__init__() initializes an already-created Python instance by assigning attributes from the arguments passed to the class call. When Python evaluates Person('Ada'), __new__ creates the object, then __init__ receives it as self before the instance is returned. section 9.2 section 9.5
Use __init__() to:
- Create instance attributes such as self.name or self.flavor. section 9.2
- Accept required arguments or provide defaults when constructing instances. section 9.5
- Initialize shared state in a parent class and extend it in subclasses. section 9.15 section 9.27
Initialize and extend an instance
class Person: def __init__(self, name): self.name = name class Employee(Person): def __init__(self, name, id_number): super().__init__(name) self.id_number = id_number e = Employee('Ada', 1) The subclass must initialize its own attributes and does not automatically execute the parent’s __init__(). Calling super().__init__(name) reuses the parent initialization without naming the parent class directly. section 9.15 section 9.27
class Employee(Person)
def __init__(self, name, id_number):
super().__init__(name) #1
self.id_number = id_number
Compare subclass initialization with super().__init__ and see where each attribute is assigned. section 9.15
The first parameter is conventionally named self; Python passes the instance there, but self is not a reserved word. The method is commonly called a constructor, although __new__ performs object creation and __init__ performs initialization. section 9.2 section 9.5
