Last item does not pass in Python either, probably due to rounding error.

Got it, thanks a lot! And this is why the problem were starting for the random items: before that, you were checking my solution against a pre-defined solution. When the items are random, you do need to check my solution against a solution generated by your script, which is why it failed.

In that case, I should have created a deep copy of the list, and ran my code on this particular list. I do agree that tuples are a better option, as they are immutables. In that case, the Test cases should be changed to reflect this :).

Very good kata otherwise, and thanks a lot for your reactivity!

There seems to be an issue with the validation of the first random passes (copying the exact same configuration of items in a Test case returns a Pass for my code, so I definitely think the validation is to blame here). Furthermore, the suggestion for the validation is an empty list (e.g. "[0, 8, 0, 0, 0, 0, 0] should equal []"), which further hints at an error.