That's not a type. A NaN is a floating point number (of type float). I used numpy because that's the easiest way to get a NaN.

This is part of the floating point standard.

This was never about None, which is a completely different thing.