python pandas 用最大值填充 NaN 或毯子
我有一个大数据框的问题。这是一个小片段。我想用最大值填充最后一列 E，如果有一些值或让它为空。那就是数据：
d = {'A': [4000074, 4000074, 4000074, 4000074, 4000074, 4000074, 4000074, 4000074, 4000074,
4000074, 4000074, 4000074, 4000074, 4000074, 4000074, 4000074, 4000074, 4000074],
'B': ['SP000796746', 'SP000796746', 'SP000796746', 'SP000796746', 'SP000796746','SP000796746',
'SP000796746', 'SP000796746', 'SP000796746', 'SP000796746', 'SP000796746', 'SP000796746',
'SP000796746', 'SP000796746', 'SP000796746', 'SP000796746', 'SP000796746', 'SP000796746'],
'C': [201926, 201926, 201926, 201926, 201926, 201926, 201909,201909, 201909, 201909, 201909,
201909, 201933, 201933, 201933, 201933, 201933, 201933],
'D': [-1, 0, 1, 2, 3, 4, -1, 0, 1, 2, 3, 4, -1, 0, 1, 2, 3, 4],
'E': [np.nan, 1000, 1000, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, 3000, 3000, np.nan]}
它看起来像这样：
A B C D E
0 4000074 SP000796746 201926 -1 NaN
1 4000074 SP000796746 201926 0 1000.0
2 4000074 SP000796746 201926 1 1000.0
3 4000074 SP000796746 201926 2 NaN
4 4000074 SP000796746 201926 3 NaN
5 4000074 SP000796746 201926 4 NaN
6 4000074 SP000796746 201909 -1 NaN
7 4000074 SP000796746 201909 0 NaN
8 4000074 SP000796746 201909 1 NaN
9 4000074 SP000796746 201909 2 NaN
10 4000074 SP000796746 201909 3 NaN
11 4000074 SP000796746 201909 4 NaN
12 4000074 SP000796746 201933 -1 NaN
13 4000074 SP000796746 201933 0 NaN
14 4000074 SP000796746 201933 1 NaN
15 4000074 SP000796746 201933 2 3000.0
16 4000074 SP000796746 201933 3 3000.0
17 4000074 SP000796746 201933 4 NaN
慕无忌16237181回答
-
料青山看我应如是
您可以使用D 列中的新 -1 和 中的组groupby.transform来完成。然后是原来的专栏。maxcumsumfillnadf['E'] = df['E'].fillna(df['E'].groupby(df['D'].eq(-1).cumsum()).transform('max'))编辑：用零填充，你可以这样做：mask = df['E'].groupby(df['D'].eq(-1).cumsum()).transform('any')df.loc[mask, 'E'] = df.loc[mask, 'E'].fillna(0)
随时随地看视频慕课网APP
相关分类
Python