Data Structures
struct	MATRIX
struct	VECTOR
Enumerations
enum	MatrixType { REAL, COMPLEX, BOOL }
enum	MatrixOrder { ROW_MAJOR, COL_MAJOR }
Functions
MATRIX	fftw (fits2d function)
MATRIX	fftwReverse (MATRIX function, bool shift)
MATRIX	fftwShift (MATRIX m)
MATRIX	ifftwShift (MATRIX m)
MATRIX	linearEqSolve (MATRIX A, MATRIX B)
MATRIX	LinearEqSolveNRHS (MATRIX A, MATRIX B)
void	linearEqSolveCleanup ()
MATRIX	linearEqSolveCholesky (MATRIX A, MATRIX B)
MATRIX	linearEqSolveSvd (MATRIX A, MATRIX B)
MASK	newMask (int r, int c, int m)
MATRIX	newMatrix (int r, int c)
MATRIX	newMatrixI (int r, int c, void *Idata, MatrixType mType)
MATRIX	newComplexMatrix (int r, int c)
MATRIX	newMatrixG (int r, int c, MatrixType mType)
MATRIX	newMatrixGI (int r, int c, void *Idata, MatrixType mType)
VECTOR	newVector (int n)
void	freeMatrix (MATRIX m)
void	freeVector (VECTOR v)
MATRIX	newRandomMatrix (int r, int c)
MATRIX	multiplyMM2 (char TransA, char TransB, MATRIX A, MATRIX B)
MATRIX	multiplyMMFull (char TransA, char TransB, Float alpha, Float beta, MATRIX A, MATRIX B)
MATRIX	multiplyScalar (MATRIX m, Float scalar)
MATRIX	multiplyScalarZ (MATRIX m, FloatComplex scalar)
Float *	toRowMajor (Float *A, int R, int C)
Float *	toColumnMajor (Float *AT, int R, int C)
MATRIX	toColumnMajor2 (MATRIX m)
MATRIX	toRowMajor2 (MATRIX m)
MATRIX	toMatrix (fits2d f)
VECTOR	toVector (fits2d f)
fits2d	toFitsObject (VECTOR v)
fits2d	MatrixToFitsObject (MATRIX m)
fits3d	MatrixToFits3dObject (MATRIX m)
MATRIX	expandDiagonalMatrix (VECTOR v)
MATRIX	transpose (MATRIX m)
MATRIX	conjugate (MATRIX m)
MATRIX	matrix_copy (MATRIX m)
MATRIX	inverseD (MATRIX m)
MATRIX	inverse (MATRIX m)
MATRIX	Cholesky (MATRIX m)
MATRIX	ConvertToReal (MATRIX m)
MATRIX	ConvertToComplex (MATRIX m)
bool	isIdentity (MATRIX m)
VECTOR	MatrixToVector (MATRIX m)
MATRIX	VectorToMatrix_Column (VECTOR v)
void	insertVector (Float *to, VECTOR from)
void	APlusLambdaB (MATRIX A, Float lambda, MATRIX B, MATRIX C)
FloatComplex	Norm2 (MATRIX m)
MATRIX	BItimesV (MATRIX B, MATRIX V)
void	splitMatrix (MATRIX m, MATRIX real, MATRIX imag)
void	mergeMatrix (MATRIX *m, MATRIX real, MATRIX imag)
void	setToNaN (VECTOR v)
void	setToNaNMatrix (MATRIX m)
VECTOR	readText (char *inputFilename)
void	print_matrix (char desc, int m, int n, Float a, int lda, MatrixOrder order)
void	print_matrix2 (char *desc, MATRIX m, MatrixOrder order)
void	print_matrix2toFile (char *desc, MATRIX m, MatrixOrder order)
int	tdSetup1d (MATRIX aM, MATRIX ad, MATRIX aR, MATRIX aCI, MATRIX aC)
void	tdinvert3dFinish (Float lambda, TempData tempData, bool debug, BaseDirectory debugFilename)
void	tdinvert1dFinish (Float lambda, int nStep, Model model, bool debug, BaseDirectory debugFilename)
void	tdCleanup ()

Enumeration Type Documentation

enum MatrixOrder

MatrixOrder designates the layout of data in a matrix.

