Floating-Point Comparison

Utilities for comparing floating-point values with tolerance. More...

Functions
CHECKASM_API int	checkasm_float_near_ulp (float a, float b, unsigned max_ulp)
	Compare floats using ULP (Units in Last Place) tolerance.
CHECKASM_API int	checkasm_float_near_abs_eps (float a, float b, float eps)
	Compare floats using absolute epsilon tolerance.
CHECKASM_API int	checkasm_float_near_abs_eps_ulp (float a, float b, float eps, unsigned max_ulp)
	Compare floats using both epsilon and ULP tolerances.
CHECKASM_API int	checkasm_float_near_ulp_array (const float *a, const float *b, unsigned max_ulp, int len)
	Compare float arrays using ULP tolerance.
CHECKASM_API int	checkasm_float_near_abs_eps_array (const float *a, const float *b, float eps, int len)
	Compare float arrays using absolute epsilon tolerance.
CHECKASM_API int	checkasm_float_near_abs_eps_array_ulp (const float *a, const float *b, float eps, unsigned max_ulp, int len)
	Compare float arrays using both epsilon and ULP tolerances.
CHECKASM_API int	checkasm_double_near_abs_eps (double a, double b, double eps)
	Compare doubles using absolute epsilon tolerance.
CHECKASM_API int	checkasm_double_near_abs_eps_array (const double *a, const double *b, double eps, unsigned len)
	Compare double arrays using absolute epsilon tolerance.

Detailed Description

Utilities for comparing floating-point values with tolerance.

These functions compare floating-point values allowing for acceptable differences due to rounding, precision loss, or different computation orders.

Function Documentation

checkasm_double_near_abs_eps()

CHECKASM_API int checkasm_double_near_abs_eps	(	double	a,
		double	b,
		double	eps )

Compare doubles using absolute epsilon tolerance.

Parameters

[in]	a	First value
[in]	b	Second value
[in]	eps	Maximum acceptable absolute difference

Returns

Non-zero if |a-b| <= eps, 0 otherwise

checkasm_double_near_abs_eps_array()

CHECKASM_API int checkasm_double_near_abs_eps_array	(	const double *	a,
		const double *	b,
		double	eps,
		unsigned	len )

Compare double arrays using absolute epsilon tolerance.

Parameters

[in]	a	First array
[in]	b	Second array
[in]	eps	Maximum acceptable absolute difference per element
[in]	len	Number of elements to compare

Returns

Non-zero if all elements are within tolerance, 0 otherwise

checkasm_float_near_abs_eps()

CHECKASM_API int checkasm_float_near_abs_eps	(	float	a,
		float	b,
		float	eps )

Compare floats using absolute epsilon tolerance.

Parameters

[in]	a	First value
[in]	b	Second value
[in]	eps	Maximum acceptable absolute difference

Returns

Non-zero if |a-b| < eps, 0 otherwise

checkasm_float_near_abs_eps_array()

CHECKASM_API int checkasm_float_near_abs_eps_array	(	const float *	a,
		const float *	b,
		float	eps,
		int	len )

Compare float arrays using absolute epsilon tolerance.

Parameters

[in]	a	First array
[in]	b	Second array
[in]	eps	Maximum acceptable absolute difference per element
[in]	len	Number of elements to compare

Returns

Non-zero if all elements are within tolerance, 0 otherwise

checkasm_float_near_abs_eps_array_ulp()

CHECKASM_API int checkasm_float_near_abs_eps_array_ulp	(	const float *	a,
		const float *	b,
		float	eps,
		unsigned	max_ulp,
		int	len )

Compare float arrays using both epsilon and ULP tolerances.

Parameters

[in]	a	First array
[in]	b	Second array
[in]	eps	Maximum acceptable absolute difference per element
[in]	max_ulp	Maximum acceptable ULP distance
[in]	len	Number of elements to compare

Returns

Non-zero if all elements are within tolerance, 0 otherwise

checkasm_float_near_abs_eps_ulp()

CHECKASM_API int checkasm_float_near_abs_eps_ulp	(	float	a,
		float	b,
		float	eps,
		unsigned	max_ulp )

Compare floats using both epsilon and ULP tolerances.

Parameters

[in]	a	First value
[in]	b	Second value
[in]	eps	Maximum acceptable absolute difference
[in]	max_ulp	Maximum acceptable ULP distance

Returns

Non-zero if within either tolerance, 0 otherwise

checkasm_float_near_ulp()

CHECKASM_API int checkasm_float_near_ulp	(	float	a,
		float	b,
		unsigned	max_ulp )

Compare floats using ULP (Units in Last Place) tolerance.

Parameters

[in]	a	First value
[in]	b	Second value
[in]	max_ulp	Maximum acceptable ULP distance

Returns

Non-zero if values are within tolerance, 0 otherwise

checkasm_float_near_ulp_array()

CHECKASM_API int checkasm_float_near_ulp_array	(	const float *	a,
		const float *	b,
		unsigned	max_ulp,
		int	len )

Compare float arrays using ULP tolerance.

Parameters

[in]	a	First array
[in]	b	Second array
[in]	max_ulp	Maximum acceptable ULP distance
[in]	len	Number of elements to compare

Returns

Non-zero if all elements are within tolerance, 0 otherwise