164902,\n",
       " 3215439378,\n",
       " 3215417419,\n",
       " 3215129654,\n",
       " 3215412289,\n",
       " 3215119017,\n",
       " 3215151576,\n",
       " 3215363364,\n",
       " 3215232872,\n",
       " 3215100187,\n",
       " 3215311956,\n",
       " 3215379763,\n",
       " 3215351593,\n",
       " 3215409860,\n",
       " 3215442394,\n",
       " 3215353990,\n",
       " 3215277734,\n",
       " 3215229457,\n",
       " 3215130835,\n",
       " 3215092287,\n",
       " 3215137548,\n",
       " 3215410873,\n",
       " 3215186260,\n",
       " 3215314246,\n",
       " 3215455747,\n",
       " 3215125088,\n",
       " 3215388049,\n",
       " 3215387904,\n",
       " 3215486254,\n",
       " 3215349760,\n",
       " 3215149581,\n",
       " 3215325668,\n",
       " 3215374088,\n",
       " 3215303752,\n",
       " 3215266899,\n",
       " 3215095314,\n",
       " 3215216831,\n",
       " 3215222186,\n",
       " 3215164824,\n",
       " 3215310710,\n",
       " 3215266966,\n",
       " 3215415225,\n",
       " 3215265780,\n",
       " 3215418853,\n",
       " 3215273920,\n",
       " 3215171733,\n",
       " 3215236033,\n",
       " 3215274732,\n",
       " 3215282518,\n",
       " 3215365416,\n",
       " 3215485464,\n",
       " 3215489360,\n",
       " 3215270915,\n",
       " 3215306738,\n",
       " 3215260577,\n",
       " 3215141466,\n",
       " 3215162054,\n",
       " 3215395730,\n",
       " 3215265050,\n",
       " 3215312365,\n",
       " 3215397328,\n",
       " 3215110538,\n",
       " 3215214818,\n",
       " 3215435057,\n",
       " 3215154136,\n",
       " 3215356322,\n",
       " 3215126746,\n",
       " 3215249783,\n",
       " 3215393413,\n",
       " 3215163349,\n",
       " 3215282633,\n",
       " 3215152262,\n",
       " 3215121064,\n",
       " 3215328313,\n",
       " 3215209313,\n",
       " 3215435386,\n",
       " 3215486232,\n",
       " 3215337531,\n",
       " 3215102476,\n",
       " 3215253869,\n",
       " 3215422166,\n",
       " 3215263577,\n",
       " 3215285461,\n",
       " 3215114749,\n",
       " 3215096854,\n",
       " 3215446534,\n",
       " 3215119491,\n",
       " 3215205453,\n",
       " 3215135325,\n",
       " 3215415463,\n",
       " 3215322189,\n",
       " 3215345510,\n",
       " 3215138506,\n",
       " 3215492929,\n",
       " 3215267165,\n",
       " 3215123467,\n",
       " 3215358156,\n",
       " 3215094142,\n",
       " 3215328883,\n",
       " 3215250045,\n",
       " 3215410384,\n",
       " 3215168749,\n",
       " 3215343292,\n",
       " 3215107300,\n",
       " 3215450701,\n",
       " 3215239658,\n",
       " 3215427391,\n",
       " 3215403005,\n",
       " 3215129173,\n",
       " 3215301736,\n",
       " 3215121346,\n",
       " 3215136149,\n",
       " 3215315656,\n",
       " 3215435380,\n",
       " 3215260528,\n",
       " 3215131741,\n",
       " 3215127161,\n",
       " 3215352898,\n",
       " 3215115754,\n",
       " 3215383518,\n",
       " 3215323457,\n",
       " 3215465807,\n",
       " 3215145418,\n",
       " 3215380602,\n",
       " 3215391171,\n",
       " 3215191784,\n",
       " 3215358722,\n",
       " 3215433160,\n",
       " 3215290038,\n",
       " 3215193538,\n",
       " 3215209365,\n",
       " 3215253420,\n",
       " 3215491015,\n",
       " 3215217135,\n",
       " 3215379611,\n",
       " 3215413988,\n",
       " 3215253204,\n",
       " 3215404742,\n",
       " 3215384989,\n",
       " 3215199127,\n",
       " 3215120092,\n",
       " 3215108762,\n",
       " 3215213377,\n",
       " 3215349948,\n",
       " 3215306421,\n",
       " 3215202477,\n",
       " 3215464043,\n",
       " 3215268057,\n",
       " 3215370518,\n",
       " 3215439438,\n",
       " 3215371756,\n",
       " 3215477162,\n",
       " 3215293625,\n",
       " 3215296744,\n",
       " 3215377187,\n",
       " 3215251365,\n",
       " 3215263061,\n",
       " 3215358458,\n",
       " 3215429839,\n",
       " 3215209805,\n",
       " 3215279179,\n",
       " 3215405966,\n",
       " 3215258832,\n",
       " 3215179518,\n",
       " 3215104961,\n",
       " 3215107873,\n",
       " 3215366168,\n",
       " 3215430628,\n",
       " 3215422313,\n",
       " 3215365394,\n",
       " 3215485880,\n",
       " 3215457207,\n",
       " 3215380931,\n",
       " 3215445032,\n",
       " 3215490011,\n",
       " 3215137211,\n",
       " 3215203727,\n",
       " 3215165184,\n",
       " 3215489708,\n",
       " 3215452967,\n",
       " 3215329371,\n",
       " 3215127826,\n",
       " 3215406396,\n",
       " 3215230963,\n",
       " 3215197322,\n",
       " 3215109921,\n",
       " 3215336230,\n",
       " 3215447774,\n",
       " 3215473864,\n",
       " 3215130557,\n",
       " 3215267128,\n",
       " 3215317064,\n",
       " 3215407823,\n",
       " 3215334574,\n",
       " 3215235851,\n",
       " 3215432287,\n",
       " 3215279599,\n",
       " 3215277575,\n",
       " 3215352751,\n",
       " 3215193905,\n",
       " 3215403773,\n",
       " 3215470994,\n",
       " 3215241393,\n",
       " 3215475777,\n",
       " 3215378128,\n",
       " 3215255293,\n",
       " 3215408793,\n",
       " 3215186419,\n",
       " 3215365233,\n",
       " 3215319925,\n",
       " 3215206639,\n",
       " 3215101069,\n",
       " 3215110880,\n",
       " 3215279096,\n",
       " 3215236308,\n",
       " 3215381288,\n",
       " 3215292947,\n",
       " 3215199673,\n",
       " 3215409332,\n",
       " 3215099900,\n",
       " 3215319240,\n",
       " 3215115194,\n",
       " 3215341642,\n",
       " 3215288694,\n",
       " 3215263248,\n",
       " 3215248395,\n",
       " 3215424333,\n",
       " 3215388328,\n",
       " 3215153855,\n",
       " 3215113703,\n",
       " 3215378253,\n",
       " 3215478594,\n",
       " 3215333633,\n",
       " 3215314457,\n",
       " 3215264390,\n",
       " 3215185323,\n",
       " 3215416462,\n",
       " 3215294519,\n",
       " 3215100773,\n",
       " 3215419450,\n",
       " 3215149559,\n",
       " 3215207589,\n",
       " 3215443285,\n",
       " 3215294238,\n",
       " 3215230801,\n",
       " 3215420148,\n",
       " 3215412427,\n",
       " 3215231453,\n",
       " 3215373592,\n",
       " 3215246152,\n",
       " 3215408093,\n",
       " 3215095140,\n",
       " 3215231193,\n",
       " 3215402750,\n",
       " 3215338649,\n",
       " 3215453826,\n",
       " 3215184198,\n",
       " 3215315286,\n",
       " 3215370517,\n",
       " 3215360704,\n",
       " 3215172975,\n",
       " 3215131925,\n",
       " 3215134357,\n",
       " 3215311164,\n",
       " 3215452739,\n",
       " 3215393161,\n",
       " 3215293239,\n",
       " 3215114502,\n",
       " 3215224556,\n",
       " 3215273884,\n",
       " 3215363634,\n",
       " 3215460930,\n",
       " 3215487728,\n",
       " 3215273633,\n",
       " 3215360865,\n",
       " 3215431682,\n",
       " 3215109989,\n",
       " 3215237594,\n",
       " 3215269721,\n",
       " 3215094857,\n",
       " 3215383204,\n",
       " 3215374815,\n",
       " 3215128099,\n",
       " 3215241740,\n",
       " 3215344197,\n",
       " 3215303641,\n",
       " 3215156007,\n",
       " 3215091817,\n",
       " 3215275030,\n",
       " 3215389491,\n",
       " 3215141012,\n",
       " 3215369987,\n",
       " 3215215211,\n",
       " 3215231482,\n",
       " 3215237574,\n",
       " 3215368109,\n",
       " 3215428944,\n",
       " 3215401494,\n",
       " 3215236425,\n",
       " 3215453523,\n",
       " 3215429486,\n",
       " 3215325280,\n",
       " 3215195387,\n",
       " 3215351934,\n",
       " 3215360397,\n",
       " 3215281754,\n",
       " 3215188607,\n",
       " 3215248934,\n",
       " 3215379022,\n",
       " 3215248543,\n",
       " 3215204076,\n",
       " 3215469804,\n",
       " 3215228885,\n",
       " 3215142848,\n",
       " 3215429599,\n",
       " 3215271563,\n",
       " 3215123800,\n",
       " 3215141137,\n",
       " 3215221304,\n",
       " 3215146919,\n",
       " 3215138954,\n",
       " 3215226005,\n",
       " 3215418104,\n",
       " 3215485789,\n",
       " 3215364262,\n",
       " 3215384549,\n",
       " 3215354207,\n",
       " 3215221726,\n",
       " 3215397936,\n",
       " 3215253751,\n",
       " 3215335055,\n",
       " 3215178825,\n",
       " 3215351093,\n",
       " 3215263210,\n",
       " 3215311818,\n",
       " 3215111810,\n",
       " 3215195203,\n",
       " 3215457131,\n",
       " 3215224003,\n",
       " 3215323074,\n",
       " 3215256552,\n",
       " 3215290260,\n",
       " 3215100967,\n",
       " 3215406255,\n",
       " 3215317707,\n",
       " 3215471566,\n",
       " 3215299694,\n",
       " 3215334730,\n",
       " 3215233853,\n",
       " 3215233472,\n",
       " 3215139044,\n",
       " 3215385425,\n",
       " 3215242754,\n",
       " 3215168054,\n",
       " 3215470801,\n",
       " 3215290794,\n",
       " 3215202214,\n",
       " 3215220375,\n",
       " 3215139027,\n",
       " 3215158933,\n",
       " 3215429365,\n",
       " 3215249296,\n",
       " 3215412550,\n",
       " 3215158795,\n",
       " 3215428180,\n",
       " 3215248555,\n",
       " 3215092943,\n",
       " 3215477813,\n",
       " 3215357991,\n",
       " 3215121936,\n",
       " 3215451491,\n",
       " 3215179481,\n",
       " 3215339823,\n",
       " 3215213742,\n",
       " 3215291849,\n",
       " 3215359386,\n",
       " 3215463125,\n",
       " 3215320771,\n",
       " 3215156866,\n",
       " 3215146289,\n",
       " 3215436023,\n",
       " 3215120403,\n",
       " 3215356956,\n",
       " 3215349109,\n",
       " 3215293231,\n",
       " 3215295277,\n",
       " 3215332926,\n",
       " 3215468557,\n",
       " 3215099428,\n",
       " 3215320538,\n",
       " 3215097367,\n",
       " 3215410767,\n",
       " 3215319194,\n",
       " 3215189993,\n",
       " 3215450560,\n",
       " 3215415472,\n",
       " 3215473056,\n",
       " 3215218922,\n",
       " 3215398154,\n",
       " 3215196752,\n",
       " 3215116970,\n",
       " 3215154857,\n",
       " 3215378775,\n",
       " 3215422840,\n",
       " 3215438030,\n",
       " 3215467864,\n",
       " 3215148515,\n",
       " 3215127261,\n",
       " 3215403415,\n",
       " 3215371556,\n",
       " 3215373819,\n",
       " 3215289869,\n",
       " 3215261385,\n",
       " 3215113646,\n",
       " 3215093449,\n",
       " 3215397566,\n",
       " 3215394731,\n",
       " 3215202644,\n",
       " 3215122425,\n",
       " 3215232510,\n",
       " 3215376702,\n",
       " 3215271022,\n",
       " 3215410326,\n",
       " 3215238449,\n",
       " 3215219242,\n",
       " 3215314765,\n",
       " 3215415407,\n",
       " 3215454340,\n",
       " 3215159145,\n",
       " 3215496824,\n",
       " 3215153285,\n",
       " 3215421241,\n",
       " 3215158522,\n",
       " 3215285625,\n",
       " 3215409861,\n",
       " 3215146832,\n",
       " 3215350288,\n",
       " 3215138718,\n",
       " 3215216041,\n",
       " 3215430085,\n",
       " 3215162220,\n",
       " 3215495154,\n",
       " 3215260875,\n",
       " 3215240410,\n",
       " 3215315676,\n",
       " 3215259776,\n",
       " 3215322753,\n",
       " 3215232566,\n",
       " 3215155195,\n",
       " 3215219753,\n",
       " 3215153655,\n",
       " 3215443368,\n",
       " 3215100159,\n",
       " 3215362400,\n",
       " 3215227158,\n",
       " 3215114738,\n",
       " 3215377124,\n",
       " 3215387761,\n",
       " 3215490323,\n",
       " 3215224622,\n",
       " 3215465088,\n",
       " 3215280953,\n",
       " 3215356704,\n",
       " 3215432469,\n",
       " 3215209576,\n",
       " 3215320141,\n",
       " 3215150494,\n",
       " 3215334861,\n",
       " 3215168791,\n",
       " 3215310210,\n",
       " 3215460927,\n",
       " 3215440346,\n",
       " 3215470802,\n",
       " 3215129573,\n",
       " 3215453894,\n",
       " 3215327057,\n",
       " 3215229668,\n",
       " 3215248682,\n",
       " 3215159362,\n",
       " 3215139650,\n",
       " 3215291485,\n",
       " 3215450401,\n",
       " 3215217478,\n",
       " 3215373808,\n",
       " 3215293059,\n",
       " 3215134209,\n",
       " 3215205881,\n",
       " 3215303809,\n",
       " 3215264353,\n",
       " 3215268085,\n",
       " 3215283260,\n",
       " 3215344640,\n",
       " 3215184925,\n",
       " 3215335004,\n",
       " 3215168443,\n",
       " 3215193183,\n",
       " 3215278547,\n",
       " 3215183829,\n",
       " 3215184150,\n",
       " 3215163824,\n",
       " 3215363725,\n",
       " 3215404047,\n",
       " 3215237753,\n",
       " 3215487652,\n",
       " 3215429858,\n",
       " 3215166954,\n",
       " 3215133584,\n",
       " 3215178851,\n",
       " 3215448807,\n",
       " 3215273126,\n",
       " 3215168647,\n",
       " 3215236617,\n",
       " 3215250424,\n",
       " 3215435433,\n",
       " 3215106537,\n",
       " 3215425535,\n",
       " 3215403559,\n",
       " 3215279143,\n",
       " 3215273425,\n",
       " 3215236699,\n",
       " 3215420829,\n",
       " 3215249423,\n",
       " 3215204448,\n",
       " 3215354835,\n",
       " 3215110230,\n",
       " 3215354404,\n",
       " 3215269421,\n",
       " 3215475433,\n",
       " 3215122079,\n",
       " 3215120445,\n",
       " 3215483256,\n",
       " 3215270003,\n",
       " 3215239180,\n",
       " 3215449129,\n",
       " 3215466202,\n",
       " 3215446800,\n",
       " 3215404824,\n",
       " 3215216964,\n",
       " 3215292094,\n",
       " 3215244813,\n",
       " 3215257022,\n",
       " 3215138667,\n",
       " 3215116447,\n",
       " 3215186293,\n",
       " 3215168235,\n",
       " 3215324033,\n",
       " 3215203845,\n",
       " 3215198821,\n",
       " 3215303652,\n",
       " 3215328046,\n",
       " 3215373453,\n",
       " 3215449613,\n",
       " 3215349133,\n",
       " 3215230728,\n",
       " 3215465483,\n",
       " 3215360232,\n",
       " 3215242866,\n",
       " 3215186737,\n",
       " 3215303349,\n",
       " 3215229664,\n",
       " 3215358878,\n",
       " 3215290969,\n",
       " 3215115985,\n",
       " 3215322709,\n",
       " 3215364163,\n",
       " 3215160302,\n",
       " 3215280768,\n",
       " 3215296742,\n",
       " 3215491599,\n",
       " 3215105569,\n",
       " 3215125998,\n",
       " 3215160079,\n",
       " 3215387762,\n",
       " 3215469210,\n",
       " 3215489589,\n",
       " 3215101312,\n",
       " 3215443055,\n",
       " 3215355623,\n",
       " 3215140780,\n",
       " 3215091713,\n",
       " 3215246979,\n",
       " 3215198239,\n",
       " 3215464565,\n",
       " 3215210891,\n",
       " 3215418615,\n",
       " 3215189899,\n",
       " 3215408690,\n",
       " 3215398951,\n",
       " 3215107442,\n",
       " 3215379218,\n",
       " 3215114940,\n",
       " 3215131372,\n",
       " 3215429005,\n",
       " 3215228880,\n",
       " 3215464416,\n",
       " 3215268798,\n",
       " 3215197421,\n",
       " 3215230662,\n",
       " 3215367208,\n",
       " 3215416423,\n",
       " 3215233122,\n",
       " 3215176893,\n",
       " 3215128471,\n",
       " 3215220038,\n",
       " 3215316757,\n",
       " 3215470163,\n",
       " 3215130943,\n",
       " 3215468280,\n",
       " 3215196602,\n",
       " 3215426455,\n",
       " 3215350260,\n",
       " 3215192607,\n",
       " 3215232226,\n",
       " 3215162029,\n",
       " 3215244712,\n",
       " 3215466440,\n",
       " 3215199992,\n",
       " 3215482756,\n",
       " 3215271541,\n",
       " 3215195629,\n",
       " 3215270783,\n",
       " 3215259663,\n",
       " 3215436836,\n",
       " 3215256304,\n",
       " 3215491011,\n",
       " 3215381245,\n",
       " 3215359319,\n",
       " 3215222660,\n",
       " 3215449248,\n",
       " 3215110552,\n",
       " 3215117067,\n",
       " 3215374344,\n",
       " 3215358157,\n",
       " 3215287506,\n",
       " 3215458174,\n",
       " 3215110350,\n",
       " 3215113122,\n",
       " 3215485265,\n",
       " 3215374250,\n",
       " 3215113325,\n",
       " 3215265759,\n",
       " 3215322186,\n",
       " 3215265172,\n",
       " 3215421174,\n",
       " 3215379819,\n",
       " 3215093702,\n",
       " 3215096117,\n",
       " 3215492042,\n",
       " 3215326780,\n",
       " 3215223250,\n",
       " 3215335663,\n",
       " 3215118408,\n",
       " 3215090968,\n",
       " 3215420674,\n",
       " 3215379283,\n",
       " 3215242491,\n",
       " 3215124675,\n",
       " 3215494997,\n",
       " 3215356775,\n",
       " 3215244219,\n",
       " 3215254617,\n",
       " 3215232682,\n",
       " 3215208828,\n",
       " 3215145084,\n",
       " 3215358667,\n",
       " 3215447392,\n",
       " 3215289158,\n",
       " 3215161481,\n",
       " 3215132824,\n",
       " 3215173685,\n",
       " 3215376877,\n",
       " 3215336686,\n",
       " 3215149242,\n",
       " 3215308356,\n",
       " 3215463785,\n",
       " 3215143402,\n",
       " 3215347623,\n",
       " 3215123239,\n",
       " 3215205312,\n",
       " ...]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['match_id'].values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[60.09383010864258,\n",
       " 22.98058319091797,\n",
       " 14.018461227416992,\n",
       " 12.276541709899902,\n",
       " 10.960655212402344,\n",
       " 10.190336227416992,\n",
       " 9.797115325927734,\n",
       " 8.665862083435059,\n",
       " 8.291083335876465,\n",
       " 6.753566265106201,\n",
