Ah! just what I was wondering. Thanks for answering. The description should state this.

At least in Python, the test cases for 121 and 180 are definitely wrong:

test.assert_equals(get_best_combination(121), "201")
test.assert_equals(get_best_combination(180), "300")

The correct solution must be "21" for 121 and "3" for 180.

see discussion below, that I closed too early

i think it's related to this : scipy.ndimage.rotate by180 degrees is not equivalent to two 90 degree rotations

Codewars' scipy rotate version does not have this patch

it also works on my local interpreter. i've inspected the source code for scipy.rotate() on both codewars and my machine and there are some changes. sorry for closing the issue too early

It also fails in replit.com for the same sizes. I think the problem is with rotate from scipy, why don't use another thing? The other import already has some method you can use and works.

We can't see your code, read this: https://docs.codewars.com/training/troubleshooting/#post-discourse

'it works on my PC' and 'i dont know where the error comes from' make for a bad issue report. there are thousands of Python completions on this kata, so such an error in the tests is unlikely. Please show us some code and provide more details ;-)

The descrition should advise that if the given number is prime, then itself is the nearest prime.

In the description, it´s unclear from where to retrive data (table schema - numbers), and also the output schema (in the graphic seems that column should be integer and it is n).

The description is unnecessarily tricky and misleading. To avoid this error you should retrieve both company_id from products and company_name from companies.

I agree. I don´t understand why this suggestion is still unresolved ? Only the kata´s sensei can do it ?

The print sentence was for testing purposes in my PC :-)