Removing Duplicates From a List in Python

Let’s go over a few idiomatic ways to remove duplicates from lists in Python.

Method #1 - Create a new list (simplest) 🔗

This is the easiest algorithm to code, but because it requires creating a new list, also requires more memory and is a bit slower.

def remove_duplicates(original):
  deduped = []
  for item in original:
    if item not in deduped:
      deduped.append(item)
  return deduped

We take advantage of Python’s in keyword here, only adding each item to the final list if it isn’t already present.

Method #2 - Create a new list with syntactic sugar (less code, harder to understand) 🔗

def remove_duplicates(original):
  deduped = []
  [deduped.append(item) for item in original if item not in deduped]
  return deduped

This is the same exact code from a performance standpoint but only uses one line. If you’re into code golf, then this might be your solution.

Method #3 - Use the built-in “set” data structure (fast, loses order) 🔗

A set() is a group of values that doesn’t contain any duplicates. By casting a list into a set and back, you remove all duplicates. The main drawback here is that you’ll lose your ordering.

def remove_duplicates(original):
  return list(set(original))

This method will be faster in most circumstances than the previous two because each transfer is O(n) in big-o notation terms. A group of two O(n) operations is faster than one O(n^2) operation. As a bonus, it even uses less code.

Method #4 - Use an ordered dictionary (fast, maintains order) 🔗

By using the collections libraries’ OrderedDict type, we can maintain the ordering of the list while maintaining the same Big-O that we had with a set().

from collections import OrderedDict

def remove_duplicates(original):
  return list(OrderedDict.fromkeys(original))