       " 5.982767105102539,\n",
       " 5.009283542633057,\n",
       " 4.787799835205078,\n",
       " 4.737585544586182,\n",
       " 4.151325702667236,\n",
       " 3.8586721420288086,\n",
       " 3.7200565338134766,\n",
       " 3.653451919555664,\n",
       " 3.6120126247406006,\n",
       " 3.55796217918396,\n",
       " 3.5249321460723877,\n",
       " 3.4550528526306152,\n",
       " 3.4301578998565674,\n",
       " 2.9261271953582764,\n",
       " 2.8132457733154297,\n",
       " 2.767648220062256,\n",
       " 2.758488178253174,\n",
       " 2.543581247329712,\n",
       " 2.5158803462982178,\n",
       " 2.4831578731536865,\n",
       " 2.4762089252471924,\n",
       " 2.392141342163086,\n",
       " 2.38230299949646,\n",
       " 2.3297765254974365,\n",
       " 2.2398288249969482,\n",
       " 2.229811429977417,\n",
       " 2.1894421577453613,\n",
       " 2.1700241565704346,\n",
       " 2.150897741317749,\n",
       " 2.0817346572875977,\n",
       " 2.0602142810821533,\n",
       " 2.044538974761963,\n",
       " 2.038116693496704,\n",
       " 2.0010886192321777,\n",
       " 1.953024983406067,\n",
       " 1.9425238370895386,\n",
       " 1.919196367263794,\n",
       " 1.8698474168777466,\n",
       " 1.8231494426727295,\n",
       " 1.7869755029678345,\n",
       " 1.7188692092895508,\n",
       " 1.7181131839752197,\n",
       " 1.7047621011734009,\n",
       " 1.6976909637451172,\n",
       " 1.691680669784546,\n",
       " 1.6713063716888428,\n",
       " 1.6653168201446533,\n",
       " 1.6202446222305298,\n",
       " 1.6132309436798096,\n",
       " 1.6028668880462646,\n",
       " 1.581583023071289,\n",
       " 1.5775842666625977,\n",
       " 1.5626850128173828,\n",
       " 1.5582425594329834,\n",
       " 1.5492855310440063,\n",
       " 1.5385805368423462,\n",
       " 1.5317155122756958,\n",
       " 1.5211113691329956,\n",
       " 1.5076344013214111,\n",
       " 1.4967833757400513,\n",
       " 1.4829853773117065,\n",
       " 1.4817631244659424,\n",
       " 1.4691892862319946,\n",
       " 1.4565132856369019,\n",
       " 1.4442218542099,\n",
       " 1.4434107542037964,\n",
       " 1.4333841800689697,\n",
       " 1.4322049617767334,\n",
       " 1.427970290184021,\n",
       " 1.424288272857666,\n",
       " 1.4241111278533936,\n",
       " 1.4210336208343506,\n",
       " 1.4192612171173096,\n",
       " 1.4180662631988525,\n",
       " 1.418045997619629,\n",
       " 1.4131355285644531,\n",
       " 1.3996341228485107,\n",
       " 1.3977913856506348,\n",
       " 1.392609715461731,\n",
       " 1.3883095979690552,\n",
       " 1.382023572921753,\n",
       " 1.3814785480499268,\n",
       " 1.3794400691986084,\n",
       " 1.364009976387024,\n",
       " 1.3631293773651123,\n",
       " 1.3609627485275269,\n",
       " 1.358451247215271,\n",
       " 1.3547123670578003,\n",
       " 1.3397060632705688,\n",
       " 1.3350133895874023,\n",
       " 1.3328020572662354,\n",
       " 1.3265687227249146,\n",
       " 1.3234529495239258,\n",
       " 1.3199751377105713,\n",
       " 1.2998747825622559,\n",
       " 1.297800898551941,\n",
       " 1.2892454862594604,\n",
       " 1.2862396240234375,\n",
       " 1.283833622932434,\n",
       " 1.2751528024673462,\n",
       " 1.2714056968688965,\n",
       " 1.2680243253707886,\n",
       " 1.2384188175201416,\n",
       " 1.2283591032028198,\n",
       " 1.222191572189331,\n",
       " 1.2142605781555176,\n",
       " 1.206538200378418,\n",
       " 1.2041807174682617,\n",
       " 1.1953727006912231,\n",
       " 1.1952720880508423,\n",
       " 1.182854175567627,\n",
       " 1.181726336479187,\n",
       " 1.179439902305603,\n",
       " 1.171888828277588,\n",
       " 1.1665818691253662,\n",
       " 1.16591215133667,\n",
       " 1.162890076637268,\n",
       " 1.1562278270721436,\n",
       " 1.1520692110061646,\n",
       " 1.144974946975708,\n",
       " 1.1360812187194824,\n",
       " 1.1236978769302368,\n",
       " 1.120529055595398,\n",
       " 1.118119239807129,\n",
       " 1.117451548576355,\n",
       " 1.1165120601654053,\n",
       " 1.1132597923278809,\n",
       " 1.1122854948043823,\n",
       " 1.1096220016479492,\n",
       " 1.1047585010528564,\n",
       " 1.1041306257247925,\n",
       " 1.1024683713912964,\n",
       " 1.1014271974563599,\n",
       " 1.0916112661361694,\n",
       " 1.0913853645324707,\n",
       " 1.0900408029556274,\n",
       " 1.084211826324463,\n",
       " 1.083559513092041,\n",
       " 1.0815109014511108,\n",
       " 1.080784797668457,\n",
       " 1.079131007194519,\n",
       " 1.0745054483413696,\n",
       " 1.0729814767837524,\n",
       " 1.072069764137268,\n",
       " 1.0666135549545288,\n",
       " 1.064724326133728,\n",
       " 1.0637664794921875,\n",
       " 1.0632541179656982,\n",
       " 1.0616685152053833,\n",
       " 1.058560848236084,\n",
       " 1.051201343536377,\n",
       " 1.0496363639831543,\n",
       " 1.0495151281356812,\n",
       " 1.049206018447876,\n",
       " 1.0463898181915283,\n",
       " 1.0434927940368652,\n",
       " 1.0426044464111328,\n",
       " 1.0395288467407227,\n",
       " 1.0386379957199097,\n",
       " 1.0352888107299805,\n",
       " 1.0337138175964355,\n",
       " 1.033097743988037,\n",
       " 1.032279133796692,\n",
       " 1.0288360118865967,\n",
       " 1.0225169658660889,\n",
       " 1.0223387479782104,\n",
       " 1.0201956033706665,\n",
       " 1.0195045471191406,\n",
       " 1.0138232707977295,\n",
       " 1.0096408128738403,\n",
       " 1.0063974857330322,\n",
       " 1.0060631036758423,\n",
       " 1.0054899454116821,\n",
       " 1.0046826601028442,\n",
       " 1.0020874738693237,\n",
       " 1.0011916160583496,\n",
       " 0.999028205871582,\n",
       " 0.9979208707809448,\n",
       " 0.9951856136322021,\n",
       " 0.9943534135818481,\n",
       " 0.9923041462898254,\n",
       " 0.9922865629196167,\n",
       " 0.9902551770210266,\n",
       " 0.9878281354904175,\n",
       " 0.9870291948318481,\n",
       " 0.9827179312705994,\n",
       " 0.9821138381958008,\n",
       " 0.9803048968315125,\n",
       " 0.9795767068862915,\n",
       " 0.9786282777786255,\n",
       " 0.9723061919212341,\n",
       " 0.970488965511322,\n",
       " 0.9695155620574951,\n",
       " 0.9688208103179932,\n",
       " 0.9679350256919861,\n",
       " 0.9669782519340515,\n",
       " 0.9644799828529358,\n",
       " 0.9643834233283997,\n",
       " 0.9639016389846802,\n",
       " 0.9597914218902588,\n",
       " 0.9596630334854126,\n",
       " 0.9576871991157532,\n",
       " 0.9555385112762451,\n",
       " 0.9550148248672485,\n",
       " 0.9534093141555786,\n",
       " 0.9518917202949524,\n",
       " 0.9480692744255066,\n",
       " 0.945451021194458,\n",
       " 0.9449933171272278,\n",
       " 0.9447908401489258,\n",
       " 0.9431512951850891,\n",
       " 0.9390014410018921,\n",
       " 0.9389370083808899,\n",
       " 0.9385173320770264,\n",
       " 0.9377952814102173,\n",
       " 0.9375945329666138,\n",
       " 0.936521053314209,\n",
       " 0.9340055584907532,\n",
       " 0.9331918358802795,\n",
       " 0.9330150485038757,\n",
       " 0.9317139983177185,\n",
       " 0.9304264783859253,\n",
       " 0.9288665056228638,\n",
       " 0.9260517358779907,\n",
       " 0.9249129891395569,\n",
       " 0.9222087264060974,\n",
       " 0.9214598536491394,\n",
       " 0.9209437966346741,\n",
       " 0.9209313988685608,\n",
       " 0.9208395481109619,\n",
       " 0.9197496175765991,\n",
       " 0.9185452461242676,\n",
       " 0.9154036641120911,\n",
       " 0.9132921695709229,\n",
       " 0.913219690322876,\n",
       " 0.9126948118209839,\n",
       " 0.9088218212127686,\n",
       " 0.9086437821388245,\n",
       " 0.9084234237670898,\n",
       " 0.9055785536766052,\n",
       " 0.9055338501930237,\n",
       " 0.9030809998512268,\n",
       " 0.9009914398193359,\n",
       " 0.8999032974243164,\n",
       " 0.899492621421814,\n",
       " 0.8974443078041077,\n",
       " 0.8955872058868408,\n",
       " 0.8940939903259277,\n",
       " 0.8897515535354614,\n",
       " 0.8891883492469788,\n",
       " 0.8884326219558716,\n",
       " 0.8873404860496521,\n",
       " 0.8867751955986023,\n",
       " 0.8864952921867371,\n",
       " 0.886139452457428,\n",
       " 0.8853210210800171,\n",
       " 0.8835608959197998,\n",
       " 0.8831455707550049,\n",
       " 0.8814792633056641,\n",
       " 0.8786903023719788,\n",
       " 0.8786889314651489,\n",
       " 0.8786466121673584,\n",
       " 0.878573477268219,\n",
       " 0.876968264579773,\n",
       " 0.8747633695602417,\n",
       " 0.8735519647598267,\n",
       " 0.8730534911155701,\n",
       " 0.8729946613311768,\n",
       " 0.8721534609794617,\n",
       " 0.8719340562820435,\n",
       " 0.8715412616729736,\n",
       " 0.8697844743728638,\n",
       " 0.8684020638465881,\n",
       " 0.8658937811851501,\n",
       " 0.8647394180297852,\n",
       " 0.8627612590789795,\n",
       " 0.8623327016830444,\n",
       " 0.8616533875465393,\n",
       " 0.8615354299545288,\n",
       " 0.8612295389175415,\n",
       " 0.8595142364501953,\n",
       " 0.8593857288360596,\n",
       " 0.8583488464355469,\n",
       " 0.8570705652236938,\n",
       " 0.8564101457595825,\n",
       " 0.8563953042030334,\n",
       " 0.8560832738876343,\n",
       " 0.8552137017250061,\n",
       " 0.8547729253768921,\n",
       " 0.8529301881790161,\n",
       " 0.851856529712677,\n",
       " 0.8514204025268555,\n",
       " 0.8499778509140015,\n",
       " 0.8469444513320923,\n",
       " 0.8467068672180176,\n",
       " 0.846032440662384,\n",
       " 0.845173716545105,\n",
       " 0.8448122143745422,\n",
       " 0.8432825803756714,\n",
       " 0.8417072296142578,\n",
       " 0.8406412601470947,\n",
       " 0.8391038179397583,\n",
       " 0.8389773368835449,\n",
       " 0.8380493521690369,\n",
       " 0.8373729586601257,\n",
       " 0.8368411064147949,\n",
       " 0.836216926574707,\n",
       " 0.8358647227287292,\n",
       " 0.8336787223815918,\n",
       " 0.8333789110183716,\n",
       " 0.8329710364341736,\n",
       " 0.8316418528556824,\n",
       " 0.8302108645439148,\n",
       " 0.830048143863678,\n",
       " 0.8297022581100464,\n",
       " 0.8295499682426453,\n",
       " 0.8286826014518738,\n",
       " 0.8277784585952759,\n",
       " 0.827299177646637,\n",
       " 0.8271003365516663,\n",
       " 0.8269991278648376,\n",
       " 0.8262494802474976,\n",
       " 0.8259848952293396,\n",
       " 0.8258390426635742,\n",
       " 0.8255735635757446,\n",
       " 0.8250579833984375,\n",
       " 0.8247339129447937,\n",
       " 0.8241198658943176,\n",
       " 0.8225634694099426,\n",
       " 0.8194668889045715,\n",
       " 0.8185660243034363,\n",
       " 0.8180004358291626,\n",
       " 0.8172050714492798,\n",
       " 0.8170410990715027,\n",
       " 0.8164354562759399,\n",
       " 0.8154846429824829,\n",
       " 0.8146878480911255,\n",
       " 0.8123286962509155,\n",
       " 0.8110998868942261,\n",
       " 0.8109608292579651,\n",
       " 0.810543417930603,\n",
       " 0.8094967007637024,\n",
       " 0.8084927797317505,\n",
       " 0.8075140118598938,\n",
       " 0.8060798048973083,\n",
       " 0.8060481548309326,\n",
       " 0.8048883676528931,\n",
       " 0.8044706583023071,\n",
       " 0.8039005994796753,\n",
       " 0.8037306070327759,\n",
       " 0.8031567335128784,\n",
       " 0.8023133277893066,\n",
       " 0.7998040914535522,\n",
       " 0.7996733784675598,\n",
       " 0.7978358864784241,\n",
       " 0.7977395057678223,\n",
       " 0.7967928051948547,\n",
       " 0.7952989935874939,\n",
       " 0.7938899993896484,\n",
       " 0.7934744358062744,\n",
       " 0.7932054400444031,\n",
       " 0.792657196521759,\n",
       " 0.792655348777771,\n",
       " 0.7925220727920532,\n",
       " 0.7915612459182739,\n",
       " 0.7906990051269531,\n",
       " 0.7903063893318176,\n",
       " 0.7900872230529785,\n",
       " 0.7893491983413696,\n",
       " 0.7878237962722778,\n",
       " 0.7857347726821899,\n",
       " 0.7855825424194336,\n",
       " 0.7842476963996887,\n",
       " 0.7831416130065918,\n",
       " 0.7821829915046692,\n",
       " 0.7811717391014099,\n",
       " 0.7788980007171631,\n",
       " 0.7783818244934082,\n",
       " 0.7776341438293457,\n",
       " 0.7757152915000916,\n",
       " 0.7753698825836182,\n",
       " 0.7746968865394592,\n",
       " 0.7746821045875549,\n",
       " 0.7739043831825256,\n",
       " 0.7726438641548157,\n",
       " 0.7714319229125977,\n",
       " 0.770415723323822,\n",
       " 0.7700993418693542,\n",
       " 0.7694905996322632,\n",
       " 0.7671429514884949,\n",
       " 0.7669565677642822,\n",
       " 0.7655788660049438,\n",
       " 0.765575647354126,\n",
       " 0.7641459107398987,\n",
       " 0.7632062435150146,\n",
       " 0.7624366879463196,\n",
       " 0.7618311047554016,\n",
       " 0.761588990688324,\n",
       " 0.7593628168106079,\n",
       " 0.7590985894203186,\n",
       " 0.7585266828536987,\n",
       " 0.7580933570861816,\n",
       " 0.7577105760574341,\n",
       " 0.7575187087059021,\n",
       " 0.7572482824325562,\n",
       " 0.7569153308868408,\n",
       " 0.7563158273696899,\n",
       " 0.7544280290603638,\n",
       " 0.7543737888336182,\n",
       " 0.7539798021316528,\n",
       " 0.7536249160766602,\n",
       " 0.7536008954048157,\n",
       " 0.7532666325569153,\n",
       " 0.7531739473342896,\n",
       " 0.752880871295929,\n",
       " 0.7528793811798096,\n",
       " 0.7526946067810059,\n",
       " 0.751673698425293,\n",
       " 0.7499583959579468,\n",
       " 0.7496762275695801,\n",
       " 0.7489606142044067,\n",
       " 0.7485086917877197,\n",
       " 0.7477176785469055,\n",
       " 0.7476752400398254,\n",
       " 0.7472490668296814,\n",
       " 0.7472139596939087,\n",
       " 0.7466698884963989,\n",
       " 0.7460474967956543,\n",
       " 0.7460435628890991,\n",
       " 0.744686484336853,\n",
       " 0.7445034384727478,\n",
       " 0.7432817220687866,\n",
       " 0.7422952055931091,\n",
       " 0.7414557337760925,\n",
       " 0.7412257194519043,\n",
       " 0.7403615713119507,\n",
       " 0.7403483986854553,\n",
       " 0.7383480072021484,\n",
       " 0.7383477091789246,\n",
       " 0.7380225658416748,\n",
       " 0.7366001009941101,\n",
       " 0.7361823320388794,\n",
       " 0.7343788743019104,\n",
       " 0.733960747718811,\n",
       " 0.732728123664856,\n",
       " 0.7322820425033569,\n",
       " 0.7319163680076599,\n",
       " 0.7304933071136475,\n",
       " 0.730053722858429,\n",
       " 0.7288756966590881,\n",
       " 0.7281818985939026,\n",
       " 0.7281191945075989,\n",
       " 0.727988064289093,\n",
       " 0.7277131676673889,\n",
       " 0.726827085018158,\n",
       " 0.7263097763061523,\n",
       " 0.7255817651748657,\n",
       " 0.7255740761756897,\n",
       " 0.7253108024597168,\n",
       " 0.7246365547180176,\n",
       " 0.7239059209823608,\n",
       " 0.7233874797821045,\n",
       " 0.7230355143547058,\n",
       " 0.723025918006897,\n",
       " 0.7225368022918701,\n",
       " 0.7217145562171936,\n",
       " 0.7210047841072083,\n",
       " 0.7206993699073792,\n",
       " 0.7205953001976013,\n",
       " 0.7205812335014343,\n",
       " 0.719957172870636,\n",
       " 0.7198128700256348,\n",
       " 0.7192848324775696,\n",
       " 0.7190936803817749,\n",
       " 0.7181803584098816,\n",
       " 0.7170781493186951,\n",
       " 0.7161182165145874,\n",
       " 0.7156762480735779,\n",
       " 0.7153546810150146,\n",
       " 0.7148224711418152,\n",
       " 0.7141834497451782,\n",
       " 0.7140740752220154,\n",
       " 0.7137510776519775,\n",
       " 0.7134366035461426,\n",
       " 0.7130857706069946,\n",
       " 0.7130664587020874,\n",
       " 0.7126566767692566,\n",
       " 0.7121343612670898,\n",
       " 0.7115955352783203,\n",
       " 0.7110877633094788,\n",
       " 0.7108469009399414,\n",
       " 0.7104701399803162,\n",
       " 0.7101932168006897,\n",
       " 0.7093899846076965,\n",
       " 0.7090077996253967,\n",
       " 0.7088366150856018,\n",
       " 0.7085787653923035,\n",
       " 0.7084393501281738,\n",
       " 0.7083696126937866,\n",
       " 0.7082640528678894,\n",
       " 0.7068285942077637,\n",
       " 0.7065154314041138,\n",
       " 0.7061051726341248,\n",
       " 0.7060731053352356,\n",
       " 0.706041693687439,\n",
       " 0.7057188153266907,\n",
       " 0.704667866230011,\n",
       " 0.7045057415962219,\n",
       " 0.7043561339378357,\n",
       " 0.7041466236114502,\n",
       " 0.7040076851844788,\n",
       " 0.7035585641860962,\n",
       " 0.7030852437019348,\n",
       " 0.7029399871826172,\n",
       " 0.7028341293334961,\n",
       " 0.7018325328826904,\n",
       " 0.7014175057411194,\n",
       " 0.7009782791137695,\n",
       " 0.7006656527519226,\n",
       " 0.6993235945701599,\n",
       " 0.6986374855041504,\n",
       " 0.6984686255455017,\n",
       " 0.6980193257331848,\n",
       " 0.6978328227996826,\n",
       " 0.6967384219169617,\n",
       " 0.6964960694313049,\n",
       " 0.6964261531829834,\n",
       " 0.6963183879852295,\n",
       " 0.6954740881919861,\n",
       " 0.6952798366546631,\n",
       " 0.695203423500061,\n",
