Indexing with multiple keys-CodePudding

  PageId    VolumePred  ConversionPred  OSBrowser
    955761  37200.0      27625.0    (11, 16)
    955764  30157.0      21155.0    (11, 16)
    955764  25258.0      17836.0    (11, 16)
    1184903 35492.0      17768.0    (11, 16)
    955764  24683.0      16965.0    (11, 16)
    955764  8.0              0.0    (32, 16)
    955761  4.0              0.0    (33, 16)
    955761  4.0              0.0    (33, 16)

I try to sum of VolumePred and ConversionPred by PageID and OSBrowser.

So here for example , the needed result should be :

  PageId    VolumePred  ConversionPred  OSBrowser
    955761  37200.0      27625.0       (11, 16)
    955764  55415.0      38991.0       (11, 16)
   1184903  35492.0      17768.0       (11, 16)
    955764  24683.0      16965.0       (11, 16)
    955764  8.0              0.0       (32, 16)
    955761  8.0              0.0       (33, 16)

I tryied with this code :

subdata.groupby(['PageId', 'OSBrowser'])['ConversionPred', 'VolumePred'].agg('sum')

but i get this warning :

/tmp/ipykernel_16715/1460716585.py:1: FutureWarning: Indexing with multiple keys (implicitly converted to a tuple of keys) will be deprecated, use a list instead.
  subdata.groupby(['PageId', 'OSBrowser'])['ConversionPred', 'VolumePred'].agg('sum')

Any idea please? thanks

CodePudding user response：

You're really close; you just need to add an extra set of square brackets:

subdata.groupby(['PageId', 'OSBrowser'])[['ConversionPred', 'VolumePred']].agg('sum')