[Chicago] 2 unique keys in a python dict?

[John Hunter]

On Thu, Oct 9, 2008 at 1:02 PM, Massimo Di Pierro
<mdipierro at cs.depaul.edu> wrote:
> Not quite the same. I believe Lukasz is not looking into adding up the
> column. He is looking into grouping records by transaction id and than
> summing up records within the same transaction.

I see, if you want to group by id

  import matplotlib.mlab as mlab
  r = mlab.csv2rec('myfile.csv')
  stats = ( ('paid', np.sum, 'paidsum')
  key = ('id', )
  rgroup = mlab.rec_groupby(r, key, stats)
  print rgroup.paidsum

The stats variable is a sequence of (inputname, groupby_function, outputname)