How can I aggregate on multiple columns in pandas?

Question

Now i have some data looks like below :

    song_id                             artist_id                     0 days 1 days 2 days  
1   0919b5ed4ce2649f61bcc6c21fadab12    0c80008b0a28d356026f4b1097041689    0   0   0
2   8a0777df37bf6a0f3384d63a47d4d21b    0c80008b0a28d356026f4b1097041689    0   1   0
3   b61bc45712ee40c3f4a37dd4d063ad52    0c80008b0a28d356026f4b1097041689    0   0   0
4   a2fbe29da3a760d7467b8a7b3247a9c8    0c80008b0a28d356026f4b1097041689    0   0   1
5   b5e92cb9ff2126189c19305cf148b25d    0c80008b0a28d356026f4b1097041689    0   0   0

And I want to group them by artist_id and aggregate the sum on 0 days , 1 days and 2 days , and get result like that .

        artist_id                      0 days 1 days 2 days
0       0c80008b0a28d356026f4b1097041689    0   1   1

I tried

df.groupby('artist_id').sum()

But it raise an error .

TypeError: Cannot compare type 'Timedelta' with type 'str'

df.info shows:

<class 'pandas.core.frame.DataFrame'>
Int64Index: 10842 entries, 0 to 10841
Columns: 185 entries, song_id to 182 days 00:00:00
dtypes: float64(183), object(2)
memory usage: 15.4+ MB

How can I solve it using pandas's way ?

Any help is welcomed .

What does df.info() show, if you have non-numeric dtypes then sum will fail — EdChum
– EdChum, Commented Apr 4, 2016 at 12:29
@jezrael I have updated the question. Thanks for commenting . — KIDJourney
– KIDJourney, Commented Apr 4, 2016 at 12:45
@EdChum I have updated the question. Thanks for commenting . — KIDJourney
– KIDJourney, Commented Apr 4, 2016 at 12:45

jezrael · Accepted Answer · 2016-04-04 12:56:51Z

1

You can use astype:

df.columns = df.columns.astype(str)

edited Apr 4, 2016 at 12:56

answered Apr 4, 2016 at 12:48

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

KIDJourney Over a year ago

Thanks for your answer , the sum() function failed because there are timedelta in columns and they can't be aggregated .

jezrael Over a year ago

:) I understand, so I add next possible solution with astype.

KIDJourney · Accepted Answer · 2016-04-04 12:49:44Z

0

Thanks everyone .

After applying

df.columns = map(str,df.columns)

The step

df.groupby('artist_id').sum()

works .

answered Apr 4, 2016 at 12:49

KIDJourney

1,2501 gold badge12 silver badges22 bronze badges

Collectives™ on Stack Overflow

How can I aggregate on multiple columns in pandas?

2 Answers 2

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related