BIAS::Matrix< T > Class Template Reference
[Matrix and Vector Classes]

matrix class with arbitrary size, indexing is row major. More...

#include <Base/Math/Matrix.hh>

Inheritance diagram for BIAS::Matrix< T >:

Collaboration diagram for BIAS::Matrix< T >:

List of all members.

Public Types

typedef const T * const_iterator

typedef const T & const_reference

typedef T element_type

typedef T * iterator

typedef T * pointer

typedef T & reference

typedef Subscript size_type

typedef T value_type

Public Member Functions

Matrix< T > Adjoint () const

computes the adjoint matrix

Subscript dim (Subscript d) const

void GaussJordan ()

use the Gauss Jordan Algrithm to transform the matrix to reduced row echelon form.

template<>

void GetAbsMaxMin (double &max, double &min) const

template<>

bool IsIdentity (double eps) const

bool IsIdentity (double eps=0.0) const

Checks if the matrix a an identity.

template<>

bool IsZero (double eps) const

bool IsZero (double eps=0.0) const

Checks if the matrix is a null matrix.

void Kronecker (Matrix< T > const B, Matrix< T > &dest) const

Kronecker-product with matrix, result in dest.

Subscript lbound () const

Matrix (const Vector< T > &v)

constructs Nx1 Matrix from N-Vector

Matrix (const TNT::Matrix< T > &A)

Matrix (const Matrix< T > &A)

Matrix (const int rows, const int cols, const std::string &s)

DEPRECATED due to instantiation problem.

Matrix (int rows, int cols, const MatrixInitType &i)

init with standard form

Matrix (int rows, int cols, const T value)

assignment with a constant values for all elements (=set)

Matrix (int rows, int cols)

Matrix (const Matrix3x3< T > &mat)

Matrix ()

Matrix< T > & newsize (Subscript M, Subscript N)

Subscript num_cols () const

Subscript num_rows () const

const_reference operator() (Subscript i, Subscript j) const

reference operator() (Subscript i, Subscript j)

const_reference operator() (Subscript i) const

reference operator() (Subscript i)

const T * operator[] (Subscript i) const

T * operator[] (Subscript i)

std::ostream & Print (std::ostream &s, const bool intCastOutput=false, const bool binaryOutput=false, const bool forceFullPrecision=false) const

std::istream & Read (std::istream &s, const bool intCastInput=false, const bool binaryInput=false)

JW.

Subscript size () const

void SwapRows (const int i, const int r)

swaps two rows

Matrix< T > Transpose () const

transpose function, storing data destination matrix

void Vec (Vector< T > &dest) const

vec-operator returns the elements of the matrix columwise as vector

virtual ~Matrix ()

Arithmetic

void Add (const T &scalar, Matrix< T > &dest) const

addition function, storing data destination matrix

void AddIP (const Matrix< T > &arg)

Adds arg to this.

void AddIP (const T &scalar)

in place addition function

void Divide (const T &scalar, Matrix< T > &dest) const

division function, storing data destination matrix

void DivideIP (const T &scalar)

in place division function

void GetSystemMatrix (Matrix< T > &dest) const

compute square system matrix dest = A^T * A

void Mult (const Vector< T > &arg, Vector< T > &result) const

matrix vector multiplication, result is not allocated

void Mult (const Matrix< T > &arg)

in Place matrix multiplication this is equal to M = M * arg, but faster

void Mult (const Matrix< T > &arg, Matrix< T > &result) const

matrix multiplication, result is not allocated

void Multiply (const T &scalar, Matrix< T > &dest) const

multiplication function, storing data destination matrix

void MultiplyIP (const T &scalar)

in place multiplication function

void MultLeft (const Vector< T > &arg, Vector< T > &result) const

vector matrix multiplication result=arg*this.

void MultLeft (const Matrix< T > &arg)

in Place matrix multiplication this is equal to M = arg*M, but faster

Matrix< T > & operator= (const Matrix< T > &mat)

Matrix< T > & operator= (const TNT::Matrix< T > &mat)

assignment operators calling corresponding operator from base class if appropriate

void Sub (const T &scalar, Matrix< T > &dest) const

substraction function, storing data destination matrix

void SubIP (const T &scalar)

in place subtraction function

void SubIP (const Matrix< T > &arg)

Subtracts arg from this
this -= arg.

