notebook.community

Edit and run



In [1]:

    
%matplotlib inline

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import glob
import datetime
import itertools
from time import sleep
import os
import seaborn as sns
from itertools import chain



In [2]:

    
print datetime.datetime.now()
validFilePaths = []
for f in os.listdir("data/anomaly_data"):
    filePath = os.path.join("data/anomaly_data", f)
    if os.path.isdir(filePath):
        continue
    if os.stat(filePath).st_size <= 3:
        continue
    validFilePaths.append(filePath)
    
numF = int(1 * len(validFilePaths))
print 'Using this many files {0}'.format(numF)
validFilePaths = np.random.choice(validFilePaths, numF, replace=False)
df_list = (pd.read_csv(f) for f in validFilePaths)
df = pd.concat(df_list, ignore_index=True)
df = df[df['radiant_win'].notnull()]









    



2017-06-20 21:51:22.869660
Using this many files 4088



In [3]:

    
df2 = pd.DataFrame(np.loadtxt('data/goodAnomaliesResidual.csv', delimiter = ','), 
                 columns = ["match_id", "residual"])

df = df.merge(df2, left_on='match_id', right_on='match_id')

matchDictionary = {}
heroColumns = [filter(lambda x: z in x, df.columns) for z in ['hero_id']][0]

# for each row, this function will return all the heroes that the match belongs in
# for example, if the match had an AM and Spectre, we should place this match in each of those bags
def heroesForMatch(row):
    pass



In [4]:

    
df = df[~df['match_id'].duplicated()]



In [5]:

    
df = df.sort_values(by='residual', ascending=False)



In [6]:

    
columns = df.columns
numericalFeatures = ['match_id', 'positive_votes', 'negative_votes', 'first_blood_time', 'radiant_win',
                    'duration', 'kills', 'deaths', 'assists', 'kpm', 'kda', 'hero_dmg',
                    'gpm', 'hero_heal', 'xpm', 'totalgold', 'totalxp', 'lasthits', 'denies',
                    'tower_kills', 'courier_kills', 'observer_uses', 'sentry_uses',
                    'ancient_kills', 'camps_stacked', 'abandons', 'residual']
numFeatures = [filter(lambda x: z in x, columns) for z in numericalFeatures]
numFeatures = list(chain(*numFeatures))
df = df[numFeatures]



In [7]:

    
df['match_id'].values.tolist()









    Out[7]:





