Question about primitive variables.

[Nirelg]

Hi all,

I have started to learn what primitive variables are vs Referenced ones.

The problem is, that once I have read that python is fully referenced, object orientated language, which have no primitive data types, I now read the following online:

"In Python there are four primitive variable types, which are listed below.
Integers
Floats
Strings
Characters"

So, can anyone please tell my what the correct answer is?
Does python has primitive variable types?
or even Strings and integers are referenced variables?

Thanks in advance,
Nirel.

[micseydel]

What's the source? What different, other than semantics, are they claiming? The difference can be important in other languages.

I wouldn't say Python has primitives. Even ints are referenced, and there's even an optimization for small ints where this can be observed in CPython (the standard Python implementation)

Output:
>>> 2**2 is 2**2
True
>>> 2**32 is 2**32
False

[ichabod801]

In Python the distinction is generally made between mutable and immutable objects. Mutable objects can be changed in place, and are stored by reference. Immutable objects must be replaced, and are stored by value. (There are exceptions to this, but they're not that big a deal). So, the immutable types (what you would call primitive) include ints, floats, complex, strs, tuples, bytes, and frozensets.

[Nirelg]

What's the source? What different, other than semantics, are they claiming? The difference can be important in other languages.

I wouldn't say Python has primitives. Even ints are referenced, and there's even an optimization for small ints where this can be observed in CPython (the standard Python implementation)

Output:
>>> 2**2 is 2**2
True
>>> 2**32 is 2**32
False

OK, firstly I must say your example gave me a new question lol.
why "2**32 is 2**32" gives us output of "False"?
I mean, it is not only the same value but it is also the same type, shouldn't it be True? like the first example?

regarding the primitives I can give you for example someone which answer a question online:
http://stackoverflow.com/questions/63916...-primitive
stating "there are no primitive types in Python"

on the other hand there is this link:
https://en.wikiversity.org/wiki/Python_C...data_types
where they stated:
"In Python there are four primitive variable types, which are listed below.
Integers
Floats
Strings
Characters"

which made me not understand why some say this way and some the other way,
but by your answer and ichabod801 answer I guess they did a mistake on this site?

In Python the distinction is generally made between mutable and immutable objects. Mutable objects can be changed in place, and are stored by reference. Immutable objects must be replaced, and are stored by value. (There are exceptions to this, but they're not that big a deal). So, the immutable types (what you would call primitive) include ints, floats, complex, strs, tuples, bytes, and frozensets.

OK, so I think I understand what you are saying.
You are basically saying that everything in python is an object, therefore we distinct between mutable and immutable objects.

[wavic]

I think you can add the booleans to the list too. I don't call this things primitives or something else. I just checked out in translate.google.com what primitives mean. I don't care about the term. I just use them.

[ichabod801]

I think you can add the booleans to the list too.

I thought about that, because they are a subclass of int. However, they are singletons, which seems to put them in another category.

@Nirelg, the term 'singleton' relates to your question about the is operator. The is operator check the id() of a value, which is where it is stored in memory. So two ints with the same value may not 'is' the same because they are stored in different locations. Singleton means there is only ever one instance of True or False. They are always stored in the same location. None is another singleton.

[wavic]

This illustrate it.

In [1]: a = 42

In [2]: b = 42

In [3]: id(a)
Out[3]: 10915680

In [4]: id(b)
Out[4]: 10915680

[Nirelg]

I think you can add the booleans to the list too.

I thought about that, because they are a subclass of int. However, they are singletons, which seems to put them in another category.

@Nirelg, the term 'singleton' relates to your question about the is operator. The is operator check the id() of a value, which is where it is stored in memory. So two ints with the same value may not 'is' the same because they are stored in different locations. Singleton means there is only ever one instance of True or False. They are always stored in the same location. None is another singleton.

Ohh, now I got it.
But why would they have a different id?
I used the following code:

print type(2**11)
print type(2**2)
print id(2**11)
print id(2**11)
print id(2**2)
print id(2**2)

and the output is:

Output:
<type 'int'>
<type 'int'>
46064336
46064324
45314908
45314908

They are both int after all, so why when I write twice 2**2 it gives me the same id yet when I do the same for 2**11 it gives them different ids?

[snippsat]

They are both int after all, so why when I write twice 2**2 it gives me the same id yet when I do the same for 2**11 it gives them different ids?

Integers between -5 and 256 are cached in Python.
This optimization strategy makes sense because small integers pop up all over the place,
and given that each integer takes 24 bytes, it saves a lot of memory for a typical program.

So id() show memory address.

>>> help(id)
Help on built-in function id in module builtins:

id(...)
    id(object) -> integer
    
    Return the identity of an object.  This is guaranteed to be unique among
    simultaneously existing objects.  (Hint: it's the object's memory address.)

>>> id(10)
491765920
>>> id(10)
491765920
>>> id(10)
491765920
>>> id(100)
491767360
>>> id(100)
491767360
>>> id(100)
491767360
>>> # Now over 256
>>> id(1000)
58284576
>>> id(1000)
51896640

[Ofnuts]

The canonical video(*) about this: https://www.youtube.com/watch?v=_AEJHKGk9ns

(*) On the whole, I deeply despise YouTube tutorial videos. But this one is too good to pass.