Pandas MultiIndex: Divide all columns by one column

Question

I have a data frame results of the form

                  TOTEXPPQ      TOTEXPCQ     FINLWT21
year quarter                                         
13   1        9.183392e+09  5.459961e+09  1271559.398
     2        2.907887e+09  1.834126e+09   481169.672

and I was trying to divide all (the first two) columns by the last one. My attempt was

weights = results.pop('FINLWT21')
results/weights

But I get

ValueError: cannot join with no level specified and no overlapping names

Which I don't get: There are overlapping names in the index:

weights.head()
year  quarter
13    1          1271559.398
      2           481169.672

Is there perhaps a better way to do this division? Do I need to reset the index?

Answer 1

You have to specify the axis for the divide (with the div method):

In [11]: results.div(weights, axis=0)
Out[11]:
                 TOTEXPPQ     TOTEXPCQ
year quarter
13   1        7222.149445  4293.909517
     2        6043.371329  3811.807158

The default is axis=1 and the result columns and weights' index names do not overlap, hence the error message.