[3215497967,
 3215097220,
 3215312675,
 3215162338,
 3215181105,
 3215377994,
 3215271631,
 3215289100,
 3215279863,
 3215253055,
 3215448302,
 3215416370,
 3215261563,
 3215294948,
 3215155134,
 3215350255,
 3215341221,
 3215306187,
 3215302636,
 3215316820,
 3215190109,
 3215355847,
 3215130854,
 3215259897,
 3215394671,
 3215152225,
 3215274711,
 3215366609,
 3215449992,
 3215376405,
 3215268540,
 3215420902,
 3215332931,
 3215327547,
 3215354992,
 3215254939,
 3215149372,
 3215370025,
 3215201012,
 3215272070,
 3215223639,
 3215422492,
 3215260456,
 3215354934,
 3215377592,
 3215445972,
 3215101148,
 3215174827,
 3215384305,
 3215188370,
 3215297782,
 3215493515,
 3215119808,
 3215360521,
 3215102154,
 3215271055,
 3215403667,
 3215217460,
 3215482564,
 3215466626,
 3215255292,
 3215095789,
 3215108689,
 3215330831,
 3215468550,
 3215446132,
 3215384555,
 3215470445,
 3215318005,
 3215238829,
 3215293087,
 3215391855,
 3215439074,
 3215401870,
 3215440922,
 3215092973,
 3215369622,
 3215116689,
 3215236348,
 3215145375,
 3215129374,
 3215199686,
 3215481084,
 3215191852,
 3215459217,
 3215189407,
 3215462703,
 3215113055,
 3215469573,
 3215428974,
 3215323020,
 3215169878,
 3215359797,
 3215419225,
 3215221028,
 3215488213,
 3215198158,
 3215275453,
 3215204587,
 3215260718,
 3215153289,
 3215120631,
 3215400918,
 3215119120,
 3215419414,
 3215443584,
 3215276839,
 3215464545,
 3215118187,
 3215247802,
 3215343954,
 3215145695,
 3215365977,
 3215162352,
 3215344482,
 3215166724,
 3215457284,
 3215144772,
 3215254006,
 3215471576,
 3215267691,
 3215203102,
 3215460086,
 3215132768,
 3215444643,
 3215327292,
 3215107898,
 3215193832,
 3215135123,
 3215361786,
 3215242906,
 3215404311,
 3215472295,
 3215359384,
 3215261245,
 3215419991,
 3215164884,
 3215482334,
 3215344415,
 3215191160,
 3215269902,
 3215132609,
 3215229099,
 3215424121,
 3215214645,
 3215280848,
 3215324165,
 3215366819,
 3215121935,
 3215217456,
 3215478948,
 3215287068,
 3215328836,
 3215189005,
 3215180352,
 3215267872,
 3215325036,
 3215403728,
 3215386000,
 3215354548,
 3215107983,
 3215193319,
 3215174289,
 3215405510,
 3215486107,
 3215308046,
 3215375609,
 3215370507,
 3215356164,
 3215394202,
 3215475044,
 3215314163,
 3215232157,
 3215449600,
 3215150608,
 3215387413,
 3215116601,
 3215476161,
 3215498524,
 3215223648,
 3215347649,
 3215288708,
 3215250234,
 3215123893,
 3215102237,
 3215294572,
 3215169250,
 3215362362,
 3215401727,
 3215453213,
 3215312256,
 3215332342,
 3215312958,
 3215223737,
 3215285061,
 3215413686,
 3215324947,
 3215327320,
 3215267032,
 3215454284,
 3215357642,
 3215488755,
 3215224918,
 3215115434,
 3215486952,
 3215459746,
 3215330354,
 3215345885,
 3215129922,
 3215300804,
 3215475617,
 3215272277,
 3215423386,
 3215130289,
 3215483739,
 3215153586,
 3215135265,
 3215299981,
 3215316700,
 3215448897,
 3215206674,
 3215475240,
 3215411914,
 3215494824,
 3215208665,
 3215390160,
 3215251697,
 3215466606,
 3215391093,
 3215162013,
 3215299806,
 3215321270,
 3215287399,
 3215375892,
 3215195313,
 3215144420,
 3215379626,
 3215137884,
 3215375087,
 3215303959,
 3215437756,
 3215336973,
 3215416702,
 3215205916,
 3215220146,
 3215440465,
 3215270302,
 3215094959,
 3215298099,
 3215165027,
 3215433320,
 3215298273,
 3215455368,
 3215363503,
 3215100697,
 3215283180,
 3215410568,
 3215291565,
 3215385605,
 3215288289,
 3215148792,
 3215094660,
 3215347239,
 3215352617,
 3215107489,
 3215474947,
 3215404711,
 3215147040,
 3215180855,
 3215286517,
 3215308199,
 3215399746,
 3215142113,
 3215138351,
 3215276184,
 3215233902,
 3215292517,
 3215166848,
 3215477347,
 3215246062,
 3215136191,
 3215281167,
 3215459906,
 3215480662,
 3215287547,
 3215337213,
 3215276888,
 3215228456,
 3215282906,
 3215354545,
 3215243693,
 3215131987,
 3215220488,
 3215496947,
 3215129208,
 3215335248,
 3215477407,
 3215386429,
 3215189721,
 3215355241,
 3215114144,
 3215446878,
 3215138380,
 3215391557,
 3215452559,
 3215388921,
 3215457606,
 3215245376,
 3215182714,
 3215425334,
 3215146405,
 3215303729,
 3215297974,
 3215488250,
 3215291673,
 3215165758,
 3215215978,
 3215226575,
 3215426458,
 3215298264,
 3215375401,
 3215164902,
 3215439378,
 3215417419,
 3215129654,
 3215412289,
 3215119017,
 3215151576,
 3215363364,
 3215232872,
 3215100187,
 3215311956,
 3215379763,
 3215351593,
 3215409860,
 3215442394,
 3215353990,
 3215277734,
 3215229457,
 3215130835,
 3215092287,
 3215137548,
 3215410873,
 3215186260,
 3215314246,
 3215455747,
 3215125088,
 3215388049,
 3215387904,
 3215486254,
 3215349760,
 3215149581,
 3215325668,
 3215374088,
 3215303752,
 3215266899,
 3215095314,
 3215216831,
 3215222186,
 3215164824,
 3215310710,
 3215266966,
 3215415225,
 3215265780,
 3215418853,
 3215273920,
 3215171733,
 3215236033,
 3215274732,
 3215282518,
 3215365416,
 3215485464,
 3215489360,
 3215270915,
 3215306738,
 3215260577,
 3215141466,
 3215162054,
 3215395730,
 3215265050,
 3215312365,
 3215397328,
 3215110538,
 3215214818,
 3215435057,
 3215154136,
 3215356322,
 3215126746,
 3215249783,
 3215393413,
 3215163349,
 3215282633,
 3215152262,
 3215121064,
 3215328313,
 3215209313,
 3215435386,
 3215486232,
 3215337531,
 3215102476,
 3215253869,
 3215422166,
 3215263577,
 3215285461,
 3215114749,
 3215096854,
 3215446534,
 3215119491,
 3215205453,
 3215135325,
 3215415463,
 3215322189,
 3215345510,
 3215138506,
 3215492929,
 3215267165,
 3215123467,
 3215358156,
 3215094142,
 3215328883,
 3215250045,
 3215410384,
 3215168749,
 3215343292,
 3215107300,
 3215450701,
 3215239658,
 3215427391,
 3215403005,
 3215129173,
 3215301736,
 3215121346,
 3215136149,
 3215315656,
 3215435380,
 3215260528,
 3215131741,
 3215127161,
 3215352898,
 3215115754,
 3215383518,
 3215323457,
 3215465807,
 3215145418,
 3215380602,
 3215391171,
 3215191784,
 3215358722,
 3215433160,
 3215290038,
 3215193538,
 3215209365,
 3215253420,
 3215491015,
 3215217135,
 3215379611,
 3215413988,
 3215253204,
 3215404742,
 3215384989,
 3215199127,
 3215120092,
 3215108762,
 3215213377,
 3215349948,
 3215306421,
 3215202477,
 3215464043,
 3215268057,
 3215370518,
 3215439438,
 3215371756,
 3215477162,
 3215293625,
 3215296744,
 3215377187,
 3215251365,
 3215263061,
 3215358458,
 3215429839,
 3215209805,
 3215279179,
 3215405966,
 3215258832,
 3215179518,
 3215104961,
 3215107873,
 3215366168,
 3215430628,
 3215422313,
 3215365394,
 3215485880,
 3215457207,
 3215380931,
 3215445032,
 3215490011,
 3215137211,
 3215203727,
 3215165184,
 3215489708,
 3215452967,
 3215329371,
 3215127826,
 3215406396,
 3215230963,
 3215197322,
 3215109921,
 3215336230,
 3215447774,
 3215473864,
 3215130557,
 3215267128,
 3215317064,
 3215407823,
 3215334574,
 3215235851,
 3215432287,
 3215279599,
 3215277575,
 3215352751,
 3215193905,
 3215403773,
 3215470994,
 3215241393,
 3215475777,
 3215378128,
 3215255293,
 3215408793,
 3215186419,
 3215365233,
 3215319925,
 3215206639,
 3215101069,
 3215110880,
 3215279096,
 3215236308,
 3215381288,
 3215292947,
 3215199673,
 3215409332,
 3215099900,
 3215319240,
 3215115194,
 3215341642,
 3215288694,
 3215263248,
 3215248395,
 3215424333,
 3215388328,
 3215153855,
 3215113703,
 3215378253,
 3215478594,
 3215333633,
 3215314457,
 3215264390,
 3215185323,
 3215416462,
 3215294519,
 3215100773,
 3215419450,
 3215149559,
 3215207589,
 3215443285,
 3215294238,
 3215230801,
 3215420148,
 3215412427,
 3215231453,
 3215373592,
 3215246152,
 3215408093,
 3215095140,
 3215231193,
 3215402750,
 3215338649,
 3215453826,
 3215184198,
 3215315286,
 3215370517,
 3215360704,
 3215172975,
 3215131925,
 3215134357,
 3215311164,
 3215452739,
 3215393161,
 3215293239,
 3215114502,
 3215224556,
 3215273884,
 3215363634,
 3215460930,
 3215487728,
 3215273633,
 3215360865,
 3215431682,
 3215109989,
 3215237594,
 3215269721,
 3215094857,
 3215383204,
 3215374815,
 3215128099,
 3215241740,
 3215344197,
 3215303641,
 3215156007,
 3215091817,
 3215275030,
 3215389491,
 3215141012,
 3215369987,
 3215215211,
 3215231482,
 3215237574,
 3215368109,
 3215428944,
 3215401494,
 3215236425,
 3215453523,
 3215429486,
 3215325280,
 3215195387,
 3215351934,
 3215360397,
 3215281754,
 3215188607,
 3215248934,
 3215379022,
 3215248543,
 3215204076,
 3215469804,
 3215228885,
 3215142848,
 3215429599,
 3215271563,
 3215123800,
 3215141137,
 3215221304,
 3215146919,
 3215138954,
 3215226005,
 3215418104,
 3215485789,
 3215364262,
 3215384549,
 3215354207,
 3215221726,
 3215397936,
 3215253751,
 3215335055,
 3215178825,
 3215351093,
 3215263210,
 3215311818,
 3215111810,
 3215195203,
 3215457131,
 3215224003,
 3215323074,
 3215256552,
 3215290260,
 3215100967,
 3215406255,
 3215317707,
 3215471566,
 3215299694,
 3215334730,
 3215233853,
 3215233472,
 3215139044,
 3215385425,
 3215242754,
 3215168054,
 3215470801,
 3215290794,
 3215202214,
 3215220375,
 3215139027,
 3215158933,
 3215429365,
 3215249296,
 3215412550,
 3215158795,
 3215428180,
 3215248555,
 3215092943,
 3215477813,
 3215357991,
 3215121936,
 3215451491,
 3215179481,
 3215339823,
 3215213742,
 3215291849,
 3215359386,
 3215463125,
 3215320771,
 3215156866,
 3215146289,
 3215436023,
 3215120403,
 3215356956,
 3215349109,
 3215293231,
 3215295277,
 3215332926,
 3215468557,
 3215099428,
 3215320538,
 3215097367,
 3215410767,
 3215319194,
 3215189993,
 3215450560,
 3215415472,
 3215473056,
 3215218922,
 3215398154,
 3215196752,
 3215116970,
 3215154857,
 3215378775,
 3215422840,
 3215438030,
 3215467864,
 3215148515,
 3215127261,
 3215403415,
 3215371556,
 3215373819,
 3215289869,
 3215261385,
 3215113646,
 3215093449,
 3215397566,
 3215394731,
 3215202644,
 3215122425,
 3215232510,
 3215376702,
 3215271022,
 3215410326,
 3215238449,
 3215219242,
 3215314765,
 3215415407,
 3215454340,
 3215159145,
 3215496824,
 3215153285,
 3215421241,
 3215158522,
 3215285625,
 3215409861,
 3215146832,
 3215350288,
 3215138718,
 3215216041,
 3215430085,
 3215162220,
 3215495154,
 3215260875,
 3215240410,
 3215315676,
 3215259776,
 3215322753,
 3215232566,
 3215155195,
 3215219753,
 3215153655,
 3215443368,
 3215100159,
 3215362400,
 3215227158,
 3215114738,
 3215377124,
 3215387761,
 3215490323,
 3215224622,
 3215465088,
 3215280953,
 3215356704,
 3215432469,
 3215209576,
 3215320141,
 3215150494,
 3215334861,
 3215168791,
 3215310210,
 3215460927,
 3215440346,
 3215470802,
 3215129573,
 3215453894,
 3215327057,
 3215229668,
 3215248682,
 3215159362,
 3215139650,
 3215291485,
 3215450401,
 3215217478,
 3215373808,
 3215293059,
 3215134209,
 3215205881,
 3215303809,
 3215264353,
 3215268085,
 3215283260,
 3215344640,
 3215184925,
 3215335004,
 3215168443,
 3215193183,
 3215278547,
 3215183829,
 3215184150,
 3215163824,
 3215363725,
 3215404047,
 3215237753,
 3215487652,
 3215429858,
 3215166954,
 3215133584,
 3215178851,
 3215448807,
 3215273126,
 3215168647,
 3215236617,
 3215250424,
 3215435433,
 3215106537,
 3215425535,
 3215403559,
 3215279143,
 3215273425,
 3215236699,
 3215420829,
 3215249423,
 3215204448,
 3215354835,
 3215110230,
 3215354404,
 3215269421,
 3215475433,
 3215122079,
 3215120445,
 3215483256,
 3215270003,
 3215239180,
 3215449129,
 3215466202,
 3215446800,
 3215404824,
 3215216964,
 3215292094,
 3215244813,
 3215257022,
 3215138667,
 3215116447,
 3215186293,
 3215168235,
 3215324033,
 3215203845,
 3215198821,
 3215303652,
 3215328046,
 3215373453,
 3215449613,
 3215349133,
 3215230728,
 3215465483,
 3215360232,
 3215242866,
 3215186737,
 3215303349,
 3215229664,
 3215358878,
 3215290969,
 3215115985,
 3215322709,
 3215364163,
 3215160302,
 3215280768,
 3215296742,
 3215491599,
 3215105569,
 3215125998,
 3215160079,
 3215387762,
 3215469210,
 3215489589,
 3215101312,
 3215443055,
 3215355623,
 3215140780,
 3215091713,
 3215246979,
 3215198239,
 3215464565,
 3215210891,
 3215418615,
 3215189899,
 3215408690,
 3215398951,
 3215107442,
 3215379218,
 3215114940,
 3215131372,
 3215429005,
 3215228880,
 3215464416,
 3215268798,
 3215197421,
 3215230662,
 3215367208,
 3215416423,
 3215233122,
 3215176893,
 3215128471,
 3215220038,
 3215316757,
 3215470163,
 3215130943,
 3215468280,
 3215196602,
 3215426455,
 3215350260,
 3215192607,
 3215232226,
 3215162029,
 3215244712,
 3215466440,
 3215199992,
 3215482756,
 3215271541,
 3215195629,
 3215270783,
 3215259663,
 3215436836,
 3215256304,
 3215491011,
 3215381245,
 3215359319,
 3215222660,
 3215449248,
 3215110552,
 3215117067,
 3215374344,
 3215358157,
 3215287506,
 3215458174,
 3215110350,
 3215113122,
 3215485265,
 3215374250,
 3215113325,
 3215265759,
 3215322186,
 3215265172,
 3215421174,
 3215379819,
 3215093702,
 3215096117,
 3215492042,
 3215326780,
 3215223250,
 3215335663,
 3215118408,
 3215090968,
 3215420674,
 3215379283,
 3215242491,
 3215124675,
 3215494997,
 3215356775,
 3215244219,
 3215254617,
 3215232682,
 3215208828,
 3215145084,
 3215358667,
 3215447392,
 3215289158,
 3215161481,
 3215132824,
 3215173685,
 3215376877,
 3215336686,
 3215149242,
 3215308356,
 3215463785,
 3215143402,
 3215347623,
 3215123239,
 3215205312,
 ...]



In [8]:

    
df['residual'].values.tolist()









    Out[8]:





[60.09383010864258,
 22.98058319091797,
 14.018461227416992,
 12.276541709899902,
 10.960655212402344,
 10.190336227416992,
 9.797115325927734,
 8.665862083435059,
 8.291083335876465,
 6.753566265106201,
 5.982767105102539,
 5.009283542633057,
 4.787799835205078,
 4.737585544586182,
 4.151325702667236,
 3.8586721420288086,
 3.7200565338134766,
 3.653451919555664,
 3.6120126247406006,
 3.55796217918396,
 3.5249321460723877,
 3.4550528526306152,
 3.4301578998565674,
 2.9261271953582764,
 2.8132457733154297,
 2.767648220062256,
 2.758488178253174,
 2.543581247329712,
 2.5158803462982178,
 2.4831578731536865,
 2.4762089252471924,
 2.392141342163086,
 2.38230299949646,
 2.3297765254974365,
 2.2398288249969482,
 2.229811429977417,
 2.1894421577453613,
 2.1700241565704346,
 2.150897741317749,
 2.0817346572875977,
 2.0602142810821533,
 2.044538974761963,
 2.038116693496704,
 2.0010886192321777,
 1.953024983406067,
 1.9425238370895386,
 1.919196367263794,
 1.8698474168777466,
 1.8231494426727295,
 1.7869755029678345,
 1.7188692092895508,
 1.7181131839752197,
 1.7047621011734009,
 1.6976909637451172,
 1.691680669784546,
 1.6713063716888428,
 1.6653168201446533,
 1.6202446222305298,
 1.6132309436798096,
 1.6028668880462646,
 1.581583023071289,
 1.5775842666625977,
 1.5626850128173828,
 1.5582425594329834,
 1.5492855310440063,
 1.5385805368423462,
 1.5317155122756958,
 1.5211113691329956,
 1.5076344013214111,
 1.4967833757400513,
 1.4829853773117065,
 1.4817631244659424,
 1.4691892862319946,
 1.4565132856369019,
 1.4442218542099,
 1.4434107542037964,
 1.4333841800689697,
 1.4322049617767334,
 1.427970290184021,
 1.424288272857666,
 1.4241111278533936,
 1.4210336208343506,
 1.4192612171173096,
 1.4180662631988525,
 1.418045997619629,
 1.4131355285644531,
 1.3996341228485107,
 1.3977913856506348,
 1.392609715461731,
 1.3883095979690552,
 1.382023572921753,
 1.3814785480499268,
 1.3794400691986084,
 1.364009976387024,
 1.3631293773651123,
 1.3609627485275269,
 1.358451247215271,
 1.3547123670578003,
 1.3397060632705688,
 1.3350133895874023,
 1.3328020572662354,
 1.3265687227249146,
 1.3234529495239258,
 1.3199751377105713,
 1.2998747825622559,
 1.297800898551941,
 1.2892454862594604,
 1.2862396240234375,
 1.283833622932434,
 1.2751528024673462,
 1.2714056968688965,
 1.2680243253707886,
 1.2384188175201416,
 1.2283591032028198,
 1.222191572189331,
 1.2142605781555176,
 1.206538200378418,
 1.2041807174682617,
 1.1953727006912231,
 1.1952720880508423,
 1.182854175567627,
 1.181726336479187,
 1.179439902305603,
 1.171888828277588,
 1.1665818691253662,
 1.16591215133667,
 1.162890076637268,
 1.1562278270721436,
 1.1520692110061646,
 1.144974946975708,
 1.1360812187194824,
 1.1236978769302368,
 1.120529055595398,
 1.118119239807129,
 1.117451548576355,
 1.1165120601654053,
 1.1132597923278809,
 1.1122854948043823,
 1.1096220016479492,
 1.1047585010528564,
 1.1041306257247925,
 1.1024683713912964,
 1.1014271974563599,
 1.0916112661361694,
 1.0913853645324707,
 1.0900408029556274,
 1.084211826324463,
 1.083559513092041,
 1.0815109014511108,
 1.080784797668457,
 1.079131007194519,
 1.0745054483413696,
 1.0729814767837524,
 1.072069764137268,
 1.0666135549545288,
 1.064724326133728,
 1.0637664794921875,
 1.0632541179656982,
 1.0616685152053833,
 1.058560848236084,
 1.051201343536377,
 1.0496363639831543,
 1.0495151281356812,
 1.049206018447876,
 1.0463898181915283,
 1.0434927940368652,
 1.0426044464111328,
 1.0395288467407227,
 1.0386379957199097,
 1.0352888107299805,
 1.0337138175964355,
 1.033097743988037,
 1.032279133796692,
 1.0288360118865967,
 1.0225169658660889,
 1.0223387479782104,
 1.0201956033706665,
 1.0195045471191406,
 1.0138232707977295,
 1.0096408128738403,
 1.0063974857330322,
 1.0060631036758423,
 1.0054899454116821,
 1.0046826601028442,
 1.0020874738693237,
 1.0011916160583496,
 0.999028205871582,
 0.9979208707809448,
 0.9951856136322021,
 0.9943534135818481,
 0.9923041462898254,
 0.9922865629196167,
 0.9902551770210266,
 0.9878281354904175,
 0.9870291948318481,
 0.9827179312705994,
 0.9821138381958008,
 0.9803048968315125,
 0.9795767068862915,
 0.9786282777786255,
 0.9723061919212341,
 0.970488965511322,
 0.9695155620574951,
 0.9688208103179932,
 0.9679350256919861,
 0.9669782519340515,
 0.9644799828529358,
 0.9643834233283997,
 0.9639016389846802,
 0.9597914218902588,
 0.9596630334854126,
 0.9576871991157532,
 0.9555385112762451,
 0.9550148248672485,
 0.9534093141555786,
 0.9518917202949524,
 0.9480692744255066,
 0.945451021194458,
 0.9449933171272278,
 0.9447908401489258,
 0.9431512951850891,
 0.9390014410018921,
 0.9389370083808899,
 0.9385173320770264,
 0.9377952814102173,
 0.9375945329666138,
 0.936521053314209,
 0.9340055584907532,
 0.9331918358802795,
 0.9330150485038757,
 0.9317139983177185,
 0.9304264783859253,
 0.9288665056228638,
 0.9260517358779907,
 0.9249129891395569,
 0.9222087264060974,
 0.9214598536491394,
 0.9209437966346741,
 0.9209313988685608,
 0.9208395481109619,
 0.9197496175765991,
 0.9185452461242676,
 0.9154036641120911,
 0.9132921695709229,
 0.913219690322876,
 0.9126948118209839,
 0.9088218212127686,
 0.9086437821388245,
 0.9084234237670898,
 0.9055785536766052,
 0.9055338501930237,
 0.9030809998512268,
 0.9009914398193359,
 0.8999032974243164,
 0.899492621421814,
 0.8974443078041077,
 0.8955872058868408,
 0.8940939903259277,
 0.8897515535354614,
 0.8891883492469788,
 0.8884326219558716,
 0.8873404860496521,
 0.8867751955986023,
 0.8864952921867371,
 0.886139452457428,
 0.8853210210800171,
 0.8835608959197998,
 0.8831455707550049,
 0.8814792633056641,
 0.8786903023719788,
 0.8786889314651489,
 0.8786466121673584,
 0.878573477268219,
 0.876968264579773,
 0.8747633695602417,
 0.8735519647598267,
 0.8730534911155701,
 0.8729946613311768,
 0.8721534609794617,
 0.8719340562820435,
 0.8715412616729736,
 0.8697844743728638,
 0.8684020638465881,
 0.8658937811851501,
 0.8647394180297852,
 0.8627612590789795,
 0.8623327016830444,
 0.8616533875465393,
 0.8615354299545288,
 0.8612295389175415,
 0.8595142364501953,
 0.8593857288360596,
 0.8583488464355469,
 0.8570705652236938,
 0.8564101457595825,
 0.8563953042030334,
 0.8560832738876343,
 0.8552137017250061,
 0.8547729253768921,
 0.8529301881790161,
 0.851856529712677,
 0.8514204025268555,
 0.8499778509140015,
 0.8469444513320923,
 0.8467068672180176,
 0.846032440662384,
 0.845173716545105,
 0.8448122143745422,
 0.8432825803756714,
 0.8417072296142578,
 0.8406412601470947,
 0.8391038179397583,
 0.8389773368835449,
 0.8380493521690369,
 0.8373729586601257,
 0.8368411064147949,
 0.836216926574707,
 0.8358647227287292,
 0.8336787223815918,
 0.8333789110183716,
 0.8329710364341736,
 0.8316418528556824,
 0.8302108645439148,
 0.830048143863678,
 0.8297022581100464,
 0.8295499682426453,
 0.8286826014518738,
 0.8277784585952759,
 0.827299177646637,
 0.8271003365516663,
 0.8269991278648376,
 0.8262494802474976,
 0.8259848952293396,
 0.8258390426635742,
 0.8255735635757446,
 0.8250579833984375,
 0.8247339129447937,
 0.8241198658943176,
 0.8225634694099426,
 0.8194668889045715,
 0.8185660243034363,
 0.8180004358291626,
 0.8172050714492798,
 0.8170410990715027,
 0.8164354562759399,
 0.8154846429824829,
 0.8146878480911255,
 0.8123286962509155,
 0.8110998868942261,
 0.8109608292579651,
 0.810543417930603,
 0.8094967007637024,
 0.8084927797317505,
 0.8075140118598938,
 0.8060798048973083,
 0.8060481548309326,
 0.8048883676528931,
 0.8044706583023071,
 0.8039005994796753,
 0.8037306070327759,
 0.8031567335128784,
 0.8023133277893066,
 0.7998040914535522,
 0.7996733784675598,
 0.7978358864784241,
 0.7977395057678223,
 0.7967928051948547,
 0.7952989935874939,
 0.7938899993896484,
 0.7934744358062744,
 0.7932054400444031,
 0.792657196521759,
 0.792655348777771,
 0.7925220727920532,
 0.7915612459182739,
 0.7906990051269531,
 0.7903063893318176,
 0.7900872230529785,
 0.7893491983413696,
 0.7878237962722778,
 0.7857347726821899,
 0.7855825424194336,
 0.7842476963996887,
 0.7831416130065918,
 0.7821829915046692,
 0.7811717391014099,
 0.7788980007171631,
 0.7783818244934082,
 0.7776341438293457,
 0.7757152915000916,
 0.7753698825836182,
 0.7746968865394592,
 0.7746821045875549,
 0.7739043831825256,
 0.7726438641548157,
 0.7714319229125977,
 0.770415723323822,
 0.7700993418693542,
 0.7694905996322632,
 0.7671429514884949,
 0.7669565677642822,
 0.7655788660049438,
 0.765575647354126,
 0.7641459107398987,
 0.7632062435150146,
 0.7624366879463196,
 0.7618311047554016,
 0.761588990688324,
 0.7593628168106079,
 0.7590985894203186,
 0.7585266828536987,
 0.7580933570861816,
 0.7577105760574341,
 0.7575187087059021,
 0.7572482824325562,
 0.7569153308868408,
 0.7563158273696899,
 0.7544280290603638,
 0.7543737888336182,
 0.7539798021316528,
 0.7536249160766602,
 0.7536008954048157,
 0.7532666325569153,
 0.7531739473342896,
 0.752880871295929,
 0.7528793811798096,
 0.7526946067810059,
 0.751673698425293,
 0.7499583959579468,
 0.7496762275695801,
 0.7489606142044067,
 0.7485086917877197,
 0.7477176785469055,
 0.7476752400398254,
 0.7472490668296814,
 0.7472139596939087,
 0.7466698884963989,
 0.7460474967956543,
 0.7460435628890991,
 0.744686484336853,
 0.7445034384727478,
 0.7432817220687866,
 0.7422952055931091,
 0.7414557337760925,
 0.7412257194519043,
 0.7403615713119507,
 0.7403483986854553,
 0.7383480072021484,
 0.7383477091789246,
 0.7380225658416748,
 0.7366001009941101,
 0.7361823320388794,
 0.7343788743019104,
 0.733960747718811,
 0.732728123664856,
 0.7322820425033569,
 0.7319163680076599,
 0.7304933071136475,
 0.730053722858429,
 0.7288756966590881,
 0.7281818985939026,
 0.7281191945075989,
 0.727988064289093,
 0.7277131676673889,
 0.726827085018158,
 0.7263097763061523,
 0.7255817651748657,
 0.7255740761756897,
 0.7253108024597168,
 0.7246365547180176,
 0.7239059209823608,
 0.7233874797821045,
 0.7230355143547058,
 0.723025918006897,
 0.7225368022918701,
 0.7217145562171936,
 0.7210047841072083,
 0.7206993699073792,
 0.7205953001976013,
 0.7205812335014343,
 0.719957172870636,
 0.7198128700256348,
 0.7192848324775696,
 0.7190936803817749,
 0.7181803584098816,
 0.7170781493186951,
 0.7161182165145874,
 0.7156762480735779,
 0.7153546810150146,
 0.7148224711418152,
 0.7141834497451782,
 0.7140740752220154,
 0.7137510776519775,
 0.7134366035461426,
 0.7130857706069946,
 0.7130664587020874,
 0.7126566767692566,
 0.7121343612670898,
 0.7115955352783203,
 0.7110877633094788,
 0.7108469009399414,