Enumerator:

ROW_MAJOR
COL_MAJOR

enum MatrixType

Type of data kept in matrix: either Real (Float) or complex (FloatComplex)

Enumerator:

REAL
COMPLEX
BOOL

Function Documentation

void APlusLambdaB	(	MATRIX	A,
		Float	lambda,
		MATRIX	B,
		MATRIX	C
	)

Calculate A + lambda * B = C

Parameters:

`[in]`	A	A matrix.
`[in]`	B	A matrix.
`[in]`	lambda	A scalar value.
`[in,out]`	C	The result matrix.

MATRIX BItimesV	(	MATRIX	B,
		MATRIX	V
	)

Calculate $F=B^{-1} V$ by solving BF = V for F then solving system of linear equations k times where k is the number of columns in V.

Parameters:

`[in]`	B	A matrix.
`[in]`	V	A matrix.

Returns:: The resulting matrix, F.

MATRIX Cholesky ( MATRIX m )

Cholesky factorization of a symmetric, positive, definite matrix.

Parameters:

[in] m is a matrix to factor.

Returns:: An upper triangular matrix that is the Cholesky factorization of the matrix, m.

MATRIX conjugate ( MATRIX m )

Take the complex conjugate of a matrix. Returns a new matrix.

Parameters:

[in] m Take complex conjugate of this matrix.

Returns:: New matrix containing conjugate of original matrix.

MATRIX ConvertToComplex ( MATRIX m )

Convert a real matrix to a complex matrix.

Parameters:

[in] m A real matrix.

Returns:: A complex matrix.

MATRIX ConvertToReal ( MATRIX m )

Convert a complex matrix to a real matrix. If the imaginary part of the complex matrix elements is nearly zero, the conversion proceeds. Otherwise, an error is thrown.

Parameters:

[in] m A complex matrix.

Returns:: A real matrix.

MATRIX expandDiagonalMatrix ( VECTOR v )

Expand a compact representation of a diagonal matrix.

Parameters:

[in] v A vector containing the diagonal elements of a matrix.

Returns:: A diagonal matrix with diagonal elements taken from input vector.

MATRIX fftw ( fits2d function )

fftw Performs a forward Fourier transform using the Fasted Fourier Transform in the West (FFTW) library.

Parameters:

[in] function A fits2d object which contains data values for a function to transform.

Returns:: A MATRIX object containing the transformed data.

MATRIX fftwReverse	(	MATRIX	function,
		bool	shift
	)

fftwReverse Performs an inverse Fourier transform using the Fasted Fourier Transform in the West (FFTW) library.

Parameters:

`[in]`	function	A MATRIX object which contains data values for a function to transform.
`[in]`	shift	A boolean value whether to shift the data values. true means shift the values. false means no shift.

Returns:: A MATRIX object containing the transformed data.

MATRIX fftwShift ( MATRIX m )

fftwShift shifts data values after a reverse transform. Because of the way FFTW performs the Fourier transform, the data sometimes needs to be shifted.

Parameters:

[in] m A MATRIX containing data to shift.

Returns:: A MATRIX containing shifted data.

void freeMatrix ( MATRIX m )

Free memory that stores matrix data. The matrix structure is not released.

Parameters:

[in] m matrix with data to free.

void freeVector ( VECTOR v )

Free memory that stores vector data. The vector structure is not released.

Parameters:

[in] v Vector with data to free.

MATRIX ifftwShift ( MATRIX m )

ifftwShift (use for odd size matrix) shifts data values after a reverse transform. Because of the way FFTW performs the Fourier transform, the data sometimes needs to be shifted.

Parameters:

[in] m A MATRIX containing data to shift.

Returns:: A MATRIX containing shifted data.

void insertVector	(	Float *	to,
		VECTOR	from
	)

Insert a vector into an array.

Parameters:

`[in,out]`	to	An array.
`[in]`	from	A vector of data to insert.

MATRIX inverse ( MATRIX m )

Inverse of a general matrix.

Parameters:

[in] m A matrix.

Returns:: The inverse of the matrix. If an error occurs during the inversion, the matrix returned has data = NULL, and dimensions are -1.

MATRIX inverseD ( MATRIX m )

Inverse of a diagonal matrix is 1/aii, where aii are the diagonal elements

Parameters:

[in] m A diagonal matrix.

Returns:: The inverse of the matrix

bool isIdentity ( MATRIX m )

Identify a matrix as the identity matrix.

Parameters:

[in] m A real matrix.

