Collections

The collections module contains a few very useful dictionaries:

type	description
Counter	dictionary for counting things
defaultdict	dictionary with a preset value
OrderedDict	dictionary preserving insertion order

Counter

The Counter is a special dictionary for counting things.

from collections import Counter

# generate 100 random names
names = ["Adam", "Bea", "Charlie", "Danielle",
         "Eve", "Frantz", "Gustav", "Helena"]
from random import choice
data = [choice(names) for i in range(100)]

You create a Counter by giving it an iterable:

c = Counter(data)

The most important new method is most_common. The rest works like a normal dictionary.

print(c)
print(c.most_common(3))
print(c.get('Adam'))

---

Defaultdict

A defaultdict inserts a default value for each key that is used the first time. It therefore never throws a KeyError when requesting a value.

Creating a defaultdict

A defaultdict is created with a function that creates the default valued.

The following dictionary defaults to a zero integer:

from collections import defaultdict

d = defaultdict(int)
d['Adam'] = 33
d['Eve'] = 55

print(d['Adam'])  # -> 33 like a normal dict
print(d['Guido']) # -> 0  because it's a new key

Most times a defaultdict is initialized with a data, but any Python function works. Try:

from random import random

d = defaultdict(random)
d['dummy']

Collecting lists

A common use case is to create a dict of lists. This becomes very easy with a defaultdict.

The following example sorts names by their initial:

names = [
        "Athene", "Ada", "Hypathia",
        "Anna", "Helena", "Thetis"
]

d = defaultdict(list)

for n in names:
    key = n[0]
    d[key].append(n)

This results in the following data:

defaultdict(list,
            {'A': ['Athene', 'Ada', 'Anna'],
             'H': ['Hypathia', 'Helena'],
             'T': ['Thetis']})

---

OrderedDict

OrderedDict is a special kind of dictionary that preserves the insertion order.

from collections import OrderedDict

od = OrderedDict()

names = ["Adam", "Bea", "Charlie", "Danielle", "Eve", "Frantz", "Gustav", "Helena"]
for i, name in enumerate(names):
     od[i] = name

With an OrderedDict, you have a guarantee that the output of the following command is always the same (which you don’t have with a normal dictionary):

print(od)

You have an additional method for changing the order:

od.move_to_end(2)
print(od)