STAT_MULTIVAR Class Reference

Class for accumulating and computing multivariate statistics. More...

#include <mi32/statmultivar.h>

List of all members.

Public Member Functions
	STAT_MULTIVAR ()
	~STAT_MULTIVAR ()
ERRVALUE	Accumulate (const double *values)
ERRVALUE	ComputeEigen ()
ERRVALUE	ComputeMultilinearRegression (MIMATRIX &CoeffRet, UINT32 TgtVar, const UINT32 *SrcVars=0, UINT32 NumSrcVars=0) const
ERRVALUE	Finish ()
const MIMATRIX &	GetCorrelation () const
const MIMATRIX &	GetCorrespondence () const
const MIMATRIX &	GetCovariance () const
const MIMATRIX &	GetEigenValues () const
const MIMATRIX &	GetEigenVectors () const
const double *	GetMaximums () const
const double *	GetMeans () const
const double *	GetMinimums () const
UINT64	GetNumSamples () const
ERRVALUE	Start (UINT32 NumVars)

---

Detailed Description

Class for accumulating and computing multivariate statistics.

---

Constructor & Destructor Documentation

STAT_MULTIVAR::STAT_MULTIVAR

(

)

Default constructor.

STAT_MULTIVAR::~STAT_MULTIVAR

(

)

Destructor.

---

Member Function Documentation

ERRVALUE STAT_MULTIVAR::Accumulate

(

const double *

values

)

Accumulate values for single sample.

ERRVALUE STAT_MULTIVAR::ComputeEigen

(

)

Compute eigenvectors, eigenvalues, correspondences.

ERRVALUE STAT_MULTIVAR::ComputeMultilinearRegression	(	MIMATRIX &	CoeffRet,
		UINT32	TgtVar,
		const UINT32 *	SrcVars = 0,
		UINT32	NumSrcVars = 0
	)			const

Compute multilinear regression coefficients for specified target variable.

The coefficients are returned in a vector MIMATRIX with (NumSrcVars+1) elements. The first element (index = 0) will be the "offset" value. Subsequent elements (starting with index = 1) correspond to the source variables in sequence. Must call Finish() before using or error will result.

Parameters:

	CoeffRet	Coefficients returned
	TgtVar	Target variable index
	SrcVars	Source variables, 0 to use all except specified target
	NumSrcVars	Number of source variables, 0 to use all except specified target

ERRVALUE STAT_MULTIVAR::Finish

(

)

Finish accumulation, compute mean, std deviation, correlation, covariance.

const MIMATRIX& STAT_MULTIVAR::GetCorrelation

(

)

const [inline]

Get correlation matrix.

const MIMATRIX& STAT_MULTIVAR::GetCorrespondence

(

)

const [inline]

Get matrix of correspondence between input variable and principal component.

Must call ComputeEigen before using this method.

const MIMATRIX& STAT_MULTIVAR::GetCovariance

(

)

const [inline]

Get covariance matrix.

const MIMATRIX& STAT_MULTIVAR::GetEigenValues

(

)

const [inline]

Get eigenvalues matrix.

Must call ComputeEigen before using this method.

const MIMATRIX& STAT_MULTIVAR::GetEigenVectors

(

)

const [inline]

Get eigenvectors matrix.

Must call ComputeEigen before using this method.

const double* STAT_MULTIVAR::GetMaximums

(

)

const [inline]

Get maximum values.

const double* STAT_MULTIVAR::GetMeans

(

)

const [inline]

Get mean values.

const double* STAT_MULTIVAR::GetMinimums

(

)

const [inline]

Get minimum values.

UINT64 STAT_MULTIVAR::GetNumSamples

(

)

const [inline]

Get number of samples.

ERRVALUE STAT_MULTIVAR::Start

(

UINT32

NumVars

)

Start accumulation, resets all statistics.

---

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

• mi32/statmultivar.h