gov.sandia.cognition.math.matrix.decomposition

Class AbstractEigenDecomposition

java.lang.Object

gov.sandia.cognition.math.matrix.decomposition.AbstractEigenDecomposition

All Implemented Interfaces:

EigenDecomposition

Direct Known Subclasses:

EigenDecompositionRightMTJ

@CodeReview(reviewer="Jonathan McClain",date="2006-05-16",changesNeeded=true,comments={"Comments indicated by triple slashes","Some refactoring needed here.","Another review needed after refactoring."},response={@CodeReviewResponse(respondent="Kevin R. Dixon",date="2006-05-17",moreChangesNeeded=true,comments="Fixes from J.T.\'s code review"),@CodeReviewResponse(respondent="Jonathan McClain",date="2006-05-17",moreChangesNeeded=false,comments="Rechecking after changes were made. Looks good now.")})
public abstract class AbstractEigenDecomposition
extends java.lang.Object
implements EigenDecomposition

Abstract partial implementation of the EigenDecomposition interface

Since:

1.0

Author:

Kevin R. Dixon

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	AbstractEigenDecomposition(ComplexNumber[] eigenValues, Matrix eigenVectorsRealPart, Matrix eigenVectorsImaginaryPart) Stores the given eigenvalues and eigenvectors internally, the eigenvalues and eigenvectors will not be sorted.
protected	AbstractEigenDecomposition(ComplexNumber[] eigenValues, Matrix eigenVectorsRealPart, Matrix eigenVectorsImaginaryPart, boolean sort) Creates a new eigendecomposition using the given eigenvalues and eigenvectors...

Method Summary

Modifier and Type	Method and Description
ComplexNumber	getEigenValue(int index) gets the indexed eigenvalue
ComplexNumber[]	getEigenValues() Getter for eigenValues
Matrix	getEigenVectorsImaginaryPart() gets the imaginary part of the eienvector, where the ith eigenvector is the ith column
Matrix	getEigenVectorsRealPart() getter for eigenvectorsrealPart
ComplexNumber	getLogDeterminant() Computes the natural logarithm determinant from the collection of eigenvalues
protected void	setEigenDecomposition(ComplexNumber[] eigenValues, Matrix eigenVectorsRealPart, Matrix eigenVectorsImaginaryPart, boolean sort) Sets the eigen decomposition for this
protected void	setEigenValues(ComplexNumber[] eigenValues) setter for eigenValues
void	setEigenVectorsImaginaryPart(Matrix eigenVectorsImaginaryPart) setter for the imaginary part of the eienvector, where the ith eigenvector is the ith column
protected void	setEigenVectorsRealPart(Matrix eigenVectorsRealPart) setter for eigenVectorsRealPart, where the ith eigenvector is the ith column
protected void	setUnsortedEigenDecomposition(ComplexNumber[] eigenValues, Matrix eigenVectorsRealPart, Matrix eigenVectorsImaginaryPart) Creates a new eigendecomposition using the given eigenvalues and eigenvectors...
protected void	sortAndSetEigenDecomposition(ComplexNumber[] eigenValues, Matrix eigenVectorsRealPart, Matrix eigenVectorsImaginaryPart) Sorts the eigendecomposition in descending order of the value of the magnitudes of the eigenvalues

Methods inherited from class java.lang.Object

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

Constructor Detail

AbstractEigenDecomposition

protected AbstractEigenDecomposition(ComplexNumber[] eigenValues,
                                     Matrix eigenVectorsRealPart,
                                     Matrix eigenVectorsImaginaryPart)

Stores the given eigenvalues and eigenvectors internally, the eigenvalues and eigenvectors will not be sorted.

Parameters:

eigenValues - array of complex-valued eigenvalue

eigenVectorsRealPart - matrix of the real parts of the eigenvectors

eigenVectorsImaginaryPart - matrix of the imaginary parts of the eigenvectors

AbstractEigenDecomposition

protected AbstractEigenDecomposition(ComplexNumber[] eigenValues,
                                     Matrix eigenVectorsRealPart,
                                     Matrix eigenVectorsImaginaryPart,
                                     boolean sort)

Creates a new eigendecomposition using the given eigenvalues and eigenvectors... this does not specify if the eigenvectors are the right or left eigenvectors. That should be done in a subclass implementation.

