The Sparse class. More...

#include <matpackII.h>

Public Member Functions
	Sparse ()
	Default constructor. More...

	Sparse (Index r, Index c)
	Constructor setting size. More...

void	split (Index offset, Index nrows)
	Reduce matrix to the row range [offset, offset + nrows]. More...

void	insert_row (Index r, Vector v)
	Insert row function. More...

void	insert_elements (Index nnz, const ArrayOfIndex &rowind, const ArrayOfIndex &colind, ConstVectorView data)
	Insert vector of elements with given row and column indices. More...

void	resize (Index r, Index c)
	Resize function. More...

bool	empty () const
	Returns true if variable size is zero. More...

Index	nrows () const
	Returns the number of rows. More...

Index	ncols () const
	Returns the number of columns. More...

Index	nnz () const
	Returns the number of nonzero elements. More...

Vector	diagonal () const
	Diagonal elements as vector. More...

Numeric &	rw (Index r, Index c)
	Read and write index operator. More...

Numeric	ro (Index r, Index c) const
	Read only index operator. More...

Numeric	operator() (Index r, Index c) const
	Plain index operator. More...

Sparse &	operator+= (const Sparse &x)
	Add sparse matrix. More...

Sparse &	operator-= (const Sparse &x)
	Subtract sparse matrix. More...

Sparse &	operator*= (Numeric x)
	Scale matrix. More...

Sparse &	operator/= (Numeric x)
	Scale matrix by reciprocal. More...

	operator Matrix () const
	Convert to dense matrix. More...

void	list_elements (Vector &values, ArrayOfIndex &row_indices, ArrayOfIndex &column_indices) const
	List elements in matrix. More...

Numeric *	get_element_pointer ()

int *	get_column_index_pointer ()

int *	get_row_start_pointer ()

Static Public Member Functions
static Sparse	diagonal (ConstVectorView v)
	Create a sparse matrix from a vector. More...

Private Attributes
Eigen::SparseMatrix< Numeric, Eigen::RowMajor >	matrix
	The actual matrix. More...

Friends
std::ostream &	operator<< (std::ostream &os, const Sparse &v)
	Output operator for Sparse. More...

void	abs (Sparse &A, const Sparse &B)
	Absolute value of sparse matrix elements. More...

void	mult (VectorView y, const Sparse &M, ConstVectorView x)
	Sparse matrix - Vector multiplication. More...

void	transpose_mult (VectorView y, const Sparse &M, ConstVectorView x)
	Sparse matrix - Vector multiplication. More...

void	mult (MatrixView A, const Sparse &B, const ConstMatrixView &C)
	SparseMatrix - Matrix multiplication. More...

void	mult (MatrixView A, const ConstMatrixView &B, const Sparse &C)
	Matrix - SparseMatrix multiplication. More...

void	mult (Sparse &A, const Sparse &B, const Sparse &C)
	Sparse - Sparse multiplication. More...

void	add (Sparse &A, const Sparse &B, const Sparse &C)
	Sparse - Sparse addition. More...

void	sub (Sparse &A, const Sparse &B, const Sparse &C)
	Sparse - Sparse subtraction. More...

void	transpose (Sparse &A, const Sparse &B)
	Transpose of sparse matrix. More...

void	id_mat (Sparse &A)
	Sparse identity matrix. More...

Detailed Description

The Sparse class.

Wrapper class for Eigen sparse matrices. The storage format used is compressed row storage. Thus inserting of elements in a single row ordered by column index is performed in constant time, if the rows are themselves inserted in increasing order.

Definition at line 60 of file matpackII.h.

Constructor & Destructor Documentation

◆ Sparse() [1/2]

Sparse::Sparse ( )

Default constructor.

Author: Stefan Buehler sbueh.nosp@m.ler@.nosp@m.ltu.s.nosp@m.e

Date: Tue Jul 15 15:05:40 2003

Definition at line 148 of file matpackII.cc.

◆ Sparse() [2/2]

Sparse::Sparse	(	Index	r,
		Index	c
	)

Constructor setting size.

Parameters

r	Row dimension of new sparse matrix.
c	Column dimension of new sparse matrix.

Definition at line 157 of file matpackII.cc.

Member Function Documentation

◆ diagonal() [1/2]

Sparse Sparse::diagonal ( ConstVectorView v )

static

Create a sparse matrix from a vector.

Parameters

v	vector containing the diagonal elements.

Returns: Sparse matrix with the elements of the given vector on the diagonal.

