CODEWITHSUNDEEP
pythonpandasgroup-bypandas-groupby
Anirban
Anirban

Reputation: 33

Access specific group in group by operation with pandas

    A   B
0   1  12
1   1  13
2   1  15
3   2  16
4   2  19
5   2  20
6   3  32
7   3  29
8   3  25
9   4   3
10  4   5
11  4   7

I have a pandas dataframe as shown above.I want to group column 'A' by values like 1 2 3 4 and subtract the values of column B of all the groups from group 4. How can I do it?

My result should look like:

    
A   B   C
1   12  9
1   13  8
1   15  8
2   16  13
2   19  14
2   20  13
3   32  29
3   29  24
3   25  18
4   3   0
4   5   0
4   7   0

Upvotes: 2

Views: 245

Answers (1)

cs95
cs95

Reputation: 402493

You can use groupby and apply. To get the group for A=4, call GroupBy.get_group(4).

g = df.groupby('A')
df['C'] = g.apply(lambda x: x['B'] - g.get_group(4)['B'].values).values
df

    A   B   C
0   1  12   9
1   1  13   8
2   1  15   8
3   2  16  13
4   2  19  14
5   2  20  13
6   3  32  29
7   3  29  24
8   3  25  18
9   4   3   0
10  4   5   0
11  4   7   0

Upvotes: 4

Related Questions