Input / Output

bool BinRead (const char *filename) const

This method reads a matrix from a given file in binary format.

bool BinWrite (const char *filename) const

This method writes a matrix to a given file in binary format.

bool Load (const std::string &filename)

method to load directly from a given filename.

bool Save (const std::string &filename) const

method to save directly to a given filename.

int WriteMatlab (std::ostream &ostr, const std::string &name) const

Write the matrix in Matlab format to the given stream.

Set Functions

virtual void clear ()

stl conform interface destroys Matrix JW

void Fill (const T &scalar)

Takes the elements of a Vector and put them as diagonal elements into a matrix.

void MakeSymmetric ()

componentwise: this = 0.5(this + this^T) yields symmetric matrix only allowed for square shaped matrices

void SetIdentity ()

Converts matrix to identity matrix.

void SetZero ()

Sets all values to zero.

Get Functions

T DetSquare () const

void GetAbsMaxMin (T &max, T &min) const

return biggest and smallest absolute values

Vector< T > GetCol (const int &col) const

return a copy of column "col", zero based counting

unsigned int GetCols () const

const T * GetData () const

T * GetData ()

get the pointer to the data array of the matrix (for faster direct memeory access)

T ** GetDataArray ()

const T ** GetDataArray () const

returns zero based arry for data access

T ** GetDataArray1 () const

returns 1 based array to data access

T * GetDataLast ()

const T * GetDataLast () const

Get a pointer to the last data element Do not use this on unitilized matrices with <= on pointers because on an unitilized matrix GetData() will be NULL and last Element < (typically FFFFF8) We don't want an if here for performance reasons of the innerst loop.

T GetMax () const

Returns the maximum value of the matrix elements.

void GetMaxMin (T &max, T &min) const

return biggest and smallest entry

T GetMean () const

Returns the mean value of the matrix elements.

T GetMin () const

Returns the minimum value of the matrix elements.

int GetNumElements () const

Returns the total number of elements.

Vector< T > GetRow (const int &row) const

return a copy of row "row" of this matrix, zero based counting

unsigned int GetRows () const

double NormFrobenius () const

Return Frobenius norm = sqrt(trace(A^t * A)).

T NormL1 () const

Return the L1 norm: |a| + |b| + |c| + .

double NormL2 () const

Return the L2 norm: a^2 + b^2 + c^2 + .

void Set (const int row, const int col, const Vector< T > &data)

Copies the contents of data into this at the position indicated by row and col.

void Set (const int row, const int col, const Matrix< T > &data)

Copies the contents of data into this at the position indicated by row and col.

void SetCol (const int row, const Vector< T > &data)

set a col of matrix from vector

void SetRow (const int row, const Vector< T > &data)

set a row of matrix from vector

void SetTranspose (const int row, const int col, const Vector< T > &data)

Copies the contents of data into this at the position indicated by row and col.

T Trace () const

Scaling and Normalization

T Normalize ()

Normalizes the matrix by the entry with the biggest absolute value by dividing all elements with this one.

void NormalizeCols ()

Normalizes each coloumn to L_2 norm one.

void NormalizeColsToOne (const int norm_row_index)

divides each column of the matrix through the element norm_row_index.

void NormalizeRows ()

Normalizes each row to L2 norm one.

void NormalizeRowsToOne (const int norm_col_index)

divides each row of the matrix through the element norm_col_index.

void ScaleCol (int NoCol, T scale)

Scales column NoCol with scale.

void ScaleRow (int NoRow, T scale)

Scales row NoRow with scale.