 0.7104701399803162,
 0.7101932168006897,
 0.7093899846076965,
 0.7090077996253967,
 0.7088366150856018,
 0.7085787653923035,
 0.7084393501281738,
 0.7083696126937866,
 0.7082640528678894,
 0.7068285942077637,
 0.7065154314041138,
 0.7061051726341248,
 0.7060731053352356,
 0.706041693687439,
 0.7057188153266907,
 0.704667866230011,
 0.7045057415962219,
 0.7043561339378357,
 0.7041466236114502,
 0.7040076851844788,
 0.7035585641860962,
 0.7030852437019348,
 0.7029399871826172,
 0.7028341293334961,
 0.7018325328826904,
 0.7014175057411194,
 0.7009782791137695,
 0.7006656527519226,
 0.6993235945701599,
 0.6986374855041504,
 0.6984686255455017,
 0.6980193257331848,
 0.6978328227996826,
 0.6967384219169617,
 0.6964960694313049,
 0.6964261531829834,
 0.6963183879852295,
 0.6954740881919861,
 0.6952798366546631,
 0.695203423500061,
 0.6944896578788757,
 0.6942644119262695,
 0.6942504644393921,
 0.6940212249755859,
 0.693926990032196,
 0.693915843963623,
 0.69270259141922,
 0.6923828721046448,
 0.6921383738517761,
 0.6912344694137573,
 0.6909181475639343,
 0.6908226013183594,
 0.6903275847434998,
 0.6885342001914978,
 0.6880930662155151,
 0.687547504901886,
 0.6873355507850647,
 0.6869430541992188,
 0.6868414878845215,
 0.6866506934165955,
 0.6856019496917725,
 0.6853703260421753,
 0.684950053691864,
 0.6844768524169922,
 0.6843836903572083,
 0.6843326687812805,
 0.6840593218803406,
 0.6839963793754578,
 0.6833631992340088,
 0.6820196509361267,
 0.6815665364265442,
 0.6803432703018188,
 0.6799181699752808,
 0.679553210735321,
 0.678919792175293,
 0.6787182688713074,
 0.6786744594573975,
 0.6782424449920654,
 0.6780423521995544,
 0.6777185201644897,
 0.6777063608169556,
 0.6776937246322632,
 0.6772192120552063,
 0.6769409775733948,
 0.6760160326957703,
 0.6759694814682007,
 0.6755587458610535,
 0.675423800945282,
 0.675076961517334,
 0.6750593781471252,
 0.6749622821807861,
 0.674652636051178,
 0.674209713935852,
 0.6741954684257507,
 0.6737287640571594,
 0.6736929416656494,
 0.6729382276535034,
 0.6728609204292297,
 0.6727287173271179,
 0.6725167632102966,
 0.6722366809844971,
 0.6722153425216675,
 0.6721881628036499,
 0.6720960140228271,
 0.671940267086029,
 0.6713663339614868,
 0.6698254346847534,
 0.6695573329925537,
 0.6687107682228088,
 0.6684702634811401,
 0.6678615808486938,
 0.6676300764083862,
 0.6671851277351379,
 0.6671191453933716,
 0.6668809652328491,
 0.6667686700820923,
 0.6666104197502136,
 0.6665686368942261,
 0.6664994955062866,
 0.6663007736206055,
 0.6657471656799316,
 0.6657465696334839,
 0.6657367944717407,
 0.6655275225639343,
 0.6652972102165222,
 0.6651871204376221,
 0.6651492118835449,
 0.6650216579437256,
 0.6648292541503906,
 0.6644583940505981,
 0.6641172766685486,
 0.6637612581253052,
 0.6635695099830627,
 0.6633908748626709,
 0.6629290580749512,
 0.6624594330787659,
 0.662200927734375,
 0.6617655158042908,
 0.6606206893920898,
 0.6597539186477661,
 0.6595001816749573,
 0.659498393535614,
 0.6579702496528625,
 0.657727062702179,
 0.6574667096138,
 0.656964123249054,
 0.6567689776420593,
 0.6556574702262878,
 0.655053436756134,
 0.6549975872039795,
 0.6542130708694458,
 0.6539175510406494,
 0.6536197662353516,
 0.6534443497657776,
 0.6531872153282166,
 0.6529080867767334,
 0.652268648147583,
 0.6522124409675598,
 0.6521428227424622,
 0.652125895023346,
 0.6517173647880554,
 0.6509376764297485,
 0.6509096622467041,
 0.6508827805519104,
 0.6505467891693115,
 0.6492860317230225,
 0.6486479043960571,
 0.6486145853996277,
 0.6485013365745544,
 0.6484130620956421,
 0.6481355428695679,
 0.647615909576416,
 0.6474666595458984,
 0.6465640068054199,
 0.6460104584693909,
 0.6459749937057495,
 0.6459360122680664,
 0.6455346345901489,
 0.6451497077941895,
 0.6451070308685303,
 0.6449517607688904,
 0.6449416875839233,
 0.6448999047279358,
 0.6447224020957947,
 0.6444844603538513,
 0.6444458365440369,
 0.6442970633506775,
 0.6436285972595215,
 0.6432618498802185,
 0.6431841850280762,
 0.6425811052322388,
 0.6417860984802246,
 0.6417686343193054,
 0.6416600346565247,
 0.6415826082229614,
 0.6414430141448975,
 0.6410348415374756,
 0.640630841255188,
 0.6404160261154175,
 0.6401026248931885,
 0.63993239402771,
 0.639882504940033,
 0.6395856142044067,
 0.639421820640564,
 0.6380833983421326,
 0.6379858255386353,
 0.6375442743301392,
 0.637353777885437,
 0.6371648907661438,
 0.6369909048080444,
 0.6369869112968445,
 0.6367626786231995,
 0.6362351179122925,
 0.636012613773346,
 0.6358704566955566,
 0.6356680393218994,
 0.6356462240219116,
 0.6355707049369812,
 0.6354865431785583,
 0.6353418827056885,
 0.6348565816879272,
 0.6345191597938538,
 0.6345074772834778,
 0.6344534754753113,
 0.6339588761329651,
 0.6336132884025574,
 0.6334683895111084,
 0.6332551836967468,
 0.6330251693725586,
 0.6329081058502197,
 0.6327518224716187,
 0.6318038702011108,
 0.6317944526672363,
 0.6317700147628784,
 0.6317121982574463,
 0.6316589713096619,
 0.6316245198249817,
 0.6313509941101074,
 0.6306886076927185,
 0.6305879950523376,
 0.6305497884750366,
 0.6305269002914429,
 0.630096971988678,
 0.6296277046203613,
 0.6295402646064758,
 0.6295005679130554,
 0.6292697787284851,
 0.6290936470031738,
 0.6286728382110596,
 0.6286492943763733,
 0.6285806894302368,
 0.6283563375473022,
 0.6279318332672119,
 0.6277845501899719,
 0.6277395486831665,
 0.6270812749862671,
 0.6268585920333862,
 0.6256161332130432,
 0.6254564523696899,
 0.6253827214241028,
 0.6252259016036987,
 0.6251668334007263,
 0.6251053810119629,
 0.6239646077156067,
 0.6235501170158386,
 0.623266875743866,
 0.6227879524230957,
 0.6223676800727844,
 0.6216578483581543,
 0.620872974395752,
 0.6207531690597534,
 0.6204898357391357,
 0.6199707388877869,
 0.619939923286438,
 0.619571328163147,
 0.6194667220115662,
 0.6189941167831421,
 0.6188552975654602,
 0.6185175776481628,
 0.6182923316955566,
 0.6180881261825562,
 0.6179910898208618,
 0.6179143786430359,
 0.6177471876144409,
 0.6172986030578613,
 0.6171382069587708,
 0.6170961856842041,
 0.6170250177383423,
 0.6169363260269165,
 0.61668461561203,
 0.6163849234580994,
 0.616259753704071,
 0.6160246133804321,
 0.6159703731536865,
 0.6155744194984436,
 0.6147223114967346,
 0.6145497560501099,
 0.6140955686569214,
 0.6140169501304626,
 0.6139309406280518,
 0.6138129830360413,
 0.613811731338501,
 0.6133521199226379,
 0.6131906509399414,
 0.6131694316864014,
 0.6130494475364685,
 0.612968921661377,
 0.6129478812217712,
 0.6116432547569275,
 0.6116387844085693,
 0.6113128662109375,
 0.6108748316764832,
 0.6107308864593506,
 0.6106475591659546,
 0.610262393951416,
 0.6101744174957275,
 0.6101717948913574,
 0.6101569533348083,
 0.6100338697433472,
 0.6099213361740112,
 0.6097745299339294,
 0.6095500588417053,
 0.6094909310340881,
 0.6094450354576111,
 0.6090566515922546,
 0.6087564826011658,
 0.6087318658828735,
 0.6083419919013977,
 0.6081241369247437,
 0.607871413230896,
 0.6077539920806885,
 0.6068829894065857,
 0.6065410375595093,
 0.6062430739402771,
 0.6058253645896912,
 0.6055247783660889,
 0.6053788065910339,
 0.6053472757339478,
 0.6052625775337219,
 0.6048938632011414,
 0.604708194732666,
 0.6043591499328613,
 0.6043391823768616,
 0.6043362021446228,
 0.6042796969413757,
 0.6038782596588135,
 0.6036362051963806,
 0.6035915613174438,
 0.6033117175102234,
 0.6032079458236694,
 0.6029709577560425,
 0.6028017997741699,
 0.6026512384414673,
 0.6024478673934937,
 0.602216899394989,
 0.6018061637878418,
 0.6017376184463501,
 0.6012206673622131,
 0.6007962226867676,
 0.6006240248680115,
 0.6005715131759644,
 0.6003936529159546,
 0.6003291606903076,
 0.6000486612319946,
 0.5996347665786743,
 0.5994622707366943,
 0.599257230758667,
 0.5991039276123047,
 0.5984320640563965,
 0.5983387231826782,
 0.5982711911201477,
 0.5978343486785889,
 0.5977898240089417,
 0.5977823734283447,
 0.5976536273956299,
 0.5973724722862244,
 0.5973426699638367,
 0.597022533416748,
 0.5968469381332397,
 0.5965945720672607,
 0.5965850353240967,
 0.5964047312736511,
 0.5962276458740234,
 0.5960531830787659,
 0.5958200097084045,
 0.595647394657135,
 0.5950970649719238,
 0.5949172377586365,
 0.5947761535644531,
 0.5942966938018799,
 0.5942276120185852,
 0.5939619541168213,
 0.5939454436302185,
 0.5937827229499817,
 0.5937712788581848,
 0.593761682510376,
 0.5936322212219238,
 0.5930981636047363,
 0.5929948091506958,
 0.592664361000061,
 0.5925729870796204,
 0.5924862623214722,
 0.5924296379089355,
 0.5921010375022888,
 0.591584324836731,
 0.5915680527687073,
 0.5915402173995972,
 0.5914790630340576,
 0.5913110375404358,
 0.5911639928817749,
 0.5910633206367493,
 0.5909237265586853,
 0.5906891226768494,
 0.5906646847724915,
 0.5903602838516235,
 0.5902671217918396,
 0.5898656845092773,
 0.5893831849098206,
 0.5891802310943604,
 0.5889486074447632,
 0.5889463424682617,
 0.5887507200241089,
 0.588686466217041,
 0.5885319113731384,
 0.588383138179779,
 0.5882313847541809,
 0.5881727933883667,
 0.5880797505378723,
 0.5878171324729919,
 0.587637186050415,
 0.5874754786491394,
 0.5874388217926025,
 0.5873831510543823,
 0.5872762203216553,
 0.5871644020080566,
 0.5870528817176819,
 0.5870521664619446,
 0.5866469740867615,
 0.5865029096603394,
 0.5861585140228271,
 0.5860657691955566,
 0.585838794708252,
 0.5858339071273804,
 0.5855178833007812,
 0.5853433609008789,
 0.5853050351142883,
 0.584908664226532,
 0.5847539901733398,
 0.5840827226638794,
 0.5840543508529663,
 0.5838108062744141,
 0.5837247967720032,
 0.5834543704986572,
 0.583451509475708,
 0.5830873250961304,
 0.5830100774765015,
 0.5828806757926941,
 0.5824500322341919,
 0.5822259187698364,
 0.5821603536605835,
 0.5813983678817749,
 0.5812755823135376,
 0.5811508893966675,
 0.5811348557472229,
 0.5810601115226746,
 0.5809980034828186,
 0.5809274315834045,
 0.5805932879447937,
 0.5805398225784302,
 0.5800670385360718,
 0.5799999833106995,
 0.579505443572998,
 0.5792343020439148,
 0.5792225003242493,
 0.5791121125221252,
 0.5790188312530518,
 0.5788261890411377,
 0.5787473320960999,
 0.5786319375038147,
 0.5783180594444275,
 0.578109622001648,
 0.5780721306800842,
 0.5780323147773743,
 0.5780043601989746,
 0.5771296620368958,
 0.5771205425262451,
 0.5770235657691956,
 0.5770174264907837,
 0.5769822597503662,
 0.5766937732696533,
 0.5766468048095703,
 0.5765354037284851,
 0.575879693031311,
 0.5758066773414612,
 0.5757375955581665,
 0.5753865838050842,
 0.5751538276672363,
 0.5751133561134338,
 0.5749435424804688,
 ...]



