如何用大pandas的groupby输出来填充？

df['D'].fillna(df.groupby(['A','B','C'])['D'].transform('mean'))

apply

In [2400]: dfOut[2400]:   A  B  C    D0  1  1  1  1.01  1  1  1  NaN2  1  1  1  3.03  3  3  3  5.0In [2401]: df['D'].fillna(df.groupby(['A','B','C'])['D'].transform('mean'))Out[2401]:0    1.01    2.02    3.03    5.0Name: D, dtype: float64In [2402]: df['D'] = df['D'].fillna(df.groupby(['A','B','C'])['D'].transform('mean'))In [2403]: dfOut[2403]:   A  B  C    D0  1  1  1  1.01  1  1  1  2.02  1  1  1  3.03  3  3  3  5.0

细节

In [2396]: df.shapeOut[2396]: (10000, 4)In [2398]: %timeit df['D'].fillna(df.groupby(['A','B','C'])['D'].transform('mean'))100 loops, best of 3: 3.44 ms per loopIn [2397]: %timeit df.groupby(['A','B','C'])['D'].apply(lambda x: x.fillna(x.mean()))100 loops, best of 3: 5.34 ms per loop