Closed.

You should avoid converting number to string many times.

See https://en.wikipedia.org/wiki/Fibonacci_number#Negafibonacci for an explanation of Fibonacci numbers for negative numbers.

And make sure that you compute the power correctly: put parenthesis around the -1, like in (-1) ** ((n.abs)+1) * fib(n.abs) if n < 0

This is expected.

same here, as ex:
fib -38
Test Passed: Value == -39088169
fib -25
Expected: 75025, instead got: -75025
fib -88
Test Passed: Value == -1100087778366101931

I just solved the C# version. Issue seems solved.

Looks like a error in the some test cases : while n is negative - expected result is positive number for some test case
Ex:fib(-6) expected 8

There was an unused var on the backend test. I removed it, could you see if it solved the issue?

I removed the unused var. Might work now.

He defines a Random variable which is never used inside his submit-testcases "private Random random = new Random();" and since CW throws warnings, all these katas with unused variables will not work (or other warnings too - it's not really an internal error, only a compiler warning)... Here the example tests work, because there are only static tests without unused variables;-)...

C# solution the same thing, as below :
Compilation succeeded - 1 warning(s)
{"stdout":"","stderr":"/tmp/csharp11653-23-v5rpbw/fixture.cs(8,20): warning CS0414: The private field `KataSolution.random' is assigned but its value is never used\n","wallTime":0}

Can you explain me why the warning interrupts the compilation? I how i can fix this?