__slots__ for optimizing classes

"Let's talk about how to optimize the memory usage and the attribute lookup time of our Python classes. How are class attributes stored by default? Here we have a class called Point in a points.py file: And here's an instance of this Point class: Normally, classes store their attributes in a dictionary called __dict__. We have a class here where every instance has x, y, and z attributes."

"To use __slots__, we need to define a __slots__ attribute on our class that points to a tuple of strings that represent valid attributes names for each instance of our class. Let's add __slots__ to our Point class: This instance of our Point class has an x attribute (just as before): We can change the value of this attribute (just as before): But if we try to make a new attribute (an attribute that isn't x, y, or z) we'll get an AttributeError:"

Classes store instance attributes in a per-instance dictionary named __dict__, allowing arbitrary attributes to be added and producing new key-value pairs. Defining __slots__ as a tuple of valid attribute names replaces per-instance __dict__ storage with fixed-slot storage for those attribute names. Instances of classes with __slots__ allow only the declared attributes and raise AttributeError on attempts to assign undeclared ones. __slots__ prevents dynamic attribute expansion and uses a compact, fixed layout similar to a list of slots. __slots__ is typically used to save memory and can improve attribute lookup time.

Read at Pythonmorsels