List, Dict & Set Comprehensions

Comprehensions are Python's concise syntax for building new collections from existing iterables. They are faster than equivalent for loops (the iteration is done in C under the hood) and, when kept simple, dramatically improve readability.

List Comprehension

Equivalent for loop for comparison:

Dict Comprehension

Set Comprehension

Nested Comprehensions

You can nest comprehensions to work with 2D structures. Keep readability in mind — more than two levels is usually a sign to use a regular loop.

Filtering with if

The if clause comes at the end of the comprehension and acts as a filter — only items passing the condition are included:

Walrus Operator (:=) in Comprehensions

Python 3.8+ allows capturing a sub-expression with := (walrus operator) to avoid computing it twice:

The walrus operator is especially useful when the filtering predicate and the value you want to keep are the same computation.

Performance: Comprehensions vs map/filter

Comprehensions are generally the preferred Pythonic style, but map/filter with simple callables can be marginally faster for very large datasets:

In practice the difference is small. Prefer comprehensions for readability unless profiling shows a bottleneck.

Readability Guidelines

Do use comprehensions when:

• The logic is a simple transformation or filter
• The result fits on one line (≤ 79 characters)
• There are at most two for clauses

Avoid comprehensions when:

• The loop body has side effects (printing, writing to a file, modifying external state)
• The logic is complex enough that a for loop with comments would be clearer
• You need the overhead of exception handling inside the loop