Python数据框使用其他列中的信息填充NaN值

Question

I tried to solve this problem on my own, but I unfortunately haven't made much progress and would really appreciate anyone who can help me out.

My current dataframe contains 3 columns: 2 healthy columns and one column with some missing values, denoted as NaN.

df
Out[18]: 
  x1  x2   x3
0  A   1  2.0
1  B   0  NaN
2  A   0  1.0
3  A   1  2.0
4  A   0  NaN
5  B   1  1.0
6  A   1  1.0
7  B   0  2.0
8  B   0  2.0

I would like to fill the missing values in 'x3' by taking the median value of groupby of 'x1' and 'x2'.

groupby_df = df.groupby(['x1', 'x2'])['x3'].median()

groupby_df
Out[22]: 
x1  x2
A   0     1.0
    1     2.0
B   0     2.0
    1     1.0

So, for instance, the NaN value corresponding to (B, 0) would be replaced by 2 and (A,0) by 1. I unfortunately can't figure out this part. Is there an elegant "DataFrame way" of filling in the NaN values with the computed median using groupby?

Thank You

Answer 1

using fillna inside groupby

df['x3']=df.groupby(['x1','x2'])['x3'].apply(lambda x : x.fillna(x.median()))
df
Out[928]: 
  x1  x2   x3
0  A   1  2.0
1  B   0  2.0
2  A   0  1.0
3  A   1  2.0
4  A   0  1.0
5  B   1  1.0
6  A   1  1.0
7  B   0  2.0
8  B   0  2.0