python - pandas 最终排名状态

Question

我正在研究国内足球比赛的排名， 我有以下数据框。

df = pd.DataFrame()
df ['Season'] = ['1314','1314','1314','1314','1314','1314','1314','1314','1314','1415','1415','1415','1415','1415','1415','1415','1415','1415']
df ['Team'] = ['A','B','C','A','B','C','A','B','C','A','B','C','A','B','C','A','B','C']
df ['GW'] = [1,1,1,2,2,2,3,3,3,1,1,1,2,2,2,3,3,3]
df['Position'] = [1,2,3,3,1,2,2,3,1,2,1,3,2,1,3,3,2,1]
df = df.sort_values (['Season','Team'])
df['Position_Change']=df.groupby(['Season','Team'])['Position'].apply(lambda x : x.diff().fillna(0))

以上代码可以跟踪排名和位置的变化。 现在，我想为最先成为冠军的团队分配一个状态。这意味着在所有 GW 中，地位冠军将被分配给该团队。其他团队作为他们在比赛上周结束的位置(在此示例中，最后 GW 为 3) 我的预期输出如下:

这是原始数据集: Click to download the dataset

我们将非常感谢您的建议。 谢谢，

泽普

Answer 1

IIUC 您需要如下所示的内容:

df1 = df.groupby(['Season','Team'])['Position'].apply(lambda x : np.select([(x.iloc[-1]==1),(x.iloc[-1]==2),(x.iloc[-1]==3)],['Champion','Second','Third'])).reset_index().rename(columns={'Position':'Status'})
print(df.merge(df1,on=['Team','Season']))

    Season Team  GW  Position  Position_Change    Status
0    1314    A   1         1              0.0    Second
1    1314    A   2         3              2.0    Second
2    1314    A   3         2             -1.0    Second
3    1314    B   1         2              0.0     Third
4    1314    B   2         1             -1.0     Third
5    1314    B   3         3              2.0     Third
6    1314    C   1         3              0.0  Champion
7    1314    C   2         2             -1.0  Champion
8    1314    C   3         1             -1.0  Champion
9    1415    A   1         2              0.0     Third
10   1415    A   2         2              0.0     Third
11   1415    A   3         3              1.0     Third
12   1415    B   1         1              0.0    Second
13   1415    B   2         1              0.0    Second
14   1415    B   3         2              1.0    Second
15   1415    C   1         3              0.0  Champion
16   1415    C   2         3              0.0  Champion
17   1415    C   3         1             -2.0  Champion

基于Chat，将原代码中df1的代码替换为:

df1 = df.groupby(['Season','Team'])['Position'].apply(lambda x : np.select([(x.iloc[-1]==1),(2<=x.iloc[-1]<=4),(5<=x.iloc[-1]<=6),(7<=x.iloc[-1]<=17),(x.iloc[-1] > 17)],['Champion','UCL','UEL','Other','Relegation'])).reset_index().rename(columns={'Position':'Status'})