Protected Member Functions

void copy (const T *v)

void destroy ()

void initialize (Subscript M, Subscript N)

void set (const T &val)

Protected Attributes

Subscript m_

Subscript mn_

Subscript n_

T ** row_

T ** rowm1_

T * v_

T * vm1_

Related Functions

(Note that these are not member functions.)

template<class T >

void Conv (const Matrix< T > &left, const Matrix< T > &right, Matrix< T > &res)

zero padded convolution of two matrices

template<class T >

bool operator!= (const Matrix< T > &argl, const Matrix< T > &argr)

template<class T >

Vector< T > operator* (const Matrix< T > &A, const Vector< T > &x)

template<class T >

Matrix< T > operator* (const Matrix< T > &A, const Matrix< T > &B)

template<class T >

Matrix< T > operator* (const Matrix3x3< T > &argLeft, const Matrix< T > &argMat)

template<class T >

Matrix< T > operator* (const T &scalar, const Matrix< T > &mat)

template<class T >

Matrix< T > operator* (const Matrix< T > &mat, const T &scalar)

template<class T >

Matrix< T > & operator*= (Matrix< T > &mat, const T &scalar)

template<class T >

Matrix< T > operator+ (const Matrix< T > &A, const Matrix< T > &B)

template<class T >

Matrix< T > operator+ (const T &scalar, const Matrix< T > &mat)

template<class T >

Matrix< T > operator+ (const Matrix< T > &mat, const T &scalar)

template<class T >

Matrix< T > & operator+= (Matrix< T > &mat, const Matrix< T > &arg)

template<class T >

Matrix< T > & operator+= (Matrix< T > &mat, const T &scalar)

operators

template<class T >

Matrix< T > operator- (const Matrix< T > &A, const Matrix< T > &B)

template<class T >

Matrix< T > operator- (const T &scalar, const Matrix< T > &mat)

template<class T >

Matrix< T > operator- (const Matrix< T > &mat, const T &scalar)

template<class T >

Matrix< T > & operator-= (Matrix< T > &mat, const T &scalar)

template<class T >

Matrix< T > & operator-= (Matrix< T > &mat, const Matrix< T > &arg)

template<class T >

Matrix< T > operator/ (const Matrix< T > &mat, const T &scalar)

template<class T >

Matrix< T > & operator/= (Matrix< T > &mat, const T &scalar)

template<class T >

std::ostream & operator<< (std::ostream &s, const Matrix< T > &A)

std::ostream & operator<< (std::ostream &s, const Matrix< unsigned char > &A)

std::ostream & operator<< (std::ostream &s, const Matrix< char > &A)

template<class T >

bool operator== (const Matrix< T > &argl, const Matrix< T > &argr)

template<class T >

std::istream & operator>> (std::istream &s, Matrix< T > &A)

std::istream & operator>> (std::istream &s, Matrix< char > &A)

std::istream & operator>> (std::istream &s, Matrix< unsigned char > &A)

Detailed Description

template<class T = double>
class BIAS::Matrix< T >

matrix class with arbitrary size, indexing is row major.

Access the data by M[row][col].

class Matrix is the matrix class that should be used by common BIAS algorithms and data structures with arbitrary size

Access the data simply by M[row][col].

It is derived from TNT::Matrix to inherit the basic operations and algorithm. special implementations should be done here and NOT in TNT:Matrix because the true 'base' Matrix class "TNT::Matrix" should be interchangeable and aside from this maintained by NIST.

Note the specialized fixed size classes Matrix3x3, Matrix3x4, ... with more efficient memory layout.

moved from MIP to BIAS (11/05/2002) operators are no longer class members, they are defined outside the class.

Author:: Jan Woetzel, woelk

Definition at line 86 of file Matrix.hh.

