Three-dimensional rectangular matrix with some algebraic operations. More...

#include <matrixnd.h>

Inheritance diagram for utl::matrix3d< T >:

Collaboration diagram for utl::matrix3d< T >:

Public Member Functions
	matrix3d ()
	Constructor creating an empty matrix.

	matrix3d (vector< T > &vec)
	Constructor using an external vector for matrix element storage.

	matrix3d (size_t n_1, size_t n_2, size_t n_3)
	Constructor creating a matrix of dimension \(n_1 \times n_2 \times n_3\).

	matrix3d (const matrix3d< T > &mat)
	Copy constructor.

	matrix3d (std::istream &is, std::streamsize max=10)
	Constructor to read a matrix from a stream.

virtual	~matrix3d ()
	Empty destructor.

void	resize (size_t n_1, size_t n_2, size_t n_3, bool forceClear=false)
	Resize the matrix to dimension \(n_1 \times n_2 \times n_3\). More...

matrix3d< T > &	operator= (const matrix3d< T > &A)
	Assignment operator.

T &	operator() (size_t i1, size_t i2, size_t i3)
	Index-1 based element access. More...

const T &	operator() (size_t i1, size_t i2, size_t i3) const
	Index-1 based element access. More...

vector< T >	getColumn (size_t i2, size_t i3) const
	Extract a column from the matrix.

void	fillColumn (size_t i2, size_t i3, const std::vector< T > &data)
	Fill a column of the matrix.

T	trace (size_t i1) const
	Return the trace of the i1'th sub-matrix.

matrix3d< T > &	operator+= (const matrix3d< T > &X)
	Add the given matrix X to *this.

matrix3d< T > &	operator-= (const matrix3d< T > &X)
	Subtract the given matrix X from *this.

matrix3d< T > &	add (const matrix3d< T > &X, T alfa=T(1))
	Add the given matrix X scaled by alfa to *this.

matrix3d< T > &	multiply (T c)
	Multiplication of this matrix by a scalar c.

bool	multiply (const matrix< T > &A, const matrix3d< T > &B, bool transA=false, bool addTo=false)
	Matrix-matrix multiplication. More...

bool	multiplyMat (const std::vector< T > &A, const matrix3d< T > &B, bool addTo=false)
	Matrix-matrix multiplication. More...

Public Member Functions inherited from utl::matrixBase< T >
size_t	dim (short int d=1) const
	Query dimensions.

size_t	size () const
	Query total matrix size.

bool	empty () const
	Check if the matrix is empty.

bool	zero (T tol=T(0)) const
	Check if the matrix elements are all zero.

const vector< T > &	toVec () const
	Type casting to a one-dimensional utl::vector, for access.

	operator const std::vector< T > & () const
	Type casting to a one-dimensional std::vector, for access.

	operator std::vector< T > & ()
	Type casting to a one-dimensional vector, for update.

T *	ptr (size_t c=0)
	Access through pointer.

const T *	ptr (size_t c=0) const
	Reference through pointer.

std::vector< T >::iterator	begin ()
	Iterator to the start of the matrix elements.

std::vector< T >::iterator	end ()
	Iterator to the end of the matrix elements.

void	clear ()
	Clears the matrix and sets its dimension to zero.

void	fill (T s)
	Fill the matrix with a scalar value.

void	fill (const T *values, size_t n=0)
	Fill the matrix with data from an array.

matrixBase< T > &	add (const matrixBase< T > &A, const T &alfa)
	Add the given matrix A scaled by alfa to *this.

matrixBase< T > &	multiply (const T &c)
	Multiplication of this matrix by a scalar c.

T	norm2 (int inc=1) const
	Return the Euclidean norm of the matrix. More...

T	asum (int inc=1) const
	Return the sum of the absolute value of the matrix elements. More...

T	sum (int inc=1) const
	Return the sum of the matrix elements. More...

Protected Member Functions
virtual void	clearIfNrowChanged (size_t n1, size_t n2, size_t)
	Clears the content if any of the first two dimensions changed.

Protected Member Functions inherited from utl::matrixBase< T >
	matrixBase ()
	The constructor is protected to allow sub-class instances only.

	matrixBase (vector< T > &vec)
	Constructor using an external vector for matrix element storage.

	matrixBase (size_t n_1, size_t n_2, size_t n_3=1, size_t n_4=1)
	Constructor creating a matrix of dimension \(n_1 \times n_2 \times n_3 \times n_4\).

	matrixBase (const matrixBase< T > &mat, bool copyContent=true)
	Copy constructor. More...

void	redim (size_t n_1, size_t n_2, size_t n_3, size_t n_4, bool forceClear)
	Resize the matrix to dimension \(n_1 \times n_2 \times n_3 \times n_4\). More...

