Crate umath

library providing a fast float wrapper.

let mut f = FFloat::new(5.0);
f *= 7.0;
assert_eq!(*f, 35.0);

Modules

• generic_float
  provides generic float traits. this is the best way to make a function possibly take a FFloat.

Structs

• FFloat
  Float wrapper that uses ffast-math. This float also implements Ord, Hash, and Eq, as it is not allowed to be NAN.

Traits

• Float
  Generic float trait, implemented by {FFloat, f32, f64}. Takes a “base” argument, intended to be set to {f32, f64}. The main purpose of this is to be taken (generically) by optionally fast functions.

Type Aliases

• FF32
  Type alias for FFloat<f32>. (fast float 32 bits)
• FF64
  Type alias for FFloat<f64>. (fast float 64 bits)