Practical Programming, Third Edition

Memory Addresses: How Python Keeps Track of Values

Back in Values, Variables, and Computer Memory, you learned that Python keeps track of each value in a separate object and that each object has a memory address. You can discover the actual memory address of an object using built-in function id:

>>> help(id)

Help on built-in function id in module builtins:

id(obj, /)

Return the identity of an object.

This is guaranteed to be unique among simultaneously existing objects.

(CPython uses the object's memory address.)

How cool is that? Let’s try it:

>>> id(-9)

4301189552

>>> id(23.1)

4298223160

>>> shoe_size = 8.5

>>> id(shoe_size)

4298223112

>>> fahrenheit = 77.7

>>> id(fahrenheit)

4298223064

The addresses you get will probably be different from what’s listed here since values get stored wherever there happens to be free space. Function objects also have memory addresses:

>>> id(abs)

4297868712

>>> id(round)

4297871160

A cache is a collection of data. Because small integers—up to about 250 or so, depending on the version of Python you’re using—are so common, Python creates those objects as it starts up and reuses the same objects whenever it can. This speeds up operations involving these values. The function id reveals this:

	>>> i = 3
	>>> j = 3
	>>> k = 4 - 1
	>>> id(i)
	4296861792
	>>> id(j)
	4296861792
	>>> id(k)
	4296861792

What that means is that variables i, j, and k refer to the exact same object. This is called aliasing.

Larger integers and all floating-point values aren’t necessarily cached:

	>>> i = 30000000000
	>>> j = 30000000000
	>>> id(i)
	4301190928
	>>> id(j)
	4302234864
	>>> f = 0.0
	>>> g = 0.0
	>>> id(f)
	4298223040
	>>> id(g)
	4298223016

Python decides for itself when to cache a value. The only reason you need to be aware of it is so that you aren’t surprised when it happens; the output of your program is not affected by when Python decides to cache.