1

Given three dicts d1, d2 and d3:

d1

{'a':1,'b':2,'c':3, 'd':0)

d2

{'b':76}

d3

{'a': 45, 'c':0}

There are some key names that are common to more than one dict (and in reality, they will represent the same real-life object). Others such as 'd' in d1 only exist in d2. I want to group all dicts together, first summing the values of the common keys first, ending up with:

{'a':46, 'b':78, 'c':3, 'd': 0}

If every dict were the same size and contained the same keys, I could do something like:

summedAndCombined = {}
    for k,v in d1.items():
        summedAndCombined[k] = d1[k]+d2[k]+d3[k]

But this will break down as soon as it reaches a key that is in d1 but not in the others. How do we achieve this?

UPDATE

Not a duplicate. collections.Counter almost works, but the key d is missing from the resulting Counter if the value of key d is zero, which it is above.

In [128]: d1 = {'a':1,'b':2,'c':3, 'd':0}

In [129]: d2 = {'b':76}

In [130]: d3 = {'a': 45, 'c':0}

In [131]: from collections import Counter

In [132]: Counter(d1) + Counter(d2) + Counter(d3)
Out[132]: Counter({'b': 78, 'a': 46, 'c': 3})
Pyderman
  • 14,809
  • 13
  • 61
  • 106

3 Answers3

2

You could use update instead of + with Counter if you want the 0 keys to persist:

>>> c = Counter()
>>> for d in d1, d2, d3:
...     c.update(d)
...     
>>> c
Counter({'b': 78, 'a': 46, 'c': 3, 'd': 0})

(This is probably a dup, but I can't find it right now.)

DSM
  • 342,061
  • 65
  • 592
  • 494
2

collections.defaultdict to the rescue

import collections
d = collections.defaultdict(int)
for thing in [d1, d2, d3]:
    for k, v in thing.items():
        d[k] += v
wwii
  • 23,232
  • 7
  • 37
  • 77
1

UNTESTED:

def merge_dicts(*dicts):
    res = {}
    for key in set(sum(map(list, dicts), [])):
        res[key] = 0
        for dct in dicts:
            res[key] += dct.get(key, 0)
    return res

Example usage:

merge_dicts(d1, d2, d3)
kirbyfan64sos
  • 10,377
  • 6
  • 54
  • 75