How to sum duplicates values in list?

Question

I have a simple list of strings:

1 test1
10 test1
2 test1
12 test2
87 test2
12 test1
75 test1
43 test1

How to get list of strings:

1+10+2 test1
12+87 test2
12+75+43 test1

?

Were you looking to add all the `test1` values into one string? — Sukrit Kalra, Jul 22 '13 at 15:47
See: http://stackoverflow.com/questions/2161752/how-to-count-the-frequency-of-the-elements-in-a-list — Jon Clements, Jul 22 '13 at 15:47
@aaronman: So how would you do use a list comprehension here? It is possible to collapse my solution into a list comprehension, but it is not going to be very readable, really. — Martijn Pieters, Jul 22 '13 at 15:50
@MartijnPieters unfortunately I'm not really a python guy but I know list comps are a good tool from haskell — aaronman, Jul 22 '13 at 15:53
@aaronman: Yes, list comprehensions are a great tool, if applied in the right places. :-) I've added a list comprehension to my answer to produce only the sums. — Martijn Pieters, Jul 22 '13 at 15:58

Martijn Pieters · Answer 1 · 2013-07-22T15:57:41.640

Use itertools.groupby() to group your numbers per test.. value:

from itertools import groupby

for key, group in groupby(list_of_strings, lambda k: k.split(None, 1)[-1]):
    # loop over `group` to see all strings that have the same last word.
    print '+'.join([s.split(None, 1)[0] for s in group]), key

group is an iterable that yields all contiguous strings that have the same second value (split on whitespace); your input has two test1 groups that are contiguous, for example.

Demo:

>>> list_of_strings = '''\
... 1 test1
... 10 test1
... 2 test1
... 12 test2
... 87 test2
... 12 test1
... 75 test1
... 43 test1
... '''.splitlines()
>>> from itertools import groupby
>>> for key, group in groupby(list_of_strings, lambda k: k.split(None, 1)[-1]):
...     # loop over `group` to see all strings that have the same last word.
...     print '+'.join([s.split(None, 1)[0] for s in group]), key
... 
1+10+2 test1
12+87 test2
12+75+43 test1

or, if you wanted to actually sum the values as integers:

for key, group in groupby(list_of_strings, lambda k: k.split(None, 1)[-1]):
    print sum(int(s.split(None, 1)[0]) for s in group), key

which prints

13 test1
99 test2
130 test1

If you just need the sums, make it a list comprehension:

sums = [sum(int(s.split(None, 1)[0]) for s in g) for _, g in groupby(L, lambda k: k.split(None, 1)[-1])]

How to sum duplicates values in list?

1 Answers1