Not only does this kata not do a thorough job of testing, I suspect the solution tested against to be wrong. I just ran all of the other verified solutions against this:

convert(0.909, '09')  # expect 10/11
convert(0.90, '90')  # expect 10/11

and none of them give the right result for the second line. To be fair, one works correctly when passing in 0.9090 and another with 0.909090, but partial credit isn't quite the same thing as correct, is it?

Although my inital solution accidentally passes on this test, there are similar cases where it failed. I know how to patch my code, if only there were a way to edit a submission without resubmitting a new one. That' just brilliant. I've resubmitted my patched code and forked too. Not sure the difference.

I haven't been on codewars long, but I feel like any kata that requires importing modules shouldn't be 8 or 7 kyu. Or there should be a mention of the module to import in the description.