Member Typedef Documentation

template<class T >

typedef const T* TNT::Matrix< T >::const_iterator [inherited]

Definition at line 77 of file cmat.h.

template<class T >

typedef const T& TNT::Matrix< T >::const_reference [inherited]

Definition at line 78 of file cmat.h.

template<class T >

typedef T TNT::Matrix< T >::element_type [inherited]

Definition at line 73 of file cmat.h.

template<class T >

typedef T* TNT::Matrix< T >::iterator [inherited]

Definition at line 75 of file cmat.h.

template<class T >

typedef T* TNT::Matrix< T >::pointer [inherited]

Definition at line 74 of file cmat.h.

template<class T >

typedef T& TNT::Matrix< T >::reference [inherited]

Definition at line 76 of file cmat.h.

template<class T >

typedef Subscript TNT::Matrix< T >::size_type [inherited]

Definition at line 71 of file cmat.h.

template<class T >

typedef T TNT::Matrix< T >::value_type [inherited]

Definition at line 72 of file cmat.h.

Constructor & Destructor Documentation

template<class T >

BIAS::Matrix< T >::~Matrix ( ) [inline, virtual]

Reimplemented from TNT::Matrix< T >.

Definition at line 60 of file Matrix.cpp.

template<class T = double>

BIAS::Matrix< T >::Matrix ( ) [inline]

Reimplemented from TNT::Matrix< T >.

Definition at line 91 of file Matrix.hh.

template<class T >

BIAS::Matrix< T >::Matrix ( const Matrix3x3< T > & mat ) [inline, explicit]

Definition at line 69 of file Matrix.cpp.

References r.

template<class T = double>

BIAS::Matrix< T >::Matrix	(	int	rows,
		int	cols
	)			`[inline]`

Author:: Jan Woetzel

Definition at line 97 of file Matrix.hh.

template<class T = double>

BIAS::Matrix< T >::Matrix	(	int	rows,
		int	cols,
		const T	value
	)			`[inline]`

assignment with a constant values for all elements (=set)

Author:: Jan Woetzel

Definition at line 101 of file Matrix.hh.

template<class T = double>

BIAS::Matrix< T >::Matrix	(	int	rows,
		int	cols,
		const MatrixInitType &	i
	)			`[inline]`

init with standard form

Author:: koeser

Definition at line 106 of file Matrix.hh.

template<class T >

BIAS::Matrix< T >::Matrix	(	const int	rows,
		const int	cols,
		const std::string &	s
	)			`[inline, explicit]`

DEPRECATED due to instantiation problem.

Author:: Jan Woetzel replacement for the above to be called as Matrix<>(2,3, "1 2 3 4 5 6")
Jan Woetzel

Definition at line 64 of file Matrix.cpp.

template<class T = double>

BIAS::Matrix< T >::Matrix ( const Matrix< T > & A ) [inline]

Author:: Jan Woetzel

Definition at line 130 of file Matrix.hh.

template<class T = double>

BIAS::Matrix< T >::Matrix ( const TNT::Matrix< T > & A ) [inline]

Author:: Jan Woetzel

Reimplemented from TNT::Matrix< T >.

Definition at line 133 of file Matrix.hh.

template<class T = double>

BIAS::Matrix< T >::Matrix ( const Vector< T > & v ) [inline]

constructs Nx1 Matrix from N-Vector

Definition at line 136 of file Matrix.hh.

Member Function Documentation

template<class T >

void BIAS::Matrix< T >::Add	(	const T &	scalar,
		Matrix< T > &	dest
	)			const `[inline]`

addition function, storing data destination matrix

implementation

Author:: Ingo Thomsen tested

Definition at line 668 of file Matrix.hh.

References BIAS::Matrix< T >::GetData(), BIAS::Matrix< T >::GetDataLast(), TNT::Matrix< T >::newsize(), TNT::Matrix< T >::num_cols(), TNT::Matrix< T >::num_rows(), and T.

