Velamuvi AIAcademy

Lesson 2 of 16 · Pythonic Code

Dictionary and Set Comprehensions

Build dictionaries and sets in one line with comprehensions. Learn to transform, filter and invert dictionaries, and remove duplicates with set comprehensions.

  • 7 min read
  • Intermediate
  • Reviewed
In this lesson you will learn
  • How to build a dictionary with a comprehension
  • How to transform and filter an existing dictionary
  • How to build a set of unique values
  • How to tell the different comprehension brackets apart

From lists to dictionaries

A list comprehension uses square brackets. Swap them for curly brackets and write key: value at the front, and you get a dictionary comprehension.

Try it yourselfPython
words = ["data", "model", "training"]

lengths = {word: len(word) for word in words}

print(lengths)

For every word, the comprehension creates one entry: the word is the key and its length is the value.

Building a dictionary from two lists

zip() pairs up items from two lists. It fits naturally into a dictionary comprehension.

Try it yourselfPython
names = ["Ada", "Grace", "Alan"]
scores = [92, 85, 78]

results = {name: score for name, score in zip(names, scores)}

print(results)
print(results["Grace"])

Transforming a dictionary

Loop over .items() to work with both the key and the value.

Try it yourselfPython
prices = {"apple": 30, "banana": 10, "cherry": 120}

discounted = {item: price * 0.9 for item, price in prices.items()}

print(discounted)

Filtering a dictionary

Add if at the end, exactly as in a list comprehension.

Try it yourselfPython
stock = {"apples": 8, "pears": 0, "kiwis": 10, "plums": 0}

available = {item: count for item, count in stock.items() if count > 0}

print(available)

Swapping keys and values

Sometimes you need to look things up the other way round. Swap the key and the value.

Try it yourselfPython
codes = {"IN": "India", "JP": "Japan", "BR": "Brazil"}

by_name = {country: code for code, country in codes.items()}

print(by_name)
print(by_name["Japan"])

This only works well if every value is unique. If two keys share a value, the later one wins and the other is lost.

Set comprehensions

Curly brackets without a colon build a set: a collection of unique values with no fixed order.

Try it yourselfPython
emails = ["ada@uni.edu", "grace@navy.mil", "alan@uni.edu", "linus@uni.edu"]

domains = {email.split("@")[1] for email in emails}

print(domains)
print(len(domains))

Four emails, but only two different domains. A set keeps each value once, so duplicates disappear automatically. The order a set prints in may differ from run to run, so never rely on it.

To get a sorted list from a set, pass it to sorted():

Try it yourselfPython
tags = ["python", "AI", "Python", "ml", "ai", "ML"]

unique = sorted({tag.lower() for tag in tags})

print(unique)

Which brackets make what?

You writeYou get
[x for x in items]a list
{x for x in items}a set
{k: v for k, v in pairs}a dictionary

Round brackets, (x for x in items), make something different called a generator. You will meet generators in Advanced Python.

Common mistakes

Practice exercise

Use a dictionary comprehension to keep only the students who passed.

Your solutionPython
marks = {"Ada": 92, "Grace": 58, "Alan": 78, "Linus": 45}

# Build a dictionary called passed with only the students who scored 60 or more.
# Each value should be "pass". Then print it. Output must be exactly:
# {'Ada': 'pass', 'Alan': 'pass'}
Knowledge check

What does {len(w) for w in ['ai', 'ml', 'data']} produce?

Summary

  • {key: value for item in items} builds a dictionary.
  • Loop over .items() to transform or filter an existing dictionary.
  • {value for item in items} builds a set of unique values.
  • An empty set is set(), because {} is an empty dictionary.

Next: unpacking, a neat way to pull values out of lists and tuples.

Finished reading? Tick it off to track your progress.