Imputing the range of values with median

Question

I have column 'X' in a dataframe. I want to impute the negative values and value above 10 with median.

Below is my sample data

index   X
0    -3
1     5
2     7
3     6
4     0
5     8
6     6
7    -2
8     9
9     2465

Below is the code that I have tried:

median = df.loc[(df['X']<10) & (df['X']>=0), 'X'].median()
df.loc[(df['X'] > 10) & (df['X']<0), 'X'] = np.nan
df['X'].fillna(median,inplace=True)

There is still no change in 'X' column even after applying the above codes.

Answer 1

Use Series.where if need median of filtered values:

mask = (df['X']<10) & (df['X']>=0)
df['X'] = df['X'].where(mask, df.loc[mask, 'X'].median())
print (df)
   X
0  6
1  5
2  7
3  6
4  0
5  8
6  6
7  6
8  9
9  6

Or median of all values:

mask = (df['X']<10) & (df['X']>=0)
df['X'] = df['X'].where(mask, df['X'].median())

Answer 2

可能使用：

df.loc[(df['X'] > 0) & (df['X'] < 10), 'X'] = df['X'].median()