org.ejml.simple

Class SimpleSVD<T extends SimpleMatrix>

java.lang.Object

org.ejml.simple.SimpleSVD<T>

public class SimpleSVD<T extends SimpleMatrix>
extends java.lang.Object

Wrapper around SVD for simple matrix. See SingularValueDecomposition for more details.

SVD is defined as the following decomposition:

A = U * W * V ^T

where A is m by n, and U and V are orthogonal matrices, and W is a diagonal matrix

Tolerance for singular values is Math.max(mat.numRows,mat.numCols) * W.get(0,0) * UtilEjml.EPS; where W.get(0,0) is the largest singular value.

Constructor Summary

Constructors

Constructor and Description
SimpleSVD(DenseMatrix64F mat, boolean compact)

Method Summary

Methods

Modifier and Type	Method and Description
double	getSingleValue(int index) Returns the specified singular value.
SingularValueDecomposition	getSVD() Returns the underlying decomposition that this is a wrapper around.
T	getU() Returns the orthogonal 'U' matrix.
T	getV() Returns the orthogonal 'V' matrix.
T	getW() Returns a diagonal matrix with the singular values.
int	nullity() The nullity of the decomposed matrix.
SimpleMatrix	nullSpace() Computes the null space from an SVD.
double	quality() Computes the quality of the computed decomposition.
int	rank() Returns the rank of the decomposed matrix.

Methods inherited from class java.lang.Object

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

Constructor Detail

SimpleSVD

public SimpleSVD(DenseMatrix64F mat,
         boolean compact)

Method Detail

getU

public T getU()

Returns the orthogonal 'U' matrix.

Returns:

An orthogonal m by m matrix.

getW

public T getW()

Returns a diagonal matrix with the singular values. The singular values are ordered from largest to smallest.

Returns:

Diagonal matrix with singular values along the diagonal.

getV

public T getV()

Returns the orthogonal 'V' matrix.

Returns:

An orthogonal n by n matrix.

quality

public double quality()

Computes the quality of the computed decomposition. A value close to or less than 1e-15 is considered to be within machine precision.

This function must be called before the original matrix has been modified or else it will produce meaningless results.

Returns:

Quality of the decomposition.

nullSpace

public SimpleMatrix nullSpace()

Computes the null space from an SVD. For more information see SingularOps.nullSpace(org.ejml.interfaces.decomposition.SingularValueDecomposition<org.ejml.data.DenseMatrix64F>, org.ejml.data.DenseMatrix64F, double).

Returns:

Null space vector.

getSingleValue

public double getSingleValue(int index)

Returns the specified singular value.

Parameters:

index - Which singular value is to be returned.

Returns:

A singular value.

rank

public int rank()

Returns the rank of the decomposed matrix.

Returns:

The matrix's rank

See Also:

SingularOps.rank(org.ejml.interfaces.decomposition.SingularValueDecomposition, double)

nullity

public int nullity()

The nullity of the decomposed matrix.

Returns:

The matrix's nullity

See Also:

SingularOps.nullity(org.ejml.interfaces.decomposition.SingularValueDecomposition, double)

getSVD

public SingularValueDecomposition getSVD()

Returns the underlying decomposition that this is a wrapper around.

Returns:

SingularValueDecomposition