       " 0.6944896578788757,\n",
       " 0.6942644119262695,\n",
       " 0.6942504644393921,\n",
       " 0.6940212249755859,\n",
       " 0.693926990032196,\n",
       " 0.693915843963623,\n",
       " 0.69270259141922,\n",
       " 0.6923828721046448,\n",
       " 0.6921383738517761,\n",
       " 0.6912344694137573,\n",
       " 0.6909181475639343,\n",
       " 0.6908226013183594,\n",
       " 0.6903275847434998,\n",
       " 0.6885342001914978,\n",
       " 0.6880930662155151,\n",
       " 0.687547504901886,\n",
       " 0.6873355507850647,\n",
       " 0.6869430541992188,\n",
       " 0.6868414878845215,\n",
       " 0.6866506934165955,\n",
       " 0.6856019496917725,\n",
       " 0.6853703260421753,\n",
       " 0.684950053691864,\n",
       " 0.6844768524169922,\n",
       " 0.6843836903572083,\n",
       " 0.6843326687812805,\n",
       " 0.6840593218803406,\n",
       " 0.6839963793754578,\n",
       " 0.6833631992340088,\n",
       " 0.6820196509361267,\n",
       " 0.6815665364265442,\n",
       " 0.6803432703018188,\n",
       " 0.6799181699752808,\n",
       " 0.679553210735321,\n",
       " 0.678919792175293,\n",
       " 0.6787182688713074,\n",
       " 0.6786744594573975,\n",
       " 0.6782424449920654,\n",
       " 0.6780423521995544,\n",
       " 0.6777185201644897,\n",
       " 0.6777063608169556,\n",
       " 0.6776937246322632,\n",
       " 0.6772192120552063,\n",
       " 0.6769409775733948,\n",
       " 0.6760160326957703,\n",
       " 0.6759694814682007,\n",
       " 0.6755587458610535,\n",
       " 0.675423800945282,\n",
       " 0.675076961517334,\n",
       " 0.6750593781471252,\n",
       " 0.6749622821807861,\n",
       " 0.674652636051178,\n",
       " 0.674209713935852,\n",
       " 0.6741954684257507,\n",
       " 0.6737287640571594,\n",
       " 0.6736929416656494,\n",
       " 0.6729382276535034,\n",
       " 0.6728609204292297,\n",
       " 0.6727287173271179,\n",
       " 0.6725167632102966,\n",
       " 0.6722366809844971,\n",
       " 0.6722153425216675,\n",
       " 0.6721881628036499,\n",
       " 0.6720960140228271,\n",
       " 0.671940267086029,\n",
       " 0.6713663339614868,\n",
       " 0.6698254346847534,\n",
       " 0.6695573329925537,\n",
       " 0.6687107682228088,\n",
       " 0.6684702634811401,\n",
       " 0.6678615808486938,\n",
       " 0.6676300764083862,\n",
       " 0.6671851277351379,\n",
       " 0.6671191453933716,\n",
       " 0.6668809652328491,\n",
       " 0.6667686700820923,\n",
       " 0.6666104197502136,\n",
       " 0.6665686368942261,\n",
       " 0.6664994955062866,\n",
       " 0.6663007736206055,\n",
       " 0.6657471656799316,\n",
       " 0.6657465696334839,\n",
       " 0.6657367944717407,\n",
       " 0.6655275225639343,\n",
       " 0.6652972102165222,\n",
       " 0.6651871204376221,\n",
       " 0.6651492118835449,\n",
       " 0.6650216579437256,\n",
       " 0.6648292541503906,\n",
       " 0.6644583940505981,\n",
       " 0.6641172766685486,\n",
       " 0.6637612581253052,\n",
       " 0.6635695099830627,\n",
       " 0.6633908748626709,\n",
       " 0.6629290580749512,\n",
       " 0.6624594330787659,\n",
       " 0.662200927734375,\n",
       " 0.6617655158042908,\n",
       " 0.6606206893920898,\n",
       " 0.6597539186477661,\n",
       " 0.6595001816749573,\n",
       " 0.659498393535614,\n",
       " 0.6579702496528625,\n",
       " 0.657727062702179,\n",
       " 0.6574667096138,\n",
       " 0.656964123249054,\n",
       " 0.6567689776420593,\n",
       " 0.6556574702262878,\n",
       " 0.655053436756134,\n",
       " 0.6549975872039795,\n",
       " 0.6542130708694458,\n",
       " 0.6539175510406494,\n",
       " 0.6536197662353516,\n",
       " 0.6534443497657776,\n",
       " 0.6531872153282166,\n",
       " 0.6529080867767334,\n",
       " 0.652268648147583,\n",
       " 0.6522124409675598,\n",
       " 0.6521428227424622,\n",
       " 0.652125895023346,\n",
       " 0.6517173647880554,\n",
       " 0.6509376764297485,\n",
       " 0.6509096622467041,\n",
       " 0.6508827805519104,\n",
       " 0.6505467891693115,\n",
       " 0.6492860317230225,\n",
       " 0.6486479043960571,\n",
       " 0.6486145853996277,\n",
       " 0.6485013365745544,\n",
       " 0.6484130620956421,\n",
       " 0.6481355428695679,\n",
       " 0.647615909576416,\n",
       " 0.6474666595458984,\n",
       " 0.6465640068054199,\n",
       " 0.6460104584693909,\n",
       " 0.6459749937057495,\n",
       " 0.6459360122680664,\n",
       " 0.6455346345901489,\n",
       " 0.6451497077941895,\n",
       " 0.6451070308685303,\n",
       " 0.6449517607688904,\n",
       " 0.6449416875839233,\n",
       " 0.6448999047279358,\n",
       " 0.6447224020957947,\n",
       " 0.6444844603538513,\n",
       " 0.6444458365440369,\n",
       " 0.6442970633506775,\n",
       " 0.6436285972595215,\n",
       " 0.6432618498802185,\n",
       " 0.6431841850280762,\n",
       " 0.6425811052322388,\n",
       " 0.6417860984802246,\n",
       " 0.6417686343193054,\n",
       " 0.6416600346565247,\n",
       " 0.6415826082229614,\n",
       " 0.6414430141448975,\n",
       " 0.6410348415374756,\n",
       " 0.640630841255188,\n",
       " 0.6404160261154175,\n",
       " 0.6401026248931885,\n",
       " 0.63993239402771,\n",
       " 0.639882504940033,\n",
       " 0.6395856142044067,\n",
       " 0.639421820640564,\n",
       " 0.6380833983421326,\n",
       " 0.6379858255386353,\n",
       " 0.6375442743301392,\n",
       " 0.637353777885437,\n",
       " 0.6371648907661438,\n",
       " 0.6369909048080444,\n",
       " 0.6369869112968445,\n",
       " 0.6367626786231995,\n",
       " 0.6362351179122925,\n",
       " 0.636012613773346,\n",
       " 0.6358704566955566,\n",
       " 0.6356680393218994,\n",
       " 0.6356462240219116,\n",
       " 0.6355707049369812,\n",
       " 0.6354865431785583,\n",
       " 0.6353418827056885,\n",
       " 0.6348565816879272,\n",
       " 0.6345191597938538,\n",
       " 0.6345074772834778,\n",
       " 0.6344534754753113,\n",
       " 0.6339588761329651,\n",
       " 0.6336132884025574,\n",
       " 0.6334683895111084,\n",
       " 0.6332551836967468,\n",
       " 0.6330251693725586,\n",
       " 0.6329081058502197,\n",
       " 0.6327518224716187,\n",
       " 0.6318038702011108,\n",
       " 0.6317944526672363,\n",
       " 0.6317700147628784,\n",
       " 0.6317121982574463,\n",
       " 0.6316589713096619,\n",
       " 0.6316245198249817,\n",
       " 0.6313509941101074,\n",
       " 0.6306886076927185,\n",
       " 0.6305879950523376,\n",
       " 0.6305497884750366,\n",
       " 0.6305269002914429,\n",
       " 0.630096971988678,\n",
       " 0.6296277046203613,\n",
       " 0.6295402646064758,\n",
       " 0.6295005679130554,\n",
       " 0.6292697787284851,\n",
       " 0.6290936470031738,\n",
       " 0.6286728382110596,\n",
       " 0.6286492943763733,\n",
       " 0.6285806894302368,\n",
       " 0.6283563375473022,\n",
       " 0.6279318332672119,\n",
       " 0.6277845501899719,\n",
       " 0.6277395486831665,\n",
       " 0.6270812749862671,\n",
       " 0.6268585920333862,\n",
       " 0.6256161332130432,\n",
       " 0.6254564523696899,\n",
       " 0.6253827214241028,\n",
       " 0.6252259016036987,\n",
       " 0.6251668334007263,\n",
       " 0.6251053810119629,\n",
       " 0.6239646077156067,\n",
       " 0.6235501170158386,\n",
       " 0.623266875743866,\n",
       " 0.6227879524230957,\n",
       " 0.6223676800727844,\n",
       " 0.6216578483581543,\n",
       " 0.620872974395752,\n",
       " 0.6207531690597534,\n",
       " 0.6204898357391357,\n",
       " 0.6199707388877869,\n",
       " 0.619939923286438,\n",
       " 0.619571328163147,\n",
       " 0.6194667220115662,\n",
       " 0.6189941167831421,\n",
       " 0.6188552975654602,\n",
       " 0.6185175776481628,\n",
       " 0.6182923316955566,\n",
       " 0.6180881261825562,\n",
       " 0.6179910898208618,\n",
       " 0.6179143786430359,\n",
       " 0.6177471876144409,\n",
       " 0.6172986030578613,\n",
       " 0.6171382069587708,\n",
       " 0.6170961856842041,\n",
       " 0.6170250177383423,\n",
       " 0.6169363260269165,\n",
       " 0.61668461561203,\n",
       " 0.6163849234580994,\n",
       " 0.616259753704071,\n",
       " 0.6160246133804321,\n",
       " 0.6159703731536865,\n",
       " 0.6155744194984436,\n",
       " 0.6147223114967346,\n",
       " 0.6145497560501099,\n",
       " 0.6140955686569214,\n",
       " 0.6140169501304626,\n",
       " 0.6139309406280518,\n",
       " 0.6138129830360413,\n",
       " 0.613811731338501,\n",
       " 0.6133521199226379,\n",
       " 0.6131906509399414,\n",
       " 0.6131694316864014,\n",
       " 0.6130494475364685,\n",
       " 0.612968921661377,\n",
       " 0.6129478812217712,\n",
       " 0.6116432547569275,\n",
       " 0.6116387844085693,\n",
       " 0.6113128662109375,\n",
       " 0.6108748316764832,\n",
       " 0.6107308864593506,\n",
       " 0.6106475591659546,\n",
       " 0.610262393951416,\n",
       " 0.6101744174957275,\n",
       " 0.6101717948913574,\n",
       " 0.6101569533348083,\n",
       " 0.6100338697433472,\n",
       " 0.6099213361740112,\n",
       " 0.6097745299339294,\n",
       " 0.6095500588417053,\n",
       " 0.6094909310340881,\n",
       " 0.6094450354576111,\n",
       " 0.6090566515922546,\n",
       " 0.6087564826011658,\n",
       " 0.6087318658828735,\n",
       " 0.6083419919013977,\n",
       " 0.6081241369247437,\n",
       " 0.607871413230896,\n",
       " 0.6077539920806885,\n",
       " 0.6068829894065857,\n",
       " 0.6065410375595093,\n",
       " 0.6062430739402771,\n",
       " 0.6058253645896912,\n",
       " 0.6055247783660889,\n",
       " 0.6053788065910339,\n",
       " 0.6053472757339478,\n",
       " 0.6052625775337219,\n",
       " 0.6048938632011414,\n",
       " 0.604708194732666,\n",
       " 0.6043591499328613,\n",
       " 0.6043391823768616,\n",
       " 0.6043362021446228,\n",
       " 0.6042796969413757,\n",
       " 0.6038782596588135,\n",
       " 0.6036362051963806,\n",
       " 0.6035915613174438,\n",
       " 0.6033117175102234,\n",
       " 0.6032079458236694,\n",
       " 0.6029709577560425,\n",
       " 0.6028017997741699,\n",
       " 0.6026512384414673,\n",
       " 0.6024478673934937,\n",
       " 0.602216899394989,\n",
       " 0.6018061637878418,\n",
       " 0.6017376184463501,\n",
       " 0.6012206673622131,\n",
       " 0.6007962226867676,\n",
       " 0.6006240248680115,\n",
       " 0.6005715131759644,\n",
       " 0.6003936529159546,\n",
       " 0.6003291606903076,\n",
       " 0.6000486612319946,\n",
       " 0.5996347665786743,\n",
       " 0.5994622707366943,\n",
       " 0.599257230758667,\n",
       " 0.5991039276123047,\n",
       " 0.5984320640563965,\n",
       " 0.5983387231826782,\n",
       " 0.5982711911201477,\n",
       " 0.5978343486785889,\n",
       " 0.5977898240089417,\n",
       " 0.5977823734283447,\n",
       " 0.5976536273956299,\n",
       " 0.5973724722862244,\n",
       " 0.5973426699638367,\n",
       " 0.597022533416748,\n",
       " 0.5968469381332397,\n",
       " 0.5965945720672607,\n",
       " 0.5965850353240967,\n",
       " 0.5964047312736511,\n",
       " 0.5962276458740234,\n",
       " 0.5960531830787659,\n",
       " 0.5958200097084045,\n",
       " 0.595647394657135,\n",
       " 0.5950970649719238,\n",
       " 0.5949172377586365,\n",
       " 0.5947761535644531,\n",
       " 0.5942966938018799,\n",
       " 0.5942276120185852,\n",
       " 0.5939619541168213,\n",
       " 0.5939454436302185,\n",
       " 0.5937827229499817,\n",
       " 0.5937712788581848,\n",
       " 0.593761682510376,\n",
       " 0.5936322212219238,\n",
       " 0.5930981636047363,\n",
       " 0.5929948091506958,\n",
       " 0.592664361000061,\n",
       " 0.5925729870796204,\n",
       " 0.5924862623214722,\n",
       " 0.5924296379089355,\n",
       " 0.5921010375022888,\n",
       " 0.591584324836731,\n",
       " 0.5915680527687073,\n",
       " 0.5915402173995972,\n",
       " 0.5914790630340576,\n",
       " 0.5913110375404358,\n",
       " 0.5911639928817749,\n",
       " 0.5910633206367493,\n",
       " 0.5909237265586853,\n",
       " 0.5906891226768494,\n",
       " 0.5906646847724915,\n",
       " 0.5903602838516235,\n",
       " 0.5902671217918396,\n",
       " 0.5898656845092773,\n",
       " 0.5893831849098206,\n",
       " 0.5891802310943604,\n",
       " 0.5889486074447632,\n",
       " 0.5889463424682617,\n",
       " 0.5887507200241089,\n",
       " 0.588686466217041,\n",
       " 0.5885319113731384,\n",
       " 0.588383138179779,\n",
       " 0.5882313847541809,\n",
       " 0.5881727933883667,\n",
       " 0.5880797505378723,\n",
       " 0.5878171324729919,\n",
       " 0.587637186050415,\n",
       " 0.5874754786491394,\n",
       " 0.5874388217926025,\n",
       " 0.5873831510543823,\n",
       " 0.5872762203216553,\n",
       " 0.5871644020080566,\n",
       " 0.5870528817176819,\n",
       " 0.5870521664619446,\n",
       " 0.5866469740867615,\n",
       " 0.5865029096603394,\n",
       " 0.5861585140228271,\n",
       " 0.5860657691955566,\n",
       " 0.585838794708252,\n",
       " 0.5858339071273804,\n",
       " 0.5855178833007812,\n",
       " 0.5853433609008789,\n",
       " 0.5853050351142883,\n",
       " 0.584908664226532,\n",
       " 0.5847539901733398,\n",
       " 0.5840827226638794,\n",
       " 0.5840543508529663,\n",
       " 0.5838108062744141,\n",
       " 0.5837247967720032,\n",
       " 0.5834543704986572,\n",
       " 0.583451509475708,\n",
       " 0.5830873250961304,\n",
       " 0.5830100774765015,\n",
       " 0.5828806757926941,\n",
       " 0.5824500322341919,\n",
       " 0.5822259187698364,\n",
       " 0.5821603536605835,\n",
       " 0.5813983678817749,\n",
       " 0.5812755823135376,\n",
       " 0.5811508893966675,\n",
       " 0.5811348557472229,\n",
       " 0.5810601115226746,\n",
       " 0.5809980034828186,\n",
       " 0.5809274315834045,\n",
       " 0.5805932879447937,\n",
       " 0.5805398225784302,\n",
       " 0.5800670385360718,\n",
       " 0.5799999833106995,\n",
       " 0.579505443572998,\n",
       " 0.5792343020439148,\n",
       " 0.5792225003242493,\n",
       " 0.5791121125221252,\n",
       " 0.5790188312530518,\n",
       " 0.5788261890411377,\n",
       " 0.5787473320960999,\n",
       " 0.5786319375038147,\n",
       " 0.5783180594444275,\n",
       " 0.578109622001648,\n",
       " 0.5780721306800842,\n",
       " 0.5780323147773743,\n",
       " 0.5780043601989746,\n",
       " 0.5771296620368958,\n",
       " 0.5771205425262451,\n",
       " 0.5770235657691956,\n",
       " 0.5770174264907837,\n",
       " 0.5769822597503662,\n",
       " 0.5766937732696533,\n",
       " 0.5766468048095703,\n",
       " 0.5765354037284851,\n",
       " 0.575879693031311,\n",
       " 0.5758066773414612,\n",
       " 0.5757375955581665,\n",
       " 0.5753865838050842,\n",
       " 0.5751538276672363,\n",
       " 0.5751133561134338,\n",
       " 0.5749435424804688,\n",
       " ...]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['residual'].values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "percentile99 = np.percentile(df['residual'], 99)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeoAAAFYCAYAAACcb79EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ecdfec10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.distplot(df['residual'].sample(n=10000),  kde=False, rug=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "42798"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "residuals = df['residual'].values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "60.033355519175529"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.max(residuals) - np.min(residuals)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def RadiantDireDifference(row, column):\n",
    "    faction = ['radiant', 'dire']\n",
    "    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']\n",
    "    radiantSum = 0\n",
    "    direSum = 0\n",
    "    for pos in poss:\n",
    "        radiantSum += row['radiant_'+ pos + '_' + column]\n",
    "        direSum += row['dire_'+ pos + '_' + column]\n",
    "    return radiantSum - direSum\n",
    "\n",
    "def reject_outliers(data, m=1.5):\n",
    "    return abs(data - np.mean(data)) < m * np.std(data)\n",
    "\n",
    "\n",
    "def heroKDAdeficit(row):\n",
    "    faction = ['radiant', 'dire']\n",
    "    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']\n",
    "    largestAbsoluteSoFar = -1\n",
    "    returnVal = -1\n",
    "    for pos in poss:\n",
    "        for fac in faction:\n",
    "            value = row[fac + '_' + pos + '_kills'] + \\\n",
    "                        row[fac + '_' + pos + '_assists'] - \\\n",
    "                        row[fac + '_' + pos + '_deaths']\n",
    "            if abs(value) >= largestAbsoluteSoFar:\n",
    "                largestAbsoluteSoFar = abs(value)\n",
    "                returnVal = value\n",
    "    return returnVal\n",
    "\n",
    "def heroKDdeficit(row):\n",
    "    faction = ['radiant', 'dire']\n",
    "    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']\n",
    "    largestAbsoluteSoFar = -1\n",
    "    returnVal = -1\n",
    "    for pos in poss:\n",
    "        for fac in faction:\n",
    "            value = row[fac + '_' + pos + '_kills'] - \\\n",
