Is it worth caching Python's range(start, stop, step)?

Question

In a Python program which runs a for loop over a fixed range many times, e.g.,

while some_clause:
    for i in range(0, 1000)
        pass
    ...

does it make sense to cache range:

r = range(0, 1000)
while some_clause:
    for i in r
        pass
    ...

or will it not add much benefit?

@deceze I appreciate your opinion and doing benchmarks is always a good advice. But if you check answers, there is some story behind how `range` works and its differences in python 2 and 3. So I'd argue and say that this question makes sense. — Anton Dovzhenko, Apr 01 '20 at 07:26
Wait wait wait why are you doing for i in range (100)? How about just ensure $i$ is an integer and check 0<=i<99? Sorry if this is a stupid suggestion — Gareth Ma, Apr 01 '20 at 07:37

score 8 · Accepted Answer · edited Apr 04 '20 at 22:14

It won't, a range call does almost nothing. Only the itering part, which is not optional, has a cost.

Interestingly, caching makes it slower for some reason, in the example below.

My benchmarks:

>>> timeit.timeit("""
for i in range(10000):
    pass""",number=10000)
1.7728144999991855
>>> timeit.timeit("""
for i in r:
    pass""","r=range(10000)",number=10000)
1.80037959999936

And caching it breaks readability, as the Zen of Python states:

Readability counts.

and

Explicit is better than implicit.
Simple is better than complex.

score 1 · Answer 2 · edited Apr 01 '20 at 07:58

1

If you are using python 2.*, range will return a list, and you should usexrange. xrange (2.) or range (3.) are lazy evaluated, which means it actually evaluated the next requested item when you ask for it.

So, no, no need to cache. Instantiate the range where you need it, no need for tricks and magic there, it's already implemented in Python.

edited Apr 01 '20 at 07:58

xkcdjerry

965
4
15

answered Apr 01 '20 at 07:21

Aaron_ab

3,450
3
28
42

score 0 · Answer 3 · answered Apr 01 '20 at 07:24

0

It won't benefit. If you want to enhance your loop activities refer to the comparison of : https://dev.to/duomly/loops-in-python-comparison-and-performance-4f2m

You can have an idea of how can improve things.

answered Apr 01 '20 at 07:24

Nilanka Manoj

3,527
4
17
48

Is it worth caching Python's range(start, stop, step)?

3 Answers3