Definition at line 161 of file matpackII.cc.

References i, insert_elements(), n, and ConstVectorView::nelem().

◆ diagonal() [2/2]

Vector Sparse::diagonal ( ) const

Diagonal elements as vector.

Extracts the diagonal elements from the sparse matrix. matrix.

Returns: A vector containing the diagonal elements.

Definition at line 172 of file matpackII.cc.

References i, matrix, min, ncols(), and nrows().

◆ empty()

bool Sparse::empty ( ) const

Returns true if variable size is zero.

Definition at line 61 of file matpackII.cc.

References matrix.

Referenced by check_and_add_block().

◆ get_column_index_pointer()

int* Sparse::get_column_index_pointer ( )

inline

Definition at line 123 of file matpackII.h.

References matrix.

Referenced by get_covariance_matrix_block(), and get_variable_value().

◆ get_element_pointer()

Numeric* Sparse::get_element_pointer ( )

inline

Definition at line 122 of file matpackII.h.

References matrix.

Referenced by get_covariance_matrix_block(), and get_variable_value().

◆ get_row_start_pointer()

int* Sparse::get_row_start_pointer ( )

inline

Definition at line 124 of file matpackII.h.

References abs, add, C, id_mat, M, matrix, mult, operator<<, sub, transpose, and transpose_mult.

Referenced by get_covariance_matrix_block(), and get_variable_value().

◆ insert_elements()

void Sparse::insert_elements	(	Index	nelems,
		const ArrayOfIndex &	rowind,
		const ArrayOfIndex &	colind,
		ConstVectorView	data
	)

Insert vector of elements with given row and column indices.

Efficient inserting of a vector of elements into the sparse matrix. Overwrites elements currently in the matrix. The complexity is linear in the number of elements and should therefore be the preferred way of inserting elements into the sparse matrix.

Parameters

nelems	The number of elements to insert.
rowind	A vector containing the row indices.
colind	A vector containing the column indices.
data	The vector containing the elements.

Definition at line 337 of file matpackII.cc.

References i, and matrix.

Referenced by diagonal(), DiagonalMatrix(), InteractiveWorkspace::set_sparse_variable(), and xml_read_from_stream().

◆ insert_row()

void Sparse::insert_row	(	Index	r,
		Vector	v
	)

Insert row function.

Inserts a Vector as row of elements at the given position.

The row index must agree with the size of the matrix. This function can not be used to expand the matrix. Only non-zero values will be stored. If the destination row already exist it will be overwritten.

Parameters

r	Where to insert the row
v	Vector to be inserted.

Definition at line 314 of file matpackII.cc.

References i, matrix, ncols(), ConstVectorView::nelem(), and nrows().

Referenced by antenna1d_matrix(), antenna2d_interp_response(), met_mm_polarisation_hmatrix(), sensor_responseFrequencySwitching(), sensor_responseGenericAMSU(), sensor_responsePolarisation(), spectrometer_matrix(), and test_insert_row().

◆ list_elements()

void Sparse::list_elements	(	Vector &	values,
		ArrayOfIndex &	row_indices,
		ArrayOfIndex &	column_indices
	)		const

List elements in matrix.

Writes the values in the matrix into the given Vector values and the row and column indices into the given ArrayOfIndex objects.

Parameters

values	The values contained in the matrix.
row_indices	The row indices corresponding to the values.
column_indices	The column indices corresponding to the values.

Definition at line 270 of file matpackII.cc.

References i, matrix, n, ncols(), nnz(), nrows(), and Vector::resize().

Referenced by xml_write_to_stream().

◆ ncols()

Index Sparse::ncols ( ) const

Returns the number of columns.

Definition at line 69 of file matpackII.cc.

References matrix.

Referenced by abs(), add(), antenna1d_matrix(), antenna2d_interp_response(), check_and_add_block(), Compare(), diagonal(), get_variable_value(), id_mat(), insert_row(), jacobianCalcFreqShift(), jacobianCalcFreqStretch(), list_elements(), MatrixCopySparse(), mueller_rotation(), mult(), operator Matrix(), random_fill_matrix(), Select(), sensor_checkedCalc(), sensor_responseAntenna(), sensor_responseBackend(), sensor_responseFrequencySwitching(), sensor_responseMetMM(), sensor_responseMixer(), sensor_responseMixerBackendPrecalcWeights(), sensor_responsePolarisation(), sensor_responseStokesRotation(), sensor_responseWMRF(), SparseSparseMultiply(), split(), sub(), test_identity(), test_insert_row(), transpose(), transpose_mult(), WMRFSelectChannels(), xml_write_to_stream(), and yCalc_mblock_loop_body().

