What is a "method" in Python?

Solution 1:

It's a function which is a member of a class:

class C:
    def my_method(self):
        print("I am a C")

c = C()
c.my_method()  # Prints("I am a C")

Simple as that!

(There are also some alternative kinds of method, allowing you to control the relationship between the class and the function. But I'm guessing from your question that you're not asking about that, but rather just the basics.)

Solution 2:

A method is a function that takes a class instance as its first parameter. Methods are members of classes.

class C:
    def method(self, possibly, other, arguments):
        pass # do something here

As you wanted to know what it specifically means in Python, one can distinguish between bound and unbound methods. In Python, all functions (and as such also methods) are objects which can be passed around and "played with". So the difference between unbound and bound methods is:

1) Bound methods

# Create an instance of C and call method()
instance = C()

print instance.method # prints '<bound method C.method of <__main__.C instance at 0x00FC50F8>>'
instance.method(1, 2, 3) # normal method call

f = instance.method
f(1, 2, 3) # method call without using the variable 'instance' explicitly

Bound methods are methods that belong to instances of a class. In this example, instance.method is bound to the instance called instance. Everytime that bound method is called, the instance is passed as first parameter automagically - which is called self by convention.

2) Unbound methods

print C.method # prints '<unbound method C.method>'
instance = C()
C.method(instance, 1, 2, 3) # this call is the same as...
f = C.method
f(instance, 1, 2, 3) # ..this one...

instance.method(1, 2, 3) # and the same as calling the bound method as you would usually do

When you access C.method (the method inside a class instead of inside an instance), you get an unbound method. If you want to call it, you have to pass the instance as first parameter because the method is not bound to any instance.

Knowing that difference, you can make use of functions/methods as objects, like passing methods around. As an example use case, imagine an API that lets you define a callback function, but you want to provide a method as callback function. No problem, just pass self.myCallbackMethod as the callback and it will automatically be called with the instance as first argument. This wouldn't be possible in static languages like C++ (or only with trickery).

Hope you got the point ;) I think that is all you should know about method basics. You could also read more about the classmethod and staticmethod decorators, but that's another topic.

Solution 3:

In Python, a method is a function that is available for a given object because of the object's type.

For example, if you create my_list = [1, 2, 3], the append method can be applied to my_list because it's a Python list: my_list.append(4). All lists have an append method simply because they are lists.

As another example, if you create my_string = 'some lowercase text', the upper method can be applied to my_string simply because it's a Python string: my_string.upper().

Lists don't have an upper method, and strings don't have an append method. Why? Because methods only exist for a particular object if they have been explicitly defined for that type of object, and Python's developers have (so far) decided that those particular methods are not needed for those particular objects.

To call a method, the format is object_name.method_name(), and any arguments to the method are listed within the parentheses. The method implicitly acts on the object being named, and thus some methods don't have any stated arguments since the object itself is the only necessary argument. For example, my_string.upper() doesn't have any listed arguments because the only required argument is the object itself, my_string.

One common point of confusion regards the following:

import math
math.sqrt(81)

Is sqrt a method of the math object? No. This is how you call the sqrt function from the math module. The format being used is module_name.function_name(), instead of object_name.method_name(). In general, the only way to distinguish between the two formats (visually) is to look in the rest of the code and see if the part before the period (math, my_list, my_string) is defined as an object or a module.