aggregate columns with sum and group by

Asked Mar 16 '22 at 11:07

Active Mar 16 '22 at 11:07

Viewed 13 times

I have a dataframe df_new :

PageId    OSBrowser    VolumePred   ConversionPred
    A      (11, 16)         10        40
    B      (11, 16)         20        50
    D      (11, 16)         10        10
    C      (11, 17)         30        10
    E      (11, 17)         10        10

I try to aggregate pageId by OSBrowser and calculate sum by group.

For example the expected result should be :

PageId        OSBrowser    VolumePred   ConversionPred
A, B, D       (11, 16)          40        100

C, E          (11, 17)          40        20

I try with this code :

final_df=df_new.groupby(['OSBrowser']).agg({'PageId','VolumePred':'sum'})

But I got this error :

final_df=df_new.groupby(['OSBrowser']).agg({'PageId','VolumePred':'sum'})
                                                                 ^ SyntaxError: invalid syntax

Any idea please to help me to fix it ? thanks

asked Mar 16 '22 at 11:07

buddy

1

need `{'PageId':', '.join,'VolumePred':'sum', 'ConversionPred': 'sum'}` – jezrael Mar 16 '22 at 11:08
1

@jezrael thank you for your help but I got this error when i try your code : TypeError: sequence item 0: expected str instance, float found – buddy Mar 16 '22 at 11:12
try `{'PageId': lambda x: ', '.join(x.dropna()),'VolumePred':'sum', 'ConversionPred': 'sum'}` – jezrael Mar 16 '22 at 11:14
@jezrael i am sorry but the the same error with the new code :/ – buddy Mar 16 '22 at 11:17
hmmm, `{'PageId': lambda x: ', '.join(x.dropna().astype(str)),'VolumePred':'sum', 'ConversionPred': 'sum'}` working ? – jezrael Mar 16 '22 at 11:19
1

@jezrael \o/ Thank youuuu so much it works !!! \o/ – buddy Mar 16 '22 at 11:24

aggregate columns with sum and group by

0 Answers0