Class for handling comparisons of floating-point results versus expected values in unit tests. The general form is to take a vector of real numbers and apply the comparison across all entries. More...

#include <approx.h>

Public Member Functions
template<typename T>
	Approx (const std::vector< T > &values_in, ComparisonType style_in=ComparisonType::ABSOLUTE, double tol_in=1.0e-6)
	Constructors for real number comparisons based on a vector of multiple input values. Other floating point and integer vectors can be converted to double-precision vectors.

	~Approx ()=default
	Default destructor.

int	size () const
	Get the size of the collection of numbers in a real number comparison.

double	getValue () const
	Get the expectation value, sans any tolerance.

std::vector< double >	getValues () const
	Get all values of an approximate comparison, sans any tolerance.

ComparisonType	getStyle () const
	Get the style used in this approximate comparison.

bool	test (const double test_value) const
	Test whether a real-valued scalar is the same as the value held for comparison.

template<typename T>
bool	test (const std::vector< T > &test_values) const
	Test whether a vector of real-valued scalars is the same as as vector held for comparison.


	Approx (double value_in, ComparisonType style_in=ComparisonType::ABSOLUTE, double tol_in=1.0e-6)
	Constructor for real number comparisons based on a single input value.

	Approx (double value_in, double tol_in, ComparisonType style_in=ComparisonType::ABSOLUTE)


double	getMargin () const
	Get the margin or tolerance associated with this approximate comparison.

double	getTolerance () const

double	getTol () const


void	setValue (double value_in)
	Set the values that form the basis of the approximate comparison.

void	setValues (const std::vector< double > &values_in)

void	setValue (double value_in, size_t index)


void	setMargin (double dtol_in)
	Set the tolerance of the existing object. This will not emit a new object.

void	setTolerance (double dtol_in)

void	setTol (double dtol_in)


Approx	margin (double dtol_in) const
	Set the tolerance. This is written in such a way as to mimic the Catch2 unit testing framework in some circumstances. It will not change the original object's tolerance, rather it will emit an object with the same comparison values and the desired tolerance.

Approx	tolerance (double dtol_in) const

Approx	tol (double dtol_in) const

Detailed Description

Class for handling comparisons of floating-point results versus expected values in unit tests. The general form is to take a vector of real numbers and apply the comparison across all entries.

Constructor & Destructor Documentation

◆ Approx() [1/2]

template<typename T>

stormm::testing::Approx::Approx	(	const std::vector< T > &	values_in,
		ComparisonType	style_in = ComparisonType::ABSOLUTE,
		double	tol_in = 1.0e-6 )

Constructors for real number comparisons based on a vector of multiple input values. Other floating point and integer vectors can be converted to double-precision vectors.

Parameters

values_in	The value for comparison
style_in	Type of comparison, relative or absolute
tol_in	Tolerance for deviations that will still yield a correct comparison

◆ Approx() [2/2]

stormm::testing::Approx::Approx	(	double	value_in,
		ComparisonType	style_in = ComparisonType::ABSOLUTE,
		double	tol_in = 1.0e-6 )

Constructor for real number comparisons based on a single input value.

Overloaded:

Take the comparison style ahead of the tolerance (convenient use of default tolerance)
Take the tolerance ahead of the comparison style (tolerance must be specified)

Parameters

value_in	The value for comparison
style_in	The type of comparison, relative or absolute
tol_in	Tolerance for deviations that will still yield a correct comparison

Member Function Documentation

◆ margin()

Approx stormm::testing::Approx::margin ( double dtol_in ) const

Set the tolerance. This is written in such a way as to mimic the Catch2 unit testing framework in some circumstances. It will not change the original object's tolerance, rather it will emit an object with the same comparison values and the desired tolerance.

Overloaded:

Three different names that individual developers may find most intuitive

Parameters

dtol_in The tolerance to use

◆ setMargin()

void stormm::testing::Approx::setMargin ( double dtol_in )

Set the tolerance of the existing object. This will not emit a new object.

Overloaded:

Three different names that individual developers may find most intuitive

Parameters

dtol_in The tolerance to use

◆ setValue()

void stormm::testing::Approx::setValue ( double value_in )

Set the values that form the basis of the approximate comparison.

Overloaded:

Set the approximation to use a single value
Set the target value of a particular index
Set the approximation to use multiple values

Parameters

value_in	The single value to set (this will change the length of the comparison vector to one)
values_in	The vector of values to set
index	Index of the value to change (this will be checked against the length of comparison values in the existing object)

◆ test() [1/2]

bool stormm::testing::Approx::test ( const double test_value ) const

Test whether a real-valued scalar is the same as the value held for comparison.

Parameters

test_value The scalar value to test

◆ test() [2/2]

template<typename T>

bool stormm::testing::Approx::test ( const std::vector< T > & test_values ) const

Test whether a vector of real-valued scalars is the same as as vector held for comparison.

Parameters

test_values The vector of values to test

The documentation for this class was generated from the following files:

src/UnitTesting/approx.h
src/UnitTesting/approx.cpp

Public Member Functions

Detailed Description

Constructor & Destructor Documentation

◆ Approx() [1/2]

◆ Approx() [2/2]

Member Function Documentation

◆ margin()

◆ setMargin()

◆ setValue()

◆ test() [1/2]

◆ test() [2/2]