pandas summing up columns by group

Asked Mar 03 '16 at 19:58

Active Mar 03 '16 at 19:58

Viewed 27 times

I have a dataframe like such:

   Bool  Count State
0     0      1    CA
1     1      1    CA
2     1      1    VT
3     0      1    CA

I wanted to group them all such that I get the number of all values of CA that are 0,1, etc, and I applied the following code:

df.groupby(["State","Bool"]).sum()

And it gives me:

            Count
State Bool       
CA    0         2
      1         1
VT    1         1

How do I output this in a JSON format?

asked Mar 03 '16 at 19:58

lordingtar

can't you simply use `df.groupby(["State","Bool"]).sum().to_json()`? – Julien Spronck Mar 03 '16 at 20:06
It really depends on the required format of your JSON. – Alexander Mar 03 '16 at 20:09
@JulienSpronck That gives a JSON in the format {"Count":{"["CA",0]":2,"["CA",1]":2,"["VT",0]":2,"["VT",1]":1}}, whereas I want it in the format {"State":"CA","Bool":0,"Count":2} – lordingtar Mar 03 '16 at 20:52

0 Answers0