对每一行进行排名以创建蒙版数据框

Question

样本数据帧df定义如下：

import numpy as np
import pandas as pd
df = pd.DataFrame(10*(2+np.random.randn(500, 8)), columns=list('ABCDEFGH'))

在每一行中，对前5列进行排名，并将其标记为1 ，其余标记为nan 。 df看起来像

df.head()
           A          B          C          D          E          F          G          H
0   6.598436  44.318800  18.064752  13.418329  17.145434   6.696975  14.757765   8.797826
1   3.593140  14.571717  16.292330  28.390669  35.289606  -4.273124  20.519388  25.137833
2  36.777253  34.360523  28.020462  15.356690  22.038938  14.960303  15.225555  34.691981
3  18.623122  27.184421  -5.320215  31.694895  21.156375   9.947077  20.257575  21.035659
4  11.864725  30.458160  13.509029  27.037195  20.581043  25.371691   1.094735  28.703618

所需的输出是：

df_output.head()
      A    B    C    D    E     F    G    H
0   nan    1    1    1    1   nan    1  nan
1   nan  nan    1    1    1   nan    1    1
2     1    1    1  nan    1   nan  nan    1
3   nan    1  nan    1    1   nan    1    1
4   nan    1  nan    1    1     1  nan    1

Answer 1

df_output = df.rank(1, ascending=False, method='first')
df_output[df_output > 5] = np.nan
df_output[df_output <= 5] = 1.0