Private Member Functions
bool	compatible (const matrix< T > &A, const matrix3d< T > &B, bool transA, size_t &M, size_t &N, size_t &K)
	Check dimension compatibility for matrix-matrix multiplication.

bool	compatible (const std::vector< T > &A, const matrix3d< T > &B, size_t &M, size_t &N, size_t &K)
	Check dimension compatibility for matrix-matrix multiplication, when the matrix A is represented by a one-dimensional vector.

Additional Inherited Members
Protected Attributes inherited from utl::matrixBase< T >
size_t	n [4]
	Dimension of the matrix.

vector< T > &	elem
	Actual matrix elements, stored column by column.

Detailed Description

template<class T>
class utl::matrix3d< T >

Three-dimensional rectangular matrix with some algebraic operations.

This is a 3D equivalent to the matrix class. The matrix elements are stored column-wise in a one-dimensional array, such that its pointer might be passed to Fortran subroutines requiring 3D arrays as arguments.

Member Function Documentation

◆ multiply()

template<class T >

bool utl::matrix3d< T >::multiply	(	const matrix< T > &	A,
		const matrix3d< T > &	B,
		bool	transA = `false`,
		bool	addTo = `false`
	)

inline

Matrix-matrix multiplication.

Performs one of the following operations (C = *this):

\( {\bf C} = {\bf A} {\bf B} \)
\( {\bf C} = {\bf A}^T {\bf B} \)
\( {\bf C} = {\bf C} + {\bf A} {\bf B} \)
\( {\bf C} = {\bf C} + {\bf A}^T {\bf B} \)

References K, M, and utl::matrixBase< T >::n.

◆ multiplyMat()

template<class T >

bool utl::matrix3d< T >::multiplyMat	(	const std::vector< T > &	A,
		const matrix3d< T > &	B,
		bool	addTo = `false`
	)

inline

Matrix-matrix multiplication.

Performs one of the following operations (C = *this):

\( {\bf C} = {\bf A} {\bf B} \)
\( {\bf C} = {\bf C} + {\bf A} {\bf B} \)

The matrix A is here represented by a one-dimensional vector, and its number of columns is assumed to match the first dimension of B and its number of rows is then the total vector length divided by the number of columns.

References K, M, and utl::matrixBase< T >::n.

◆ operator()() [1/2]

template<class T >

T& utl::matrix3d< T >::operator()	(	size_t	i1,
		size_t	i2,
		size_t	i3
	)

inline

Index-1 based element access.

Assuming column-wise storage as in Fortran.

◆ operator()() [2/2]

template<class T >

const T& utl::matrix3d< T >::operator()	(	size_t	i1,
		size_t	i2,
		size_t	i3
	)		const

inline

Index-1 based element access.

Assuming column-wise storage as in Fortran.

◆ resize()

template<class T >

void utl::matrix3d< T >::resize	(	size_t	n_1,
		size_t	n_2,
		size_t	n_3,
		bool	forceClear = `false`
	)

inline

Resize the matrix to dimension \(n_1 \times n_2 \times n_3\).

Will erase the previous content, but only if both the total matrix size, and n_1 or n_2 in the matrix are changed. It is therefore possible to add or remove a given number of elements in the third dimension of the matrix without loosing the contents of the first and second dimensions. If forceClear is true, the old matrix content is always erased.

References utl::matrixBase< T >::redim().

Referenced by ASMu3D::evaluateBasis(), SplineUtils::extractBasis(), ASMs1D::extractBasis(), utl::Hessian(), LRSplineFields2Dmx::hessianFE(), LRSplineFields3Dmx::hessianFE(), SplineFields2Dmx::hessianFE(), SplineFields3Dmx::hessianFE(), utl::matrix3d< T >::matrix3d(), and HDF5Reader::read3DArray().

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

src/LinAlg/matrixnd.h

Public Member Functions

Protected Member Functions

Private Member Functions

Additional Inherited Members

Detailed Description

template<class T> class utl::matrix3d< T >

Member Function Documentation

◆ multiply()

◆ multiplyMat()

◆ operator()() [1/2]

◆ operator()() [2/2]

◆ resize()

template<class T>
class utl::matrix3d< T >