What are named tuples in Python?

Python's named tuples are a very simple yet interesting feature that can make a developer's life easier. They are part of the collections module and act very similar to regular tuples, the main difference being that values stored in a named tuple can be accessed using field names instead of indexes.
For example, a point in the two-dimensional plane can be represented using two coordinates. In a regular tuple, these values would be accessed by index ([0] and [1]), but if we define a named tuple, Point, we can access them using x and y instead (although we can still use indexes, too, if we want):
Apart from the increased readability of your code, named tuples provide a few other quality of life improvements. First and foremost, they allow for default values to be specified via the defaults iterable argument. Secondly, they have the ability to automatically rename duplicate or invalid fields via the rename boolean argument. And, finally, they even provide a convenient option to specify field names as a list or comma/space-separated string.
_Where's the catch?_ you might ask. Well, it seems like there's none! The obvious parallel to dictionaries in terms of syntax doesn't seem to go any further, as named tuple instances do not have per-instance dictionaries, meaning they require as much memory as regular tuples.

from collections import namedtuple

# Regular tuple
p = (2, 4) # p[0] = 2, p[1] = 4

# Named tuple
Point = namedtuple('Point', 'x y')
q = Point(3, 5) # q.x = 3, q.y = 5

Apart from the increased readability of your code, named tuples provide a few other quality of life improvements. First and foremost, they allow for default values to be specified via the defaults iterable argument. Secondly, they have the ability to automatically rename duplicate or invalid fields via the rename boolean argument. And, finally, they even provide a convenient option to specify field names as a list or comma/space-separated string.
_Where's the catch?_ you might ask. Well, it seems like there's none! The obvious parallel to dictionaries in terms of syntax doesn't seem to go any further, as named tuple instances do not have per-instance dictionaries, meaning they require as much memory as regular tuples.

From 30-seconds-of-code: named-tuples