Struct cgmath::Ulps

pub struct Ulps<A, B = A>where
    A: UlpsEq<B> + ?Sized,
    B: ?Sized,{
    pub epsilon: <A as AbsDiffEq<B>>::Epsilon,
    pub max_ulps: u32,
}

The requisite parameters for testing for approximate equality using an ULPs based comparison.

This is not normally used directly, rather via the assert_ulps_{eq|ne}! and ulps_{eq|ne}! macros.

Example

use std::f64;
use approx::Ulps;

Ulps::default().eq(&1.0, &1.0);
Ulps::default().epsilon(f64::EPSILON).eq(&1.0, &1.0);
Ulps::default().max_ulps(4).eq(&1.0, &1.0);
Ulps::default().epsilon(f64::EPSILON).max_ulps(4).eq(&1.0, &1.0);
Ulps::default().max_ulps(4).epsilon(f64::EPSILON).eq(&1.0, &1.0);

Fields

epsilon: <A as AbsDiffEq<B>>::Epsilon

The tolerance to use when testing values that are close together.

max_ulps: u32

The ULPs to tolerate when testing values that are far-apart.

Implementations

impl<A, B> Ulps<A, B>where A: UlpsEq<B> + ?Sized, B: ?Sized,

pub fn epsilon(self, epsilon: <A as AbsDiffEq<B>>::Epsilon) -> Ulps<A, B>

Replace the epsilon value with the one specified.

pub fn max_ulps(self, max_ulps: u32) -> Ulps<A, B>

Replace the max ulps value with the one specified.

pub fn eq(self, lhs: &A, rhs: &B) -> bool

Peform the equality comparison

pub fn ne(self, lhs: &A, rhs: &B) -> bool

Peform the inequality comparison

Trait Implementations

impl<A, B> Default for Ulps<A, B>where A: UlpsEq<B> + ?Sized, B: ?Sized,

fn default() -> Ulps<A, B>

Returns the “default value” for a type.