CODEWITHSUNDEEP
pythonpandasdataframegroup-bypandas-groupby
spectrum
spectrum

Reputation: 97

Change all first values in column, grouped from other column

I have 2 columns - _a, _b. 

import numpy as np 
import pandas as pd
df = pd.DataFrame({'_a':[1,1,1,2,2,3,3],'_b':[3,4,5,3,3,3,9]})
df

    _a  _b
0   1   3
1   1   4
2   1   5
3   2   3
4   2   3
5   3   3
6   3   9

I need change first value in column _b to 0, grouped by column _a Like here:

    _a  _b
0   1   0
1   1   4
2   1   5
3   2   0
4   2   3
5   3   0
6   3   9

Thank you in advance.

Upvotes: 0

Views: 72

Answers (3)

cfort
cfort

Reputation: 2776

If the first values of each group aren't duplicates, you can use this:

df.loc[df.groupby('_a').head(1).index, '_b'] = 0

output:

        _a  _b
    0   1   0
    1   1   4
    2   1   5
    3   2   0
    4   2   3
    5   3   0
    6   3   9

Upvotes: 2

hilberts_drinking_problem
hilberts_drinking_problem

Reputation: 11602

You can get indices of first occurrences for each value in df._b with np.unique:

df._b[np.unique(df._a, return_index=True)[1]] = 0

Upvotes: 0

user2285236
user2285236

Reputation:

By default, pd.Series.duplicated returns False for the first item. Making use of this, you can do:

df.loc[~df['_a'].duplicated(), '_b'] = 0

df
Out: 
   _a  _b
0   1   0
1   1   4
2   1   5
3   2   0
4   2   3
5   3   0
6   3   9

Upvotes: 3

Related Questions