From Dummy to a List pandas-CodePudding

I have a dataframe with many dummy variables. Instead of having a lot of different dummy columns, I want only one column and each row needs to contain a string with only the dummy variable equal to 1.

index  a   b   c 
0      1   1   1  
1      0   0   1

Output:

index  dummies  
0      ['a','b','c']  
1      ['c']

CodePudding user response：

dummies = df.apply(lambda x: [col for col in df.columns if x[col] == 1], axis=1)

CodePudding user response：

You can stack and use groupby:

df.where(df.eq(1)).stack().reset_index(level=1).groupby(level=0)['level_1'].agg(list)

Output:

0    [a, b, c]
1          [c]
Name: level_1, dtype: object