唐宇迪《python数据分析与机器学习实战》学习笔记 32探索性数据分析-足球赛事数据集
原始数据：链接，提取码：yypl
一、数据介绍、导入及查看
数据包含球员和裁判的信息，-的比赛数据，总共设计球员2053名，裁判3147名，特征列表如下：
1.1 数据及模块导入
#from __future__ import absolute_import, division, print_functionimport matplotlib as mplfrom matplotlib import pyplot as pltfrom matplotlib.pyplot import GridSpecimport seaborn as snsimport numpy as npimport pandas as pdimport os, sysfrom tqdm import tqdmimport warningswarnings.filterwarnings('ignore')sns.set_context("poster", font_scale=1.3)#import missingno as msno#import pandas_profilingfrom sklearn.datasets import make_blobsimport timedf = pd.read_csv("redcard.csv.gz", compression='gzip')print(df.shape)df.head()
(146028, 28)
1.2 简单的统计：(count统计非空值个数)
df.describe().T
1.3查看数据类型：
(机器学习建模时只认识‘float’和‘int’型，其他类型需要映射转换一下，这里做探索分析就不用了)
df.dtypes
playerShort object
player object
club object
leagueCountry object
birthday object
height float64
weight float64
position object
games int64
victories int64
ties int64
defeats int64
goals int64
yellowCards int64
yellowReds int64
redCards int64
photoID object
rater1 float64
rater2 float64
refNum int64
refCountry int64
Alpha_3 object
meanIAT float64
nIAT float64
seIAT float64
meanExp float64
nExp float64
seExp float64
dtype: object
1.4 查看并提取列名
all_columns = df.columns.tolist()all_columns
[‘playerShort’,
‘player’,
‘club’,
‘leagueCountry’,
‘birthday’,
‘height’,
‘weight’,
‘position’,
‘games’,
‘victories’,
‘ties’,
‘defeats’,
‘goals’,
‘yellowCards’,
‘yellowReds’,
‘redCards’,
‘photoID’,
‘rater1’,
‘rater2’,
‘refNum’,
‘refCountry’,
‘Alpha_3’,
‘meanIAT’,
‘nIAT’,
‘seIAT’,
‘meanExp’,
‘nExp’,
‘seExp’]
思考问题，加入一个运动员出现多次，计算时相当于其权重加强，所以可以用groupby解决这个问题：
print(df['height'].mean())print(np.mean(df.groupby('playerShort').height.mean()))
181.93593798236887
181.74372848007872
二、数据切分模块(Tidy Data)
数据通常具有多特征高纬度，分析时统计指标不同，因此可以将其分为几个小的数据集单项分析。例如：单看球员、裁判，看球员-裁判关系，单看国家…
2.1切分球员数据
2.1.1数据切分
#切分出球员数据，及其相关特征数据player_index = 'playerShort' #球员IDplayer_cols = [#'player', #球员名字丢弃，因为有对应ID了'birthday','height','weight','position','photoID','rater1','rater2',]
2.1.2检测及去重
#数据检测，避免复制错误、重名等影响all_cols_unique_players = df.groupby('playerShort').agg({col:'nunique' for col in player_cols})all_cols_unique_players.head() #为1就是出现一次，为2就是重复了，这里数据干净
all_cols_unique_players[all_cols_unique_players > 1].dropna().head()#去重
这里直接写了一个检测去重函数，主要是看key值重复没
def get_subgroup(dataframe, g_index, g_columns):"""Helper function that creates a sub-table from the columns and runs a quick uniqueness test."""g = dataframe.groupby(g_index).agg({col:'nunique' for col in g_columns})if g[g > 1].dropna().shape[0] != 0:print("Warning: you probably assumed this had all unique values but it doesn't.")return dataframe.groupby(g_index).agg({col:'max' for col in g_columns})
函数调用
players = get_subgroup(df, player_index, player_cols)players.head()
数据干净后就储存，这里增加储存函数：
def save_subgroup(dataframe, g_index, subgroup_name, prefix='raw_'):save_subgroup_filename = "".join([prefix, subgroup_name, ".csv.gz"])dataframe.to_csv(save_subgroup_filename, compression='gzip', encoding='UTF-8')test_df = pd.read_csv(save_subgroup_filename, compression='gzip', index_col=g_index, encoding='UTF-8')# Test that we recover what we send inif dataframe.equals(test_df):print("Test-passed: we recover the equivalent subgroup dataframe.")else:print("Warning -- equivalence test!!! Double-check.")
save_subgroup(players, player_index, "players")
Test-passed: we recover the equivalent subgroup dataframe. 储存成功
根据上面一套操作思路还可以切割其他数据：
2.2 切分俱乐部-国家关系
club_index = 'club'club_cols = ['leagueCountry']clubs = get_subgroup(df, club_index, club_cols)#检测去重print(clubs.head())print('-------------')print(clubs['leagueCountry'].value_counts())#查看所属国家save_subgroup(clubs, club_index, "clubs")#保存数据
club leagueCountry
FC Nürnberg GermanyFSV Mainz 05 Germany
1899 Hoffenheim Germany
AC Ajaccio France
AFC Bournemouth England
England 48
Spain 27
France 22
Germany 21
Name: leagueCountry, dtype: int64
Test-passed: we recover the equivalent subgroup dataframe.