Referenced by BIAS::Matrix< T >::operator+().

template<class T >

void BIAS::Matrix< T >::AddIP ( const Matrix< T > & arg ) [inline]

Adds arg to this.

Author:: grest,2004

Definition at line 907 of file Matrix.hh.

References BIASABORT, BIASERR, BIAS::Matrix< T >::GetData(), TNT::Matrix< T >::num_cols(), TNT::Matrix< T >::num_rows(), TNT::Matrix< T >::size(), and T.

template<class T = double>

void BIAS::Matrix< T >::AddIP ( const T & scalar ) [inline]

in place addition function

Author:: Ingo Thomsen tested

Definition at line 339 of file Matrix.hh.

Referenced by BIAS::Matrix< T >::operator+=(), and BIAS::Matrix< T >::operator-=().

template<class T >

Matrix< T > BIAS::Matrix< T >::Adjoint ( ) const [inline]

computes the adjoint matrix

Author:: Christian Beder

Definition at line 734 of file Matrix.hh.

References BIASASSERT, BIAS::Matrix< T >::DetSquare(), j, TNT::Matrix< T >::num_cols(), TNT::Matrix< T >::num_rows(), and r.

template<class T >

bool BIAS::Matrix< T >::BinRead ( const char * filename ) const [inline]

This method reads a matrix from a given file in binary format.

The file is not human read-/editable but provides full precision. A dimension check is performed.

Returns:: true on success

Author:: mdunda 04 2004

Definition at line 425 of file Matrix.cpp.

References BIASERR, j, TNT::Matrix< T >::m_, TNT::Matrix< T >::n_, and T.

template<class T >

bool BIAS::Matrix< T >::BinWrite ( const char * filename ) const [inline]

This method writes a matrix to a given file in binary format.

The file is not human read-/editable but provides full precision.

Returns:: true on success

Author:: mdunda 04 2004

Definition at line 460 of file Matrix.cpp.

References j, TNT::Matrix< T >::m_, TNT::Matrix< T >::n_, and T.

template<class T = double>

virtual void BIAS::Matrix< T >::clear ( ) [inline, virtual]

stl conform interface destroys Matrix JW

Reimplemented in BIAS::Matrix2x2< T >, BIAS::Matrix3x4< T >, BIAS::Matrix4x4< T >, BIAS::Matrix2x2< KLT_TYPE >, BIAS::Matrix2x2< double >, BIAS::Matrix3x4< double >, BIAS::Matrix3x4< PMATRIX_TYPE >, BIAS::Matrix4x4< double >, BIAS::Matrix4x4< QUADRIC3D_TYPE >, BIAS::Matrix4x4< PLUECKERMATRIX_TYPE >, and BIAS::Matrix4x4< PMATRIX_TYPE >.

Definition at line 449 of file Matrix.hh.

Referenced by BIAS::ExtendedKalman::Update().

template<class T >

void TNT::Matrix< T >::copy ( const T * v ) [inline, protected, inherited]

Definition at line 121 of file cmat.h.

Referenced by TNT::Matrix< BIAS::Vector< int > >::Matrix(), and TNT::Matrix< BIAS::Vector< int > >::operator=().

template<class T >

void TNT::Matrix< T >::destroy ( ) [inline, protected, inherited]

Definition at line 180 of file cmat.h.

Referenced by TNT::Matrix< BIAS::Vector< int > >::newsize(), TNT::Matrix< BIAS::Vector< int > >::operator=(), and TNT::Matrix< BIAS::Vector< int > >::~Matrix().

template<class T = double>

T BIAS::Matrix< T >::DetSquare ( ) const [inline]

Author:: koeser

Warning:: very slow, generic implementation (order "n!"), better to use matrix decomposition (see BIAS/MathAlgo/Lapack.hh)

Definition at line 293 of file Matrix.hh.