Returns:: True if the diagonal elements are unity and all other elements are zero.

MATRIX linearEqSolve	(	MATRIX	A,
		MATRIX	B
	)

Solve Ax=B for x.

Parameters:

`[in]`	A	Matrix
`[in]`	B	Matrix

Returns:: A Matrix containing the solution x.

MATRIX linearEqSolveCholesky	(	MATRIX	A,
		MATRIX	B
	)

Solve Ax=B for x.

Parameters:

`[in]`	A	Matrix
`[in]`	B	Matrix

Returns:: A Matrix containing the solution x.

void linearEqSolveCleanup ( )

Free memory allocated for LUDecomp and IPiv intermediate values.

MATRIX LinearEqSolveNRHS	(	MATRIX	A,
		MATRIX	B
	)

Using this function requires first calling linearEqSolve to calculate the LUDecomp of Matrix A. Solves set of linear equations with multiple right-hand sides, i.e. the number of columns of matrix B.

Parameters:

`[in]`	A	Matrix
`[in]`	B	Matrix

Returns:: A Matrix containing the solution x.

MATRIX linearEqSolveSvd	(	MATRIX	A,
		MATRIX	B
	)

Solve Ax=B for x, using SVD.

Parameters:

`[in]`	A	Matrix
`[in]`	B	Matrix

Returns:: A Matrix containing the solution x.

MATRIX matrix_copy ( MATRIX m )

Return a copy of a matrix. Returns a new matrix.

Parameters:

[in] m A matrix.

Returns:: New matrix containing copy of original matrix.

fits3d MatrixToFits3dObject ( MATRIX m )

Convert from a Matrix object to a fits3d object.

Parameters:

[in] m A matrix.

Returns:: The converted matrix as a fits3d object.

fits2d MatrixToFitsObject ( MATRIX m )

Convert from a Matrix object to a fits2d object.

Parameters:

[in] m A matrix.

Returns:: The converted matrix as a fits2d object; i.e. a 2-d array with different member names and always data of type Float.

VECTOR MatrixToVector ( MATRIX m )

Convert a matrix where one dimension is unity to a vector.

Parameters:

[in] m A matrix with either 1 row or 1 column.

Returns:: The matrix as a vector equivalent.

void mergeMatrix	(	MATRIX *	m,
		MATRIX	real,
		MATRIX	imag
	)

Build a complex matrix from separate real and imaginary parts.

Parameters:

`[in,out]`	m	A complex matrix.
`[in]`	real	A matrix containing real part.
`[in]`	imag	A matrix containing imaginary part.

MATRIX multiplyMM2	(	char	TransA,
		char	TransB,
		MATRIX	A,
		MATRIX	B
	)

Multiply two matrices and return a new matrix containing the product of AB. data.

Parameters:

`[in]`	TransA	Letter 'N', 'T', or 'C', for no transpose, transpose, and conjugate transpose.
`[in]`	TransB	Letter 'N', 'T', or 'C', for no transpose, transpose, and conjugate transpose.
`[in]`	A	Matrix A
`[in]`	B	Matrix B

Returns:: A new matrix AB

MATRIX multiplyMMFull	(	char	TransA,
		char	TransB,
		Float	alpha,
		Float	beta,
		MATRIX	A,
		MATRIX	B
	)

Multiply two matrices and return a new matrix containing the product of AB. data. C = alpha A x B + beta C

Parameters:

`[in]`	TransA	Letter 'N', 'T', or 'C', for no transpose, transpose, and conjugate transpose.
`[in]`	TransB	Letter 'N', 'T', or 'C', for no transpose, transpose, and conjugate transpose.
`[in]`	alpha	Multiply matrix A by scalar alpha.
`[in]`	beta	Multiply matrix C by scalar beta.
`[in]`	A	Matrix A
`[in]`	B	Matrix B

Returns:: A new matrix AB

MATRIX multiplyScalar	(	MATRIX	m,
		Float	scalar
	)

multiply matrix by a scalar value

Parameters:

`[in]`	m	A Matrix.
`[in]`	scalar	A single value.

Returns:: The resulting matrix.

MATRIX multiplyScalarZ	(	MATRIX	m,
		FloatComplex	scalar
	)

multiply matrix by a scalar value

Parameters:

`[in]`	m	A Matrix.
`[in]`	scalar	A single value.