2.3 切分裁判-国家关系
referee_index = 'refNum'referee_cols = ['refCountry']referees = get_subgroup(df, referee_index, referee_cols)#检测去重referees.head()print(referees.refCountry.nunique())#查看裁判国家数print(referees.shape)#查看形状save_subgroup(referees, referee_index, "referees")#保存数据
refNum refCountry
1 1
2 2
3 3
4 4
5 5
161
(3147, 1)
Test-passed: we recover the equivalent subgroup dataframe.
2.4切分国家关系
country_index = 'refCountry'country_cols = ['Alpha_3', 'meanIAT', 'nIAT', 'seIAT', 'meanExp', 'nExp', 'seExp']countries = get_subgroup(df, country_index, country_cols)countries = countries.rename(columns = {'Alpha_3' : 'countryName'})print(countries.shape)save_subgroup(countries, country_index, 'countries')
(161, 7)
2.5切分裁判-球员关系
dyad_index = ['refNum', 'playerShort']dyad_cols = ['games','victories','ties','defeats','goals','yellowCards','yellowReds','redCards',]dyads = get_subgroup(df, g_index=dyad_index, g_columns=dyad_cols)dyads.head()
print(dyads.shape)dyads[dyads.redCards > 1].head(10)#查看超过1次红牌的球员
(146028, 8)
save_subgroup(dyads, dyad_index, "dyads")#保存
Test-passed: we recover the equivalent subgroup dataframe.
三、缺失值可视化分析(球员数据为例)
3.1数据读取
players = pd.read_csv("raw_players.csv.gz", compression='gzip',index_col=0)players.shapeplayers.head()
这里用missingno直观的展示出缺失值位置：
详细用法点击链接
安装，cmd中：
pip install missingno
3.2缺失值分布图（ missingno.matrix）
import missingno as msnomsno.matrix(players, figsize = (16, 6))
3.3缺失值直方图（ missingno.bar）
msno.bar(players.sample(500))#取样500个
3.4缺失值热度图（ missingno.heatmap）
直观表达缺失值间的相关性
msno.heatmap(players.sample(500),figsize=(16, 7),)
photoID与rate1、rate2相关性极强。
3.5缺失值数量查询
print("All players:", len(players))print("rater1 nulls:", len(players[(players.rater1.isnull())]))print("rater2 nulls:", len(players[players.rater2.isnull()]))print("Both nulls:", len(players[(players.rater1.isnull()) & (players.rater2.isnull())]))
All players: 2053
rater1 nulls: 468
rater2 nulls: 468
Both nulls: 468
3.6缺失值处理
缺失值多的时候直接移除
其他方式查询，统计不为缺失的值，再减去
players = players[players.rater1.notnull()]print(players.shape[0])print(len(players)-players.shape[0])
1585
468
msno.matrix(players.sample(500),#上面只留不缺失的值figsize=(16, 7),width_ratios=(15, 1))#这里rate1没有缺失的了
四、特征可视化及关系分析
4.1评分展示
4.1 .1评分相关性展示
这里用crosstab将rate1和rate2按类别分组，统计各个分组中值的频数：
pd.crosstab(players.rater1, players.rater2)
为了更加直观，这里用热度图展示：
fig, ax = plt.subplots(figsize=(12, 8)) #sns画图命令更简洁sns.heatmap(pd.crosstab(players.rater1, players.rater2), cmap='Blues', annot=True, fmt='d', ax=ax) ax.set_title("Correlation between Rater 1 and Rater 2\n")fig.tight_layout()
4.1.2 评分取均值
到底用rate1还是rate2？不如干脆取均值
players['skintone'] = players[['rater1', 'rater2']].mean(axis=1)players.head()
评分均值分布情况：
sns.set_style("dark")#风格设置sns.distplot(players.skintone, kde=False);
4.2位置展示(positions)
MIDSIZE = (12, 8)fig, ax = plt.subplots(figsize=MIDSIZE)#位置值统计players.position.value_counts(dropna=False, ascending=True).plot(kind='barh', ax=ax)ax.set_ylabel("Position")ax.set_xlabel("Counts")fig.tight_layout()
这里想探索位置与被罚的关系，但是如图位置太多了，有左前锋、右前锋，都是前锋，可以合并
提取位置
position_types = players.position.unique()position_types
array([‘Center Back’, ‘Attacking Midfielder’, ‘Right Midfielder’, ‘Center Midfielder’, ‘Goalkeeper’, ‘Defensive Midfielder’,‘Left Fullback’, nan, ‘Left Midfielder’, ‘Right Fullback’,‘Center Forward’, ‘Left Winger’, ‘Right Winger’], dtype=object)