Referenced by BIAS::Matrix< T >::Adjoint(), and BIAS::Matrix< EP_TYPE >::DetSquare().

template<class T >

Subscript TNT::Matrix< T >::dim ( Subscript d ) const [inline, inherited]

Definition at line 309 of file cmat.h.

template<class T >

void BIAS::Matrix< T >::Divide	(	const T &	scalar,
		Matrix< T > &	dest


Public Types
typedef const T *	const_iterator
typedef const T &	const_reference
typedef T	element_type
typedef T *	iterator
typedef T *	pointer
typedef T &	reference
typedef Subscript	size_type
typedef T	value_type
Public Member Functions
Matrix< T >	Adjoint () const
	computes the adjoint matrix
Subscript	dim (Subscript d) const
void	GaussJordan ()
	use the Gauss Jordan Algrithm to transform the matrix to reduced row echelon form.
template<>
void	GetAbsMaxMin (double &max, double &min) const
template<>
bool	IsIdentity (double eps) const
bool	IsIdentity (double eps=0.0) const
	Checks if the matrix a an identity.
template<>
bool	IsZero (double eps) const
bool	IsZero (double eps=0.0) const
	Checks if the matrix is a null matrix.
void	Kronecker (Matrix< T > const B, Matrix< T > &dest) const
	Kronecker-product with matrix, result in dest.
Subscript	lbound () const
	Matrix (const Vector< T > &v)
	constructs Nx1 Matrix from N-Vector
	Matrix (const TNT::Matrix< T > &A)
	Matrix (const Matrix< T > &A)
	Matrix (const int rows, const int cols, const std::string &s)
	DEPRECATED due to instantiation problem.
	Matrix (int rows, int cols, const MatrixInitType &i)
	init with standard form
	Matrix (int rows, int cols, const T value)
	assignment with a constant values for all elements (=set)
	Matrix (int rows, int cols)
	Matrix (const Matrix3x3< T > &mat)
	Matrix ()
Matrix< T > &	newsize (Subscript M, Subscript N)
Subscript	num_cols () const
Subscript	num_rows () const
const_reference	operator() (Subscript i, Subscript j) const
reference	operator() (Subscript i, Subscript j)
const_reference	operator() (Subscript i) const
reference	operator() (Subscript i)
const T *	operator[] (Subscript i) const
T *	operator[] (Subscript i)
std::ostream &	Print (std::ostream &s, const bool intCastOutput=false, const bool binaryOutput=false, const bool forceFullPrecision=false) const
std::istream &	Read (std::istream &s, const bool intCastInput=false, const bool binaryInput=false)
	JW.
Subscript	size () const
void	SwapRows (const int i, const int r)
	swaps two rows
Matrix< T >	Transpose () const
	transpose function, storing data destination matrix
void	Vec (Vector< T > &dest) const
	vec-operator returns the elements of the matrix columwise as vector
virtual	~Matrix ()
Arithmetic
void	Add (const T &scalar, Matrix< T > &dest) const
	addition function, storing data destination matrix
void	AddIP (const Matrix< T > &arg)
	Adds arg to this.
void	AddIP (const T &scalar)
	in place addition function
void	Divide (const T &scalar, Matrix< T > &dest) const
	division function, storing data destination matrix
void	DivideIP (const T &scalar)
	in place division function
void	GetSystemMatrix (Matrix< T > &dest) const
	compute square system matrix dest = A^T * A
void	Mult (const Vector< T > &arg, Vector< T > &result) const
	matrix vector multiplication, result is not allocated
void	Mult (const Matrix< T > &arg)
	in Place matrix multiplication this is equal to M = M * arg, but faster
void	Mult (const Matrix< T > &arg, Matrix< T > &result) const
	matrix multiplication, result is not allocated
void	Multiply (const T &scalar, Matrix< T > &dest) const
	multiplication function, storing data destination matrix
void	MultiplyIP (const T &scalar)
	in place multiplication function
void	MultLeft (const Vector< T > &arg, Vector< T > &result) const
	vector matrix multiplication result=arg*this.
