Grouping in Python -

i have list of dictionaries (which uploaded using csv), , run "group by" equivalent based on 1 of "columns". trying group on teamid , sum "r" columns based on groupings.

i trying following code:

import itertools  key, group in itertools.groupby(batting, lambda item: item["teamid"]):     print key, sum([item["r"] item in group]) 

however, not seeing them grouped correctly. there multiple instances of same team id.

for example:

rc1 30 cl1 28 ws3 28 rc1 29 fw1 9 rc1 0 bs1 66 fw1 1 bs1 13 cl1 18 

as padric said in comment, itertools.groupby() needs ordered data want. simplest solution (as in least code edits) be:

import itertools  key_func = lambda item: item["teamid"]  key, group in itertools.groupby(sorted(batting, key=key_func), key_func):     print key, sum([item["r"] item in group]) 

if data relatively big, may want consider more efficient doesn't require duplicate sorted copy in memory. defaultdict mentioned in comment may choice.

from collections import defaultdict  d = defaultdict(int)  item in batting:   d[item['teamid']] += item.get('r', 0) or 0  team, r_sum in sorted(d.items(), key=lambda x: x[0]):   print team, r_sum 

the code may need slight adjustments python 3.