STK++ 0.9.13
STK::HDMatrixModelParameters< Array_ > Struct Template Reference

structure storing the parameters of the HD matrix valued mixture More...

#include <STK_HDMatrixGaussianParameters.h>

Public Member Functions

 HDMatrixModelParameters (int nbCluster)
 default constructor
 
 HDMatrixModelParameters (HDMatrixModelParameters const &param)
 copy constructor.
 
 ~HDMatrixModelParameters ()
 destructor
 
HDMatrixModelParametersoperator= (HDMatrixModelParameters const &param)
 copy operator.
 
void resize (Range const &rangeRows, Range const &rangeCols)
 resize and initialize the set of parameter.
 
void updateStatistics ()
 update statistics of the parameters.
 
void setStatistics ()
 Set the computed statistics.
 
void releaseStatistics ()
 Release the computed statistics.
 
template<class Array >
void setParameters (ExprBase< Array > const &params)
 Set the parameters of the mixture model.
 

Public Attributes

Array1D< Array_meank_
 array of size nbCluster with for each cluster the mean matrix
 
Array1D< CSquareXrowQk_
 array of size nbCluster with the rotation matrix for rows
 
Array1D< CVectorXrowAjk_
 array of size nbCluster with inertia in low dimensional space for rows
 
Array1D< RealrowBk_
 array of size nbCluster with remaining variance noise for rows
 
CVectorXi rowDk_
 array of size nbCluster with dimension of low dimensional space for rows
 
Array1D< CSquareXcolQk_
 array of size nbCluster with the rotation matrix for columns
 
Array1D< CVectorXcolAjk_
 array of size nbCluster with inertia in low dimensional space for columns
 
Array1D< RealcolBk_
 array of size nbCluster with variance noise for columns
 
CVectorXi colDk_
 array of size nbCluster with dimension of low dimensional space for columns
 
Array1D< Stat::Online< Array_, Real > > statMeank_
 Array of size nbCluster of the mean statistics.
 
Array1D< Stat::Online< CSquareX, Real > > statRowQk_
 Array of size nbCluster of the rotation matrix statistics for rows.
 
Array1D< Stat::Online< CVectorX, Real > > statRowAjk_
 Array of size nbCluster of the rotation matrix statistics for rows.
 
Array1D< Stat::Online< Real, Real > > statRowBk_
 Array of size nbCluster with the variance noise statistics for rows.
 
Array1D< Stat::Online< CSquareX, Real > > statColQk_
 Array of the rotation matrix statistics for rows.
 
Array1D< Stat::Online< CVectorX, Real > > statColAjk_
 Array of the standard deviation statistics.
 
Array1D< Stat::Online< Real, Real > > statColBk_
 Array of the standard deviation statistics.
 

Detailed Description

template<class Array_>
struct STK::HDMatrixModelParameters< Array_ >

structure storing the parameters of the HD matrix valued mixture

struct storing the parameters of the matrix valued mixture.

Parameters of a matrix mixture model have two Id defined in STK::Clust::HDCovarianceModel enumeration.

Definition at line 54 of file STK_HDMatrixGaussianParameters.h.

Constructor & Destructor Documentation

◆ HDMatrixModelParameters() [1/2]

template<class Array_ >
STK::HDMatrixModelParameters< Array_ >::HDMatrixModelParameters ( int  nbCluster)
inline

default constructor

Parameters
nbClusterthe number of class of the mixture

Definition at line 99 of file STK_HDMatrixGaussianParameters.h.