void	MultLeft (const Matrix< T > &arg)
	in Place matrix multiplication this is equal to M = arg*M, but faster
Matrix< T > &	operator= (const Matrix< T > &mat)
Matrix< T > &	operator= (const TNT::Matrix< T > &mat)
	assignment operators calling corresponding operator from base class if appropriate
void	Sub (const T &scalar, Matrix< T > &dest) const
	substraction function, storing data destination matrix
void	SubIP (const T &scalar)
	in place subtraction function
void	SubIP (const Matrix< T > &arg)
	Subtracts arg from this this -= arg.
Input / Output
bool	BinRead (const char *filename) const
	This method reads a matrix from a given file in binary format.
bool	BinWrite (const char *filename) const
	This method writes a matrix to a given file in binary format.
bool	Load (const std::string &filename)
	method to load directly from a given filename.
bool	Save (const std::string &filename) const
	method to save directly to a given filename.
int	WriteMatlab (std::ostream &ostr, const std::string &name) const
	Write the matrix in Matlab format to the given stream.
Set Functions
virtual void	clear ()
	stl conform interface destroys Matrix JW
void	Fill (const T &scalar)
	Takes the elements of a Vector and put them as diagonal elements into a matrix.
void	MakeSymmetric ()
	componentwise: this = 0.5(this + this^T) yields symmetric matrix only allowed for square shaped matrices
void	SetIdentity ()
	Converts matrix to identity matrix.
void	SetZero ()
	Sets all values to zero.
Get Functions
T	DetSquare () const
void	GetAbsMaxMin (T &max, T &min) const
	return biggest and smallest absolute values
Vector< T >	GetCol (const int &col) const
	return a copy of column "col", zero based counting
unsigned int	GetCols () const
const T *	GetData () const
T *	GetData ()
	get the pointer to the data array of the matrix (for faster direct memeory access)
T **	GetDataArray ()
const T **	GetDataArray () const
	returns zero based arry for data access
T **	GetDataArray1 () const
	returns 1 based array to data access
T *	GetDataLast ()
const T *	GetDataLast () const
	Get a pointer to the last data element Do not use this on unitilized matrices with <= on pointers because on an unitilized matrix GetData() will be NULL and last Element < (typically FFFFF8) We don't want an if here for performance reasons of the innerst loop.
T	GetMax () const
	Returns the maximum value of the matrix elements.
void	GetMaxMin (T &max, T &min) const
	return biggest and smallest entry
T	GetMean () const
	Returns the mean value of the matrix elements.
T	GetMin () const
	Returns the minimum value of the matrix elements.
int	GetNumElements () const
	Returns the total number of elements.
Vector< T >	GetRow (const int &row) const
	return a copy of row "row" of this matrix, zero based counting
unsigned int	GetRows () const
double	NormFrobenius () const
	Return Frobenius norm = sqrt(trace(A^t * A)).
T	NormL1 () const
	Return the L1 norm: \|a\| + \|b\| + \|c\| + .
double	NormL2 () const
	Return the L2 norm: a^2 + b^2 + c^2 + .
void	Set (const int row, const int col, const Vector< T > &data)
	Copies the contents of data into this at the position indicated by row and col.
void	Set (const int row, const int col, const Matrix< T > &data)
	Copies the contents of data into this at the position indicated by row and col.
void	SetCol (const int row, const Vector< T > &data)
	set a col of matrix from vector
void	SetRow (const int row, const Vector< T > &data)
	set a row of matrix from vector
void	SetTranspose (const int row, const int col, const Vector< T > &data)
	Copies the contents of data into this at the position indicated by row and col.
T	Trace () const
Scaling and Normalization
