Replace Nan values with the minimum value in the same parent class-CodePudding

country      count

Argentina     3
Argentina     4
Argentina    NaN
Wales         1
Wales         3
Wales        NaN

Nan values, I want to make a minimum of the values that share the same countries as the NaN value.

Output Should Be:

Argentina     3
Argentina     4
Argentina     3
Wales         1
Wales         3
Wales         1

I tried bfill method but didn't manage this. Thank You For Help

CodePudding user response：

One way is using map, groupby and fillna

map allows you to match by an index, when you create a groupby object it defaults to creating an index, (as_index=True) so it's pretty simple to map it back to your main dataframe.

df['count']  = df['count'].fillna(
          df['country'].map(df.groupby('country')['count'].min()))



print(df)

     country  count
0  Argentina    3.0
1  Argentina    4.0
2  Argentina    3.0
3      Wales    1.0
4      Wales    3.0
5      Wales    1.0

a simplier method is to use .transform which applies groupby operations without modifying the index

df['count'] = df['count'].fillna(
               df.groupby('country')['count'].transform('min'))

     country  count
0  Argentina    3.0
1  Argentina    4.0
2  Argentina    3.0
3      Wales    1.0
4      Wales    3.0
5      Wales    1.0