◆ nnz()

Index Sparse::nnz ( ) const

Returns the number of nonzero elements.

Definition at line 72 of file matpackII.cc.

References matrix.

Referenced by get_covariance_matrix_block(), get_variable_value(), list_elements(), Select(), and xml_write_to_stream().

◆ nrows()

Index Sparse::nrows ( ) const

Returns the number of rows.

Definition at line 66 of file matpackII.cc.

References matrix.

Referenced by abs(), add(), check_and_add_block(), Compare(), diagonal(), get_rowindex_for_mblock(), get_variable_value(), id_mat(), insert_row(), jacobianCalcFreqShift(), jacobianCalcFreqStretch(), jacobianCalcPointingZaInterp(), jacobianCalcPointingZaRecalc(), list_elements(), MatrixCopySparse(), mueller_rotation(), mult(), operator Matrix(), random_fill_matrix(), Select(), sensor_checkedCalc(), sensor_responseAntenna(), sensor_responseBackend(), sensor_responseFillFgrid(), sensor_responseFrequencySwitching(), sensor_responseMetMM(), sensor_responseMixer(), sensor_responseMixerBackendPrecalcWeights(), sensor_responseMultiMixerBackend(), sensor_responsePolarisation(), sensor_responseStokesRotation(), sensor_responseWMRF(), SparseSparseMultiply(), sub(), test_identity(), test_insert_row(), transpose(), transpose_mult(), WMRFSelectChannels(), xml_write_to_stream(), yCalc(), and ySimpleSpectrometer().

◆ operator Matrix()

Sparse::operator Matrix ( ) const

explicit

Convert to dense matrix.

Converts a given sparse matrix to a dense matrix. Intended mainly for testing purposes.

Returns: The dense representation of the given sparse matrix.

Definition at line 189 of file matpackII.cc.

References i, matrix, n, ncols(), and nrows().

◆ operator()()

Numeric Sparse::operator()	(	Index	r,
		Index	c
	)		const

Plain index operator.

This is the same as the .ro index operator.

Parameters

r	Row index.
c	Column index.

Returns: The data element with these indices.

Definition at line 111 of file matpackII.cc.

◆ operator*=()

Sparse & Sparse::operator*= ( Numeric x )

Scale matrix.

Parameters

x	Scalar factor to scale the matrix with.

Returns: Reference to the scaled sparse matrix object.

Definition at line 243 of file matpackII.cc.

References matrix.

◆ operator+=()

Sparse & Sparse::operator+= ( const Sparse & A )

Add sparse matrix.

Element-wise, in-situ sum of two sparse matrices. Matrix must have the same dimensions.

Parameters

A	The sparse matrix to add to the Sparse object.

Returns: Reference to the resulting sum.

Definition at line 215 of file matpackII.cc.

References matrix.

◆ operator-=()

Sparse & Sparse::operator-= ( const Sparse & A )

Subtract sparse matrix.

Element-wise, in-situ difference of two sparse matrices. Matrix must have the same dimensions.

Parameters

A	The sparse matrix to add to the Sparse object.

Returns: Reference to the resulting sum.

Definition at line 231 of file matpackII.cc.

References matrix.

◆ operator/=()

Sparse & Sparse::operator/= ( Numeric x )

Scale matrix by reciprocal.

Parameters

x	Reciprocal of the factor to scale the matrix with.

Returns: Reference to the scaled sparse matrix object.

Definition at line 255 of file matpackII.cc.

References matrix.

◆ resize()

void Sparse::resize	(	Index	r,
		Index	c
	)

Resize function.

If the size is already correct this function does nothing.

All data is lost after resizing! The new Sparse is not initialised so it will be empty.

Parameters

r	New row dimension.
c	New column dimension.

Definition at line 361 of file matpackII.cc.

Referenced by add(), antenna1d_matrix(), antenna2d_interp_response(), DiagonalMatrix(), sensor_responseAntenna(), sensor_responseBackend(), sensor_responseFrequencySwitching(), sensor_responseGenericAMSU(), sensor_responseInit(), sensor_responseMixer(), sensor_responseMixerBackendPrecalcWeights(), sensor_responsePolarisation(), sensor_responseStokesRotation(), sensor_responseWMRF(), SparseMatrixIdentity(), spectrometer_matrix(), sub(), test47(), test_identity(), test_insert_row(), test_workspace_methods(), WMRFSelectChannels(), and xml_read_from_stream().

