Principal component analysis. More...

#include <PCA.hpp>

Collaboration diagram for PCA:

Public Member Functions
	PCA (MatrixD data)
	Principal component analysis algorithm. More...

void	fit ()
	Finds the principal components of a Matrix. More...

MatrixD	transform ()
	Rotates the data set, using the eigenvectors of the covariance matrix as the new base. More...

MatrixD	transform (int numComponents)
	Rotates the data set, using the eigenvectors of the covariance matrix with the largest eigenvalues as the new base. More...

const MatrixD &	getEigenvalues () const

const MatrixD &	getEigenvectors () const

const MatrixD &	getPercentages () const

const MatrixD &	getCumPercentages () const

Private Attributes
MatrixD	X

MatrixD	eigenvalues

MatrixD	eigenvectors

MatrixD	percentages

MatrixD	cumPercentages

Detailed Description

Principal component analysis.

Definition at line 18 of file PCA.hpp.

Constructor & Destructor Documentation

◆ PCA()

PCA::PCA ( MatrixD data )

inlineexplicit

Principal component analysis algorithm.

Parameters

data	the matrix whose principal components will be found

Definition at line 28 of file PCA.hpp.

Member Function Documentation

◆ fit()

void PCA::fit ( )

inline

Finds the principal components of a Matrix.

Eigenvectors and eigenvalues are found via the Jacobi eigenvalue algorithm

Definition at line 35 of file PCA.hpp.

◆ getCumPercentages()

const MatrixD& PCA::getCumPercentages ( ) const

inline

Definition at line 91 of file PCA.hpp.

◆ getEigenvalues()

const MatrixD& PCA::getEigenvalues ( ) const

inline

Definition at line 79 of file PCA.hpp.

◆ getEigenvectors()

const MatrixD& PCA::getEigenvectors ( ) const

inline

Definition at line 83 of file PCA.hpp.

◆ getPercentages()

const MatrixD& PCA::getPercentages ( ) const

inline

Definition at line 87 of file PCA.hpp.

◆ transform() [1/2]

MatrixD PCA::transform ( )

inline

Rotates the data set, using the eigenvectors of the covariance matrix as the new base.

Returns: the original dataset rotated using the eigenvectors of the covariance matrix as the new base

Definition at line 60 of file PCA.hpp.

◆ transform() [2/2]

MatrixD PCA::transform ( int numComponents )

inline

Rotates the data set, using the eigenvectors of the covariance matrix with the largest eigenvalues as the new base.

Returns: the original dataset rotated using the eigenvectors of the covariance matrix with the largest eigenvalues as the new base

Definition at line 68 of file PCA.hpp.