    "                        row[fac + '_' + pos + '_deaths']\n",
    "            if abs(value) >= largestAbsoluteSoFar:\n",
    "                largestAbsoluteSoFar = abs(value)\n",
    "                returnVal = value\n",
    "    return returnVal\n",
    "\n",
    "def heroKD(row):\n",
    "    faction = ['radiant', 'dire']\n",
    "    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']\n",
    "    largestAbsoluteSoFar = -1\n",
    "    returnVal = -1\n",
    "    for pos in poss:\n",
    "        for fac in faction:\n",
    "            value = row[fac + '_' + pos + '_kills'] - \\\n",
    "                        row[fac + '_' + pos + '_deaths']\n",
    "            if abs(value) >= largestAbsoluteSoFar:\n",
    "                largestAbsoluteSoFar = abs(value)\n",
    "                returnVal = int(pos[3:4])\n",
    "    return returnVal\n",
    "\n",
    "def heroGoldSpent(row):\n",
    "    faction = ['radiant', 'dire']\n",
    "    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']\n",
    "    largestAbsoluteSoFar = -1\n",
    "    returnVal = -1\n",
    "    for pos in poss:\n",
    "        for fac in faction:\n",
    "            value = row[fac + '_' + pos + '_gold_spent']\n",
    "            if abs(value) >= largestAbsoluteSoFar:\n",
    "                largestAbsoluteSoFar = abs(value)\n",
    "                returnVal = int(pos[3:4])\n",
    "    return returnVal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "vote_mask = (df['positive_votes'] == 0) & (df['negative_votes'] == 0)\n",
    "def outlier_mask(data, other_mask = vote_mask, m = 2.):\n",
    "    d = np.abs(data - np.median(data))\n",
    "    mdev = np.median(d)\n",
    "    s = d/mdev if mdev else 0.\n",
    "    return np.logical_and(s<m, other_mask)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFYCAYAAAB+s6Q9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n39ZfAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ebd81790>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "kill_difference = np.abs(df.apply(RadiantDireDifference, args=('kills',), axis=1))\n",
    "kill_difference_mask = outlier_mask(kill_difference)\n",
    "sns.regplot(kill_difference[kill_difference_mask], np.log(df['residual'][kill_difference_mask]))\n",
    "ax = plt.gca()\n",
    "ax.set_xlabel(\"Kill difference\")\n",
    "ax.set_ylabel(\"Logarithm of Residual\")\n",
    "plt.savefig('graphs/Kill_Difference.png', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfcAAAFYCAYAAABOP7UcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nyGIY7kRERBbDcCciIrIYhjsREZHF/D93CRZIE+mRJwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ebbe5d10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "gold_difference = np.abs(df.apply(RadiantDireDifference, args=('gpm',), axis=1))\n",
    "gold_difference_mask = outlier_mask(gold_difference)\n",
    "sns.regplot(np.abs(gold_difference)[gold_difference_mask], df['residual'][gold_difference_mask])\n",
    "ax = plt.gca()\n",
    "ax.set_xlabel(\"GPM difference\")\n",
    "ax.set_ylabel(\"Raw Residual\")\n",
    "plt.savefig('graphs/GPM_Difference.png', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe8AAAFYCAYAAAB6RnQAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ebdfa790>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "kda = df.apply(heroKDAdeficit, axis=1)\n",
    "kda_mask = outlier_mask(kda)\n",
    "sns.regplot(kda[kda_mask], df['residual'][kda_mask])\n",
    "ax = plt.gca()\n",
    "ax.set_xlabel(\"K + A - D\")\n",
    "ax.set_ylabel(\"Raw Residual\")\n",
    "plt.savefig('graphs/MaxKAD_Difference.png', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFYCAYAAAB+s6Q9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nfVu7TlOdAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ebe01c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "kd = df.apply(heroKDdeficit, axis=1)\n",
    "kd_mask = outlier_mask(kd)\n",
    "sns.regplot(np.abs(kd)[kd_mask], df['residual'][kd_mask])\n",
    "ax = plt.gca()\n",
    "ax.set_xlabel(\"Absolute(K - D)\")\n",
    "ax.set_ylabel(\"Raw Residual\")\n",
    "plt.savefig('graphs/absolutek-d_Difference.png', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFYCAYAAAB+s6Q9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ebe93e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "direpos2kills = df['dire_pos2_kills']\n",
    "direpos2kills_mask = outlier_mask(direpos2kills)\n",
    "sns.regplot(df['dire_pos2_kills'][direpos2kills_mask], df['residual'][direpos2kills_mask])\n",
    "ax = plt.gca()\n",
    "ax.set_xlabel(\"Dire Pos 2 Kills\")\n",
    "ax.set_ylabel(\"Raw Residual\")\n",
    "plt.savefig('graphs/direpos2kills.png', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFYCAYAAAB+s6Q9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nERFpBJM2ERGRRjBpExERacT/AV1MDmDVCymJAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ebed9110>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "radiantpos1kills = df['radiant_pos1_kills']\n",
    "radiantpos1kills_mask = outlier_mask(radiantpos1kills)\n",
    "sns.regplot(df['radiant_pos1_kills'][radiantpos1kills_mask], df['residual'][radiantpos1kills_mask])\n",
    "ax = plt.gca()\n",
    "ax.set_xlabel(\"Radiant Pos 1 Kills\")\n",
    "ax.set_ylabel(\"Raw Residual\")\n",
    "plt.savefig('graphs/radiantpos1kills.png', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFYCAYAAAB+s6Q9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nm4iISBD/D81/iH/NJ351AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ebf42990>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "radiantpos1kd = df['radiant_pos1_kills'] - df['radiant_pos1_deaths']\n",
    "radiantpos1kd_mask = outlier_mask(radiantpos1kd)\n",
    "sns.regplot(radiantpos1kd[radiantpos1kd_mask], df['residual'][radiantpos1kd_mask])\n",
    "ax = plt.gca()\n",
    "ax.set_xlabel(\"Radiant Pos 1 Kills - Deaths\")\n",
    "ax.set_ylabel(\"Raw Residual\")\n",
    "plt.savefig('graphs/radiantpos1killsminusdeaths.png', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFYCAYAAAB+s6Q9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIiJSiP8fxHeMBbkA3sUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ebf9fd50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "radiantpos5kd = df['radiant_pos5_kills'] - df['radiant_pos5_deaths']\n",
    "radiantpos5kd_mask = outlier_mask(radiantpos5kd)\n",
    "sns.regplot(radiantpos5kd[radiantpos5kd_mask], df['residual'][radiantpos5kd_mask])\n",
    "ax = plt.gca()\n",
    "ax.set_xlabel(\"Radiant Pos 5 Kills - Deaths\")\n",
    "ax.set_ylabel(\"Raw Residual\")\n",
    "plt.savefig('graphs/radiantpos5killsminusdeaths.png', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFYCAYAAAB+s6Q9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAGyC0AYAwCb+P+lucfjNO0QiAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f97ebfec9d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "duration = df['duration']\n",
    "duration_mask = outlier_mask(duration)\n",
    "sns.regplot(df['duration'][duration_mask], df['residual'][duration_mask])\n",
    "ax = plt.gca()\n",
    "ax.set_xlabel(\"Duration\")\n",
    "ax.set_ylabel(\"Raw Residual\")\n",
    "plt.savefig('graphs/durationresidual.png', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>match_id</th>\n",
       "      <th>positive_votes</th>\n",
       "      <th>negative_votes</th>\n",
       "      <th>first_blood_time</th>\n",
       "      <th>radiant_win</th>\n",
       "      <th>duration</th>\n",
       "      <th>dire_pos1_ancient_kills</th>\n",
       "      <th>dire_pos1_courier_kills</th>\n",
       "      <th>dire_pos1_kills</th>\n",
       "      <th>dire_pos1_neutral_kills</th>\n",
       "      <th>...</th>\n",
       "      <th>dire_pos2_abandons</th>\n",
       "      <th>dire_pos3_abandons</th>\n",
       "      <th>dire_pos4_abandons</th>\n",
       "      <th>dire_pos5_abandons</th>\n",
       "      <th>radiant_pos1_abandons</th>\n",
       "      <th>radiant_pos2_abandons</th>\n",
       "      <th>radiant_pos3_abandons</th>\n",
       "      <th>radiant_pos4_abandons</th>\n",
       "      <th>radiant_pos5_abandons</th>\n",
       "      <th>residual</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>40056</th>\n",
       "      <td>3215497967</td>\n",
       "      <td>49</td>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "      <td>True</td>\n",
       "      <td>2748</td>\n",
       "      <td>43.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>141.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>60.093830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3222</th>\n",
       "      <td>3215097220</td>\n",
       "      <td>15</td>\n",
       "      <td>7</td>\n",
       "      <td>15</td>\n",
       "      <td>True</td>\n",
       "      <td>2470</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>108.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>22.980583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42706</th>\n",
       "      <td>3215312675</td>\n",
       "      <td>9</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>2687</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10</td>\n",
       "      <td>9.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>14.018461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14505</th>\n",
       "      <td>3215162338</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>135</td>\n",
       "      <td>False</td>\n",
       "      <td>3981</td>\n",
       "      <td>18.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>27</td>\n",
       "      <td>70.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>12.276542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8310</th>\n",
       "      <td>3215181105</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>10.960655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6186</th>\n",
       "      <td>3215377994</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>72</td>\n",
       "      <td>False</td>\n",
       "      <td>2924</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.190336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10574</th>\n",
       "      <td>3215271631</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>False</td>\n",
       "      <td>1846</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "      <td>36.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9.797115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28970</th>\n",
       "      <td>3215289100</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.665862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32230</th>\n",
       "      <td>3215279863</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.291083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1609</th>\n",
       "      <td>3215253055</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>182</td>\n",
       "      <td>True</td>\n",
       "      <td>1949</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>6.753566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41758</th>\n",
       "      <td>3215448302</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>208</td>\n",
       "      <td>False</td>\n",
       "      <td>7375</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>21</td>\n",
       "      <td>61.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.982767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10214</th>\n",
       "      <td>3215416370</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5.009284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42735</th>\n",
       "      <td>3215261563</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4.787800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9080</th>\n",
       "      <td>3215294948</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>94</td>\n",
       "      <td>True</td>\n",
       "      <td>1804</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>19.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4.737586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8192</th>\n",
       "      <td>3215155134</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>780</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4.151326</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12556</th>\n",
       "      <td>3215350255</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.858672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15967</th>\n",
       "      <td>3215341221</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>932</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>34</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.720057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21049</th>\n",
       "      <td>3215306187</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.653452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5501</th>\n",
       "      <td>3215302636</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3.612013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28073</th>\n",
       "      <td>3215316820</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3.557962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37873</th>\n",
       "      <td>3215190109</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>324</td>\n",
       "      <td>False</td>\n",
       "      <td>4877</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>24.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.524932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19538</th>\n",
       "      <td>3215355847</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3.455053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18809</th>\n",
       "      <td>3215130854</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>True</td>\n",
       "      <td>2760</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.430158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37087</th>\n",
       "      <td>3215259897</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>1394</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.926127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16205</th>\n",
       "      <td>3215394671</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.813246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37604</th>\n",
       "      <td>3215152225</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>2641</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.767648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37101</th>\n",
       "      <td>3215274711</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>858</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.758488</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24635</th>\n",
       "      <td>3215366609</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.543581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1345</th>\n",
       "      <td>3215449992</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>104</td>\n",
       "      <td>False</td>\n",
       "      <td>5045</td>\n",
       "      <td>60.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18</td>\n",
       "      <td>317.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.515880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26691</th>\n",
       "      <td>3215376405</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.483158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20347</th>\n",
       "      <td>3215337531</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>62</td>\n",
       "      <td>False</td>\n",
       "      <td>3373</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.769491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15810</th>\n",
       "      <td>3215102476</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>137</td>\n",
       "      <td>False</td>\n",
       "      <td>3312</td>\n",
       "      <td>45.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8</td>\n",
       "      <td>350.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.767143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27774</th>\n",
       "      <td>3215253869</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>292</td>\n",
       "      <td>False</td>\n",
       "      <td>5234</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.766957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27894</th>\n",
       "      <td>3215422166</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>118</td>\n",
       "      <td>False</td>\n",
       "      <td>2817</td>\n",
       "      <td>47.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10</td>\n",
       "      <td>155.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.765579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12285</th>\n",
       "      <td>3215263577</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>1143</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>35</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.765576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20308</th>\n",
       "      <td>3215285461</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "      <td>False</td>\n",
       "      <td>2682</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10</td>\n",
