add values of identical dicts in a list - python-3.x

Question

I have a list of identical dicts (same keys, same number of keys..) ..like this:
mydict_list = [
{'win32': [2, 45], 'https': [2, 13], 'nofollow': [3, 45], 'href': [4, 847]},
{'win32': [0, 5], 'https': [0, 13], 'nofollow': [1, 5], 'href': [2, 87]}
]
I want to create one dict that has the same keys and values added up nicely....like this:
{'win32': [2, 50], 'https': [2, 26], 'nofollow': [4, 50], 'href': [6, 934]}
Is there an elegant way to do it?
This is what I have working. It works and is readable also:
keys = list(mydict_list[0].keys())
final_dict = {item: [] for item in keys}
for k in list(final_dict.keys()):
v0 = v1 = 0
for d in mydict_list:
for kk, v in d.items():
if k == kk:
v0 += v[0]
v1 += v[1]
final_dict.update({k: [v0, v1]})

Answer 1

Short approach:
lst = [
{'win32': [2, 45], 'https': [2, 13], 'nofollow': [3, 45], 'href': [4, 847]},
{'win32': [0, 5], 'https': [0, 13], 'nofollow': [1, 5], 'href': [2, 87]}
]
res = lst[0] # start with the 1st dict
for d in lst[1:]:
for k, v in d.items():
res[k] = list(map(sum, zip(v, res[k])))
print(res)
The output:
{'win32': [2, 50], 'https': [2, 26], 'nofollow': [4, 50], 'href': [6, 934]}

Answer 2

Consider using a dictionary comprehension on this to make it more "pythonic" (though readability goes down significantly)
d1=[
{'win32': [2, 45], 'https': [2, 13], 'nofollow': [3, 45], 'href': [4, 847]},
{'win32': [0, 5], 'https': [0, 13], 'nofollow': [1, 5], 'href': [2, 87]}
]
Can do it in a a two-liner (and should be mostly generic for almost all cases). Assuming that all the keys are the same across each dictionary in the array.
d2={key:[d[key] for d in d1] for key in d1[0].keys()}
d3={key:[sum(d3) for d3 in list(map(list,zip(*d2[key])))] for key in d1[0].keys()}
print(d3)
{'win32': [2, 50], 'https': [2, 26], 'nofollow': [4, 50], 'href': [6, 934]}
The Explanation
Dictionary comprehensions allow you to collapse your array so that we can work with each key individually.
In the first step, we create d2 which is d1 rearranged by key.
Once you have that in place, d3 the list-map re-arranges the array so that you can sum each list element wise using the built-in sum as part of a list comprehension.
If data was stored as a numpy array, this process would probably go significantly faster (as is, it's a bit memory inefficient).