Source: đź“– Effective Python item 32
Generator expressions are much like comprehensions, except that instead of creating a container they create an iterator. The syntax used is the same but the expression must me wrapped in () instead of [].
my_nums = [1, 2, 3, 4, 5, 6]
square_container = [x**2 for x in my_nums] # comprehension
square_it = (x**2 for x in my_nums) # generator expression
Generator expressions can be used in nested combinations—that is, use an iterator already created by a generator expression inside a subsequent generator expression.
square_pairs = ((i, square) for i, square in enumerate(square_it))
The above code utilises the square_it iterator that was created in the first example, and combines it with enumerate to create another iterator that returns a tuple with the enumerate index value and the squared number at that index.
Each time the square_pairs iterator is advanced using next, the interior square_it iterator is also advanced creating a cascading effect. This is very memory-efficient as it only ever generates and evaluates the values being used for the current call, rather than generating all values and storing them in memory. The only thing to be wary of is the fact that these iterators can become exhausted once fully iterated, so they cannot be reused for further iteration unlike iterable containers.