Skip to main content

Sets Deep Dive

set is a particularly useful but easily underestimated container in Python. Whenever these keywords appear in a problem, I usually think of sets first:

  • Deduplication
  • Duplicate checking
  • Whether something has been visited
  • Intersection, union, difference

Creating Sets​

numbers = {1, 2, 3, 4}
empty = set()

Note: {} creates an empty dictionary, not an empty set.

visited = {1, 2, 3}

visited.add(4)
visited.remove(2)
visited.discard(10)
print(3 in visited)

The difference:

  • remove(x): Raises an error if the element doesn't exist
  • discard(x): Doesn't raise an error if the element doesn't exist

Set Operations​

a = {1, 2, 3, 4}
b = {3, 4, 5, 6}

print(a | b)  # union
print(a & b)  # intersection
print(a - b)  # difference
print(a ^ b)  # symmetric difference

Most common application: deduplication​

numbers = [1, 2, 2, 3, 3, 4]
unique_numbers = list(set(numbers))

If you also want to preserve the original order, plain set() isn't enough -- you'll need other methods.

Second most common application: visited markers​

In graph search, DFS, and BFS, I almost always write a visited set:

graph = {
    "A": ["B", "C"],
    "B": ["A", "D"],
    "C": ["A"],
    "D": ["B"],
}

visited = set()

frozenset​

If you need an "immutable set", you can use frozenset:

fs = frozenset([1, 2, 3])

It cannot be modified, but it can perform set operations and can also serve as a dictionary key.

A reminder I often stumble on​

Sets are unordered containers. Don't assume they preserve the order in which you inserted elements. Even if sometimes they "appear to preserve it", don't build logic around that behavior.