vtkMathUtilities Namespace Reference

Functions
template<class A>
bool	FuzzyCompare (A a, A b, A epsilon=std::numeric_limits< A >::epsilon())
	Perform a fuzzy compare of floats/doubles, specify the allowed tolerance NB: this uses an absolute tolerance.
template<class A>
A	SafeDivision (A a, A b)
	Performs safe division that catches overflow and underflow.
template<class A>
bool	NearlyEqual (A a, A b, A tol=std::numeric_limits< A >::epsilon())
	A slightly different fuzzy comparator that checks if two values are "nearly" equal based on Knuth, "The Art of Computer Programming (vol II)" NB: this uses a relative tolerance.
template<class A>
void	UpdateRange (A &min, A &max, const A &value)
	Update an existing min - max range with a new prospective value.
template<class A>
void	UpdateRangeFinite (A &min, A &max, const A &value)
	Update an existing min - max range with a new prospective value.

Function Documentation

FuzzyCompare()

template<class A>

bool vtkMathUtilities::FuzzyCompare	(	A	a,
		A	b,
		A	epsilon = std::numeric_limits<A>::epsilon() )

Perform a fuzzy compare of floats/doubles, specify the allowed tolerance NB: this uses an absolute tolerance.

Definition at line 39 of file vtkMathUtilities.h.

SafeDivision()

template<class A>

A vtkMathUtilities::SafeDivision	(	A	a,
		A	b )

Performs safe division that catches overflow and underflow.

Definition at line 48 of file vtkMathUtilities.h.

NearlyEqual()

template<class A>

bool vtkMathUtilities::NearlyEqual	(	A	a,
		A	b,
		A	tol = std::numeric_limits<A>::epsilon() )

A slightly different fuzzy comparator that checks if two values are "nearly" equal based on Knuth, "The Art of Computer Programming (vol II)" NB: this uses a relative tolerance.

Definition at line 73 of file vtkMathUtilities.h.

UpdateRange()

template<class A>

void vtkMathUtilities::UpdateRange ( A & min, A & max, const A & value )

inline

Update an existing min - max range with a new prospective value.

If the value is non NaN then the appropriate range comparisons are made and updated, otherwise the original min - max values are set.

Examples:

No change: UpdateRange(-100, 100, 20) -> (-100, 100)

Update min: UpdateRange(-100, 100, -200) -> (-200, 100)

Update max: UpdateRange(-100, 100, 200) -> (-100, 200)

Input min and max are inverted creating an invalid range so a new range with the specified value is set: UpdateRange(100, -100, 20) -> (20, 20)

Input value is NaN so the original range is set UpdateRange(-100, 100, NaN) -> (-100, 100)

Definition at line 106 of file vtkMathUtilities.h.

UpdateRangeFinite()

template<class A>

void vtkMathUtilities::UpdateRangeFinite ( A & min, A & max, const A & value )

inline

Update an existing min - max range with a new prospective value.

If the value is finite (not NaN or Inf) then the appropriate range comparisons are made and updated, otherwise the original min - max values are set.

Definition at line 133 of file vtkMathUtilities.h.