Python Filter a List

Filtering means picking out the elements of a collection that match a condition.

The condition takes one element and answers True (keep it) or False (skip it). A function that does this is called a predicate.

Filtering builds a new list. The original list is not changed.

In Python you filter with a list comprehension that has an if part: [w for w in words if len(w) > 3]. It reads almost like English: keep every w whose length is greater than 3.

Python also has a built-in filter() function that takes a predicate and a list. It works, but it returns a lazy iterator and needs a lambda for a simple condition, so the comprehension is usually clearer. This page uses comprehensions.

The simplest way to filter is a comprehension whose if part holds the condition. Python walks through the list and keeps every element that passes the test.

Dictionaries work the same way: a dict comprehension over items() copies the matching key:value pairs into a new dictionary.

If the condition also needs the position of the element, enumerate gives you the index and the value together.

To keep the elements that do not match, put not in front of the condition. Python has no separate filter_not.

A list can hold values of different types. isinstance(v, str) checks whether a value is a string. A value can also be None, which means "no value"; compare with is not None to skip such entries.

Keep only the words that are longer than 3 letters. The condition len(w) > 3 sits in the if part of the comprehension.

Loop over items() and keep the pairs whose key ends with "1" and whose value is greater than 10. endswith() checks the end of a string.

enumerate gives the index and the value. Keep a word only when its index is not 0 and it has fewer than 5 letters.

Drop the words with 3 letters or fewer. The condition len(w) = 3 is negated with not, so only the longer words stay.

The list holds None, a number, two strings and a float. isinstance(v, str) finds the strings. Each one is printed in upper case.

The list holds strings, but some entries are None. Keep the values that are not None with the is not None check.

Sometimes you want both groups: the elements that match and the ones that do not.

Python has no built-in partition for lists. A small helper does it: it loops once and appends each element to either the match list or the rest list. The helper returns both lists, because a Python function can return more than one value.

You could also write two comprehensions, one with the condition and one with its negation, but that walks the list twice; the helper does one pass.

Words longer than 3 letters go into match; all the others go into rest. Both lists are printed.

Sometimes you do not need the matching elements, only a yes or no answer. Python has built-in functions for this:

• any: does at least one element pass the test? any() answers this.
• none: does no element pass the test? Put not in front of any().
• all: does every element pass the test? all() answers this.

Both take a series of True/False values, so you pass them a generator expression such as w.endswith("e") for w in numbers. any() stops as soon as it finds a match.

Without a condition, "any element at all?" is len(numbers) > 0 and "no element at all?" is len(numbers) == 0.

On an empty list, all() returns True. There is no element that fails the test, so the test is never broken. This is called vacuous truth.

Check whether at least one word ends with "e", then check whether the list has any element at all.

Check that no word ends with "a", then check whether the list is empty.

Check whether every word ends with "e". Then call all() on an empty list, which gives True.