In [9]:

    
percentile99 = np.percentile(df['residual'], 99)



In [10]:

    
sns.distplot(df['residual'].sample(n=10000),  kde=False, rug=True)
plt.show()



In [11]:

    
len(df)









    Out[11]:





42798



In [12]:

    
plt.close()



In [13]:

    
residuals = df['residual'].values.tolist()



In [14]:

    
np.max(residuals) - np.min(residuals)









    Out[14]:





60.033355519175529



In [15]:

    
def RadiantDireDifference(row, column):
    faction = ['radiant', 'dire']
    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']
    radiantSum = 0
    direSum = 0
    for pos in poss:
        radiantSum += row['radiant_'+ pos + '_' + column]
        direSum += row['dire_'+ pos + '_' + column]
    return radiantSum - direSum

def reject_outliers(data, m=1.5):
    return abs(data - np.mean(data)) < m * np.std(data)


def heroKDAdeficit(row):
    faction = ['radiant', 'dire']
    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']
    largestAbsoluteSoFar = -1
    returnVal = -1
    for pos in poss:
        for fac in faction:
            value = row[fac + '_' + pos + '_kills'] + \
                        row[fac + '_' + pos + '_assists'] - \
                        row[fac + '_' + pos + '_deaths']
            if abs(value) >= largestAbsoluteSoFar:
                largestAbsoluteSoFar = abs(value)
                returnVal = value
    return returnVal

def heroKDdeficit(row):
    faction = ['radiant', 'dire']
    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']
    largestAbsoluteSoFar = -1
    returnVal = -1
    for pos in poss:
        for fac in faction:
            value = row[fac + '_' + pos + '_kills'] - \
                        row[fac + '_' + pos + '_deaths']
            if abs(value) >= largestAbsoluteSoFar:
                largestAbsoluteSoFar = abs(value)
                returnVal = value
    return returnVal

def heroKD(row):
    faction = ['radiant', 'dire']
    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']
    largestAbsoluteSoFar = -1
    returnVal = -1
    for pos in poss:
        for fac in faction:
            value = row[fac + '_' + pos + '_kills'] - \
                        row[fac + '_' + pos + '_deaths']
            if abs(value) >= largestAbsoluteSoFar:
                largestAbsoluteSoFar = abs(value)
                returnVal = int(pos[3:4])
    return returnVal

def heroGoldSpent(row):
    faction = ['radiant', 'dire']
    poss = ['pos1', 'pos2', 'pos3', 'pos4', 'pos5']
    largestAbsoluteSoFar = -1
    returnVal = -1
    for pos in poss:
        for fac in faction:
            value = row[fac + '_' + pos + '_gold_spent']
            if abs(value) >= largestAbsoluteSoFar:
                largestAbsoluteSoFar = abs(value)
                returnVal = int(pos[3:4])
    return returnVal



In [16]:

    
vote_mask = (df['positive_votes'] == 0) & (df['negative_votes'] == 0)
def outlier_mask(data, other_mask = vote_mask, m = 2.):
    d = np.abs(data - np.median(data))
    mdev = np.median(d)
    s = d/mdev if mdev else 0.
    return np.logical_and(s<m, other_mask)



In [17]:

    
kill_difference = np.abs(df.apply(RadiantDireDifference, args=('kills',), axis=1))
kill_difference_mask = outlier_mask(kill_difference)
sns.regplot(kill_difference[kill_difference_mask], np.log(df['residual'][kill_difference_mask]))
ax = plt.gca()
ax.set_xlabel("Kill difference")
ax.set_ylabel("Logarithm of Residual")
plt.savefig('graphs/Kill_Difference.png', bbox_inches='tight')
plt.show()



In [18]:

    
gold_difference = np.abs(df.apply(RadiantDireDifference, args=('gpm',), axis=1))
gold_difference_mask = outlier_mask(gold_difference)
sns.regplot(np.abs(gold_difference)[gold_difference_mask], df['residual'][gold_difference_mask])
ax = plt.gca()
ax.set_xlabel("GPM difference")
ax.set_ylabel("Raw Residual")
plt.savefig('graphs/GPM_Difference.png', bbox_inches='tight')
plt.show()



In [19]:

    
kda = df.apply(heroKDAdeficit, axis=1)
kda_mask = outlier_mask(kda)
sns.regplot(kda[kda_mask], df['residual'][kda_mask])
ax = plt.gca()
ax.set_xlabel("K + A - D")
ax.set_ylabel("Raw Residual")
plt.savefig('graphs/MaxKAD_Difference.png', bbox_inches='tight')
plt.show()



In [20]:

    
kd = df.apply(heroKDdeficit, axis=1)
kd_mask = outlier_mask(kd)
sns.regplot(np.abs(kd)[kd_mask], df['residual'][kd_mask])
ax = plt.gca()
ax.set_xlabel("Absolute(K - D)")
ax.set_ylabel("Raw Residual")
plt.savefig('graphs/absolutek-d_Difference.png', bbox_inches='tight')
plt.show()



In [21]:

    
direpos2kills = df['dire_pos2_kills']
direpos2kills_mask = outlier_mask(direpos2kills)
sns.regplot(df['dire_pos2_kills'][direpos2kills_mask], df['residual'][direpos2kills_mask])
ax = plt.gca()
ax.set_xlabel("Dire Pos 2 Kills")
ax.set_ylabel("Raw Residual")
plt.savefig('graphs/direpos2kills.png', bbox_inches='tight')
plt.show()



In [22]:

    
radiantpos1kills = df['radiant_pos1_kills']
radiantpos1kills_mask = outlier_mask(radiantpos1kills)
sns.regplot(df['radiant_pos1_kills'][radiantpos1kills_mask], df['residual'][radiantpos1kills_mask])
ax = plt.gca()
ax.set_xlabel("Radiant Pos 1 Kills")
ax.set_ylabel("Raw Residual")
plt.savefig('graphs/radiantpos1kills.png', bbox_inches='tight')
plt.show()



In [23]:

    
radiantpos1kd = df['radiant_pos1_kills'] - df['radiant_pos1_deaths']
radiantpos1kd_mask = outlier_mask(radiantpos1kd)
sns.regplot(radiantpos1kd[radiantpos1kd_mask], df['residual'][radiantpos1kd_mask])
ax = plt.gca()
ax.set_xlabel("Radiant Pos 1 Kills - Deaths")
ax.set_ylabel("Raw Residual")
plt.savefig('graphs/radiantpos1killsminusdeaths.png', bbox_inches='tight')
plt.show()



In [24]:

    
radiantpos5kd = df['radiant_pos5_kills'] - df['radiant_pos5_deaths']
radiantpos5kd_mask = outlier_mask(radiantpos5kd)
sns.regplot(radiantpos5kd[radiantpos5kd_mask], df['residual'][radiantpos5kd_mask])
ax = plt.gca()
ax.set_xlabel("Radiant Pos 5 Kills - Deaths")
ax.set_ylabel("Raw Residual")
plt.savefig('graphs/radiantpos5killsminusdeaths.png', bbox_inches='tight')
plt.show()



In [25]:

    
duration = df['duration']
duration_mask = outlier_mask(duration)
sns.regplot(df['duration'][duration_mask], df['residual'][duration_mask])
ax = plt.gca()
ax.set_xlabel("Duration")
ax.set_ylabel("Raw Residual")
plt.savefig('graphs/durationresidual.png', bbox_inches='tight')
plt.show()



In [26]:

    
df[df['residual'] > percentile99]









    Out[26]:






  
    
      
      match_id
      positive_votes
      negative_votes
      first_blood_time
      radiant_win
      duration
      dire_pos1_ancient_kills
      dire_pos1_courier_kills
      dire_pos1_kills
      dire_pos1_neutral_kills
      ...
      dire_pos2_abandons
      dire_pos3_abandons
      dire_pos4_abandons
      dire_pos5_abandons
      radiant_pos1_abandons
      radiant_pos2_abandons
      radiant_pos3_abandons
      radiant_pos4_abandons
      radiant_pos5_abandons
      residual
    
  
  
    
      40056
      3215497967
      49
      2
      14
      True
      2748
      43.0
      0.0
      4
      141.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      60.093830
    