◆ ro()

Numeric Sparse::ro	(	Index	r,
		Index	c
	)		const

Read only index operator.

This has to correctly handle two cases:

The data element exists. In this case the element is returned.
The element does not exist. In this case the value 0 is returned.

Parameters

r	Row index.
c	Column index.

Returns: The data element with these indices, or zero.

Author: Stefan Buehler sbueh.nosp@m.ler@.nosp@m.ltu.s.nosp@m.e

Date: Tue Jul 15 15:05:40 2003

Definition at line 130 of file matpackII.cc.

Referenced by sensor_responseStokesRotation().

◆ rw()

Numeric & Sparse::rw	(	Index	r,
		Index	c
	)

Read and write index operator.

This has to correctly handle two cases:

The data element exists. In this case the operator acts similar to the const index operator in that the element is returned.
The element does not exist. In this case it is created.

Parameters

r	Row index.
c	Column index.

Returns: The data element with these indices.

Definition at line 91 of file matpackII.cc.

Referenced by mueller_rotation(), random_fill_matrix(), test3(), test40(), and test49().

◆ split()

void Sparse::split	(	Index	offset,
		Index	nrows
	)

Reduce matrix to the row range [offset, offset + nrows].

Definition at line 296 of file matpackII.cc.

References matrix, and ncols().

Friends And Related Function Documentation

◆ abs

void abs	(	Sparse &	A,
		const Sparse &	B
	)

friend

Absolute value of sparse matrix elements.

Computes the absolute values of the elements in sparse matrix B.

The output matrix A must have been initialized with the correct size.

Parameters

A	Output: Absolute value matrix.
B	Original matrix.

Author: Mattias Ekstrom

Date: 2005-03-21

Definition at line 394 of file matpackII.cc.

Referenced by get_row_start_pointer().

◆ add

void add	(	Sparse &	A,
		const Sparse &	B,
		const Sparse &	C
	)

friend

Sparse - Sparse addition.

Calculates A = B+C, where result A is sparse, and B and C are also sparse.

Output comes first!

Dimensions of B, and C must match. A will be resized.

Parameters

A	Output: Result matrix.
B	First summand matrix.
C	Second summand matrix.

Author: Oliver Lemke

Date: 2009-09-03

Definition at line 630 of file matpackII.cc.

Referenced by get_row_start_pointer().

◆ id_mat

void id_mat ( Sparse & A )

friend

Sparse identity matrix.

Set the given Sparse matrix object to the identity matrix. The matrix must be square.

Parameters

A	The matrix to be set to the identity matrix.

Definition at line 647 of file matpackII.cc.

Referenced by get_row_start_pointer().

◆ mult [1/4]

void mult	(	VectorView	y,
		const Sparse &	M,
		ConstVectorView	x
	)

friend

Sparse matrix - Vector multiplication.

This calculates the product

y = M*x, where M is sparse.

Output comes first!

Dimensions of y, M, and x must match. No memory reallocation takes place, only the data is copied.

Parameters

y	Output: The multiplication result.
M	Matrix for multiplication (sparse).
x	Vector for multiplication.

Definition at line 417 of file matpackII.cc.

Referenced by get_row_start_pointer().

◆ mult [2/4]

void mult	(	MatrixView	A,
		const Sparse &	B,
		const ConstMatrixView &	C
	)

friend

SparseMatrix - Matrix multiplication.

Calculates the matrix product:

A = B*C, where B is sparse.

Output comes first!

Dimensions of A, B, and C must match. No memory reallocation takes place, only the data is copied.

Parameters

A	Output: Result matrix (full).
B	First matrix to multiply (sparse).
C	Second matrix to multiply (full).

Author: Stefan Buehler sbueh.nosp@m.ler@.nosp@m.ltu.s.nosp@m.e

Date: Tue Jul 15 15:05:40 2003

Definition at line 490 of file matpackII.cc.

◆ mult [3/4]

void mult	(	MatrixView	A,
		const ConstMatrixView &	B,
		const Sparse &	C
	)

friend

Matrix - SparseMatrix multiplication.

Calculates the matrix product:

A = B*C, where C is sparse.

Output comes first!

Dimensions of A, B, and C must match. No memory reallocation takes place, only the data is copied.

