List vs tuple vs set vs dict in Python

The short answer

In a set and in dict keys, items must be hashable. Numbers, strings and tuples of hashable items are. Lists, sets and dicts are not.

List: an ordered, changeable sequence

Use a list when the order matters and the contents change. It is the default choice.

scores = [72, 88, 65]
scores.append(91)          # add at the end
scores[0] = 75             # change an item
scores.remove(65)          # remove by value
print(scores)
print(scores[1:3])         # slicing gives a new list
print(sorted(scores))

[75, 88, 91]
[88, 91]
[75, 88, 91]

Tuple: a fixed record

A tuple cannot change after creation. Use it for a group of values that belong together, such as a point or a row. Python functions that return several values really return a tuple.

point = (3, 4)
x, y = point               # unpacking
print(x, y)

def min_max(values):
    return min(values), max(values)    # returns a tuple

low, high = min_max([4, 9, 2])
print(low, high)

try:
    point[0] = 10
except TypeError as error:
    print("TypeError:", error)

print(type((5)), type((5,)))           # one-item tuple needs a comma

3 4
2 9
TypeError: 'tuple' object does not support item assignment
<class 'int'> <class 'tuple'>

Set: unique items and fast membership

A set throws away duplicates and answers "is this in the collection?" in O(1) on average. Use it to remove duplicates and to compare groups.

python_class = {"Asha", "Ravi", "Meena"}
sql_class = {"Ravi", "Tom", "Meena"}

print(sorted(python_class & sql_class))   # in both
print(sorted(python_class | sql_class))   # in either
print(sorted(python_class - sql_class))   # only in Python class
print("Asha" in python_class)

print(len(set([1, 2, 2, 3, 3, 3])))       # remove duplicates

['Meena', 'Ravi']
['Asha', 'Meena', 'Ravi', 'Tom']
['Asha']
True
3

A set has no order and no index, so s[0] is an error. If you need the items in order, use sorted(s).

Dict: look things up by key

A dict maps keys to values. Use it when you want to find something by name or id instead of by position.

marks = {"Asha": 88, "Ravi": 72}
marks["Meena"] = 91             # add
marks["Ravi"] = 75              # update
print(marks["Asha"])
print(marks.get("Tom", 0))      # safe lookup with a default

for name, mark in marks.items():
    print(name, mark)

counts = {}
for ch in "banana":
    counts[ch] = counts.get(ch, 0) + 1
print(counts)

88
0
Asha 88
Ravi 75
Meena 91
{'b': 1, 'a': 3, 'n': 2}

Which one should I use?

The speed difference

Looking up an item in a list scans it one by one, so the cost grows with the size. A set or dict jumps straight to the item using its hash. For a large collection that you search many times, this is the difference between a slow solution and a fast one.

names = ["Asha", "Ravi", "Meena", "Tom"]
lookup = set(names)

print("Meena" in names)     # scans the list: O(n)
print("Meena" in lookup)    # one hash lookup: O(1) on average

True
True

Common mistakes

• {} is an empty dict, not an empty set. Write set() for an empty set.
• (5) is just the number 5. A one-item tuple is (5,).
• Putting a list in a set or as a dict key raises TypeError: unhashable type: 'list'. Convert it to a tuple first.
• Using in on a big list inside a loop. Convert the list to a set once, before the loop.
• Changing a list or dict while looping over it. Loop over a copy, or build a new collection.
• Expecting a tuple to be deeply fixed. A tuple that holds a list is fixed, but the list inside can still change.

bad_key = [1, 2]
try:
    seen = {bad_key}
except TypeError as error:
    print("TypeError:", error)

good_key = tuple(bad_key)
print({good_key})

record = (1, [2, 3])
record[1].append(4)         # allowed: the list inside is mutable
print(record)

TypeError: unhashable type: 'list'
{(1, 2)}
(1, [2, 3, 4])

• Python Collections & Data Structures
• Python Basics
• Python interview questions for freshers
• DSA in Python roadmap