`filter()`

Introduction

The filter() function constructs a lazy iterator from elements of an iterable for which a function returns True.

Like map(), filter() has largely been superseded by list comprehensions in modern Python, but you should know how it works in case you encounter it or need a functional programming approach.

Basic Usage

filter(function, iterable)

nums = [1, 2, 3, 4, 5, 6]

def is_even(n):
    return n % 2 == 0

# Returns a lazy iterator
evens = filter(is_even, nums)

# Wrap in list() to compute
print(list(evens)) # [2, 4, 6]

Filtering Falsy Values

If you pass None as the function, filter() will automatically remove any "falsy" values (0, "", False, None, []) from the iterable.

data = [1, 0, 2, "", 3, None, 4]

truthy_only = list(filter(None, data))
print(truthy_only) # [1, 2, 3, 4]

When to AVOID `filter()` in Interviews

Just like map(), if you have to write a lambda to use filter(), you should use a list comprehension instead. Comprehensions with if clauses are faster and more readable.

Avoid this:

nums = [1, 2, 3, 4, 5, 6]
evens = list(filter(lambda x: x % 2 == 0, nums))

Write this instead:

evens = [x for x in nums if x % 2 == 0]

Time and Space Complexity

Time Complexity: \(O(N)\) once consumed.
Space Complexity: \(O(1)\) memory overhead for the iterator itself.

Summary

filter() creates an iterator of elements that pass a boolean function.
filter(None, iterable) is a quick trick to remove falsy values.
In almost all other cases in an interview, use a list comprehension with an if clause instead.

filter()