[python] What is the purpose of self?



8 Answers

Let’s take a simple vector class:

class Vector:
    def __init__(self, x, y):
        self.x = x
        self.y = y

We want to have a method which calculates the length. What would it look like if we wanted to define it inside the class?

    def length(self):
        return math.sqrt(self.x ** 2 + self.y ** 2)

What should it look like when we were to define it as a global method/function?

def length_global(vector):
    return math.sqrt(vector.x ** 2 + vector.y ** 2)

So the whole structure stays the same. How can me make use of this? If we assume for a moment that we hadn’t written a length method for our Vector class, we could do this:

Vector.length_new = length_global
v = Vector(3, 4)
print(v.length_new()) # 5.0

This works because the first parameter of length_global, can be re-used as the self parameter in length_new. This would not be possible without an explicit self.


Another way of understanding the need for the explicit self is to see where Python adds some syntactical sugar. When you keep in mind, that basically, a call like

v_instance.length()

is internally transformed to

Vector.length(v_instance)

it is easy to see where the self fits in. You don't actually write instance methods in Python; what you write is class methods which must take an instance as a first parameter. And therefore, you’ll have to place the instance parameter somewhere explicitly.

Question

What is the purpose of the self word in Python? I understand it refers to the specific object created from that class, but I can't see why it explicitly needs to be added to every function as a parameter. To illustrate, in Ruby I can do this:

class myClass
    def myFunc(name)
        @name = name
    end
end

Which I understand, quite easily. However in Python I need to include self:

class myClass:
    def myFunc(self, name):
        self.name = name

Can anyone talk me through this? It is not something I've come across in my (admittedly limited) experience.




I'm surprised nobody has brought up Lua. Lua also uses the 'self' variable however it can be omitted but still used. C++ does the same with 'this'. I don't see any reason to have to declare 'self' in each function but you should still be able to use it just like you can with lua and C++. For a language that prides itself on being brief it's odd that it requires you to declare the self variable.




Python is not a language built for Object Oriented Programming unlike Java or C++.

When calling a static method in Python, one simply writes a method with regular arguments inside it.

class Animal():
    def staticMethod():
        print "This is a static method"

However, an object method, which requires you to make a variable, which is an Animal, in this case, needs the self argument

class Animal():
    def objectMethod(self):
        print "This is an object method which needs an instance of a class"

The self method is also used to refer to a variable field within the class.

class Animal():
    #animalName made in constructor
    def Animal(self):
        self.animalName = "";


    def getAnimalName(self):
        return self.animalName

In this case, self is referring to the animalName variable of the entire class. REMEMBER: If you have a variable within a method, self will not work. That variable is simply existent only while that method is running. For defining fields (the variables of the entire class), you have to define them OUTSIDE the class methods.

If you don't understand a single word of what I am saying, then Google "Object Oriented Programming." Once you understand this, you won't even need to ask that question :).




I will demonstrate with code that does not use classes:

def state_init(state):
    state['field'] = 'init'

def state_add(state, x):
    state['field'] += x

def state_mult(state, x):
    state['field'] *= x

def state_getField(state):
    return state['field']

myself = {}
state_init(myself)
state_add(myself, 'added')
state_mult(myself, 2)

print( state_getField(myself) )
#--> 'initaddedinitadded'

Classes are just a way to avoid passing in this "state" thing all the time (and other nice things like initializing, class composition, the rarely-needed metaclasses, and supporting custom methods to override operators).

Now let's demonstrate the above code using the built-in python class machinery, to show how it's basically the same thing.

class State(object):
    def __init__(self):
        self.field = 'init'
    def add(self, x):
        self.field += x
    def mult(self, x):
        self.field *= x

s = State()
s.add('added')    # self is implicitly passed in
s.mult(2)         # self is implicitly passed in
print( s.field )

[migrated my answer from duplicate closed question]




The following excerpts are from the Python documentation about self:

As in Modula-3, there are no shorthands [in Python] for referencing the object’s members from its methods: the method function is declared with an explicit first argument representing the object, which is provided implicitly by the call.

Often, the first argument of a method is called self. This is nothing more than a convention: the name self has absolutely no special meaning to Python. Note, however, that by not following the convention your code may be less readable to other Python programmers, and it is also conceivable that a class browser program might be written that relies upon such a convention.

For more information, see the Python documentation tutorial on classes.




self is an object reference to the object itself, therefore, they are same. Python methods are not called in the context of the object itself. self in Python may be used to deal with custom object models or something.




Is because by the way python is designed the alternatives would hardly work. Python is designed to allow methods or functions to be defined in a context where both implicit this (a-la Java/C++) or explicit @ (a-la ruby) wouldn't work. Let's have an example with the explicit approach with python conventions:

def fubar(x):
    self.x = x

class C:
    frob = fubar

Now the fubar function wouldn't work since it would assume that self is a global variable (and in frob as well). The alternative would be to execute method's with a replaced global scope (where self is the object).

The implicit approach would be

def fubar(x)
    myX = x

class C:
    frob = fubar

This would mean that myX would be interpreted as a local variable in fubar (and in frob as well). The alternative here would be to execute methods with a replaced local scope which is retained between calls, but that would remove the posibility of method local variables.

However the current situation works out well:

 def fubar(self, x)
     self.x = x

 class C:
     frob = fubar

here when called as a method frob will receive the object on which it's called via the self parameter, and fubar can still be called with an object as parameter and work the same (it is the same as C.frob I think).




When objects are instantiated, the object itself is passed into the self parameter.

Because of this, the object’s data is bound to the object. Below is an example of how you might like to visualize what each object’s data might look. Notice how ‘self’ is replaced with the objects name. I'm not saying this example diagram below is wholly accurate but it hopefully with serve a purpose in visualizing the use of self.

The Object is passed into the self parameter so that the object can keep hold of its own data.

Although this may not be wholly accurate, think of the process of instantiating an object like this: When an object is made it uses the class as a template for its own data and methods. Without passing it's own name into the self parameter, the attributes and methods in the class would remain as a general template and would not be referenced to (belong to) the object. So by passing the object's name into the self parameter it means that if 100 objects are instantiated from the one class, they can all keep track of their own data and methods.

See the illustration below:




it's an explicit reference to the class instance object.




Related