#include <deal.II/base/floating_point_comparator.h>

Public Types
enum class	ComparisonResult { less , equal , greater }

using	ScalarNumber = typename ::internal::VectorizedArrayTrait< Number >::value_type

Public Member Functions
	FloatingPointComparator (const ScalarNumber tolerance, const bool use_absolute_tolerance=true, const std::bitset< width > &mask=std::bitset< width >().flip())

	FloatingPointComparator (const FloatingPointComparator &rhs)=default

	FloatingPointComparator (FloatingPointComparator &&rhs) noexcept=default

template<typename T >
bool	operator() (const T &object1, const T &object2) const

template<typename T >
ComparisonResult	compare (const std::vector< T > &v1, const std::vector< T > &v2) const

template<std::size_t dim, typename T >
ComparisonResult	compare (const std::array< T, dim > &t1, const std::array< T, dim > &t2) const

template<int rank, int dim, typename T >
ComparisonResult	compare (const Tensor< rank, dim, T > &t1, const Tensor< rank, dim, T > &t2) const

template<typename T >
ComparisonResult	compare (const Table< 2, T > &t1, const Table< 2, T > &t2) const

ComparisonResult	compare (const ScalarNumber s1, const ScalarNumber s2) const

ComparisonResult	compare (const VectorizedArray< ScalarNumber, width > &v1, const VectorizedArray< ScalarNumber, width > &v2) const

Static Public Attributes
static constexpr std::size_t	width

Private Attributes
const ScalarNumber	tolerance

const bool	use_absolute_tolerance

const std::bitset< width >	mask

Detailed Description

template<typename Number>
struct FloatingPointComparator< Number >

A class that is used to compare floating point arrays (e.g. std::vector, Tensor<1,dim>, etc.). The most common use case of this comparator is for detecting arrays with the same content for the sake of compression. The idea of this class is to consider two arrays as equal if they are the same within a given tolerance. We use this comparator class within a std::map<> of the given arrays. Note that this comparison operator does not satisfy all the mathematical properties one usually wants to have (consider e.g. the numbers a=0, b=0.1, c=0.2 with tolerance 0.15; the operator gives a<c, but neither a<b? nor b<c? is satisfied). This is not a problem in the use cases for this class, but be careful when using it in other contexts.

Definition at line 44 of file floating_point_comparator.h.

Member Typedef Documentation

◆ ScalarNumber

template<typename Number >

using FloatingPointComparator< Number >::ScalarNumber = typename ::internal::VectorizedArrayTrait<Number>::value_type

Definition at line 46 of file floating_point_comparator.h.

Member Enumeration Documentation

◆ ComparisonResult

template<typename Number >

enum class FloatingPointComparator::ComparisonResult

strong

An enum to decode whether a particular comparison returns less, larger, or equal. This is needed beyond the regular operator() because of efficiency reasons in nested checks as supported by this class.

Enumerator
less
equal
greater

Definition at line 56 of file floating_point_comparator.h.

Constructor & Destructor Documentation

◆ FloatingPointComparator() [1/3]

template<typename Number >

FloatingPointComparator< Number >::FloatingPointComparator	(	const ScalarNumber	tolerance,
		const bool	use_absolute_tolerance = `true`,
		const std::bitset< width > &	mask = `std::bitset<width>().flip()`
	)

Constructor.

Definition at line 139 of file floating_point_comparator.h.

◆ FloatingPointComparator() [2/3]

template<typename Number >

FloatingPointComparator< Number >::FloatingPointComparator ( const FloatingPointComparator< Number > & rhs )

default

◆ FloatingPointComparator() [3/3]

template<typename Number >

FloatingPointComparator< Number >::FloatingPointComparator ( FloatingPointComparator< Number > && rhs )

defaultnoexcept

Member Function Documentation

◆ operator()()

template<typename Number >

template<typename T >

bool FloatingPointComparator< Number >::operator()	(	const T &	object1,
		const T &	object2
	)		const

Compare two objects of the same type and return true in case the first object is less than the second one. This function calls the compare() functions below, so only types supported by compare() will work with this function.

Definition at line 155 of file floating_point_comparator.h.

◆ compare() [1/6]

template<typename Number >

template<typename T >

FloatingPointComparator< Number >::ComparisonResult FloatingPointComparator< Number >::compare	(	const std::vector< T > &	v1,
		const std::vector< T > &	v2
	)		const

Compare two vectors of numbers (not necessarily of the same length), where vectors of different lengths are first sorted by their length and then by the entries.

Definition at line 166 of file floating_point_comparator.h.

◆ compare() [2/6]

template<typename Number >

template<std::size_t dim, typename T >

FloatingPointComparator< Number >::ComparisonResult FloatingPointComparator< Number >::compare	(	const std::array< T, dim > &	t1,
		const std::array< T, dim > &	t2
	)		const

Compare two arrays.

Definition at line 189 of file floating_point_comparator.h.

◆ compare() [3/6]

template<typename Number >

template<int rank, int dim, typename T >

FloatingPointComparator< Number >::ComparisonResult FloatingPointComparator< Number >::compare	(	const Tensor< rank, dim, T > &	t1,
		const Tensor< rank, dim, T > &	t2
	)		const

Compare two tensors.

Definition at line 206 of file floating_point_comparator.h.

◆ compare() [4/6]

template<typename Number >

template<typename T >

FloatingPointComparator< Number >::ComparisonResult FloatingPointComparator< Number >::compare	(	const Table< 2, T > &	t1,
		const Table< 2, T > &	t2
	)		const

Compare two tables.

Definition at line 223 of file floating_point_comparator.h.

◆ compare() [5/6]

template<typename Number >

FloatingPointComparator< Number >::ComparisonResult FloatingPointComparator< Number >::compare	(	const ScalarNumber	s1,
		const ScalarNumber	s2
	)		const

Compare two scalar numbers.

Definition at line 243 of file floating_point_comparator.h.

◆ compare() [6/6]

template<typename Number >

FloatingPointComparator< Number >::ComparisonResult FloatingPointComparator< Number >::compare	(	const VectorizedArray< ScalarNumber, width > &	v1,
		const VectorizedArray< ScalarNumber, width > &	v2
	)		const