如何创建显示三个选项之一的百分比的熊猫数据透视表？

Question

我具有以下结构的体育数据：

season, country, league, hometeam, awayteam, status

状态可以是H，D或A

我想为每个赛季，国家和联赛创建一个数据透视表，将H表示为总比赛数的百分比。

import pandas as pd

grouped_success_rate = df_data.groupby(["Season", "Country", "League"])

homewins_per_league = grouped_success_rate.apply(lambda x: x[(x["STATUS"] == "H")].shape[0]).unstack("Season")
homewins_per_league.fillna(0, inplace=True)
homewins_per_league["TOTAL"] = homewins_per_league.apply(lambda x: sum(x), axis=1)

total_matches_per_league = grouped_success_rate.apply(lambda x: len(x["STATUS"])).unstack("Season")
total_matches_per_league.fillna(0, inplace=True)
total_matches_per_league["TOTAL"] = total_matches_per_league.apply(lambda x: sum(x), axis=1)

homewins_rate_per_league = (homewins_per_league / total_matches_per_league).applymap(lambda x: round(x, 3))

如您所见，我将数据框分组，但是随后我必须为特定选项和总计创建一个单独的数据框。 有没有一种方法可以创建两个数据框？

Answer 1

我认为最简单的方法是先向DataFrame添加一列：

df_data['is_H'] = df_data['Status'] == 'H'
df_data.groupby(["Season", "Country", "League"])['is_H'].mean()