    
      3222
      3215097220
      15
      7
      15
      True
      2470
      23.0
      0.0
      7
      108.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      22.980583
    
    
      42706
      3215312675
      9
      4
      0
      False
      2687
      0.0
      0.0
      10
      9.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      14.018461
    
    
      14505
      3215162338
      0
      0
      135
      False
      3981
      18.0
      21.0
      27
      70.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      12.276542
    
    
      8310
      3215181105
      0
      0
      0
      True
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      1
      0
      10.960655
    
    
      6186
      3215377994
      0
      0
      72
      False
      2924
      NaN
      NaN
      28
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      10.190336
    
    
      10574
      3215271631
      7
      1
      3
      False
      1846
      4.0
      0.0
      5
      36.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      9.797115
    
    
      28970
      3215289100
      0
      0
      0
      False
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      1
      0
      0
      0
      0
      0
      0
      8.665862
    
    
      32230
      3215279863
      0
      0
      0
      False
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      1
      0
      0
      0
      0
      0
      0
      8.291083
    
    
      1609
      3215253055
      5
      0
      182
      True
      1949
      0.0
      0.0
      4
      10.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      6.753566
    
    
      41758
      3215448302
      0
      0
      208
      False
      7375
      4.0
      1.0
      21
      61.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      5.982767
    
    
      10214
      3215416370
      0
      0
      0
      False
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      1
      0
      5.009284
    
    
      42735
      3215261563
      0
      0
      0
      True
      0
      NaN
      NaN
      0
      NaN
      ...
      1
      0
      0
      0
      0
      0
      0
      0
      0
      4.787800
    
    
      9080
      3215294948
      4
      1
      94
      True
      1804
      0.0
      0.0
      4
      19.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      4.737586
    
    
      8192
      3215155134
      0
      0
      0
      True
      780
      0.0
      0.0
      0
      0.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      4.151326
    
    
      12556
      3215350255
      0
      0
      0
      True
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      1
      0
      0
      0
      0
      0
      0
      3.858672
    
    
      15967
      3215341221
      0
      0
      0
      False
      932
      NaN
      NaN
      34
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      3.720057
    
    
      21049
      3215306187
      0
      0
      0
      True
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      1
      0
      0
      0
      0
      0
      0
      0
      3.653452
    
    
      5501
      3215302636
      0
      0
      0
      False
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      1
      0
      3.612013
    
    
      28073
      3215316820
      0
      0
      0
      False
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      1
      0
      3.557962
    
    
      37873
      3215190109
      0
      0
      324
      False
      4877
      3.0
      0.0
      6
      24.0
      ...
      0
      0
      0
      0
      0
      1
      0
      0
      0
      3.524932
    
    
      19538
      3215355847
      0
      0
      0
      False
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      1
      0
      3.455053
    
    
      18809
      3215130854
      0
      0
      9
      True
      2760
      0.0
      0.0
      3
      0.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      3.430158
    
    
      37087
      3215259897
      0
      0
      0
      True
      1394
      3.0
      0.0
      1
      3.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      2.926127
    
    
      16205
      3215394671
      0
      0
      0
      True
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      1
      0
      0
      0
      0
      0
      0
      2.813246
    
    
      37604
      3215152225
      0
      0
      0
      True
      2641
      NaN
      NaN
      10
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      2.767648
    
    
      37101
      3215274711
      0
      0
      0
      True
      858
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      2.758488
    
    
      24635
      3215366609
      0
      0
      0
      False
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      1
      0
      2.543581
    
    
      1345
      3215449992
      0
      0
      104
      False
      5045
      60.0
      0.0
      18
      317.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      2.515880
    
    
      26691
      3215376405
      0
      0
      0
      True
      0
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      1
      0
      0
      0
      0
      0
      0
      2.483158
    
    
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
    
    
      20347
      3215337531
      0
      0
      62
      False
      3373
      NaN
      NaN
      1
      NaN
      ...
      0
      0
      0
      0
      0
      1
      0
      1
      0
      0.769491
    
    
      15810
      3215102476
      0
      0
      137
      False
      3312
      45.0
      0.0
      8
      350.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.767143
    
    
      27774
      3215253869
      0
      0
      292
      False
      5234
      NaN
      NaN
      7
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.766957
    
    
      27894
      3215422166
      0
      0
      118
      False
      2817
      47.0
      1.0
      10
      155.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.765579
    
    
      12285
      3215263577
      0
      0
      0
      False
      1143
      NaN
      NaN
      35
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.765576
    
    
      20308
      3215285461
      0
      0
      61
      False
      2682
      20.0
      0.0
      10
      74.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.764146
    
    
      8590
      3215114749
      0
      0
      0
      False
      746
      NaN
      NaN
      3
      NaN
      ...
      0
      0
      0
      0
      1
      0
      1
      1
      1
      0.763206
    
    
      14222
      3215096854
      0
      0
      93
      True
      2936
      21.0
      0.0
      5
      208.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.762437
    
    
      8943
      3215446534
      0
      0
      218
      True
      1714
      0.0
      0.0
      1
      25.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.761831
    
    
      21170
      3215119491
      0
      0
      169
      True
      4051
      75.0
      0.0
      12
      354.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.761589
    
    
      9693
      3215205453
      0
      0
      140
      False
      2083
      NaN
      NaN
      3
      NaN
      ...
      0
      1
      0
      1
      0
      0
      0
      0
      0
      0.759363
    
    
      10663
      3215135325
      0
      0
      0
      True
      3645
      23.0
      0.0
      4
      143.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      1
      0.759099
    
    
      11652
      3215415463
      0
      0
      191
      False
      2846
      31.0
      0.0
      8
      175.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.758527
    
    
      30828
      3215322189
      0
      0
      93
      True
      2680
      0.0
      0.0
      3
      12.0
      ...
      0
      1
      0
      0
      0
      0
      0
      0
      0
      0.758093
    
    
      15394
      3215345510
      0
      0
      307
      True
      4429
      NaN
      NaN
      10
      NaN
      ...
      0
      0
      0
      1
      0
      0
      0
      1
      0
      0.757711
    
    
      7846
      3215138506
      0
      0
      160
      False
      1780
      NaN
      NaN
      4
      NaN
      ...
      1
      1
      1
      0
      0
      0
      0
      0
      0
      0.757519
    
    
      37444
      3215492929
      0
      0
      73
      False
      3048
      20.0
      1.0
      8
      115.0
      ...
      0
      1
      0
      0
      0
      0
      0
      0
      0
      0.757248
    
    
      19730
      3215267165
      0
      0
      2
      True
      2125
      NaN
      NaN
      6
      NaN
      ...
      0
      0
      0
      0
      1
      1
      1
      0
      0
      0.756915
    
    
      32004
      3215123467
      0
      0
      224
      False
      3050
      0.0
      0.0
      2
      12.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.756316
    
    
      288
      3215358156
      0
      0
      163
      True
      4185
      24.0
      0.0
      7
      55.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.754428
    
    
      36615
      3215094142
      0
      0
      203
      True
      4973
      NaN
      NaN
      4
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.754374
    
    
      11493
      3215328883
      0
      0
      3
      True
      3377
      37.0
      0.0
      12
      179.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.753980
    
    
      39744
      3215250045
      0
      0
      184
      True
      4439
      44.0
      0.0
      31
      132.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.753625
    
    
      9634
      3215410384
      0
      0
      0
      True
      344
      NaN
      NaN
      0
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.753601
    
    
      17709
      3215168749
      0
      0
      0
      True
      413
      0.0
      0.0
      0
      0.0
      ...
      0
      0
      1
      0
      0
      0
      0
      0
      0
      0.753267
    
    
      27443
      3215343292
      0
      0
      86
      True
      3411
      NaN
      NaN
      7
      NaN
      ...
      0
      0
      0
      1
      0
      0
      0
      1
      0
      0.753174
    
    
      39465
      3215107300
      0
      0
      108
      True
      2965
      8.0
      0.0
      9
      88.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.752881
    
    
      29673
      3215450701
      0
      0
      90
      False
      2907
      45.0
      1.0
      8
      197.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.752879
    
    
      22278
      3215239658
      0
      0
      126
      False
      3992
      NaN
      NaN
      10
      NaN
      ...
      0
      0
      1
      0
      0
      0
      0
      0
      0
      0.752695
    
    
      32637
      3215427391
      0
      0
      169
      True
      4418
      NaN
      NaN
      6
      NaN
      ...
      0
      1
      0
      0
      0
      0
      0
      0
      0
      0.751674
    
  

428 rows × 247 columns



In [27]:

    
df[df['residual'] < percentile99].sample(n=10, random_state=1)









    Out[27]:






  
    
      
      match_id
      positive_votes
      negative_votes
      first_blood_time
      radiant_win
      duration
      dire_pos1_ancient_kills
      dire_pos1_courier_kills
      dire_pos1_kills
      dire_pos1_neutral_kills
      ...
      dire_pos2_abandons
      dire_pos3_abandons
      dire_pos4_abandons
      dire_pos5_abandons
      radiant_pos1_abandons
      radiant_pos2_abandons
      radiant_pos3_abandons
      radiant_pos4_abandons
      radiant_pos5_abandons
      residual
    
  
  
    
      15328
      3215105241
      0
      0
      112
      False
      2777
      NaN
      NaN
      24
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.174453
    
    
      26872
      3215376151
      0
      0
      114
      True
      2399
      NaN
      NaN
      7
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.108168
    
    
      9558
      3215415736
      0
      0
      169
      False
      3833
      NaN
      NaN
      20
      NaN
      ...
      0
      0
      1
      0
      0
      0
      0
      0
      0
      0.539407
    
    
      38056
      3215109078
      0
      0
      191
      False
      3583
      1.0
      0.0
      15
      51.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.522020
    
    
      28038
      3215171445
      0
      0
      106
      True
      3963
      NaN
      NaN
      14
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.219839
    
    
      19879
      3215172311
      1
      0
      60
      True
      4322
      NaN
      NaN
      4
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.478943
    
    
      23736
      3215186932
      0
      0
      402
      True
      2640
      NaN
      NaN
      10
      NaN
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.108369
    
    
      24764
      3215409668
      0
      0
      179
      False
      2451
      NaN
      NaN
      9
      NaN
      ...
      0
      0
      0
      0
      0
      0
      1
      0
      0
      0.138435
    
    
      3126
      3215384337
      0
      0
      146
      True
      3289
      NaN
      NaN
      8
      NaN
      ...
      0
      0
      0
      0
      0
      0
      1
      0
      0
      0.514544
    
    
      5952
      3215141313
      0
      0
      126
      False
      2572
      3.0
      0.0
      11
      116.0
      ...
      0
      0
      0
      0
      0
      0
      0
      0
      0
      0.341837
    
  

10 rows × 247 columns



In [28]:

    
biggestAnom = np.loadtxt('data/anomalies.csv', delimiter = ',')
theirOutput = np.loadtxt('data/output.csv', delimiter = ",")



