TIL: Python Benchmarking

I’ll be the first to admit that I’ve never really cared about performance when writing scripts in the past. My main goal is to accomplsh some piece of work. If it takes 5 seconds longer to do it, but it works, that’s totally alright with me. But what about the cases where those extra 5 seconds aren’t okay? I finally played around with benchmarking today in pythong.

There’s really two ways that I played around with: timing entire portions of my code, and timing individual commands. The first gives me an idea of how long it’s taking to get through my actual code. The later gives a better idea of how many cpu cycles are spent trying to do things one way vs doing them another.

Timing Code Blocks

The way that I went about doing this was pretty simple: grab the time at the start, grab the time at the end, subtract and get your total time. So something simple like this:

1
2
3
4
5
6

from datetime import datetime as DT

startTime = DT.now()

# Do your stuff
print("Function runtime: {}".format(DT.now()-startTime))

This would give you the run time for that specific block of code/function call. Super easy. Already knew about this but had never really used it before.

Timing specific commands

Python has a built in timing tool called timeit that lets you check how long it takes to perform a specific action. This helped greatly in finding what was causing me the most issues with my traffic generator (spoilers: it was using random).

To check how long it takes to run a specific item, from the command line you can run somthing like this:

python -m timeit 'int(1234.2345)'

This will return output showing the performance report:

10000000 loops, best of 3: 0.136 usec per loop

If you need to check somthing using packages, you can use the -s command to use them. So here’s an example showing how I was able to fix performance problems with random:

1
2
3
4
5

python -m timeit -s 'import random' 'random.randint(1, 10)'
1000000 loops, best of 3: 1.2 usec per loop

python -m timeit -s 'import random' 'round(random.random()*10)'
1000000 loops, best of 3: 0.243 usec per loop

With just a simple tweak I’m able to shave off an entire microsecond. Not a lot on its own, but when you’re doing thosands of operations per minute for a long durration, it all adds up.

💚