Package jdistlib.accelerator

Class SingularValueDecomposition

java.lang.Object

jdistlib.accelerator.SingularValueDecomposition

public final class SingularValueDecomposition
extends Object

Immutable FP64 thin singular-value decomposition A = U*S*Vt.

Constructor Summary

Constructors

Constructor	Description
SingularValueDecomposition(int rows, int columns, double[] singularValues, double[] left, double[] rightTransposed)	Creates a thin SVD with descending singular values.

Method Summary

Modifier and Type	Method	Description
int	columns()
int	components()
double[]	leftSingularVectors()	Returns row-major thin U with shape rows by components.
int	rank()	Returns numerical rank using the standard dimension-scaled machine threshold.
double[]	rightSingularVectorsTransposed()	Returns row-major thin V-transpose with shape components by columns.
int	rows()
double[]	singularValues()	Returns singular values in descending order.

Methods inherited from class java.lang.Object

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

Constructor Details

SingularValueDecomposition

public SingularValueDecomposition(int rows,
 int columns,
 double[] singularValues,
 double[] left,
 double[] rightTransposed)

Creates a thin SVD with descending singular values.

Method Details

rows

public int rows()

columns

public int columns()

components

public int components()

singularValues

public double[] singularValues()

Returns singular values in descending order.

leftSingularVectors

public double[] leftSingularVectors()

Returns row-major thin U with shape rows by components.

rightSingularVectorsTransposed

public double[] rightSingularVectorsTransposed()

Returns row-major thin V-transpose with shape components by columns.

rank

public int rank()

Returns numerical rank using the standard dimension-scaled machine threshold.