       "      <td>74.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.764146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8590</th>\n",
       "      <td>3215114749</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>746</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.763206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14222</th>\n",
       "      <td>3215096854</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>93</td>\n",
       "      <td>True</td>\n",
       "      <td>2936</td>\n",
       "      <td>21.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "      <td>208.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.762437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8943</th>\n",
       "      <td>3215446534</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>218</td>\n",
       "      <td>True</td>\n",
       "      <td>1714</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>25.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.761831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21170</th>\n",
       "      <td>3215119491</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>169</td>\n",
       "      <td>True</td>\n",
       "      <td>4051</td>\n",
       "      <td>75.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>354.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.761589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9693</th>\n",
       "      <td>3215205453</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>140</td>\n",
       "      <td>False</td>\n",
       "      <td>2083</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.759363</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10663</th>\n",
       "      <td>3215135325</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>3645</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>143.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.759099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11652</th>\n",
       "      <td>3215415463</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>191</td>\n",
       "      <td>False</td>\n",
       "      <td>2846</td>\n",
       "      <td>31.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8</td>\n",
       "      <td>175.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.758527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30828</th>\n",
       "      <td>3215322189</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>93</td>\n",
       "      <td>True</td>\n",
       "      <td>2680</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>12.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.758093</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15394</th>\n",
       "      <td>3215345510</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>307</td>\n",
       "      <td>True</td>\n",
       "      <td>4429</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.757711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7846</th>\n",
       "      <td>3215138506</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>160</td>\n",
       "      <td>False</td>\n",
       "      <td>1780</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.757519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37444</th>\n",
       "      <td>3215492929</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>73</td>\n",
       "      <td>False</td>\n",
       "      <td>3048</td>\n",
       "      <td>20.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8</td>\n",
       "      <td>115.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.757248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19730</th>\n",
       "      <td>3215267165</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>True</td>\n",
       "      <td>2125</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.756915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32004</th>\n",
       "      <td>3215123467</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>224</td>\n",
       "      <td>False</td>\n",
       "      <td>3050</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>12.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.756316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>288</th>\n",
       "      <td>3215358156</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>163</td>\n",
       "      <td>True</td>\n",
       "      <td>4185</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>55.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.754428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36615</th>\n",
       "      <td>3215094142</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>203</td>\n",
       "      <td>True</td>\n",
       "      <td>4973</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.754374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11493</th>\n",
       "      <td>3215328883</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>True</td>\n",
       "      <td>3377</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>179.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.753980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39744</th>\n",
       "      <td>3215250045</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>184</td>\n",
       "      <td>True</td>\n",
       "      <td>4439</td>\n",
       "      <td>44.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>31</td>\n",
       "      <td>132.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.753625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9634</th>\n",
       "      <td>3215410384</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>344</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.753601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17709</th>\n",
       "      <td>3215168749</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>413</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.753267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27443</th>\n",
       "      <td>3215343292</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>86</td>\n",
       "      <td>True</td>\n",
       "      <td>3411</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.753174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39465</th>\n",
       "      <td>3215107300</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>108</td>\n",
       "      <td>True</td>\n",
       "      <td>2965</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9</td>\n",
       "      <td>88.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.752881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29673</th>\n",
       "      <td>3215450701</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>90</td>\n",
       "      <td>False</td>\n",
       "      <td>2907</td>\n",
       "      <td>45.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8</td>\n",
       "      <td>197.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.752879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22278</th>\n",
       "      <td>3215239658</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>126</td>\n",
       "      <td>False</td>\n",
       "      <td>3992</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.752695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32637</th>\n",
       "      <td>3215427391</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>169</td>\n",
       "      <td>True</td>\n",
       "      <td>4418</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.751674</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>428 rows × 247 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         match_id  positive_votes  negative_votes  first_blood_time  \\\n",
       "40056  3215497967              49               2                14   \n",
       "3222   3215097220              15               7                15   \n",
       "42706  3215312675               9               4                 0   \n",
       "14505  3215162338               0               0               135   \n",
       "8310   3215181105               0               0                 0   \n",
       "6186   3215377994               0               0                72   \n",
       "10574  3215271631               7               1                 3   \n",
       "28970  3215289100               0               0                 0   \n",
       "32230  3215279863               0               0                 0   \n",
       "1609   3215253055               5               0               182   \n",
       "41758  3215448302               0               0               208   \n",
       "10214  3215416370               0               0                 0   \n",
       "42735  3215261563               0               0                 0   \n",
       "9080   3215294948               4               1                94   \n",
       "8192   3215155134               0               0                 0   \n",
       "12556  3215350255               0               0                 0   \n",
       "15967  3215341221               0               0                 0   \n",
       "21049  3215306187               0               0                 0   \n",
       "5501   3215302636               0               0                 0   \n",
       "28073  3215316820               0               0                 0   \n",
       "37873  3215190109               0               0               324   \n",
       "19538  3215355847               0               0                 0   \n",
       "18809  3215130854               0               0                 9   \n",
       "37087  3215259897               0               0                 0   \n",
       "16205  3215394671               0               0                 0   \n",
       "37604  3215152225               0               0                 0   \n",
       "37101  3215274711               0               0                 0   \n",
       "24635  3215366609               0               0                 0   \n",
       "1345   3215449992               0               0               104   \n",
       "26691  3215376405               0               0                 0   \n",
       "...           ...             ...             ...               ...   \n",
       "20347  3215337531               0               0                62   \n",
       "15810  3215102476               0               0               137   \n",
       "27774  3215253869               0               0               292   \n",
       "27894  3215422166               0               0               118   \n",
       "12285  3215263577               0               0                 0   \n",
       "20308  3215285461               0               0                61   \n",
       "8590   3215114749               0               0                 0   \n",
       "14222  3215096854               0               0                93   \n",
       "8943   3215446534               0               0               218   \n",
       "21170  3215119491               0               0               169   \n",
       "9693   3215205453               0               0               140   \n",
       "10663  3215135325               0               0                 0   \n",
       "11652  3215415463               0               0               191   \n",
       "30828  3215322189               0               0                93   \n",
       "15394  3215345510               0               0               307   \n",
       "7846   3215138506               0               0               160   \n",
       "37444  3215492929               0               0                73   \n",
       "19730  3215267165               0               0                 2   \n",
       "32004  3215123467               0               0               224   \n",
       "288    3215358156               0               0               163   \n",
       "36615  3215094142               0               0               203   \n",
       "11493  3215328883               0               0                 3   \n",
       "39744  3215250045               0               0               184   \n",
       "9634   3215410384               0               0                 0   \n",
       "17709  3215168749               0               0                 0   \n",
       "27443  3215343292               0               0                86   \n",
       "39465  3215107300               0               0               108   \n",
       "29673  3215450701               0               0                90   \n",
       "22278  3215239658               0               0               126   \n",
       "32637  3215427391               0               0               169   \n",
       "\n",
       "      radiant_win  duration  dire_pos1_ancient_kills  dire_pos1_courier_kills  \\\n",
       "40056        True      2748                     43.0                      0.0   \n",
       "3222         True      2470                     23.0                      0.0   \n",
       "42706       False      2687                      0.0                      0.0   \n",
       "14505       False      3981                     18.0                     21.0   \n",
       "8310         True         0                      NaN                      NaN   \n",
       "6186        False      2924                      NaN                      NaN   \n",
       "10574       False      1846                      4.0                      0.0   \n",
       "28970       False         0                      NaN                      NaN   \n",
       "32230       False         0                      NaN                      NaN   \n",
       "1609         True      1949                      0.0                      0.0   \n",
       "41758       False      7375                      4.0                      1.0   \n",
       "10214       False         0                      NaN                      NaN   \n",
       "42735        True         0                      NaN                      NaN   \n",
       "9080         True      1804                      0.0                      0.0   \n",
       "8192         True       780                      0.0                      0.0   \n",
       "12556        True         0                      NaN                      NaN   \n",
       "15967       False       932                      NaN                      NaN   \n",
       "21049        True         0                      NaN                      NaN   \n",
       "5501        False         0                      NaN                      NaN   \n",
       "28073       False         0                      NaN                      NaN   \n",
       "37873       False      4877                      3.0                      0.0   \n",
       "19538       False         0                      NaN                      NaN   \n",
       "18809        True      2760                      0.0                      0.0   \n",
       "37087        True      1394                      3.0                      0.0   \n",
       "16205        True         0                      NaN                      NaN   \n",
       "37604        True      2641                      NaN                      NaN   \n",
       "37101        True       858                      NaN                      NaN   \n",
       "24635       False         0                      NaN                      NaN   \n",
       "1345        False      5045                     60.0                      0.0   \n",
       "26691        True         0                      NaN                      NaN   \n",
       "...           ...       ...                      ...                      ...   \n",
       "20347       False      3373                      NaN                      NaN   \n",
       "15810       False      3312                     45.0                      0.0   \n",
       "27774       False      5234                      NaN                      NaN   \n",
       "27894       False      2817                     47.0                      1.0   \n",
       "12285       False      1143                      NaN                      NaN   \n",
       "20308       False      2682                     20.0                      0.0   \n",
       "8590        False       746                      NaN                      NaN   \n",
       "14222        True      2936                     21.0                      0.0   \n",
       "8943         True      1714                      0.0                      0.0   \n",
       "21170        True      4051                     75.0                      0.0   \n",
       "9693        False      2083                      NaN                      NaN   \n",
       "10663        True      3645                     23.0                      0.0   \n",
       "11652       False      2846                     31.0                      0.0   \n",
       "30828        True      2680                      0.0                      0.0   \n",
       "15394        True      4429                      NaN                      NaN   \n",
       "7846        False      1780                      NaN                      NaN   \n",
       "37444       False      3048                     20.0                      1.0   \n",
       "19730        True      2125                      NaN                      NaN   \n",