合并为4类
defense = ['Center Back','Defensive Midfielder', 'Left Fullback', 'Right Fullback', ]midfield = ['Right Midfielder', 'Center Midfielder', 'Left Midfielder',]forward = ['Attacking Midfielder', 'Left Winger', 'Right Winger', 'Center Forward']keeper = 'Goalkeeper'#df.isin(['b','c']) 判断后者是否在前者中players.loc[players['position'].isin(defense), 'position_agg'] = "Defense"players.loc[players['position'].isin(midfield), 'position_agg'] = "Midfield"players.loc[players['position'].isin(forward), 'position_agg'] = "Forward"players.loc[players['position'].eq(keeper), 'position_agg'] = "Keeper"
展示合并结果
MIDSIZE = (12, 8)fig, ax = plt.subplots(figsize=MIDSIZE)players['position_agg'].value_counts(dropna=False, ascending=True).plot(kind='barh', ax=ax)ax.set_ylabel("position_agg")ax.set_xlabel("Counts")fig.tight_layout()
4.3 两两特征展示及等级划分
4.3.1身高、体重、肤色间的关系
from pandas.plotting import scatter_matrixfig, ax = plt.subplots(figsize=(10, 10))scatter_matrix(players[['height', 'weight', 'skintone']], alpha=0.2, diagonal='hist', ax=ax);
如图展示，可以观察发现(体重、身高)和肤色间基本无相关性。身高体重间有正相关性。
4.3.2身高-体重关系
fig, ax = plt.subplots(figsize=MIDSIZE) sns.regplot('weight', 'height', data=players, ax=ax,scatter_kws={"color":"darkred",'s':15,'alpha':1}) #颜色、大小、透明度设置ax.set_ylabel("Height [cm]")ax.set_xlabel("Weight [kg]")fig.tight_layout()
4.3.3身高、体重等级划分
都是离散值，具体连续值在这里分析时无太大用，但是可以按照等级划分。
#身高划分height_categories = ["vlow_height","low_height","mid_height","high_height","vhigh_height",]players['heightclass'] = pd.qcut(players['height'],len(height_categories),height_categories)#pd.qcut（切分数据，份数，类别名称）#体重划分weight_categories = ["vlow_weight","low_weight","mid_weight","high_weight","vhigh_weight",]players['weightclass'] = pd.qcut(players['weight'],len(weight_categories),weight_categories)players.head()
4.3.4 肤色等级划分
print (players['skintone'])pd.qcut(players['skintone'], 3)players['skintoneclass'] = pd.qcut(players['skintone'], 3)#切为3份players
0 0.125
1 0.125
2 0.250
3 0.000
4 0.250
…
2047 0.250
2048 0.000
2050 0.875
2051 0.000
2052 0.000
Name: skintone, Length: 1585, dtype: float64
五、报表可视化分析（强大的pandas_profiling）
安装：
pip install pandas_profiling
Github示例：/pandas-profiling/pandas-profiling
生成分析表：
import pandas_profilingpandas_profiling.ProfileReport(players)
通过pandas_profiling生成一大堆可视化分析，这里截图部分展示：
数据概况：
特征缺陷警告
各个特征的详细展示（这里只展示部分）
特征间的相互关联
统计
缺失值展示
六、特征创造、选择及组合
6.1年龄计算
面展示的球员生日并无大用，但是可以通过其知道年龄。
players.birthday.head()
playerShort
aaron-hughes 08.11.1979
aaron-hunt 04.09.1986
aaron-lennon 16.04.1987
aaron-ramsey 26.12.1990
abdelhamid-el-kaoutari 17.03.1990
Name: birthday, dtype: object
players['birth_date'] = pd.to_datetime(players.birthday, format='%d.%m.%Y')players['age_years'] = ((pd.to_datetime("-01-01") - players['birth_date']).dt.days)/365.25 #闰年多的一天均分players['age_years']
playerShort
aaron-hughes 33.149897
aaron-hunt 26.327173
aaron-lennon 25.713895
aaron-ramsey 22.017796
abdelhamid-el-kaoutari 22.795346
…
zoltan-gera 33.697467
zoltan-stieber 24.210815
zoumana-camara 33.749487
zubikarai 28.848734
zurutuza 26.455852
Name: age_years, Length: 1585, dtype: float64
players.head()
6.2选择有用特征并展示
players_cleaned_variables = [#'birthday','height','weight',#'position',#'photoID',#'rater1',#'rater2','skintone','position_agg','weightclass','heightclass','skintoneclass',#'birth_date','age_years']pandas_profiling.ProfileReport(players[players_cleaned_variables])
