How to group data by each two (3,7) hours groups in pandas?

Question

I am aggregating data by each hour in each day:

                    hokkey_common_en  hokkey_negative_en  hokkey_Positive_en  
time                                                                           
2014-02-06 00:00:00               733                   1                  12  
2014-02-06 00:30:00               739                   3                  22  
2014-02-06 01:00:00               767                   2                  15  
2014-02-06 01:30:00               950                   8                  27  

d = frame.groupby([frame.index.day, frame.index.hour], sort=True).sum()

print d

 hokkey_common_en  hokkey_negative_en  hokkey_Positive_en  
6 0               1472                   4                  34  
  1               1717                  10                  42  
  2               1549                  20                  46  

but how should I generate a sum for each two hours (tree, seven) in a day?

Answer 1

You can just resample by '2H':

In [11]: df.resample('2H', how='sum')
Out[11]: 
            hokkey_common_en  hokkey_negative_en  jotkey_positive_en
time                                                                
2014-02-06              3189                  14                  76