       "32004       False      3050                      0.0                      0.0   \n",
       "288          True      4185                     24.0                      0.0   \n",
       "36615        True      4973                      NaN                      NaN   \n",
       "11493        True      3377                     37.0                      0.0   \n",
       "39744        True      4439                     44.0                      0.0   \n",
       "9634         True       344                      NaN                      NaN   \n",
       "17709        True       413                      0.0                      0.0   \n",
       "27443        True      3411                      NaN                      NaN   \n",
       "39465        True      2965                      8.0                      0.0   \n",
       "29673       False      2907                     45.0                      1.0   \n",
       "22278       False      3992                      NaN                      NaN   \n",
       "32637        True      4418                      NaN                      NaN   \n",
       "\n",
       "       dire_pos1_kills  dire_pos1_neutral_kills    ...      \\\n",
       "40056                4                    141.0    ...       \n",
       "3222                 7                    108.0    ...       \n",
       "42706               10                      9.0    ...       \n",
       "14505               27                     70.0    ...       \n",
       "8310                 0                      NaN    ...       \n",
       "6186                28                      NaN    ...       \n",
       "10574                5                     36.0    ...       \n",
       "28970                0                      NaN    ...       \n",
       "32230                0                      NaN    ...       \n",
       "1609                 4                     10.0    ...       \n",
       "41758               21                     61.0    ...       \n",
       "10214                0                      NaN    ...       \n",
       "42735                0                      NaN    ...       \n",
       "9080                 4                     19.0    ...       \n",
       "8192                 0                      0.0    ...       \n",
       "12556                0                      NaN    ...       \n",
       "15967               34                      NaN    ...       \n",
       "21049                0                      NaN    ...       \n",
       "5501                 0                      NaN    ...       \n",
       "28073                0                      NaN    ...       \n",
       "37873                6                     24.0    ...       \n",
       "19538                0                      NaN    ...       \n",
       "18809                3                      0.0    ...       \n",
       "37087                1                      3.0    ...       \n",
       "16205                0                      NaN    ...       \n",
       "37604               10                      NaN    ...       \n",
       "37101                0                      NaN    ...       \n",
       "24635                0                      NaN    ...       \n",
       "1345                18                    317.0    ...       \n",
       "26691                0                      NaN    ...       \n",
       "...                ...                      ...    ...       \n",
       "20347                1                      NaN    ...       \n",
       "15810                8                    350.0    ...       \n",
       "27774                7                      NaN    ...       \n",
       "27894               10                    155.0    ...       \n",
       "12285               35                      NaN    ...       \n",
       "20308               10                     74.0    ...       \n",
       "8590                 3                      NaN    ...       \n",
       "14222                5                    208.0    ...       \n",
       "8943                 1                     25.0    ...       \n",
       "21170               12                    354.0    ...       \n",
       "9693                 3                      NaN    ...       \n",
       "10663                4                    143.0    ...       \n",
       "11652                8                    175.0    ...       \n",
       "30828                3                     12.0    ...       \n",
       "15394               10                      NaN    ...       \n",
       "7846                 4                      NaN    ...       \n",
       "37444                8                    115.0    ...       \n",
       "19730                6                      NaN    ...       \n",
       "32004                2                     12.0    ...       \n",
       "288                  7                     55.0    ...       \n",
       "36615                4                      NaN    ...       \n",
       "11493               12                    179.0    ...       \n",
       "39744               31                    132.0    ...       \n",
       "9634                 0                      NaN    ...       \n",
       "17709                0                      0.0    ...       \n",
       "27443                7                      NaN    ...       \n",
       "39465                9                     88.0    ...       \n",
       "29673                8                    197.0    ...       \n",
       "22278               10                      NaN    ...       \n",
       "32637                6                      NaN    ...       \n",
       "\n",
       "       dire_pos2_abandons  dire_pos3_abandons  dire_pos4_abandons  \\\n",
       "40056                   0                   0                   0   \n",
       "3222                    0                   0                   0   \n",
       "42706                   0                   0                   0   \n",
       "14505                   0                   0                   0   \n",
       "8310                    0                   0                   0   \n",
       "6186                    0                   0                   0   \n",
       "10574                   0                   0                   0   \n",
       "28970                   0                   0                   1   \n",
       "32230                   0                   0                   1   \n",
       "1609                    0                   0                   0   \n",
       "41758                   0                   0                   0   \n",
       "10214                   0                   0                   0   \n",
       "42735                   1                   0                   0   \n",
       "9080                    0                   0                   0   \n",
       "8192                    0                   0                   0   \n",
       "12556                   0                   0                   1   \n",
       "15967                   0                   0                   0   \n",
       "21049                   0                   1                   0   \n",
       "5501                    0                   0                   0   \n",
       "28073                   0                   0                   0   \n",
       "37873                   0                   0                   0   \n",
       "19538                   0                   0                   0   \n",
       "18809                   0                   0                   0   \n",
       "37087                   0                   0                   0   \n",
       "16205                   0                   0                   1   \n",
       "37604                   0                   0                   0   \n",
       "37101                   0                   0                   0   \n",
       "24635                   0                   0                   0   \n",
       "1345                    0                   0                   0   \n",
       "26691                   0                   0                   1   \n",
       "...                   ...                 ...                 ...   \n",
       "20347                   0                   0                   0   \n",
       "15810                   0                   0                   0   \n",
       "27774                   0                   0                   0   \n",
       "27894                   0                   0                   0   \n",
       "12285                   0                   0                   0   \n",
       "20308                   0                   0                   0   \n",
       "8590                    0                   0                   0   \n",
       "14222                   0                   0                   0   \n",
       "8943                    0                   0                   0   \n",
       "21170                   0                   0                   0   \n",
       "9693                    0                   1                   0   \n",
       "10663                   0                   0                   0   \n",
       "11652                   0                   0                   0   \n",
       "30828                   0                   1                   0   \n",
       "15394                   0                   0                   0   \n",
       "7846                    1                   1                   1   \n",
       "37444                   0                   1                   0   \n",
       "19730                   0                   0                   0   \n",
       "32004                   0                   0                   0   \n",
       "288                     0                   0                   0   \n",
       "36615                   0                   0                   0   \n",
       "11493                   0                   0                   0   \n",
       "39744                   0                   0                   0   \n",
       "9634                    0                   0                   0   \n",
       "17709                   0                   0                   1   \n",
       "27443                   0                   0                   0   \n",
       "39465                   0                   0                   0   \n",
       "29673                   0                   0                   0   \n",
       "22278                   0                   0                   1   \n",
       "32637                   0                   1                   0   \n",
       "\n",
       "       dire_pos5_abandons  radiant_pos1_abandons  radiant_pos2_abandons  \\\n",
       "40056                   0                      0                      0   \n",
       "3222                    0                      0                      0   \n",
       "42706                   0                      0                      0   \n",
       "14505                   0                      0                      0   \n",
       "8310                    0                      0                      0   \n",
       "6186                    0                      0                      0   \n",
       "10574                   0                      0                      0   \n",
       "28970                   0                      0                      0   \n",
       "32230                   0                      0                      0   \n",
       "1609                    0                      0                      0   \n",
       "41758                   0                      0                      0   \n",
       "10214                   0                      0                      0   \n",
       "42735                   0                      0                      0   \n",
       "9080                    0                      0                      0   \n",
       "8192                    0                      0                      0   \n",
       "12556                   0                      0                      0   \n",
       "15967                   0                      0                      0   \n",
       "21049                   0                      0                      0   \n",
       "5501                    0                      0                      0   \n",
       "28073                   0                      0                      0   \n",
       "37873                   0                      0                      1   \n",
       "19538                   0                      0                      0   \n",
       "18809                   0                      0                      0   \n",
       "37087                   0                      0                      0   \n",
       "16205                   0                      0                      0   \n",
       "37604                   0                      0                      0   \n",
       "37101                   0                      0                      0   \n",
       "24635                   0                      0                      0   \n",
       "1345                    0                      0                      0   \n",
       "26691                   0                      0                      0   \n",
       "...                   ...                    ...                    ...   \n",
       "20347                   0                      0                      1   \n",
       "15810                   0                      0                      0   \n",
       "27774                   0                      0                      0   \n",
       "27894                   0                      0                      0   \n",
       "12285                   0                      0                      0   \n",
       "20308                   0                      0                      0   \n",
       "8590                    0                      1                      0   \n",
       "14222                   0                      0                      0   \n",
       "8943                    0                      0                      0   \n",
       "21170                   0                      0                      0   \n",
       "9693                    1                      0                      0   \n",
       "10663                   0                      0                      0   \n",
       "11652                   0                      0                      0   \n",
       "30828                   0                      0                      0   \n",
       "15394                   1                      0                      0   \n",
       "7846                    0                      0                      0   \n",
       "37444                   0                      0                      0   \n",
       "19730                   0                      1                      1   \n",
       "32004                   0                      0                      0   \n",
       "288                     0                      0                      0   \n",
       "36615                   0                      0                      0   \n",
       "11493                   0                      0                      0   \n",
       "39744                   0                      0                      0   \n",
       "9634                    0                      0                      0   \n",
       "17709                   0                      0                      0   \n",
       "27443                   1                      0                      0   \n",
       "39465                   0                      0                      0   \n",
       "29673                   0                      0                      0   \n",
       "22278                   0                      0                      0   \n",
       "32637                   0                      0                      0   \n",
       "\n",
       "       radiant_pos3_abandons  radiant_pos4_abandons  radiant_pos5_abandons  \\\n",
       "40056                      0                      0                      0   \n",
       "3222                       0                      0                      0   \n",
       "42706                      0                      0                      0   \n",
       "14505                      0                      0                      0   \n",
       "8310                       0                      1                      0   \n",
       "6186                       0                      0                      0   \n",
       "10574                      0                      0                      0   \n",
       "28970                      0                      0                      0   \n",
       "32230                      0                      0                      0   \n",
       "1609                       0                      0                      0   \n",
       "41758                      0                      0                      0   \n",
       "10214                      0                      1                      0   \n",
       "42735                      0                      0                      0   \n",
