sufficient imo, or at least in an acceptable state now

^That was approved by someone.

Leaving this open since I'm not sure if that is sufficient or I need to go further.

Approved by someone.

approved

Python fork

Approximately halves the amount of time needed to generate all the random tests. Leaves test parameters and author solution unmodified. I also removed the redunant [::] on all the inputs passed to the user, strings are immutable in Python.

Fork fixing the slow Python test case generation

Runtime for author solution is down to 6.5s from 11.5s.

Most of the runtime is spent generating the tests, not running the refsol.

The most naive optimisation possible (switching to using random.choices instead of repeated calls to random.choice) halves the runtime of the kata with refsol + authorsol.

Awesome Kata! I popped in back on Codewars after a long while and such performance challenge was exactly what I needed :) Thought it will be done in couple minutes, took a bit longer :D I agree with other comment that it would be good to mention in description performance will be important to pass random tests.

So that's what that meant.. Nice one! :)

The deck is preloaded.

I get how this should work in priniciple, but how is it passing any tests with only this portion of code? Where is it getting the "deck" from ?

Thanks for answer, its really help me understand. Never thought about that solution. Its really great.

How it works ? i dont understand :(

It worked just fine for me. Read this: https://github.com/codewars/codewars.com/wiki/Troubleshooting-your-solution

Running into the same issue. Run the "failed" random tests in a seperate IDE and works fine