Python sum of second and third element of tuple in list of tuples grouped by first

Question

How do you get the sum of second and third values in a list of tuples grouped by the first value?

I.e:

list_of_tuples = [(1, 3, 1), (1, 2, 4), (2, 1, 0), (2, 2, 0)]

expected_output = [(1, 5, 5), (2, 3, 0)]

I found several great answers on StackOverflow doing this found tuples with two values, but couldn't figure out how to adjust them for summing both second and third values.

One of the good answers for just the second value was this:

def sum_pairs(pairs):
sums = {}
for pair in pairs:
    sums.setdefault(pair[0], 0)
    sums[pair[0]] += pair[1]
return sums.items()

score 3 · Accepted Answer · answered Oct 27 '17 at 20:21

3

You could also do it like this:

list_of_tuples = [(1, 3, 1), (1, 2, 4), (2, 1, 0), (2, 2, 0)]

# create empty dictionary to store data
sums = {}

# iterate over list of typles
for pair in list_of_tuples:

  # create new item in dictionary if it didnt exist
  if pair[0] not in sums: sums[pair[0]] = [pair[0], 0 ,0]

  # sum the values
  sums[pair[0]][1] += pair[1]
  sums[pair[0]][2] += pair[2]

#print resulting tuple   
print(tuple(sums.values()))

answered Oct 27 '17 at 20:21

Vico

579
3
13

1

No guarantee that the initial ordering will be preserved (at least in Python < 3.6) – Moses Koledoye Oct 27 '17 at 20:32

Moses Koledoye · Answer 2 · 2017-10-27T20:21:13.703

1

You can use itertools.groupby to group based on the first item, and then take the cumulative sums of all the last two items in each group:

from itertools import groupby

list_of_tuples = [(1, 3, 1), (1, 2, 4), (2, 1, 0), (2, 2, 0)]
lst = [(k,)+tuple(sum(x) for x in zip(*g))[1:] 
                         for k, g in groupby(list_of_tuples, lambda x: x[0])]
print(lst)
# [(1, 5, 5), (2, 3, 0)]

edited Oct 27 '17 at 20:21

answered Oct 27 '17 at 20:13

Moses Koledoye

77,341
8
133
139

Note, the tuples need to be grouped already by the first element, so you'd have to sort first, for example, unless you can make that assumption. – juanpa.arrivillaga Oct 27 '17 at 20:19
@juanpa.arrivillaga No assumptions: *in a list of tuples grouped by the first value* – Moses Koledoye Oct 27 '17 at 20:19
Aha! Missed that. – juanpa.arrivillaga Oct 27 '17 at 20:20

score 1 · Answer 3 · answered Oct 27 '17 at 20:17

Use a defaultdict as a grouper:

>>> from collections import defaultdict
>>> grouper = defaultdict(lambda: (0,0))
>>> list_of_tuples = [(1, 3, 1), (1, 2, 4), (2, 1, 0), (2, 2, 0)]
>>> for a, b, c in list_of_tuples:
...     x, y = grouper[a]
...     grouper[a] = (x + b, y + c)
...
>>> grouper
defaultdict(<function <lambda> at 0x102b240d0>, {1: (5, 5), 2: (3, 0)})

Now, you can always get a list of tuples back like this:

>>> [(k, a, b) for k, (a, b) in grouper.items()]
[(1, 5, 5), (2, 3, 0)]

Python sum of second and third element of tuple in list of tuples grouped by first

3 Answers3