How to aggregate multiple columns in pandas groupby

Question

I have created a pandas dataframe mn using following input:

keyA     state n1    n2     d1  d2
key1     CA   100   1000    1   2
key2     FL   200   2000    2   4
key1     CA   300   3000    3   6
key1     AL   400   4000    4   8
key2     FL   500   5000    5   2
key1     NY   600   6000    6   4
key2     CA   700   7000    7   6

Have created a sum object as follows :

s = mn.groupby(['keyA','state'], as_index=False).sum()

How do I iterate the sum object s, so I can get following output:

The v1 column in the result below is computed as s['n1']/s['d1']

The v2 column in the result below is computed as s['n2']/s['d2']

keyA state  v1  v2
'key1','AL',100,500
'key1','CA',100,500
'key1','NY',100,1500
'key2','CA',100,1166
'key2','FL',100,1166

TomAugspurger · Accepted Answer · 2014-03-07 21:35:16Z

10

Pretty much just write it like your pseudocode.

In [14]: s = mn.groupby(['keyA','state'], as_index=False).sum()

In [15]: s['v1'] = s['n1'] / s['d1']

In [16]: s['v2'] = s['n2'] / s['d2']

In [17]: s[['keyA', 'state', 'v1', 'v2']]
Out[17]: 
   keyA state   v1           v2
0  key1    AL  100   500.000000
1  key1    CA  100   500.000000
2  key1    NY  100  1500.000000
3  key2    CA  100  1166.666667
4  key2    FL  100  1166.666667

[5 rows x 4 columns]

I think you have a typo in your example data by the way. The second n1 header should be n2.

edited Mar 7, 2014 at 21:35

answered Mar 7, 2014 at 20:18

TomAugspurger

29k8 gold badges89 silver badges71 bronze badges

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

Collectives™ on Stack Overflow

How to aggregate multiple columns in pandas groupby

1 Answer 1

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related