Returns:: The resulting matrix.

MATRIX newComplexMatrix	(	int	r,
		int	c
	)

Creates a new matrix of FloatComplex values and allocates space for data.

Parameters:

`[in]`	r	The number of rows
`[in]`	c	The number of columns

Returns:: A new matrix

MASK newMask	(	int	r,
		int	c,
		int	m
	)

Creates a new 3D matrix and allocates space for the data of type boolean

Parameters:

`[in]`	r	number of rows
`[in]`	c	number of columns
`[in]`	m	number of masks; corresponds to the number of maps

Returns:: A new matrix

MATRIX newMatrix	(	int	r,
		int	c
	)

Creates a new matrix and allocates space for the data of type Float

Parameters:

`[in]`	r	number of rows
`[in]`	c	number of columns

Returns:: A new matrix

MATRIX newMatrixG	(	int	r,
		int	c,
		MatrixType	mType
	)

Creates a new matrix of FloatComplex values and allocates space for data.

Parameters:

`[in]`	r	The number of rows
`[in]`	c	The number of columns
`[in]`	mType	REAL, COMPLEX, BOOL, see MatrixType

Returns:: A new matrix

MATRIX newMatrixGI	(	int	r,
		int	c,
		void *	Idata,
		MatrixType	mType
	)

Creates a new matrix and initializes the data to array sent

Parameters:

`[in]`	r	number of rows
`[in]`	c	number of columns
`[in]`	Idata	point to initial data values
`[in]`	mType	MatrixType either REAL, COMPLEX, or BOOL

Returns:: A new matrix

MATRIX newMatrixI	(	int	r,
		int	c,
		void *	Idata,
		MatrixType	mType
	)

Creates a new matrix and initializes the data to array sent

Parameters:

`[in]`	r	number of rows
`[in]`	c	number of columns
`[in]`	Idata	point to initial data values
`[in]`	mType	MatrixType either REAL or COMPLEX

Returns:: A new matrix

MATRIX newRandomMatrix	(	int	r,
		int	c
	)

Creates a new matrix of random numbers and allocates space for the data of type FloatComplex. NOTE: the seed for the random number generator is set separately.

Parameters:

`[in]`	r	number of rows
`[in]`	c	number of columns

Returns:: A new matrix

VECTOR newVector ( int n )

Creates a new VECTOR.

Parameters:

[in] n the size of the vector

Returns:: A new VECTOR object.

FloatComplex Norm2 ( MATRIX m )

Calculate the ||M||_2 of a matrix, M.

Parameters:

[in] m A matrix.

Returns:: The second norm of matrix.

void print_matrix	(	char *	desc,
		int	m,
		int	n,
		Float *	a,
		int	lda,
		MatrixOrder	order
	)

Print a matrix.

Parameters:

`[in]`	desc	A description or name of the matrix.
`[in]`	m	The number of rows in the matrix.
`[in]`	n	The number of columns in the matrix.
`[in]`	a	An array of the data in the matrix.
`[in]`	lda	Leading dimension of the matrix.
`[in]`	order	Layout of the matrix data, either row-major or column-major.

void print_matrix2	(	char *	desc,
		MATRIX	m,
		MatrixOrder	order
	)

Print a matrix.

Parameters:

`[in]`	desc	A description or name of the matrix.
`[in]`	m	A matrix.
`[in]`	order	Layout of the matrix data, either row-major or column-major.

void print_matrix2toFile	(	char *	desc,
		MATRIX	m,
		MatrixOrder	order
	)

Print a matrix to a file.

Parameters:

`[in]`	desc	A description or name of the matrix. This is also the name of the file.
`[in]`	m	A matrix.
`[in]`	order	Layout of the matrix data, either row-major or column-major.

VECTOR readText ( char * inputFilename )

Read a vector from a file Format of the file can be eol, tab, comma, space, separated list of numbers

Parameters:

[in] inputFilename Name of file to read.

Returns:: Vector object containing data from file.

void setToNaN ( VECTOR v )

Set a vector to NaN.

Parameters:

[in,out] v A vector.

void setToNaNMatrix ( MATRIX m )

Set a matrix to NaN.

Parameters:

[in,out] m A matrix.

void splitMatrix	(	MATRIX	m,
		MATRIX *	real,
		MATRIX *	imag
	)

Split a complex matrix into real and imaginary parts.

