It should be 15/5，18/6，33/11，36/12and 69/13.The index of sequence anOverAverage(n) start with 1. But the Araay in code start with 0（in most programming languages）

i copied over the C tests suite, with better messages, to NASM. the NASM tests suite was comparing floating-point numbers with strict equality.

I could not find any solution actually doing this because of built-in copy or const ref semantics but added a test for it anyways:
https://www.codewars.com/kumite/67346c955b6178ce3d28a71e?sel=67346c955b6178ce3d28a71e

your issue report is unclear. this behavior is normal and is the same across most languages on Codewars: the fixed and random tests are each in their own tests group / it block and are run separately.

duplicate suggestion

kak dela drug

That's on me. First I approved the transaltion with a bug, then I spotted the bug, then I fixed it in full tests, and forgot about sample tests.

Now the sample tests should be fixed too (remember to reset the trainer to see new sample tests).

Reason for the bug: C sample test's tester function should replace tolerance 1e10 with 1e-10

It's part of the kata to make that optional 3rd parameter working. While I might agree that it is hard to parse this task from the description this error message is not a kata issue but intended.

Appears to be resolved?

Fixed

.

fixed