结果内容和五类似，这里选择指定特征展示，就不一一放图了
保存
players[players_cleaned_variables].to_csv("cleaned_players.csv.gz", compression='gzip')
6.4数据组合
clean_players= pd.read_csv("cleaned_players.csv.gz", compression='gzip')players = pd.read_csv("raw_players.csv.gz", compression='gzip')referees = pd.read_csv("raw_referees.csv.gz", compression='gzip')#通过球员新的子表取过滤其他子表agg_dyads = pd.read_csv("raw_dyads.csv.gz", compression='gzip', index_col=[0, 1])print(agg_dyads.shape)agg_dyads.head(10)
(146028, 8)
判断逻辑
踢球总场次 = 胜场+平局+败局
all(agg_dyads['games'] == agg_dyads.victories + agg_dyads.ties + agg_dyads.defeats)
True
数据增加
agg_dyads['totalRedCards'] = agg_dyads['yellowReds'] + agg_dyads['redCards']agg_dyads.rename(columns={'redCards': 'strictRedCards'}, inplace=True)agg_dyads.head()
clean_players.head()
agg_dyads.head()
agg_dyads.reset_index().head()
agg_dyads.reset_index().set_index('playerShort').head()
数据组合（agg_dyads+clean_players）
player_dyad = (clean_players.merge(agg_dyads.reset_index().set_index('playerShort'),left_index = True,right_index = True))player_dyad.head()
过滤通过clean_players过滤agg_dyads得到clean_dyads
clean_dyads = (agg_dyads.reset_index()[agg_dyads.reset_index().playerShort.isin(set(clean_players.index))]).set_index(['refNum', 'playerShort'])clean_dyads.head()
clean_dyads.shape,agg_dyads.shape,player_dyad.shape
((124621, 9), (146028, 9), (124621, 18))
七、红牌与肤色的关系
提取出要用的特征
j = 0out = [0 for _ in range(sum(clean_dyads['games']))]for index, row in clean_dyads.reset_index().iterrows():n = row['games']d = row['totalRedCards']ref = row['refNum']player = row['playerShort']for _ in range(n):row['totalRedCards'] = 1 if (d - _) > 0 else 0rowlist = list([ref, player, row['totalRedCards']])out[j] = rowlistj += 1tidy_dyads = pd.DataFrame(out, columns=['refNum','playerShort','redcard']).set_index(['refNum', 'playerShort'])print(tidy_dyads.redcard.sum())print(clean_dyads.games.sum())print(tidy_dyads.shape)tidy_dyads.head()
3092
373067
(373067, 1)
tidy_dyads.redcard.sum()# 3092clean_referees = referees.reset_index()[referees.reset_index().refNum.isin(tidy_dyads.reset_index().refNum.unique())].set_index('refNum')clean_referees.shape, referees.shape# ((2978, 1), (3147, 1))clean_countries = countries.reset_index()[countries.reset_index().refCountry.isin(clean_referees.refCountry.unique())].set_index('refCountry')clean_countries.shape, countries.shape# ((160, 7), (161, 7))tidy_dyads.to_csv('cleaned_dyads.csv.gz', compression = 'gzip')tidy_dyads.head()
裁判给红牌的分布
(tidy_dyads.groupby(level=0).sum().sort_values('redcard', ascending=False).rename(columns={'redcard':'total redcards given'})).head()
运动员收获的红牌分布
tidy_dyads.groupby(level=1).sum().sort_values(ascending=False, by='redcard').head()
红牌与肤色的关系
player_ref_game = (tidy_dyads.reset_index().set_index('playerShort').merge(clean_players,left_index=True,right_index=True))player_ref_game.head()
bootstrap = pd.concat([player_ref_game.sample(replace=True, n=10000).groupby('skintone').mean() for _ in range(100)])player_ref_game.sample(replace=True,n=10000).groupby('skintone').mean()
fig, ax = plt.subplots(figsize=(12,8)) ax = sns.regplot(bootstrap.index.values,y='redcard',data=bootstrap,lowess=True,scatter_kws={"color":"darkred",'s':15,'alpha':1},x_jitter=(0.125 / 4.0))ax.set_xlabel("Skintone");
