2

Let's take the following Python code:

a = 3
print(a)

As far as I know, a is a reference to an object of class int — correct me if I am wrong. Diagrammatically, it should look something like this, as far as I know.

    The object of 'int' class
       containing value 3

        |---|
a ----> | 3 |
        |---|

Does the 'int' object in memory only contain the value 3 or it contains some space for other variables as well?

Our professor told us that it looks something like below:

   The object of class 'int'
        |--------|
        |   ___  |
a ----> |   |3|  |
        |   ---  |
        |--------|

So, is the remaining space utilized by some other variables?

Improve this question
1
  • 1
    Using sys.getsizeof() to check the size 3 has a size of 28 bytes. So yes, there is other information contained within the int object. Commented May 4, 2020 at 21:04

2 Answers 2

Reset to default
1

Any object in Python inherits from the object class, which means that it's treated as any Python object, not as a primitive value as you'd expect from a language like C. This is what people mean when they say that in Python, everything is an object.

For example, the int class has a method __str__ that dictates what its string representation looks like:

>>> a = 3
>>> a.__str__()
'3'

You can use dir(a) to get a list of every method associated with a.

So to answer your question, yes - the int object does not only allocates the space needed for the primitive value itself, but also for the whole Python object associated with it.

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

12 Comments

All instances consist of a pointer to their type and a reference count. There is some additional bookkeeping info whose details I don't recall, as well as space required for the actual value. 0, for instance, is 24 bytes on a 64-bit machine. Numbers with an absolute value under 2**30 require an additional 4 bytes, and the size grows in 4-byte increments as the magnitude of the number requires.
@NirajRaut By design, the primitive value is "kept away" from the Python user. From a C standpoint, all objects in Python are structs named PyObject whose value you can't (normally) access directly. So a is really a reference to the PyObject, not its value. This link may help clarify this further: realpython.com/pointers-in-python.
(And most of this is specific to the CPython implementation of the Python language. Python itself doesn't expose how an object is actually structured in memory.)
@NirajRaut Yes, and note, Python doesn't used fixed-size primitive int types, but rather, arbitrary length int objects. Essentially, in the PyObject struct there is a pointer to a primitive array to hold the raw bytes necessary to encode integers of arbitrary size. Here is a pretty decent blog post: codementor.io/@arpitbhayani/…
@chepner: No extra bookkeeping info is needed for integers; what you listed is everything. On a typical 64-bit machine an int object has an 8-byte type pointer, an 8-byte reference count, an 8-byte combination limb count + sign, and 4 bytes per 30-bit limb, and that's all. For objects that need to be tracked by the cyclic garbage collector (e.g., lists) there is additional bookkeeping needed, but an int can never be involved in a reference cycle, so that extra overhead doesn't apply.
|
0

I am not sure what you are asking but you can use type() to check what class an object or variable is:

type(a)

OUTPUT:

<class 'int'>
Improve this answer

1 Comment

This doesn't answer my question.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.