T	Normalize ()
	Normalizes the matrix by the entry with the biggest absolute value by dividing all elements with this one.
void	NormalizeCols ()
	Normalizes each coloumn to L_2 norm one.
void	NormalizeColsToOne (const int norm_row_index)
	divides each column of the matrix through the element norm_row_index.
void	NormalizeRows ()
	Normalizes each row to L2 norm one.
void	NormalizeRowsToOne (const int norm_col_index)
	divides each row of the matrix through the element norm_col_index.
void	ScaleCol (int NoCol, T scale)
	Scales column NoCol with scale.
void	ScaleRow (int NoRow, T scale)
	Scales row NoRow with scale.
Protected Member Functions
void	copy (const T *v)
void	destroy ()
void	initialize (Subscript M, Subscript N)
void	set (const T &val)
Protected Attributes
Subscript	m_
Subscript	mn_
Subscript	n_
T **	row_
T **	rowm1_
T *	v_
T *	vm1_
Related Functions
(Note that these are not member functions.)
template<class T >
void	Conv (const Matrix< T > &left, const Matrix< T > &right, Matrix< T > &res)
	zero padded convolution of two matrices
template<class T >
bool	operator!= (const Matrix< T > &argl, const Matrix< T > &argr)
template<class T >
Vector< T >	operator* (const Matrix< T > &A, const Vector< T > &x)
template<class T >
Matrix< T >	operator* (const Matrix< T > &A, const Matrix< T > &B)
template<class T >
Matrix< T >	operator* (const Matrix3x3< T > &argLeft, const Matrix< T > &argMat)
template<class T >
Matrix< T >	operator* (const T &scalar, const Matrix< T > &mat)
template<class T >
Matrix< T >	operator* (const Matrix< T > &mat, const T &scalar)
template<class T >
Matrix< T > &	operator*= (Matrix< T > &mat, const T &scalar)
template<class T >
Matrix< T >	operator+ (const Matrix< T > &A, const Matrix< T > &B)
template<class T >
Matrix< T >	operator+ (const T &scalar, const Matrix< T > &mat)
template<class T >
Matrix< T >	operator+ (const Matrix< T > &mat, const T &scalar)
template<class T >
Matrix< T > &	operator+= (Matrix< T > &mat, const Matrix< T > &arg)
template<class T >
Matrix< T > &	operator+= (Matrix< T > &mat, const T &scalar)
	operators
template<class T >
Matrix< T >	operator- (const Matrix< T > &A, const Matrix< T > &B)
template<class T >
Matrix< T >	operator- (const T &scalar, const Matrix< T > &mat)
template<class T >
Matrix< T >	operator- (const Matrix< T > &mat, const T &scalar)
template<class T >
Matrix< T > &	operator-= (Matrix< T > &mat, const T &scalar)
template<class T >
Matrix< T > &	operator-= (Matrix< T > &mat, const Matrix< T > &arg)
template<class T >
Matrix< T >	operator/ (const Matrix< T > &mat, const T &scalar)
template<class T >
Matrix< T > &	operator/= (Matrix< T > &mat, const T &scalar)
template<class T >
std::ostream &	operator<< (std::ostream &s, const Matrix< T > &A)
std::ostream &	operator<< (std::ostream &s, const Matrix< unsigned char > &A)
std::ostream &	operator<< (std::ostream &s, const Matrix< char > &A)
template<class T >
bool	operator== (const Matrix< T > &argl, const Matrix< T > &argr)
template<class T >
std::istream &	operator>> (std::istream &s, Matrix< T > &A)
std::istream &	operator>> (std::istream &s, Matrix< char > &A)
std::istream &	operator>> (std::istream &s, Matrix< unsigned char > &A)

BIAS::Matrix< T > Class Template Reference [Matrix and Vector Classes]

Public Types

Public Member Functions

Protected Member Functions

Protected Attributes

Related Functions

Detailed Description

template<class T = double> class BIAS::Matrix< T >

Member Typedef Documentation

Constructor & Destructor Documentation

Member Function Documentation

BIAS::Matrix< T > Class Template Reference
[Matrix and Vector Classes]

template<class T = double>
class BIAS::Matrix< T >