TIL: Python's defaultdict takes a factory function
Create a dictionary-like object that takes a factory function so you have lots of control over what the default value can be.
I've never really paid attention to this object but maybe I should have. It takes a single argument of a callable function. If you put in Python types it sets the default value to those types. For example, if I use an int at the instantiating argument then it gives us a zero.
>>> from collections import defaultdict
>>>
>>> mydict = defaultdict(int)
>>> print(mydict['anykey'])
0
Note that defaultdict also act like regular dictionaries, in that you can set keys. So mydict['me'] = 'danny' will work as you expect it to with a standard dictionary.
It gets more interesting if we pass in a more dynamic function. In the exmaple below we use random.randint and a lambda to make the default value be a random number between 1 and 100.
>>> from random import randint
>>>
>>> random_values = defaultdict(lambda: randint(1,100))
Let's try it out!
>>> for i in range(5):
>>> print(random_values[i])
>>> print(random_values)
29
90
56
42
70
defaultdict(<function <lambda> at 0x72d292bb6de0>, {0: 29, 1: 90, 2: 56, 3: 42, 4: 70})
Attribution goes to Laksman Prasad, who pointing this out and encouraging me to closer look at defaultdict.