In [29]:

    
biggestAnom = np.column_stack([df['match_id'].values, biggestAnom])
theirOutput = np.column_stack([df['match_id'].values, theirOutput])



In [30]:

    
dictToEndAllDicts = {'match_id': [], 'column': [], 'actual': [], 'predicted': []}



In [31]:

    
def whichIndex(row1, row2):
    largestError = -1
    index = -1
    for i in range(1, len(row1)):
        if abs(row1[i] - row2[i]) > largestError:
            largestError = abs(row1[i] - row2[i])
            index = i
    return index
columns = list(df.columns)
columns.remove('match_id')
mask = np.where(df['residual'] >= percentile99)[0]

biggestAnom = biggestAnom[mask, :]
theirOutput = theirOutput[mask, :]

for i in range(len(biggestAnom)):
    index = whichIndex(biggestAnom[i, :], theirOutput[i, :])
    dictToEndAllDicts['match_id'].append(int(biggestAnom[i, 0]))
    dictToEndAllDicts['column'].append(columns[index])
    dictToEndAllDicts['actual'].append(biggestAnom[i, index])
    dictToEndAllDicts['predicted'].append(theirOutput[i, index])



In [32]:

    
tableOfAnomalies = pd.DataFrame.from_dict(dictToEndAllDicts)



In [33]:

    
tableOfAnomalies.to_csv('data/tableOfAnomaliesAndTheirCause.csv', index = False)



In [34]:

    
tableOfAnomalies









    Out[34]:






  
    
      
      actual
      column
      match_id
      predicted
    
  
  
    
      0
      49.000000
      negative_votes
      3215497967
      1.000000
    
    
      1
      15.000000
      negative_votes
      3215097220
      1.000000
    
    
      2
      9.000000
      negative_votes
      3215312675
      1.000000
    
    
      3
      0.997236
      dire_pos5_courier_kills
      3215162338
      0.254108
    
    
      4
      0.967832
      radiant_pos3_gpm
      3215181105
      0.215278
    
    
      5
      0.986188
      dire_pos5_kda
      3215377994
      0.675354
    
    
      6
      7.000000
      negative_votes
      3215271631
      1.000000
    
    
      7
      0.915403
      dire_pos4_gpm
      3215289100
      0.162801
    
    
      8
      0.915403
      dire_pos4_gpm
      3215279863
      0.155206
    
    
      9
      5.000000
      negative_votes
      3215253055
      1.000000
    
    
      10
      0.977942
      radiant_pos1_sentry_uses
      3215448302
      0.730816
    
    
      11
      0.023621
      dire_pos3_gpm
      3215416370
      0.595808
    
    
      12
      0.243854
      radiant_pos3_gpm
      3215261563
      0.536931
    
    
      13
      4.000000
      negative_votes
      3215294948
      1.000000
    
    
      14
      0.996014
      radiant_pos5_kills
      3215155134
      0.729708
    
    
      15
      0.172037
      radiant_pos4_gpm
      3215350255
      0.478153
    
    
      16
      0.765351
      radiant_pos1_kpm
      3215341221
      0.473024
    
    
      17
      0.796921
      radiant_pos2_gpm
      3215306187
      0.594073
    
    
      18
      0.995038
      dire_pos4_gpm
      3215302636
      0.616405
    
    
      19
      0.995249
      dire_pos5_gpm
      3215316820
      0.661066
    
    
      20
      0.991397
      dire_pos3_sentry_uses
      3215190109
      0.725191
    
    
      21
      0.994378
      dire_pos2_gpm
      3215355847
      0.634865
    
    
      22
      0.992697
      radiant_pos3_kills
      3215130854
      0.734208
    
    
      23
      0.995991
      radiant_pos1_kills
      3215259897
      0.731750
    
    
      24
      0.420948
      dire_pos1_totalgold
      3215394671
      0.155802
    
    
      25
      0.853668
      dire_pos4_deaths
      3215152225
      0.677466
    
    
      26
      0.970540
      radiant_pos2_kpm
      3215274711
      0.548053
    
    
      27
      0.417960
      dire_pos3_xpm
      3215366609
      0.140712
    
    
      28
      0.837322
      radiant_pos2_courier_kills
      3215449992
      0.687711
    
    
      29
      0.417725
      radiant_pos2_xpm
      3215376405
      0.142942
    
    
      ...
      ...
      ...
      ...
      ...
    
    
      398
      0.508021
      dire_pos1_totalxp
      3215337531
      0.461330
    
    
      399
      0.987752
      radiant_pos5_hero_heal
      3215102476
      0.764587
    
    
      400
      0.602995
      dire_pos4_lasthits
      3215253869
      0.546018
    
    
      401
      0.068665
      dire_pos5_courier_kills
      3215422166
      0.011190
    
    
      402
      0.602511
      radiant_pos3_deaths
      3215263577
      0.545937
    
    
      403
      0.647322
      radiant_pos2_courier_kills
      3215285461
      0.583701
    
    
      404
      0.977324
      residual
      3215114749
      0.947976
    
    
      405
      0.728203
      dire_pos5_courier_kills
      3215096854
      0.634661
    
    
      406
      0.242480
      radiant_pos3_sentry_uses
      3215446534
      0.059556
    
    
      407
      0.586169
      dire_pos3_observer_uses
      3215119491
      0.541193
    
    
      408
      -0.022676
      residual
      3215205453
      -0.055943
    
    
      409
      0.568367
      radiant_pos3_ancient_kills
      3215135325
      0.522728
    
    
      410
      0.117480
      radiant_pos3_sentry_uses
      3215415463
      0.008837
    
    
      411
      0.668367
      radiant_pos3_tower_kills
      3215322189
      0.594004
    
    
      412
      0.976364
      radiant_pos1_abandons
      3215345510
      0.948507
    
    
      413
      -0.022676
      residual
      3215138506
      -0.061634
    
    
      414
      0.653263
      dire_pos1_abandons
      3215492929
      0.580061
    
    
      415
      0.612837
      radiant_pos5_totalgold
      3215267165
      0.538579
    
    
      416
      0.754499
      dire_pos2_observer_uses
      3215123467
      0.658702
    
    
      417
      0.032951
      dire_pos5_courier_kills
      3215358156
      0.000771
    
    
      418
      0.786506
      radiant_pos5_hero_dmg
      3215094142
      0.659190
    
    
      419
      0.711157
      radiant_pos1_courier_kills
      3215328883
      0.623964
    
    
      420
      0.032951
      dire_pos5_courier_kills
      3215250045
      -0.003149
    
    
      421
      0.336952
      radiant_pos2_kpm
      3215410384
      0.213571
    
    
      422
      0.613161
      dire_pos1_kda
      3215168749
      0.543565
    
    
      423
      0.976364
      radiant_pos1_abandons
      3215343292
      0.948496
    
    
      424
      0.641713
      radiant_pos1_ancient_kills
      3215107300
      0.576132
    
    
      425
      0.117480
      radiant_pos3_sentry_uses
      3215450701
      0.033871
    
    
      426
      0.863760
      dire_pos1_xpm
      3215239658
      0.700287
    
    
      427
      0.523910
      radiant_pos1_totalgold
      3215427391
      0.481955
    
  

428 rows × 4 columns



In [ ]:

