Closed markuswntr closed 8 months ago
Real
refines FloatingPoint
which refines SignedNumeric
with the constraint where Magnitude == Self
.
See the declaration of FloatingPoint in the standard library.
Thanks so much for the feedback!
I did not know this is a constraint on FloatingPoint
– closing this PR then.
Currently, the type of
Complex.magnitude
is equal to the underlyingReal
type of the complex number. This is fine for Float[16|32|64] et al, but I believe incorrect, sinceReal
does not necessary requireMagnitude
to always be of the same type.