serialPython.py

import math
import numpy as np
from timebudget import timebudget

iterations_count = round(1e7)

def complex_operation(input_index):
    print("Complex operation. Input index: {:2d}".format(input_index))

    [math.exp(i) * math.sinh(i) for i in [1] * iterations_count]

@timebudget
def run_complex_operations(operation, input):
    for i in input:
        operation(i)

input = range(10)
run_complex_operations(complex_operation, input)

While that would likely work, I coded it the original way so that the rest of the tutorial can easily build off this code snippet.

Sounds good! My previous suggestion makes little sense.

Which Mac processor are you using there btw?

Not M1. Honestly would like to do a more extensive set of testing when i have time with different processors, cores, and more as well as include more technologies: https://twitter.com/rasbt/status/1438557627169099777

Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz, 2592 Mhz, 6 Core(s), 12 Logical Processor(s), this is the current Processor I'm using maybe I could run your scripts and add results?

I have a better idea! Why don't you make another blog post with blocking or non blocking IO operations to have a more comprehensive set of benchmarks? I say this so that you can make your own opinion of each library and share that :)

Rather than adding to current blog, I will write another blog in the future with some more libraries as it definitely seems there are a ton of other libraries to include (joblib, dask, mpire to name a few) with a lot of different use cases, integrations, syntaxes (following each libraries best practices is important) and data sizes. Honestly, I really want to vary data sizes and see how well each library performs. Problem is making this sort of blog may take a bit of time to make, but should be fun :)

I have a better idea! Why don't you make another blog post with blocking or non blocking IO operations to have a more comprehensive set of benchmarks? I say this so that you can make your own opinion of each library and share that :)

Rather than adding to current blog, I will write another blog in the future with some more libraries as it definitely seems there are a ton of other libraries to include (joblib, dask, mpire to name a few) with a lot of different use cases, integrations, syntaxes (following each libraries best practices is important) and data sizes. Honestly, I really want to vary data sizes and see how well each library performs. Problem is making this sort of blog may take a bit of time to make, but should be fun :)

I am not realy sure of what operations would classify as blocking and non- io blocking ?

Oh dont worry about it then! It is not always easy to find that out unfortunately :(

Just suggesting you write. Every benchmark isn't perfect so I was suggesting you try your own :)

Yea , I will try and see if I get time :(