@trashy_incel although you're right, the point "2000 is not the twentieth century." is still there.

Read the error message. It tells you exactly what's wrong, and it's in your code, not the tests.

Did you know there's a Troubleshooting Guide in the Docs?

round and abs, this passes random tests:

def geo_mean(nums, arith_mean):
    mean = ... # correct result
    return -mean

Rounding too? Just k!ll me now.

The random generator should just not fscking generate negative numbers. I just fixed this for JS. Generating a completely random mean is also bullsh!t. Anyone fixing this, please have a look at how JS does it. ( Testing should probably be updated for the new test framework and explicitly importing things as modules as well. )

Why do these problems surface after five years of silence ?!?

@Johan - I've raised the issue above; it seems like there are 4 threads for this same discussion so posting here to close out the sub-thread.

Proof would be a generated testcase and an expected return value that are bullsh!t, separately or in combination. Your, or anyone's, solution code has nothing to do with that. Please stop posting solutions.

@Johan I raised an issue myself above.

@joshred83 - the idea is to use data from tests to show that the error is with the kata; if you read my post hopefully it's clear what the "reporting" should look like.

Hey @joshred83 - it seems you had the same thing as @PDrich101 so see my comment to him just a few minutes ago:

https://www.codewars.com/kata/57c6b44f58da9ea6c20003da/discuss#63a60d1d9803f25147a80faf

TL;DR - basically don't use Numpy

I tried so hard to tell you how to post a useful Issue. And you gave us nothing to work with.

• Please follow the Troubleshooting Guide. Carefully.

If it doesn't say "post an Issue", do not post an Issue. If you do post an Issue, include relevant information. Input, actual output, expected output, and language, at least. There is no need to post your code if the kata really is wrong.

Using prod from the math module solved it for me.