	match_id	positive_votes	negative_votes	first_blood_time	radiant_win	duration	dire_pos1_ancient_kills	dire_pos1_courier_kills	dire_pos1_kills	dire_pos1_neutral_kills	...	dire_pos2_abandons	dire_pos3_abandons	dire_pos4_abandons	dire_pos5_abandons	radiant_pos1_abandons	radiant_pos2_abandons	radiant_pos3_abandons	radiant_pos4_abandons	radiant_pos5_abandons	residual
40056	3215497967	49	2	14	True	2748	43.0	0.0	4	141.0	...	0	0	0	0	0	0	0	0	0	60.093830
3222	3215097220	15	7	15	True	2470	23.0	0.0	7	108.0	...	0	0	0	0	0	0	0	0	0	22.980583
42706	3215312675	9	4	0	False	2687	0.0	0.0	10	9.0	...	0	0	0	0	0	0	0	0	0	14.018461
14505	3215162338	0	0	135	False	3981	18.0	21.0	27	70.0	...	0	0	0	0	0	0	0	0	0	12.276542
8310	3215181105	0	0	0	True	0	NaN	NaN	0	NaN	...	0	0	0	0	0	0	0	1	0	10.960655
6186	3215377994	0	0	72	False	2924	NaN	NaN	28	NaN	...	0	0	0	0	0	0	0	0	0	10.190336
10574	3215271631	7	1	3	False	1846	4.0	0.0	5	36.0	...	0	0	0	0	0	0	0	0	0	9.797115
28970	3215289100	0	0	0	False	0	NaN	NaN	0	NaN	...	0	0	1	0	0	0	0	0	0	8.665862
32230	3215279863	0	0	0	False	0	NaN	NaN	0	NaN	...	0	0	1	0	0	0	0	0	0	8.291083
1609	3215253055	5	0	182	True	1949	0.0	0.0	4	10.0	...	0	0	0	0	0	0	0	0	0	6.753566
41758	3215448302	0	0	208	False	7375	4.0	1.0	21	61.0	...	0	0	0	0	0	0	0	0	0	5.982767
10214	3215416370	0	0	0	False	0	NaN	NaN	0	NaN	...	0	0	0	0	0	0	0	1	0	5.009284
42735	3215261563	0	0	0	True	0	NaN	NaN	0	NaN	...	1	0	0	0	0	0	0	0	0	4.787800
9080	3215294948	4	1	94	True	1804	0.0	0.0	4	19.0	...	0	0	0	0	0	0	0	0	0	4.737586
8192	3215155134	0	0	0	True	780	0.0	0.0	0	0.0	...	0	0	0	0	0	0	0	0	0	4.151326
12556	3215350255	0	0	0	True	0	NaN	NaN	0	NaN	...	0	0	1	0	0	0	0	0	0	3.858672
15967	3215341221	0	0	0	False	932	NaN	NaN	34	NaN	...	0	0	0	0	0	0	0	0	0	3.720057
21049	3215306187	0	0	0	True	0	NaN	NaN	0	NaN	...	0	1	0	0	0	0	0	0	0	3.653452
5501	3215302636	0	0	0	False	0	NaN	NaN	0	NaN	...	0	0	0	0	0	0	0	1	0	3.612013
28073	3215316820	0	0	0	False	0	NaN	NaN	0	NaN	...	0	0	0	0	0	0	0	1	0	3.557962
37873	3215190109	0	0	324	False	4877	3.0	0.0	6	24.0	...	0	0	0	0	0	1	0	0	0	3.524932
19538	3215355847	0	0	0	False	0	NaN	NaN	0	NaN	...	0	0	0	0	0	0	0	1	0	3.455053
18809	3215130854	0	0	9	True	2760	0.0	0.0	3	0.0	...	0	0	0	0	0	0	0	0	0	3.430158
37087	3215259897	0	0	0	True	1394	3.0	0.0	1	3.0	...	0	0	0	0	0	0	0	0	0	2.926127
16205	3215394671	0	0	0	True	0	NaN	NaN	0	NaN	...	0	0	1	0	0	0	0	0	0	2.813246
37604	3215152225	0	0	0	True	2641	NaN	NaN	10	NaN	...	0	0	0	0	0	0	0	0	0	2.767648
37101	3215274711	0	0	0	True	858	NaN	NaN	0	NaN	...	0	0	0	0	0	0	0	0	0	2.758488
24635	3215366609	0	0	0	False	0	NaN	NaN	0	NaN	...	0	0	0	0	0	0	0	1	0	2.543581
1345	3215449992	0	0	104	False	5045	60.0	0.0	18	317.0	...	0	0	0	0	0	0	0	0	0	2.515880
26691	3215376405	0	0	0	True	0	NaN	NaN	0	NaN	...	0	0	1	0	0	0	0	0	0	2.483158
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
20347	3215337531	0	0	62	False	3373	NaN	NaN	1	NaN	...	0	0	0	0	0	1	0	1	0	0.769491
15810	3215102476	0	0	137	False	3312	45.0	0.0	8	350.0	...	0	0	0	0	0	0	0	0	0	0.767143
27774	3215253869	0	0	292	False	5234	NaN	NaN	7	NaN	...	0	0	0	0	0	0	0	0	0	0.766957
27894	3215422166	0	0	118	False	2817	47.0	1.0	10	155.0	...	0	0	0	0	0	0	0	0	0	0.765579
12285	3215263577	0	0	0	False	1143	NaN	NaN	35	NaN	...	0	0	0	0	0	0	0	0	0	0.765576
20308	3215285461	0	0	61	False	2682	20.0	0.0	10	74.0	...	0	0	0	0	0	0	0	0	0	0.764146
8590	3215114749	0	0	0	False	746	NaN	NaN	3	NaN	...	0	0	0	0	1	0	1	1	1	0.763206
14222	3215096854	0	0	93	True	2936	21.0	0.0	5	208.0	...	0	0	0	0	0	0	0	0	0	0.762437
8943	3215446534	0	0	218	True	1714	0.0	0.0	1	25.0	...	0	0	0	0	0	0	0	0	0	0.761831
21170	3215119491	0	0	169	True	4051	75.0	0.0	12	354.0	...	0	0	0	0	0	0	0	0	0	0.761589
9693	3215205453	0	0	140	False	2083	NaN	NaN	3	NaN	...	0	1	0	1	0	0	0	0	0	0.759363
10663	3215135325	0	0	0	True	3645	23.0	0.0	4	143.0	...	0	0	0	0	0	0	0	0	1	0.759099
11652	3215415463	0	0	191	False	2846	31.0	0.0	8	175.0	...	0	0	0	0	0	0	0	0	0	0.758527
30828	3215322189	0	0	93	True	2680	0.0	0.0	3	12.0	...	0	1	0	0	0	0	0	0	0	0.758093
15394	3215345510	0	0	307	True	4429	NaN	NaN	10	NaN	...	0	0	0	1	0	0	0	1	0	0.757711
7846	3215138506	0	0	160	False	1780	NaN	NaN	4	NaN	...	1	1	1	0	0	0	0	0	0	0.757519
37444	3215492929	0	0	73	False	3048	20.0	1.0	8	115.0	...	0	1	0	0	0	0	0	0	0	0.757248
19730	3215267165	0	0	2	True	2125	NaN	NaN	6	NaN	...	0	0	0	0	1	1	1	0	0	0.756915
32004	3215123467	0	0	224	False	3050	0.0	0.0	2	12.0	...	0	0	0	0	0	0	0	0	0	0.756316
288	3215358156	0	0	163	True	4185	24.0	0.0	7	55.0	...	0	0	0	0	0	0	0	0	0	0.754428
36615	3215094142	0	0	203	True	4973	NaN	NaN	4	NaN	...	0	0	0	0	0	0	0	0	0	0.754374
11493	3215328883	0	0	3	True	3377	37.0	0.0	12	179.0	...	0	0	0	0	0	0	0	0	0	0.753980
39744	3215250045	0	0	184	True	4439	44.0	0.0	31	132.0	...	0	0	0	0	0	0	0	0	0	0.753625
9634	3215410384	0	0	0	True	344	NaN	NaN	0	NaN	...	0	0	0	0	0	0	0	0	0	0.753601
17709	3215168749	0	0	0	True	413	0.0	0.0	0	0.0	...	0	0	1	0	0	0	0	0	0	0.753267
27443	3215343292	0	0	86	True	3411	NaN	NaN	7	NaN	...	0	0	0	1	0	0	0	1	0	0.753174
39465	3215107300	0	0	108	True	2965	8.0	0.0	9	88.0	...	0	0	0	0	0	0	0	0	0	0.752881
29673	3215450701	0	0	90	False	2907	45.0	1.0	8	197.0	...	0	0	0	0	0	0	0	0	0	0.752879
22278	3215239658	0	0	126	False	3992	NaN	NaN	10	NaN	...	0	0	1	0	0	0	0	0	0	0.752695
32637	3215427391	0	0	169	True	4418	NaN	NaN	6	NaN	...	0	1	0	0	0	0	0	0	0	0.751674

	match_id	positive_votes	first_blood_time	radiant_win	duration	dire_pos1_ancient_kills	dire_pos1_courier_kills	dire_pos1_kills	dire_pos1_neutral_kills	...	dire_pos4_abandons	radiant_pos3_abandons	residual
15328	3215105241	0	112	False	2777	NaN	NaN	24	NaN	...	0	0	0.174453
26872	3215376151	0	114	True	2399	NaN	NaN	7	NaN	...	0	0	0.108168
9558	3215415736	0	169	False	3833	NaN	NaN	20	NaN	...	1	0	0.539407
38056	3215109078	0	191	False	3583	1.0	0.0	15	51.0	...	0	0	0.522020
28038	3215171445	0	106	True	3963	NaN	NaN	14	NaN	...	0	0	0.219839
19879	3215172311	1	60	True	4322	NaN	NaN	4	NaN	...	0	0	0.478943
23736	3215186932	0	402	True	2640	NaN	NaN	10	NaN	...	0	0	0.108369
24764	3215409668	0	179	False	2451	NaN	NaN	9	NaN	...	0	1	0.138435
3126	3215384337	0	146	True	3289	NaN	NaN	8	NaN	...	0	1	0.514544
5952	3215141313	0	126	False	2572	3.0	0.0	11	116.0	...	0	0	0.341837

	actual	column	match_id	predicted
0	49.000000	negative_votes	3215497967	1.000000
1	15.000000	negative_votes	3215097220	1.000000
2	9.000000	negative_votes	3215312675	1.000000
3	0.997236	dire_pos5_courier_kills	3215162338	0.254108
4	0.967832	radiant_pos3_gpm	3215181105	0.215278
5	0.986188	dire_pos5_kda	3215377994	0.675354
6	7.000000	negative_votes	3215271631	1.000000
7	0.915403	dire_pos4_gpm	3215289100	0.162801
8	0.915403	dire_pos4_gpm	3215279863	0.155206
9	5.000000	negative_votes	3215253055	1.000000
10	0.977942	radiant_pos1_sentry_uses	3215448302	0.730816
11	0.023621	dire_pos3_gpm	3215416370	0.595808
12	0.243854	radiant_pos3_gpm	3215261563	0.536931
13	4.000000	negative_votes	3215294948	1.000000
14	0.996014	radiant_pos5_kills	3215155134	0.729708
15	0.172037	radiant_pos4_gpm	3215350255	0.478153
16	0.765351	radiant_pos1_kpm	3215341221	0.473024
17	0.796921	radiant_pos2_gpm	3215306187	0.594073
18	0.995038	dire_pos4_gpm	3215302636	0.616405
19	0.995249	dire_pos5_gpm	3215316820	0.661066
20	0.991397	dire_pos3_sentry_uses	3215190109	0.725191
21	0.994378	dire_pos2_gpm	3215355847	0.634865
22	0.992697	radiant_pos3_kills	3215130854	0.734208
23	0.995991	radiant_pos1_kills	3215259897	0.731750
24	0.420948	dire_pos1_totalgold	3215394671	0.155802
25	0.853668	dire_pos4_deaths	3215152225	0.677466
26	0.970540	radiant_pos2_kpm	3215274711	0.548053
27	0.417960	dire_pos3_xpm	3215366609	0.140712
28	0.837322	radiant_pos2_courier_kills	3215449992	0.687711
29	0.417725	radiant_pos2_xpm	3215376405	0.142942
...	...	...	...	...
398	0.508021	dire_pos1_totalxp	3215337531	0.461330
399	0.987752	radiant_pos5_hero_heal	3215102476	0.764587
400	0.602995	dire_pos4_lasthits	3215253869	0.546018
401	0.068665	dire_pos5_courier_kills	3215422166	0.011190
402	0.602511	radiant_pos3_deaths	3215263577	0.545937
403	0.647322	radiant_pos2_courier_kills	3215285461	0.583701
404	0.977324	residual	3215114749	0.947976
405	0.728203	dire_pos5_courier_kills	3215096854	0.634661
406	0.242480	radiant_pos3_sentry_uses	3215446534	0.059556
407	0.586169	dire_pos3_observer_uses	3215119491	0.541193
408	-0.022676	residual	3215205453	-0.055943
409	0.568367	radiant_pos3_ancient_kills	3215135325	0.522728
410	0.117480	radiant_pos3_sentry_uses	3215415463	0.008837
411	0.668367	radiant_pos3_tower_kills	3215322189	0.594004
412	0.976364	radiant_pos1_abandons	3215345510	0.948507
413	-0.022676	residual	3215138506	-0.061634
414	0.653263	dire_pos1_abandons	3215492929	0.580061
415	0.612837	radiant_pos5_totalgold	3215267165	0.538579
416	0.754499	dire_pos2_observer_uses	3215123467	0.658702
417	0.032951	dire_pos5_courier_kills	3215358156	0.000771
418	0.786506	radiant_pos5_hero_dmg	3215094142	0.659190
419	0.711157	radiant_pos1_courier_kills	3215328883	0.623964
420	0.032951	dire_pos5_courier_kills	3215250045	-0.003149
421	0.336952	radiant_pos2_kpm	3215410384	0.213571
422	0.613161	dire_pos1_kda	3215168749	0.543565
423	0.976364	radiant_pos1_abandons	3215343292	0.948496
424	0.641713	radiant_pos1_ancient_kills	3215107300	0.576132
425	0.117480	radiant_pos3_sentry_uses	3215450701	0.033871
426	0.863760	dire_pos1_xpm	3215239658	0.700287
427	0.523910	radiant_pos1_totalgold	3215427391	0.481955