       "9080                       0                      0                      0   \n",
       "8192                       0                      0                      0   \n",
       "12556                      0                      0                      0   \n",
       "15967                      0                      0                      0   \n",
       "21049                      0                      0                      0   \n",
       "5501                       0                      1                      0   \n",
       "28073                      0                      1                      0   \n",
       "37873                      0                      0                      0   \n",
       "19538                      0                      1                      0   \n",
       "18809                      0                      0                      0   \n",
       "37087                      0                      0                      0   \n",
       "16205                      0                      0                      0   \n",
       "37604                      0                      0                      0   \n",
       "37101                      0                      0                      0   \n",
       "24635                      0                      1                      0   \n",
       "1345                       0                      0                      0   \n",
       "26691                      0                      0                      0   \n",
       "...                      ...                    ...                    ...   \n",
       "20347                      0                      1                      0   \n",
       "15810                      0                      0                      0   \n",
       "27774                      0                      0                      0   \n",
       "27894                      0                      0                      0   \n",
       "12285                      0                      0                      0   \n",
       "20308                      0                      0                      0   \n",
       "8590                       1                      1                      1   \n",
       "14222                      0                      0                      0   \n",
       "8943                       0                      0                      0   \n",
       "21170                      0                      0                      0   \n",
       "9693                       0                      0                      0   \n",
       "10663                      0                      0                      1   \n",
       "11652                      0                      0                      0   \n",
       "30828                      0                      0                      0   \n",
       "15394                      0                      1                      0   \n",
       "7846                       0                      0                      0   \n",
       "37444                      0                      0                      0   \n",
       "19730                      1                      0                      0   \n",
       "32004                      0                      0                      0   \n",
       "288                        0                      0                      0   \n",
       "36615                      0                      0                      0   \n",
       "11493                      0                      0                      0   \n",
       "39744                      0                      0                      0   \n",
       "9634                       0                      0                      0   \n",
       "17709                      0                      0                      0   \n",
       "27443                      0                      1                      0   \n",
       "39465                      0                      0                      0   \n",
       "29673                      0                      0                      0   \n",
       "22278                      0                      0                      0   \n",
       "32637                      0                      0                      0   \n",
       "\n",
       "        residual  \n",
       "40056  60.093830  \n",
       "3222   22.980583  \n",
       "42706  14.018461  \n",
       "14505  12.276542  \n",
       "8310   10.960655  \n",
       "6186   10.190336  \n",
       "10574   9.797115  \n",
       "28970   8.665862  \n",
       "32230   8.291083  \n",
       "1609    6.753566  \n",
       "41758   5.982767  \n",
       "10214   5.009284  \n",
       "42735   4.787800  \n",
       "9080    4.737586  \n",
       "8192    4.151326  \n",
       "12556   3.858672  \n",
       "15967   3.720057  \n",
       "21049   3.653452  \n",
       "5501    3.612013  \n",
       "28073   3.557962  \n",
       "37873   3.524932  \n",
       "19538   3.455053  \n",
       "18809   3.430158  \n",
       "37087   2.926127  \n",
       "16205   2.813246  \n",
       "37604   2.767648  \n",
       "37101   2.758488  \n",
       "24635   2.543581  \n",
       "1345    2.515880  \n",
       "26691   2.483158  \n",
       "...          ...  \n",
       "20347   0.769491  \n",
       "15810   0.767143  \n",
       "27774   0.766957  \n",
       "27894   0.765579  \n",
       "12285   0.765576  \n",
       "20308   0.764146  \n",
       "8590    0.763206  \n",
       "14222   0.762437  \n",
       "8943    0.761831  \n",
       "21170   0.761589  \n",
       "9693    0.759363  \n",
       "10663   0.759099  \n",
       "11652   0.758527  \n",
       "30828   0.758093  \n",
       "15394   0.757711  \n",
       "7846    0.757519  \n",
       "37444   0.757248  \n",
       "19730   0.756915  \n",
       "32004   0.756316  \n",
       "288     0.754428  \n",
       "36615   0.754374  \n",
       "11493   0.753980  \n",
       "39744   0.753625  \n",
       "9634    0.753601  \n",
       "17709   0.753267  \n",
       "27443   0.753174  \n",
       "39465   0.752881  \n",
       "29673   0.752879  \n",
       "22278   0.752695  \n",
       "32637   0.751674  \n",
       "\n",
       "[428 rows x 247 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['residual'] > percentile99]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>match_id</th>\n",
       "      <th>positive_votes</th>\n",
       "      <th>negative_votes</th>\n",
       "      <th>first_blood_time</th>\n",
       "      <th>radiant_win</th>\n",
       "      <th>duration</th>\n",
       "      <th>dire_pos1_ancient_kills</th>\n",
       "      <th>dire_pos1_courier_kills</th>\n",
       "      <th>dire_pos1_kills</th>\n",
       "      <th>dire_pos1_neutral_kills</th>\n",
       "      <th>...</th>\n",
       "      <th>dire_pos2_abandons</th>\n",
       "      <th>dire_pos3_abandons</th>\n",
       "      <th>dire_pos4_abandons</th>\n",
       "      <th>dire_pos5_abandons</th>\n",
       "      <th>radiant_pos1_abandons</th>\n",
       "      <th>radiant_pos2_abandons</th>\n",
       "      <th>radiant_pos3_abandons</th>\n",
       "      <th>radiant_pos4_abandons</th>\n",
       "      <th>radiant_pos5_abandons</th>\n",
       "      <th>residual</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15328</th>\n",
       "      <td>3215105241</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>112</td>\n",
       "      <td>False</td>\n",
       "      <td>2777</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.174453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26872</th>\n",
       "      <td>3215376151</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>114</td>\n",
       "      <td>True</td>\n",
       "      <td>2399</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.108168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9558</th>\n",
       "      <td>3215415736</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>169</td>\n",
       "      <td>False</td>\n",
       "      <td>3833</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.539407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38056</th>\n",
       "      <td>3215109078</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>191</td>\n",
       "      <td>False</td>\n",
       "      <td>3583</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15</td>\n",
       "      <td>51.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.522020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28038</th>\n",
       "      <td>3215171445</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>106</td>\n",
       "      <td>True</td>\n",
       "      <td>3963</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.219839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19879</th>\n",
       "      <td>3215172311</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>60</td>\n",
       "      <td>True</td>\n",
       "      <td>4322</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.478943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23736</th>\n",
       "      <td>3215186932</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>402</td>\n",
       "      <td>True</td>\n",
       "      <td>2640</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.108369</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24764</th>\n",
       "      <td>3215409668</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>179</td>\n",
       "      <td>False</td>\n",
       "      <td>2451</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.138435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3126</th>\n",
       "      <td>3215384337</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>146</td>\n",
       "      <td>True</td>\n",
       "      <td>3289</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.514544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5952</th>\n",
       "      <td>3215141313</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>126</td>\n",
       "      <td>False</td>\n",
       "      <td>2572</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11</td>\n",
       "      <td>116.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.341837</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 247 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         match_id  positive_votes  negative_votes  first_blood_time  \\\n",
       "15328  3215105241               0               0               112   \n",
       "26872  3215376151               0               0               114   \n",
       "9558   3215415736               0               0               169   \n",
       "38056  3215109078               0               0               191   \n",
       "28038  3215171445               0               0               106   \n",
       "19879  3215172311               1               0                60   \n",
       "23736  3215186932               0               0               402   \n",
       "24764  3215409668               0               0               179   \n",
       "3126   3215384337               0               0               146   \n",
       "5952   3215141313               0               0               126   \n",
       "\n",
       "      radiant_win  duration  dire_pos1_ancient_kills  dire_pos1_courier_kills  \\\n",
       "15328       False      2777                      NaN                      NaN   \n",
       "26872        True      2399                      NaN                      NaN   \n",
       "9558        False      3833                      NaN                      NaN   \n",
       "38056       False      3583                      1.0                      0.0   \n",
       "28038        True      3963                      NaN                      NaN   \n",
       "19879        True      4322                      NaN                      NaN   \n",
       "23736        True      2640                      NaN                      NaN   \n",
       "24764       False      2451                      NaN                      NaN   \n",
       "3126         True      3289                      NaN                      NaN   \n",
       "5952        False      2572                      3.0                      0.0   \n",
       "\n",
       "       dire_pos1_kills  dire_pos1_neutral_kills    ...     dire_pos2_abandons  \\\n",
       "15328               24                      NaN    ...                      0   \n",
       "26872                7                      NaN    ...                      0   \n",
       "9558                20                      NaN    ...                      0   \n",
       "38056               15                     51.0    ...                      0   \n",
       "28038               14                      NaN    ...                      0   \n",
       "19879                4                      NaN    ...                      0   \n",
       "23736               10                      NaN    ...                      0   \n",
       "24764                9                      NaN    ...                      0   \n",
       "3126                 8                      NaN    ...                      0   \n",
       "5952                11                    116.0    ...                      0   \n",
       "\n",
       "       dire_pos3_abandons  dire_pos4_abandons  dire_pos5_abandons  \\\n",
       "15328                   0                   0                   0   \n",
       "26872                   0                   0                   0   \n",
       "9558                    0                   1                   0   \n",
       "38056                   0                   0                   0   \n",
       "28038                   0                   0                   0   \n",
       "19879                   0                   0                   0   \n",
       "23736                   0                   0                   0   \n",
       "24764                   0                   0                   0   \n",
       "3126                    0                   0                   0   \n",
       "5952                    0                   0                   0   \n",
       "\n",
       "       radiant_pos1_abandons  radiant_pos2_abandons  radiant_pos3_abandons  \\\n",
       "15328                      0                      0                      0   \n",
       "26872                      0                      0                      0   \n",
       "9558                       0                      0                      0   \n",
       "38056                      0                      0                      0   \n",
       "28038                      0                      0                      0   \n",
       "19879                      0                      0                      0   \n",
       "23736                      0                      0                      0   \n",
       "24764                      0                      0                      1   \n",
       "3126                       0                      0                      1   \n",
       "5952                       0                      0                      0   \n",
       "\n",
       "       radiant_pos4_abandons  radiant_pos5_abandons  residual  \n",
       "15328                      0                      0  0.174453  \n",
       "26872                      0                      0  0.108168  \n",
       "9558                       0                      0  0.539407  \n",
       "38056                      0                      0  0.522020  \n",
       "28038                      0                      0  0.219839  \n",
       "19879                      0                      0  0.478943  \n",
       "23736                      0                      0  0.108369  \n",
       "24764                      0                      0  0.138435  \n",
       "3126                       0                      0  0.514544  \n",
       "5952                       0                      0  0.341837  \n",
       "\n",
       "[10 rows x 247 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['residual'] < percentile99].sample(n=10, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "biggestAnom = np.loadtxt('data/anomalies.csv', delimiter = ',')\n",
    "theirOutput = np.loadtxt('data/output.csv', delimiter = \",\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "biggestAnom = np.column_stack([df['match_id'].values, biggestAnom])\n",
    "theirOutput = np.column_stack([df['match_id'].values, theirOutput])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dictToEndAllDicts = {'match_id': [], 'column': [], 'actual': [], 'predicted': []}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "def whichIndex(row1, row2):\n",
    "    largestError = -1\n",
    "    index = -1\n",
    "    for i in range(1, len(row1)):\n",
    "        if abs(row1[i] - row2[i]) > largestError:\n",
    "            largestError = abs(row1[i] - row2[i])\n",
    "            index = i\n",
    "    return index\n",
    "columns = list(df.columns)\n",
    "columns.remove('match_id')\n",
    "mask = np.where(df['residual'] >= percentile99)[0]\n",
    "\n",
    "biggestAnom = biggestAnom[mask, :]\n",
    "theirOutput = theirOutput[mask, :]\n",
