List of dicts: Add up numbers with same values

Question

I got the following list of dicts

list_of_dicts = [
                {'product': 'car', 'city': 'new york', 'quantity': 13},
                {'product': 'car', 'city': 'new york', 'quantity': 25},
                {'product': 'bus', 'city': 'miami', 'quantity': 5},
                {'product': 'container', 'city': 'atlanta', 'quantity' 5},
                {'product': 'container', 'city': 'atlanta', 'quantity' 8}
            ]

My target is, when values of 'product' and 'city' are the same, sum up the values of 'quantity'. The result should look like this:

result_list_of_dicts = [
                {'product': 'car', 'city': 'new york', 'quantity': 38},
                {'product': 'bus', 'city': 'miami', 'quantity': 5},
                {'product': 'container', 'city': 'atlanta', 'quantity' 13},
            ]

Is there a pythonic way? I tried a couple of things but I better not show them because they are really ugly.

Thank you in advance!

Could you post your attempts, regardless of their appearance? — Tzane, Oct 11 '21 at 10:38
Are you limited solely to `python` built-ins or are external modules like `numpy` or `pandas` allowed? — Daweo, Oct 11 '21 at 10:39
Do you have good reasons to store this data as dictionaries rather than just lists ? — , Oct 11 '21 at 10:41
Possible duplicate https://stackoverflow.com/questions/11692613/python-sum-values-in-dictionary — abdou_dev, Oct 11 '21 at 10:41

user2390182 · Accepted Answer · 2021-10-11T10:56:08.830

You can do the following, using only standard library utils:

from operator import itemgetter
from functools import reduce
from itertools import groupby

pc = itemgetter("product", "city")  # sorting and grouping key
q = itemgetter("quantity")
combine = lambda d1, d2: {**d1, "quantity": q(d1) + q(d2)}

[reduce(combine, g) for _, g in groupby(sorted(list_of_dicts, key=pc), key=pc)]
# [{'product': 'bus', 'city': 'miami', 'quantity': 5}, 
#  {'product': 'car', 'city': 'new york', 'quantity': 38}, 
#  {'product': 'container', 'city': 'atlanta', 'quantity': 13}]

Or, maybe even simpler and linear:

from collections import Counter

pc = itemgetter("product", "city") 
q = itemgetter("quantity")

totals = Counter()
for dct in list_of_dicts:
    totals[pc(dct)] += q(dct)

result_list_of_dicts = [
    {"product": p, "city": c, "quantity": q} for (p, c), q in totals.items()
]

Ok, amazing! Thanks a lot!! – buxxxbaum Oct 11 '21 at 15:20 — buxxxbaum, Oct 11 '21 at 15:20

score 2 · Answer 2 · answered Oct 11 '21 at 10:57

One approach using collections.Counter

from collections import Counter

list_of_dicts = [
    {'product': 'car', 'city': 'new york', 'quantity': 13},
    {'product': 'car', 'city': 'new york', 'quantity': 25},
    {'product': 'bus', 'city': 'miami', 'quantity': 5},
    {'product': 'container', 'city': 'atlanta', 'quantity': 5},
    {'product': 'container', 'city': 'atlanta', 'quantity': 8}
]

counts = sum((Counter({(d["product"], d["city"]): d["quantity"]}) for d in list_of_dicts), Counter())
result = [{"product": product, "city": city, "quantity": quantity} for (product, city), quantity in counts.items()]
print(result)

score 1 · Answer 3 · answered Oct 11 '21 at 10:46

A pandas implementation

Group by "product" and "city", sum over the groups and reset index to get original columns.

import pandas as pd

list_of_dicts = [
    {'product': 'car', 'city': 'new york', 'quantity': 13},
    {'product': 'car', 'city': 'new york', 'quantity': 25},
    {'product': 'bus', 'city': 'miami', 'quantity': 5},
    {'product': 'container', 'city': 'atlanta', 'quantity': 5},
    {'product': 'container', 'city': 'atlanta', 'quantity': 8}
]

df = pd.DataFrame(list_of_dicts)
print(df)
df = df.groupby(["product", "city"]).sum().reset_index()
print(df)
summed_dict = df.to_dict("records")
print(summed_dict)

scmanjarrez · Answer 4 · 2021-10-11T11:09:01.223

You could do it with a loop, initializing it the first time you encounter the product.

list_of_dicts = [
    {'product': 'car', 'city': 'new york', 'quantity': 13},
    {'product': 'car', 'city': 'new york', 'quantity': 25},
    {'product': 'bus', 'city': 'miami', 'quantity': 5},
    {'product': 'container', 'city': 'atlanta', 'quantity': 5},
    {'product': 'container', 'city': 'atlanta', 'quantity': 8}
]

new_dict = {}
for ld in list_of_dicts:
    if ld['product'] not in new_dict:
        new_dict[ld['product']] = {}
        new_dict[ld['product']]['city'] = ld['city']
        new_dict[ld['product']]['quantity'] = 0
    new_dict[ld['product']]['quantity'] += ld['quantity']

# print(new_dict)
# {'car': {'city': 'new york', 'quantity': 38}, 'bus': {'city': 'miami', 'quantity': 5}, 'container': {'city': 'atlanta', 'quantity': 13}}

result_list_of_dicts = [{'product': nd,
                         'city': new_dict[nd]['city'],
                         'quantity': new_dict[nd]['quantity']} for nd in new_dict]
# print(result_list_of_dicts)
# [{'product': 'car', 'city': 'new york', 'quantity': 38}, {'product': 'bus', 'city': 'miami', 'quantity': 5}, {'product': 'container', 'city': 'atlanta', 'quantity': 13}]

List of dicts: Add up numbers with same values

4 Answers4