100 : meank_(nbCluster)
101 , rowQk_(nbCluster), rowAjk_(nbCluster)
102 , rowBk_(nbCluster), rowDk_(nbCluster)
103 , colQk_(nbCluster), colAjk_(nbCluster)
104 , colBk_(nbCluster), colDk_(nbCluster)
105 , statMeank_(nbCluster)
106 , statRowQk_(nbCluster), statRowAjk_(nbCluster)
107 , statRowBk_(nbCluster)
108 , statColQk_(nbCluster), statColAjk_(nbCluster)
109 , statColBk_(nbCluster)
110 {}
CVectorXi colDk_
array of size nbCluster with dimension of low dimensional space for columns
Array1D< Stat::Online< CSquareX, Real > > statColQk_
Array of the rotation matrix statistics for rows.
Array1D< Stat::Online< Real, Real > > statColBk_
Array of the standard deviation statistics.
Array1D< CVectorX > rowAjk_
array of size nbCluster with inertia in low dimensional space for rows
Array1D< Array_ > meank_
array of size nbCluster with for each cluster the mean matrix
Array1D< Real > rowBk_
array of size nbCluster with remaining variance noise for rows
Array1D< Stat::Online< CVectorX, Real > > statRowAjk_
Array of size nbCluster of the rotation matrix statistics for rows.
Array1D< Real > colBk_
array of size nbCluster with variance noise for columns
Array1D< Stat::Online< CVectorX, Real > > statColAjk_
Array of the standard deviation statistics.
Array1D< CVectorX > colAjk_
array of size nbCluster with inertia in low dimensional space for columns
CVectorXi rowDk_
array of size nbCluster with dimension of low dimensional space for rows
Array1D< Stat::Online< Array_, Real > > statMeank_
Array of size nbCluster of the mean statistics.
Array1D< CSquareX > rowQk_
array of size nbCluster with the rotation matrix for rows
Array1D< Stat::Online< CSquareX, Real > > statRowQk_
Array of size nbCluster of the rotation matrix statistics for rows.
Array1D< Stat::Online< Real, Real > > statRowBk_
Array of size nbCluster with the variance noise statistics for rows.
Array1D< CSquareX > colQk_
array of size nbCluster with the rotation matrix for columns

◆ HDMatrixModelParameters() [2/2]

template<class Array_ >
STK::HDMatrixModelParameters< Array_ >::HDMatrixModelParameters ( HDMatrixModelParameters< Array_ > const param)
inline

copy constructor.

Parameters
paramthe parameters to copy.

Definition at line 114 of file STK_HDMatrixGaussianParameters.h.

115 : meank_(param.meank_)
116 , rowQk_(param.rowQk_), rowAjk_(param.rowAjk_)
117 , rowBk_(param.rowBk_), rowDk_(param.rowDk_)
118 , colQk_(param.colQk_), colAjk_(param.colAjk_)
119 , colBk_(param.colBk_), colDk_(param.colDk_)
120 , statMeank_(param.statMeank_)
121 , statRowQk_(param.statRowQk_), statRowAjk_(param.statRowAjk_)
122 , statRowBk_(param.statRowBk_)
123 , statColQk_(param.statColQk_), statColAjk_(param.statColAjk_)
124 , statColBk_(param.statColBk_)
125 {}

◆ ~HDMatrixModelParameters()

destructor

Definition at line 127 of file STK_HDMatrixGaussianParameters.h.

127{}

Member Function Documentation

◆ operator=()

◆ releaseStatistics()

◆ resize()

template<class Array_ >
void STK::HDMatrixModelParameters< Array_ >::resize ( Range const rangeRows,
Range const rangeCols 
)

resize and initialize the set of parameter.

Parameters
rangerange of the variables in the data set

◆ setParameters()

template<class Array_ >
template<class Array >
void STK::HDMatrixModelParameters< Array_ >::setParameters ( ExprBase< Array > const params)
inline

Set the parameters of the mixture model.

It is assumed that the array params store for each class the mean and sd parameters on two consecutive rows. The number of column of params is the number of variables.

Definition at line 199 of file STK_HDMatrixGaussianParameters.h.

200 {}

◆ setStatistics()

◆ updateStatistics()

Member Data Documentation

◆ colAjk_

◆ colBk_

◆ colDk_

array of size nbCluster with dimension of low dimensional space for columns

Definition at line 77 of file STK_HDMatrixGaussianParameters.h.

Referenced by STK::HDMatrixModelParameters< Array_ >::operator=().

◆ colQk_

◆ meank_

◆ rowAjk_

◆ rowBk_

◆ rowDk_

array of size nbCluster with dimension of low dimensional space for rows

Definition at line 67 of file STK_HDMatrixGaussianParameters.h.

Referenced by STK::HDMatrixModelParameters< Array_ >::operator=().

◆ rowQk_

◆ statColAjk_

◆ statColBk_

◆ statColQk_

◆ statMeank_

◆ statRowAjk_

◆ statRowBk_

◆ statRowQk_


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