Parameters

A	Output: Result matrix (full).
B	First matrix to multiply (sparse).
C	Second matrix to multiply (full).

Author: Stefan Buehler sbueh.nosp@m.ler@.nosp@m.ltu.s.nosp@m.e

Date: Tue Jul 15 15:05:40 2003

Definition at line 539 of file matpackII.cc.

◆ mult [4/4]

void mult	(	Sparse &	A,
		const Sparse &	B,
		const Sparse &	C
	)

friend

Sparse - Sparse multiplication.

Calculates A = B*C, where result A is sparse, and B and C are also sparse.

Output comes first!

Dimensions of A, B, and C must match. No memory reallocation takes place, only the data is copied.

Parameters

A	Output: Result matrix.
B	First product matrix.
C	Second product matrix.

Author: Mattias Ekstroem

Date: 2003-08-06

Definition at line 606 of file matpackII.cc.

◆ operator<<

std::ostream& operator<<	(	std::ostream &	os,
		const Sparse &	v
	)

friend

Output operator for Sparse.

Parameters

os	Output stream.
v	Sparse matrix to print.

Returns: Output stream.

Definition at line 375 of file matpackII.cc.

Referenced by get_row_start_pointer().

◆ sub

void sub	(	Sparse &	A,
		const Sparse &	B,
		const Sparse &	C
	)

friend

Sparse - Sparse subtraction.

Calculates A = B-C, where result A is sparse, and B and C are also sparse.

Output comes first!

Dimensions of B, and C must match. A will be resized.

Parameters

A	Output: Result matrix.
B	First subtrahend matrix.
C	Second subtrahend matrix.

Author: Oliver Lemke

Date: 2009-09-03

Definition at line 667 of file matpackII.cc.

Referenced by get_row_start_pointer().

◆ transpose

void transpose	(	Sparse &	A,
		const Sparse &	B
	)

friend

Transpose of sparse matrix.

Computes the transpose of the sparse matrix B.

The output matrix A must have been initialized with the correct size.

Parameters

A	Output: Transposed matrix.
B	Original matrix.

Author: Mattias Ekstroem

Date: 2003-08-05

Definition at line 582 of file matpackII.cc.

Referenced by get_row_start_pointer().

◆ transpose_mult

void transpose_mult	(	VectorView	y,
		const Sparse &	M,
		ConstVectorView	x
	)

friend

Sparse matrix - Vector multiplication.

This calculates the product

y = M*x, where M is sparse.

Output comes first!

Dimensions of y, M, and x must match. No memory reallocation takes place, only the data is copied.

Parameters

y	Output: The multiplication result.
M	Matrix for multiplication (sparse).
x	Vector for multiplication.

Definition at line 452 of file matpackII.cc.

Referenced by get_row_start_pointer().

Member Data Documentation

◆ matrix

Eigen::SparseMatrix<Numeric, Eigen::RowMajor> Sparse::matrix

private

The actual matrix.

Definition at line 141 of file matpackII.h.

Referenced by abs(), add(), diagonal(), empty(), get_column_index_pointer(), get_element_pointer(), get_row_start_pointer(), id_mat(), insert_elements(), insert_row(), list_elements(), mult(), ncols(), nnz(), nrows(), operator Matrix(), operator*=(), operator+=(), operator-=(), operator/=(), operator<<(), split(), sub(), transpose(), and transpose_mult().

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

Public Member Functions

Static Public Member Functions

Private Attributes

Friends

Detailed Description

Constructor & Destructor Documentation

◆ Sparse() [1/2]

◆ Sparse() [2/2]

Member Function Documentation

◆ diagonal() [1/2]

◆ diagonal() [2/2]

◆ empty()

◆ get_column_index_pointer()

◆ get_element_pointer()

◆ get_row_start_pointer()

◆ insert_elements()

◆ insert_row()

◆ list_elements()

◆ ncols()

◆ nnz()

◆ nrows()

◆ operator Matrix()

◆ operator()()

◆ operator*=()

◆ operator+=()

◆ operator-=()

◆ operator/=()

◆ resize()

◆ ro()

◆ rw()

◆ split()

Friends And Related Function Documentation

◆ abs

◆ add

◆ id_mat

◆ mult [1/4]

◆ mult [2/4]

◆ mult [3/4]

◆ mult [4/4]

◆ operator<<

◆ sub

◆ transpose

◆ transpose_mult

Member Data Documentation

◆ matrix