    "\n",
    "for i in range(len(biggestAnom)):\n",
    "    index = whichIndex(biggestAnom[i, :], theirOutput[i, :])\n",
    "    dictToEndAllDicts['match_id'].append(int(biggestAnom[i, 0]))\n",
    "    dictToEndAllDicts['column'].append(columns[index])\n",
    "    dictToEndAllDicts['actual'].append(biggestAnom[i, index])\n",
    "    dictToEndAllDicts['predicted'].append(theirOutput[i, index])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "tableOfAnomalies = pd.DataFrame.from_dict(dictToEndAllDicts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "tableOfAnomalies.to_csv('data/tableOfAnomaliesAndTheirCause.csv', index = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>column</th>\n",
       "      <th>match_id</th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>49.000000</td>\n",
       "      <td>negative_votes</td>\n",
       "      <td>3215497967</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>15.000000</td>\n",
       "      <td>negative_votes</td>\n",
       "      <td>3215097220</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9.000000</td>\n",
       "      <td>negative_votes</td>\n",
       "      <td>3215312675</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.997236</td>\n",
       "      <td>dire_pos5_courier_kills</td>\n",
       "      <td>3215162338</td>\n",
       "      <td>0.254108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.967832</td>\n",
       "      <td>radiant_pos3_gpm</td>\n",
       "      <td>3215181105</td>\n",
       "      <td>0.215278</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.986188</td>\n",
       "      <td>dire_pos5_kda</td>\n",
       "      <td>3215377994</td>\n",
       "      <td>0.675354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7.000000</td>\n",
       "      <td>negative_votes</td>\n",
       "      <td>3215271631</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.915403</td>\n",
       "      <td>dire_pos4_gpm</td>\n",
       "      <td>3215289100</td>\n",
       "      <td>0.162801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.915403</td>\n",
       "      <td>dire_pos4_gpm</td>\n",
       "      <td>3215279863</td>\n",
       "      <td>0.155206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>negative_votes</td>\n",
       "      <td>3215253055</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.977942</td>\n",
       "      <td>radiant_pos1_sentry_uses</td>\n",
       "      <td>3215448302</td>\n",
       "      <td>0.730816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.023621</td>\n",
       "      <td>dire_pos3_gpm</td>\n",
       "      <td>3215416370</td>\n",
       "      <td>0.595808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.243854</td>\n",
       "      <td>radiant_pos3_gpm</td>\n",
       "      <td>3215261563</td>\n",
       "      <td>0.536931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>4.000000</td>\n",
       "      <td>negative_votes</td>\n",
       "      <td>3215294948</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.996014</td>\n",
       "      <td>radiant_pos5_kills</td>\n",
       "      <td>3215155134</td>\n",
       "      <td>0.729708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.172037</td>\n",
       "      <td>radiant_pos4_gpm</td>\n",
       "      <td>3215350255</td>\n",
       "      <td>0.478153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.765351</td>\n",
       "      <td>radiant_pos1_kpm</td>\n",
       "      <td>3215341221</td>\n",
       "      <td>0.473024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.796921</td>\n",
       "      <td>radiant_pos2_gpm</td>\n",
       "      <td>3215306187</td>\n",
       "      <td>0.594073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.995038</td>\n",
       "      <td>dire_pos4_gpm</td>\n",
       "      <td>3215302636</td>\n",
       "      <td>0.616405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.995249</td>\n",
       "      <td>dire_pos5_gpm</td>\n",
       "      <td>3215316820</td>\n",
       "      <td>0.661066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.991397</td>\n",
       "      <td>dire_pos3_sentry_uses</td>\n",
       "      <td>3215190109</td>\n",
       "      <td>0.725191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.994378</td>\n",
       "      <td>dire_pos2_gpm</td>\n",
       "      <td>3215355847</td>\n",
       "      <td>0.634865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.992697</td>\n",
       "      <td>radiant_pos3_kills</td>\n",
       "      <td>3215130854</td>\n",
       "      <td>0.734208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.995991</td>\n",
       "      <td>radiant_pos1_kills</td>\n",
       "      <td>3215259897</td>\n",
       "      <td>0.731750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.420948</td>\n",
       "      <td>dire_pos1_totalgold</td>\n",
       "      <td>3215394671</td>\n",
       "      <td>0.155802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.853668</td>\n",
       "      <td>dire_pos4_deaths</td>\n",
       "      <td>3215152225</td>\n",
       "      <td>0.677466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.970540</td>\n",
       "      <td>radiant_pos2_kpm</td>\n",
       "      <td>3215274711</td>\n",
       "      <td>0.548053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.417960</td>\n",
       "      <td>dire_pos3_xpm</td>\n",
       "      <td>3215366609</td>\n",
       "      <td>0.140712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.837322</td>\n",
       "      <td>radiant_pos2_courier_kills</td>\n",
       "      <td>3215449992</td>\n",
       "      <td>0.687711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>0.417725</td>\n",
       "      <td>radiant_pos2_xpm</td>\n",
       "      <td>3215376405</td>\n",
       "      <td>0.142942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>398</th>\n",
       "      <td>0.508021</td>\n",
       "      <td>dire_pos1_totalxp</td>\n",
       "      <td>3215337531</td>\n",
       "      <td>0.461330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399</th>\n",
       "      <td>0.987752</td>\n",
       "      <td>radiant_pos5_hero_heal</td>\n",
       "      <td>3215102476</td>\n",
       "      <td>0.764587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>400</th>\n",
       "      <td>0.602995</td>\n",
       "      <td>dire_pos4_lasthits</td>\n",
       "      <td>3215253869</td>\n",
       "      <td>0.546018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>401</th>\n",
       "      <td>0.068665</td>\n",
       "      <td>dire_pos5_courier_kills</td>\n",
       "      <td>3215422166</td>\n",
       "      <td>0.011190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>402</th>\n",
       "      <td>0.602511</td>\n",
       "      <td>radiant_pos3_deaths</td>\n",
       "      <td>3215263577</td>\n",
       "      <td>0.545937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>403</th>\n",
       "      <td>0.647322</td>\n",
       "      <td>radiant_pos2_courier_kills</td>\n",
       "      <td>3215285461</td>\n",
       "      <td>0.583701</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>404</th>\n",
       "      <td>0.977324</td>\n",
       "      <td>residual</td>\n",
       "      <td>3215114749</td>\n",
       "      <td>0.947976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>405</th>\n",
       "      <td>0.728203</td>\n",
       "      <td>dire_pos5_courier_kills</td>\n",
       "      <td>3215096854</td>\n",
       "      <td>0.634661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>406</th>\n",
       "      <td>0.242480</td>\n",
       "      <td>radiant_pos3_sentry_uses</td>\n",
       "      <td>3215446534</td>\n",
       "      <td>0.059556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>407</th>\n",
       "      <td>0.586169</td>\n",
       "      <td>dire_pos3_observer_uses</td>\n",
       "      <td>3215119491</td>\n",
       "      <td>0.541193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>408</th>\n",
       "      <td>-0.022676</td>\n",
       "      <td>residual</td>\n",
       "      <td>3215205453</td>\n",
       "      <td>-0.055943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>409</th>\n",
       "      <td>0.568367</td>\n",
       "      <td>radiant_pos3_ancient_kills</td>\n",
       "      <td>3215135325</td>\n",
       "      <td>0.522728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>410</th>\n",
       "      <td>0.117480</td>\n",
       "      <td>radiant_pos3_sentry_uses</td>\n",
       "      <td>3215415463</td>\n",
       "      <td>0.008837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>411</th>\n",
       "      <td>0.668367</td>\n",
       "      <td>radiant_pos3_tower_kills</td>\n",
       "      <td>3215322189</td>\n",
       "      <td>0.594004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>412</th>\n",
       "      <td>0.976364</td>\n",
       "      <td>radiant_pos1_abandons</td>\n",
       "      <td>3215345510</td>\n",
       "      <td>0.948507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>413</th>\n",
       "      <td>-0.022676</td>\n",
       "      <td>residual</td>\n",
       "      <td>3215138506</td>\n",
       "      <td>-0.061634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>414</th>\n",
       "      <td>0.653263</td>\n",
       "      <td>dire_pos1_abandons</td>\n",
       "      <td>3215492929</td>\n",
       "      <td>0.580061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>415</th>\n",
       "      <td>0.612837</td>\n",
       "      <td>radiant_pos5_totalgold</td>\n",
       "      <td>3215267165</td>\n",
       "      <td>0.538579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>416</th>\n",
       "      <td>0.754499</td>\n",
       "      <td>dire_pos2_observer_uses</td>\n",
       "      <td>3215123467</td>\n",
       "      <td>0.658702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>417</th>\n",
       "      <td>0.032951</td>\n",
       "      <td>dire_pos5_courier_kills</td>\n",
       "      <td>3215358156</td>\n",
       "      <td>0.000771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>418</th>\n",
       "      <td>0.786506</td>\n",
       "      <td>radiant_pos5_hero_dmg</td>\n",
       "      <td>3215094142</td>\n",
       "      <td>0.659190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>419</th>\n",
       "      <td>0.711157</td>\n",
       "      <td>radiant_pos1_courier_kills</td>\n",
       "      <td>3215328883</td>\n",
       "      <td>0.623964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>420</th>\n",
       "      <td>0.032951</td>\n",
       "      <td>dire_pos5_courier_kills</td>\n",
       "      <td>3215250045</td>\n",
       "      <td>-0.003149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>421</th>\n",
       "      <td>0.336952</td>\n",
       "      <td>radiant_pos2_kpm</td>\n",
       "      <td>3215410384</td>\n",
       "      <td>0.213571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>422</th>\n",
       "      <td>0.613161</td>\n",
       "      <td>dire_pos1_kda</td>\n",
       "      <td>3215168749</td>\n",
       "      <td>0.543565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>423</th>\n",
       "      <td>0.976364</td>\n",
       "      <td>radiant_pos1_abandons</td>\n",
       "      <td>3215343292</td>\n",
       "      <td>0.948496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>424</th>\n",
       "      <td>0.641713</td>\n",
       "      <td>radiant_pos1_ancient_kills</td>\n",
       "      <td>3215107300</td>\n",
       "      <td>0.576132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>425</th>\n",
       "      <td>0.117480</td>\n",
       "      <td>radiant_pos3_sentry_uses</td>\n",
       "      <td>3215450701</td>\n",
       "      <td>0.033871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>426</th>\n",
       "      <td>0.863760</td>\n",
       "      <td>dire_pos1_xpm</td>\n",
       "      <td>3215239658</td>\n",
       "      <td>0.700287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>427</th>\n",
       "      <td>0.523910</td>\n",
       "      <td>radiant_pos1_totalgold</td>\n",
       "      <td>3215427391</td>\n",
       "      <td>0.481955</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>428 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        actual                      column    match_id  predicted\n",
       "0    49.000000              negative_votes  3215497967   1.000000\n",
       "1    15.000000              negative_votes  3215097220   1.000000\n",
       "2     9.000000              negative_votes  3215312675   1.000000\n",
       "3     0.997236     dire_pos5_courier_kills  3215162338   0.254108\n",
       "4     0.967832            radiant_pos3_gpm  3215181105   0.215278\n",
       "5     0.986188               dire_pos5_kda  3215377994   0.675354\n",
       "6     7.000000              negative_votes  3215271631   1.000000\n",
       "7     0.915403               dire_pos4_gpm  3215289100   0.162801\n",
       "8     0.915403               dire_pos4_gpm  3215279863   0.155206\n",
       "9     5.000000              negative_votes  3215253055   1.000000\n",
       "10    0.977942    radiant_pos1_sentry_uses  3215448302   0.730816\n",
       "11    0.023621               dire_pos3_gpm  3215416370   0.595808\n",
       "12    0.243854            radiant_pos3_gpm  3215261563   0.536931\n",
       "13    4.000000              negative_votes  3215294948   1.000000\n",
       "14    0.996014          radiant_pos5_kills  3215155134   0.729708\n",
       "15    0.172037            radiant_pos4_gpm  3215350255   0.478153\n",
       "16    0.765351            radiant_pos1_kpm  3215341221   0.473024\n",
       "17    0.796921            radiant_pos2_gpm  3215306187   0.594073\n",
       "18    0.995038               dire_pos4_gpm  3215302636   0.616405\n",
       "19    0.995249               dire_pos5_gpm  3215316820   0.661066\n",
       "20    0.991397       dire_pos3_sentry_uses  3215190109   0.725191\n",
       "21    0.994378               dire_pos2_gpm  3215355847   0.634865\n",
       "22    0.992697          radiant_pos3_kills  3215130854   0.734208\n",
       "23    0.995991          radiant_pos1_kills  3215259897   0.731750\n",
       "24    0.420948         dire_pos1_totalgold  3215394671   0.155802\n",
       "25    0.853668            dire_pos4_deaths  3215152225   0.677466\n",
       "26    0.970540            radiant_pos2_kpm  3215274711   0.548053\n",
       "27    0.417960               dire_pos3_xpm  3215366609   0.140712\n",
       "28    0.837322  radiant_pos2_courier_kills  3215449992   0.687711\n",
       "29    0.417725            radiant_pos2_xpm  3215376405   0.142942\n",
       "..         ...                         ...         ...        ...\n",
       "398   0.508021           dire_pos1_totalxp  3215337531   0.461330\n",
       "399   0.987752      radiant_pos5_hero_heal  3215102476   0.764587\n",
       "400   0.602995          dire_pos4_lasthits  3215253869   0.546018\n",
       "401   0.068665     dire_pos5_courier_kills  3215422166   0.011190\n",
       "402   0.602511         radiant_pos3_deaths  3215263577   0.545937\n",
       "403   0.647322  radiant_pos2_courier_kills  3215285461   0.583701\n",
       "404   0.977324                    residual  3215114749   0.947976\n",
       "405   0.728203     dire_pos5_courier_kills  3215096854   0.634661\n",
       "406   0.242480    radiant_pos3_sentry_uses  3215446534   0.059556\n",
       "407   0.586169     dire_pos3_observer_uses  3215119491   0.541193\n",
       "408  -0.022676                    residual  3215205453  -0.055943\n",
       "409   0.568367  radiant_pos3_ancient_kills  3215135325   0.522728\n",
       "410   0.117480    radiant_pos3_sentry_uses  3215415463   0.008837\n",
       "411   0.668367    radiant_pos3_tower_kills  3215322189   0.594004\n",
       "412   0.976364       radiant_pos1_abandons  3215345510   0.948507\n",
       "413  -0.022676                    residual  3215138506  -0.061634\n",
       "414   0.653263          dire_pos1_abandons  3215492929   0.580061\n",
       "415   0.612837      radiant_pos5_totalgold  3215267165   0.538579\n",
       "416   0.754499     dire_pos2_observer_uses  3215123467   0.658702\n",
       "417   0.032951     dire_pos5_courier_kills  3215358156   0.000771\n",
       "418   0.786506       radiant_pos5_hero_dmg  3215094142   0.659190\n",
       "419   0.711157  radiant_pos1_courier_kills  3215328883   0.623964\n",
       "420   0.032951     dire_pos5_courier_kills  3215250045  -0.003149\n",
       "421   0.336952            radiant_pos2_kpm  3215410384   0.213571\n",
       "422   0.613161               dire_pos1_kda  3215168749   0.543565\n",
       "423   0.976364       radiant_pos1_abandons  3215343292   0.948496\n",
       "424   0.641713  radiant_pos1_ancient_kills  3215107300   0.576132\n",
       "425   0.117480    radiant_pos3_sentry_uses  3215450701   0.033871\n",
       "426   0.863760               dire_pos1_xpm  3215239658   0.700287\n",
       "427   0.523910      radiant_pos1_totalgold  3215427391   0.481955\n",
       "\n",
       "[428 rows x 4 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tableOfAnomalies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}