How do I sort a dictionary by value?

Sorting Python dictionaries: under the hood

Given the unordered nature of Python dictionaries (until 3.6), sorting them into a meaningful order can feel like sorting a room full of hyperactive puppies. But hang tight! Let's untangle the leash here.

The operator way to sort

If lambdas give you headache, Python operator module is your analgesic:

import operator
# Python's operator module in disguise. Oops, given away the secret identity!
sorted_items = sorted(d.items(), key=operator.itemgetter(1))


Python's vintage way of sorting dictionaries

Prior to Python 3.7, attempting to sort a dictionary was like trying to herd cats. It refused to stay put, that's when collections.OrderedDict came to the rescue:

import collections
# It's like sorting your old pics...but these are key-value pairs...and not actually old.
sorted_dict = collections.OrderedDict(sorted(d.items(), key=lambda item: item[1]))


The sorted function Deja vu

Ever used sorted()? That's your ticket to dictionary sorting. It's like your personal butler, catering to your sorting whims with three parameters:

• iterable: The sequence (dressed in dashing tuple attire AND bubble-wrapped in dict.items())
• key: Your ordering principle, cleverly masked as a function (No, not Batman, it's just a lambda function in a key costume 🦸‍♀️).
• reverse: The switch from ascending to descending is a True away.

The sorting infinity stones

Python's sorting algorithm carries some hidden gems that include:

• A flip switch (reverse=True) to tilt the world from ascending to descending.
• The ability to sort by word count like a boss:

from collections import Counter
# Now, tell me python, which word did they use the most?
word_counts = Counter(your_text.split())
sorted_by_occurrence = sorted(word_counts.items(), key=lambda item: item[1], reverse=True)


Mastering the sorting jutsu

Sorting in Python is an art. To be a true Naruto of sorting, hone these skills:

Sorting you can trust: by keys

If keys are your thing, voila! Just omit the key function:

sorted_by_key = dict(sorted(d.items()))


Sorting you can flip: descending order

Like flipping pancakes? You can flip your dictionary order with reverse=True:

# And gravity does its thing...
sorted_desc = dict(sorted(d.items(), key=lambda item: item[1], reverse=True))


Sorting you can customize: composite values

Fun fact: Python also sorts tuples or lists like a charm:

# Pairing and sorting, we're basically matchmaking here
sorted_comp = dict(sorted(d.items(), key=lambda item: (item[1][0], item[1][1])))


The sorting wok of fire

Beware of the fire-spitting dragons while sorting:

• Memory-intensive operations in large datasets. Fight fire with fire by using generators or sorting in chunks.
• Sorting complex data types requires a bit of ingenuity and customized keys.
• Python's sort is stable: identical items retain their order. But remember, with great power comes great stability.

Sorting with swag: more scenarios, more sorting

Sorting custom objects

Every wondered if dictionaries can store custom objects? Keep wondering...because yes, they can!

# Unleashing the object sorter
sorted_custom = dict(sorted(d.items(), key=lambda item: item[1].custom_attribute))


Sorting with a safety net for nulls

Fear no more None values! The custom key function is here to clear your path:

# Clearing the None hurdles in style
sorted_handle_none = dict(sorted(d.items(), key=lambda item: (item[1] is not None, item[1])))


Bringing your SQL skills to Python sorting

ORDER BY your wishes, here's Python mimicking SQL like style:

# SQL in Python clothing. Clever, eh?
sorted_like_db = dict(sorted(d.items(), key=lambda item: (item[1]['age'], item[1]['name'])))
