In Python OOP When Calling Class Variable Should self or Class Name be used?

Question

I am very rusty on OOP. Now my question is, how exactly would I go about calling a class variable?

I know to call an __init__ variable you would do the following:

class HelloWorld:

    def __init__(self):
        self.hello = "Hello"

    def world(self):
        print self.hello + "World"

How do I call a variable not initialized in the class though? Which of the following would be correct?

class HelloWorld:

    hello = "Hello"

    def world(self):
        #Do I use self?
        print self.hello + "World"
        #Do I use the class name?
        print HelloWorld.hello + "World"

I am just confused on whether I should be using self or the class name. Most examples I see are using self, but the one on Tutorials Point uses the class's name. Please elaborate, thank you!

Yes thank you! Kinda missed that

Michael Jones
– Michael Jones

2016-07-08 21:35:55 +00:00
Commented Jul 8, 2016 at 21:35 — Michael Jones
– Michael Jones, Commented Jul 8, 2016 at 21:35

Alden · Accepted Answer · 2016-07-08 21:56:16Z

6

HelloWorld.hello does not require an instance of HelloWorld to be used. This might be used in a case where a value should be common among all instances of HelloWorld. self.hello belongs to the instance and could be different across instances of HelloWorld.

The code

class HelloWorld:

    hello = "Hello"

    def __init__(self, text):
        self.hello = text

    def world(self):
        #Do I use self?
        print self.hello + "World"
        #Do I use the class name?
        print HelloWorld.hello + "World"

print HelloWorld.hello
print HelloWorld('Goodbye').hello

MyInstance = HelloWorld('Goodbye')
MyInstance.world()

outputs

Hello
Goodbye
GoodbyeWorld
HelloWorld

edited Jul 8, 2016 at 21:56

answered Jul 8, 2016 at 21:36

Alden

2,2691 gold badge15 silver badges22 bronze badges

Sign up to request clarification or add additional context in comments.

7 Comments

Michael Jones Over a year ago

Ok so basically, whenever referring to a variable defined outside of __init__ use classname.variablename to reference it?

Alden Over a year ago

A member of a class does not need to be defined in __init__. The real answer is to understand if the variable was defined by the class or the instance. If you find yourself needing to create a HelloWorld object to access the variable, which is the case with self.hello then don't use classname.variablename. Otherwise you can use classname.variablename.

Bakuriu Over a year ago

@MichaelJones No, it depends on the behaviour you want to achieve. As I and mgilson show there are cases where you want to make that value overridable by instances/derived classes, and other times not. So it depends.

Michael Jones Over a year ago

Ok, I guess one of the biggest issues is I am not understanding this line HelloWorld('Goodbye').hello What does that do?

Alden Over a year ago

The line HelloWorld('Goodbye').hello creates a HelloWorld object and then accesses the hello member belonging to that object. It is the same as object = HelloWorld('Goodbye') and then print object.hello

|

mgilson · Accepted Answer · 2016-07-08 21:32:36Z

1

Generally, you'll want to use self here. This allows for a few nice "tricks" to be played. e.g. You can override hello in a subclass:

class GreetWorld(HelloWorld):
    hello = "Greetings"

Or you can re-set the attribute on a per-instance basis:

hw = HelloWorld()
hw.hello = 'Hey, How are you'

answered Jul 8, 2016 at 21:32

mgilson

312k70 gold badges656 silver badges722 bronze badges

Comments

Kaushal Dontula · Accepted Answer · 2016-07-08 21:42:04Z

0

Both of them will work. When you have self passed in to the method self == HelloWorld

class HelloWorld:
    hello = "hello"
    def world(self):
        print self.hello + "World"

But when self is not passed in to method i.e, static method then use

class HelloWorld:
    hello = "hello"
    @staticmethod
    def world():
        print HelloWorld.hello

answered Jul 8, 2016 at 21:42

Kaushal Dontula

2793 silver badges6 bronze badges

Comments

Bakuriu · Accepted Answer · 2016-07-08 21:48:23Z

You can use both, the results are well defined and in that case they will do the same.

Note however that this changes when you have derived classes:

class Base:
    hello = 'A'

    def greet(self):
        print(self.hello)

    def greet2(self):
        print(Base.hello)


class Derived(Base):
    hello = 'B'

d = Derived()
d.greet()   # prints B
d.greet2()  # prints A

Moreover in some cases you could have instances that use that attribute to override the class value:

class Base:
    hello = 'A'
    def __init__(self, val=None):
        if val is not None:
            self.hello = 'B'
    def greet(self):
        print(self.hello)


A = Base()
B = Base()
A.greet()    # A
B.greet('B') # B
A.greet()    # *still* A

So if you want to be sure that the value used is precisely the one of that class use the explicit class name, otherwise use self.

Collectives™ on Stack Overflow

In Python OOP When Calling Class Variable Should self or Class Name be used?

4 Answers 4

7 Comments

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

7 Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related