org.knime.base.node.mine.pca
Class EigenValue

java.lang.Object
  org.knime.base.node.mine.pca.EigenValue

All Implemented Interfaces:

Comparable<EigenValue>

public class EigenValue
extends Object
implements Comparable<EigenValue>

This class describes an eigenvalue - eigenvector pair, comparable by absolute of eigenvalue.

Constructor Summary

EigenValue(int position, double value, Matrix vector)
Create pair.

Method Summary

int	compareTo(EigenValue o)
static List<EigenValue>	createSortedList(double[][] eigenVectors, double[] eigenvalues) create list of EigenValues sorted by absolute value
static double[]	extractEVVector(EigenvalueDecomposition eig) extract the vector of eigenvalues.
int	getPosition()
static Matrix	getSortedEigenVectors(double[][] eigenVectors, double[] eigenvalues, int number) Constructs a matrix consisting of the first number eigenvectors of m, where first refers to the corresponding eigenvalues sorted by absolute value. Take care that the matrix is symmetric!
Double	getValue()
Matrix	getVector()
String	toString()

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

EigenValue

EigenValue(int position,
           double value,
           Matrix vector)

Create pair.

Parameters:

position - original position in eigenvalue matrix

value - eigenvalue

vector - eigenvector

Method Detail

getPosition

public int getPosition()

Returns:

original position in ev-matrix

compareTo

public int compareTo(EigenValue o)

Specified by:

compareTo in interface Comparable<EigenValue>

getValue

public Double getValue()

Returns:

the value

getVector

public Matrix getVector()

Returns:

the vector

toString

public String toString()

Overrides:

toString in class Object

extractEVVector

public static double[] extractEVVector(EigenvalueDecomposition eig)

extract the vector of eigenvalues.

Parameters:

eig - eigenvalue decomposition

Returns:

vector of eigenvalues

getSortedEigenVectors

public static Matrix getSortedEigenVectors(double[][] eigenVectors,
                                           double[] eigenvalues,
                                           int number)

Constructs a matrix consisting of the first number eigenvectors of m, where first refers to the corresponding eigenvalues sorted by absolute value.
Take care that the matrix is symmetric!

Parameters:

eigenVectors - list of eigenvectors

eigenvalues - list of eigenvalues

number - number of eigenvectors in return matrix

Returns:

matrix with eigenvectors columnwise

createSortedList

public static List<EigenValue> createSortedList(double[][] eigenVectors,
                                                double[] eigenvalues)

create list of EigenValues sorted by absolute value

Parameters:

eigenVectors - matrix of eigenvector (in columns)

eigenvalues - eigenvalues, same order as columns of eigenVectors

Returns:

sorted list of EigenValues