Sets are a special kind of collection which contain only unique elements. They have special methods to perform set operations like union, intersection, difference etc.

Set basics

Sets can be created from a list or tuple using the set constructor. Notice how all duplicated items are removed, and appear only once int the resulting set:

>>> set_1 = set([1, 3, 2, 3, 4, 3])
>>> print set_1
set([1, 2, 3, 4])

New syntax for creating sets

Since Python 2.7, non-empty sets can be created by enclosing a sequence of comma-separated values in curly brackets:

>>> set_2 = {1, 8, 4, 5, 6}
set([8, 1, 4, 5, 6])

Curly brackets are also used to create dictionaries. So a pair of empty curly brackets will create a dictionary, not a set:

>>> print type({})
<type 'dict'>

To create an empty set, it is still required to use the set constructor:

>>> print type(set())
<type 'set'>

Set operations

Sets have methods to perform boolean set operations with other sets: union, difference and intersection:

The union operation returns a new set with elements from both sets:

>>> print set_1
>>> print set_2
set([1, 2, 3, 4])
set([8, 1, 4, 5, 6])
>>> print set_1.union(set_2)
set([1, 2, 3, 4, 5, 6, 8])

The intersection operation returns a new set with elements which are common to both sets:

>>> print set_1.intersection(set_2)
set([1, 4])

The difference operation returns a new set with elements which are in the first set, and not in the second:

>>> print set_1.difference(set_2)
set([2, 3])
>>> print set_2.difference(set_1)
set([8, 5, 6])


For the difference operation, the order of the sets matters. For the other operations, changing the order will not change the result.