Python: Split list of dicts by key

user_5953619 0
Viewed 666

If I have a bunch of dicts in a list like:

people = [{"name": "Abe", "age": 40},
          {"name": "Bob", "age": 25},
          {"name": "Charles", "age": 32}]

I want to split it so it looks like

names = ["Abe", "Bob", "Charles"]
ages = [40, 25, 32]

I know I can do two separate list comprehensions (ex: names = [person["name"] for person in people] twice), but is there any trick to split this data in a one-liner?

python-3.x list dictionary python
4 Answers

You can use the zip trick:

>>> names, ages = zip(*(d.values() for d in people))
>>> names
('Abe', 'Bob', 'Charles')
>>> ages
(40, 25, 32)

Note that this will only work on Python 3.7 and later as dictionaries are insertion ordered in those versions.

user_20111470

You can use zip and map dict.values to the list:

names, ages = zip(*map(dict.values, people))

>>> names
('Abe', 'Bob', 'Charles')
>>> ages
(40, 25, 32)

Of course they are tuples but that's easy enough to resolve by mapping list to the unzipped result:

names, ages = map(list, zip(*map(dict.values, people)))

>>> names
['Abe', 'Bob', 'Charles']
>>> ages
[40, 25, 32]
user_2250200

Assuming your dicts share structure (same set of keys), you can aggregate values per key with:

>>> import operator
>>> keys = people[0].keys()
>>> getter = operator.itemgetter(*keys)
>>> dict(zip(keys, zip(*map(getter, people))))
{'name': ('Abe', 'Bob', 'Charles'), 'age': (40, 25, 32)}

Works in any Python (ordered dict not required), and it's handy for large number of keys/variables.

user_4045560

You can do it with list comprehension extracting the field names you want from the dictionaries:

people = [{"name": "Abe", "age": 40},
          {"name": "Bob", "age": 25},
          {"name": "Charles", "age": 32}]

names,ages = ([p[field] for p in people] for field in ("name","age"))

print(names)  # ['Abe', 'Bob', 'Charles']

print(ages)   # [40, 25, 32]
user_52375600
Related