Double package:numeric-prelude
Double-precision floating point numbers. It is desirable that this
type be at least equal in range and precision to the IEEE
double-precision type.
Only return as much digits as are contained in Double. This will
speedup further computations.