Why time.time() gives 0.0?

Question

I have a python program defined by a function myFunc(m,n)

Basically, the function contains two for loops.

def myFunc(m, n) : 
    for i in range(m) : 
        for j in range(n) : 
            # do it ...
    return

I would like to calculate the time elapsed for m,n=20,20

I do it like this:

start_time = time.time()
b = myFunc(m,n)
elapsed_time = time.time() - start_time
print elapsed_time

The result is 0.0 almost always. Why?

20 loops are too short. Try to make it artificially large. – stanleyli Oct 12 '15 at 17:06 — stanleyli, Oct 12 '15 at 17:06
It is rarely `0.0009` and almost always `0.0`. – Kira Oct 12 '15 at 17:08 — Kira, Oct 12 '15 at 17:08

score 4 · Answer 1 · edited Oct 12 '15 at 17:27

4

Because time.time() - time.time() will be 0.0 when myFunc doesn't take very long.

edited Oct 12 '15 at 17:27

cdonts

9,304
4
46
72

answered Oct 12 '15 at 17:04

Chad S.

6,252
15
25

score 3 · Accepted Answer · edited May 23 '17 at 11:52

3

Per the time.time docs,

even though the time is always returned as a floating point number, not all systems provide time with a better precision than 1 second.

So if myFunc takes less than a second and you are using such an OS, then the difference between two calls of time.time could be 0.0.

You could use timeit.default_timer instead of time.time. timeit.default_timer will choose the best timer (time.time, time.clock or time.perf_counter) for your system.

Also note that the timeit module offers more accurate ways to test the performance of code snippets.

edited May 23 '17 at 11:52

Community

1
1

answered Oct 12 '15 at 17:19

unutbu

842,883
184
1,785
1,677

Thank you. I appreciate it. – Kira Oct 12 '15 at 17:26

Why time.time() gives 0.0?

2 Answers2

Linked