熊猫，如何使用多个分组列替换python数据框中的平均值

Question

Replace a data frame values with mean using multiple grouped columns. The below snapshot is the dataframe:

    Current Loan Amount DateTime    Day Month   Year
0   611314  1-Jan-92    1   Jan 92
1   266662  2-Jan-92    2   Jan 92
2   153494  3-Jan-92    3   Jan 92
3   176242  4-Jan-92    4   Jan 92
4   321992  5-Jan-92    5   Jan 92
5   202928  6-Jan-92    6   Jan 92
6   621786  7-Jan-92    7   Jan 92
7   266794  8-Jan-92    8   Jan 92
8   202466  9-Jan-92    9   Jan 92
9   266288  10-Jan-92   10  Jan 92
10  121110  11-Jan-92   11  Jan 92
11  258104  12-Jan-92   12  Jan 92
12  161722  13-Jan-92   13  Jan 92
13  753016  14-Jan-92   14  Jan 92
14  444664  15-Jan-92   15  Jan 92
15  172282  16-Jan-92   16  Jan 92
16  275440  17-Jan-92   17  Jan 92
17  218834  18-Jan-92   18  Jan 92
18  0   19-Jan-92   19  Jan 92
19  0   20-Jan-92   20  Jan 92

I need to replace the 0.0 values which with mean of the Current Loan Amount for the year and within the same month.

I used different methods, and the below does give me the mean, but it doesnot change the dataframe and removes the rest of the columns

data = data_loan.groupby(['Year','Month'])
def replace(group):
    mask = (group==0)
    group[mask] = group[~mask].mean()
    return group
new_data = data.transform(replace)

Answer 1

import numpy as np    
data_loan['current'] = data_loan['current'].replace(0, np.nan)    
data_loan["current"] = data_loan.groupby(['Month','Year'])["current"].transform(lambda x: x.fillna(x.mean()))

This will replace 0 with mean of the group.