That's what I was afraid off, so if you match with one element, you can't re-match with the same element later on. Almost like pairing.

The element in the array only can be used once. The questions descrption dont tell it.

That's the whole point of this kata, though.

I think this might be the best solution as it gets around the ugly "not _ is None and..."s by a nice try/catch

test ([3,4],[0,25], False) This solution is not applicaple for many test cases where any sum of squares rules apply to the arrays
also list(map(lambda i:i * i),array1) is more efficient than [i*i for i in array1]

The explanation is badly worded, I recommend the second clause that readily explains what is requested

so 0.0 in the sample test is not 0, 0.0 or float(0)?

integer 0 is actually not "0"

0 == "0"
False

But yeah, those 0.0's are not 0.0's either