Merging a pandas groupby result back into DataFrame

Question

I have a DataFrame that looks like this...

   idn value  
0  ID1    25
1  ID1    30
2  ID2    30
3  ID2    50

I want to add another column to t

Answer 1

set the index of df to idn, and then use df.merge. after the merge, reset the index and rename columns

dfmax = df.groupby('idn')['value'].max()

df.set_index('idn', inplace=True)

df = df.merge(dfmax, how='outer', left_index=True, right_index=True)

df.reset_index(inplace=True)

df.columns = ['idn', 'value', 'max_value']

Answer 2

Use the transform method on a groupby object:

In [5]: df['maxval'] = df.groupby(by=['idn']).transform('max')

In [6]: df
Out[6]: 
   idn  value  maxval
0  ID1     25      30
1  ID1     30      30
2  ID2     30      50
3  ID2     50      50