Parameters:

`[in]`	m	A complex matrix.
`[in,out]`	real	A matrix containing real part.
`[in,out]`	imag	A matrix containing imaginary part.

void tdCleanup ( )

free memory used by static variables internal to inversion

void tdinvert1dFinish	(	Float	lambda,
		int	nStep,
		Model *	model,
		bool	debug,
		BaseDirectory *	debugFilename
	)

Completes the inversion calculation. By sending the individual lambda values the inversion calculation is finished.

Parameters:

`[in]`	lambda	regularization parameter
`[in]`	nStep	Nth step in lambda-parameter range
`[in]`	debug	Flag to print intermediate matrices during inversion
`[in]`	debugFilename	Base directory and filename for intermediate, debug values.
`[in,out]`	model	Model structure contains all calculated solutions including averaging kernel and error covariance.

void tdinvert3dFinish	(	Float	lambda,
		TempData *	tempData,
		bool	debug,
		BaseDirectory *	debugFilename
	)

Completes the inversion calculation. By sending the individual lambda values the inversion calculation is finished.

Parameters:

`[in]`	lambda	regularization parameter
`[in]`	debug	Flag to print intermediate matrices during inversion
`[in]`	debugFilename	Base directory and filename for intermediate, debug values.
`[in,out]`	tempData	TempData structure contains all calculated solutions including averaging kernel and error covariance for one wave number.

int tdSetup1d	(	MATRIX	aM,
		MATRIX	ad,
		MATRIX	aR,
		MATRIX	aCI,
		MATRIX	aC
	)

tdSetup calculates all the non-changing, static matrices

Parameters:

`[in]`	aM	Kernel matrix
`[in]`	ad	map data
`[in]`	aR	Regularization matrix; symmetric positive-definite matrix
`[in]`	aCI	Inverse of matrix C, a noise covariance matrix
`[in]`	aC	matrix C, a noise covariance matrix

Float* toColumnMajor	(	Float *	AT,
		int	R,
		int	C
	)

Copy a 2D array in row-major layout return the copy in column-major layout.

Parameters:

`[in]`	AT	The array of data in row-major layout.
`[in]`	R	The number of rows.
`[in]`	C	The number of columns.

Returns:: A 2D array in column-major layout.

MATRIX toColumnMajor2 ( MATRIX m )

Copy a matrix stored in row-major order and convert it to column-major order.

Parameters:

[in] m A row-major matrix

Returns:: A copy of matrix m with column-major layout of data.

fits2d toFitsObject ( VECTOR v )

Convert from a Vector object to a fits2d object.

Parameters:

[in] v A vector

Returns:: The converted vector as a fits2d object; i.e. a 2-d array with different member names and always data of type Float.

MATRIX toMatrix ( fits2d f )

Convert a fits2d object to a Matrix object.

Parameters:

[in] f A fits2d object; i.e. a 2-d array with different member names and always data of type Float.

Returns:: The converted object as a matrix.

Float* toRowMajor	(	Float *	A,
		int	R,
		int	C
	)

Copy a 2D array in column-major layout return the copy in row-major layout.

Parameters:

`[in]`	A	The array of data in column-major layout.
`[in]`	R	The number of rows.
`[in]`	C	The number of columns.

Returns:: A 2D array in row-major layout.

MATRIX toRowMajor2 ( MATRIX m )

Copy a matrix stored in column-major order and convert it to row-major order.

Parameters:

[in] m A column-major matrix

Returns:: A copy of matrix n with row-major layout of data.

VECTOR toVector ( fits2d f )

Convert a fits2d object to a Vector object.

Parameters:

[in] f A fits2d object; i.e. a 2-d array with different member names and always data of type Float.

Returns:: The converted object as a vector.

MATRIX transpose ( MATRIX m )

Transpose a matrix.

Parameters:

[in] m A matrix.

Returns:: The transpose of the matrix.

MATRIX VectorToMatrix_Column ( VECTOR v )

Convert vector to a matrix with a single column vector.

Parameters:

[in] v A vector.

Returns:: A matrix containing the vector as a column.

Matrix - Matrix creates and manipulates data in 1D and 2D matrix form. [Supporting Modules]

Data Structures

Enumerations

Functions

Enumeration Type Documentation

Function Documentation

Matrix - Matrix creates and manipulates data in 1D and 2D matrix form.
[Supporting Modules]