python pandas "groupby" and "if any" condition

user_15814444 0
Viewed 857

I have a dataframe similar to the one below, and I would like to create a new variable which contains true/false if for each project the sector "a" has been covered at least once. I'm trying with the group.by() function, and wanted to use the .transform() method but since my data is text, I don't know how to use it.

      project    sector  
    
        01         a    
        01         b    
        02         b     
        02         b     
        03         a     
        03         a     
    



 project    sector   new_col

    01         a     true
    01         b     true
    02         b     false
    02         b     false
    03         a     true
    03         a     true
pandas python
2 Answers

It's not the fastest option, but definitely should work.

new_col = your_db.groupby(['project'])['sector'].unique().apply(lambda x: 'a' in x).rename('new_col')
your_db = your_db.merge(new_col, how = 'inner', left_on = 'project', right_on = 'project')
user_103606850

You could try the following:

df['new_col'] = df.groupby('project')['sector'].transform(lambda x: (x == 'a').any() )

This will group by project and check if any 'a' is in the groups sectors

user_139631930
Related