Parameters:

eigenValues - array of complex-valued eigenvalues to store

eigenVectorsRealPart - matrix where the ith column of the matrix contains the real part of the ith eigenvector of the underlying matrix

eigenVectorsImaginaryPart - matrix where the ith column of the matrix contains the real part of the ith eigenvector of the underlying matrix

sort - if true, then the constructor will sort the eigenvectors and eigenvectors by descending magnitude of the eigenvalue. In this case, eigenValues[0] will be the largest magnitude and eigenVectors(:,0) is its corresponding eigenvector.

Method Detail

setEigenDecomposition

protected void setEigenDecomposition(ComplexNumber[] eigenValues,
                                     Matrix eigenVectorsRealPart,
                                     Matrix eigenVectorsImaginaryPart,
                                     boolean sort)

Sets the eigen decomposition for this

Parameters:

eigenValues - array of eigenvalues for the underlying matrix

eigenVectorsRealPart - real part of the eigenvectors for the underlying matrix

eigenVectorsImaginaryPart - imaginary part of the eigenvalues for the underlying matrix

sort - true to sort eigen values/vectors by descending order of magnitude of the eigenvalues

sortAndSetEigenDecomposition

protected void sortAndSetEigenDecomposition(ComplexNumber[] eigenValues,
                                            Matrix eigenVectorsRealPart,
                                            Matrix eigenVectorsImaginaryPart)

Sorts the eigendecomposition in descending order of the value of the magnitudes of the eigenvalues

Parameters:

eigenValues - array of eigenvalues for the underlying matrix

eigenVectorsRealPart - real part of the eigenvectors for the underlying matrix

eigenVectorsImaginaryPart - imaginary part of the eigenvalues for the underlying matrix

setUnsortedEigenDecomposition

protected void setUnsortedEigenDecomposition(ComplexNumber[] eigenValues,
                                             Matrix eigenVectorsRealPart,
                                             Matrix eigenVectorsImaginaryPart)

Creates a new eigendecomposition using the given eigenvalues and eigenvectors... this does not specify if the eigenvectors are the right or left eigenvectors. That should be done in a subclass implementation.

Parameters:

eigenValues - array of eigenvalues for the underlying matrix

eigenVectorsRealPart - real part of the eigenvectors for the underlying matrix

eigenVectorsImaginaryPart - imaginary part of the eigenvalues for the underlying matrix

getEigenValues

public ComplexNumber[] getEigenValues()

Getter for eigenValues

Specified by:

getEigenValues in interface EigenDecomposition

Returns:

eigenValues

setEigenValues

protected void setEigenValues(ComplexNumber[] eigenValues)

setter for eigenValues

Parameters:

eigenValues - eigenvalues to set

getEigenValue

public ComplexNumber getEigenValue(int index)

gets the indexed eigenvalue

Specified by:

getEigenValue in interface EigenDecomposition

Parameters:

index - zero-based index into the eigenvalue array

Returns:

index eigenvalue

getEigenVectorsRealPart

public Matrix getEigenVectorsRealPart()

getter for eigenvectorsrealPart

Specified by:

getEigenVectorsRealPart in interface EigenDecomposition

Returns:

real part of the eienvector, where the ith eigenvector is the ith column

setEigenVectorsRealPart

protected void setEigenVectorsRealPart(Matrix eigenVectorsRealPart)

setter for eigenVectorsRealPart, where the ith eigenvector is the ith column

Parameters:

eigenVectorsRealPart - real part of the eienvector, where the ith eigenvector is the ith column

getEigenVectorsImaginaryPart

public Matrix getEigenVectorsImaginaryPart()

gets the imaginary part of the eienvector, where the ith eigenvector is the ith column

Specified by:

getEigenVectorsImaginaryPart in interface EigenDecomposition

Returns:

imaginary part of the eienvector, where the ith eigenvector is the ith column

setEigenVectorsImaginaryPart

public void setEigenVectorsImaginaryPart(Matrix eigenVectorsImaginaryPart)

setter for the imaginary part of the eienvector, where the ith eigenvector is the ith column

Parameters:

eigenVectorsImaginaryPart - imaginary part of the eienvector, where the ith eigenvector is the ith column

getLogDeterminant

public ComplexNumber getLogDeterminant()

Computes the natural logarithm determinant from the collection of eigenvalues

Specified by:

getLogDeterminant in interface EigenDecomposition

Returns:

natural logarithm of the determinant