com.mockturtlesolutions.snifflib.linalg

Class NRQRAlgorithm

java.lang.Object

com.mockturtlesolutions.snifflib.linalg.AbstractQRAlgorithm

com.mockturtlesolutions.snifflib.linalg.NRQRAlgorithm

All Implemented Interfaces:

QRAlgorithm

public class NRQRAlgorithm
extends AbstractQRAlgorithm

This QR algorithm closely follows the one given in Numerical Recipes in Fortran by Press et al. 1992

Constructor Summary

Constructors

Constructor and Description
NRQRAlgorithm(DblMatrix X) Construct with a DblMatrix.

Method Summary

Methods

Modifier and Type	Method and Description
DblMatrix	getA()
DblMatrix	getD()
DblMatrix	getFullQ() Returns an m-by-m Q matrix when the matrix being decomposed was m-by-n.
DblMatrix	getFullR() Returns an m-by-n R matrix when the matrix being decomposed was m-by-n.
DblMatrix	getQ() Returns the economy sized m-by-n Q for the decomposition of an m-by-n matrix.
DblMatrix	getR() Get the upper triangular matrix R of the QR decomposition.
boolean	isSingular() Return true if a singular condition is encountered during.
DblMatrix	QtransDot(DblMatrix Y) Return the result of Q'*X for the input X.
DblMatrix	rsolve(DblMatrix Y) Solve the System R*b = Y for b given Y.
DblMatrix	solveSystem(DblMatrix Y) Solve a linear system (X*b = Y) returning the coeficients as a DblMatrix.
void	update() Update the QR decomposition.

Methods inherited from class java.lang.Object

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

Constructor Detail

NRQRAlgorithm

public NRQRAlgorithm(DblMatrix X)

Construct with a DblMatrix.

Method Detail

getFullR

public DblMatrix getFullR()

Returns an m-by-n R matrix when the matrix being decomposed was m-by-n. In contrast getQ() returns the economy sized m-by-n Q.

Specified by:

getFullR in interface QRAlgorithm

Specified by:

getFullR in class AbstractQRAlgorithm

getFullQ

public DblMatrix getFullQ()

Returns an m-by-m Q matrix when the matrix being decomposed was m-by-n. In contrast getQ() returns the economy sized m-by-n Q.

Specified by:

getFullQ in interface QRAlgorithm

Specified by:

getFullQ in class AbstractQRAlgorithm

getQ

public DblMatrix getQ()

Returns the economy sized m-by-n Q for the decomposition of an m-by-n matrix.

Specified by:

getQ in interface QRAlgorithm

Specified by:

getQ in class AbstractQRAlgorithm

getR

public DblMatrix getR()

Get the upper triangular matrix R of the QR decomposition.

Specified by:

getR in interface QRAlgorithm

Specified by:

getR in class AbstractQRAlgorithm

getA

public DblMatrix getA()

getD

public DblMatrix getD()

update

public void update()

Update the QR decomposition.

Specified by:

update in interface QRAlgorithm

Specified by:

update in class AbstractQRAlgorithm

isSingular

public boolean isSingular()

Return true if a singular condition is encountered during.

Specified by:

isSingular in interface QRAlgorithm

Specified by:

isSingular in class AbstractQRAlgorithm

QtransDot

public DblMatrix QtransDot(DblMatrix Y)

Return the result of Q'*X for the input X. X may be a matrix in which case the output is a matrix having the same number of rows of A and the same number of columns as X.

Specified by:

QtransDot in interface QRAlgorithm

Specified by:

QtransDot in class AbstractQRAlgorithm

solveSystem

public DblMatrix solveSystem(DblMatrix Y)

Solve a linear system (X*b = Y) returning the coeficients as a DblMatrix. Q*R = X; Q*R*b = Y R*b = Q'*Y Form Q'*Y then solve for b by backsubstitution.

Specified by:

solveSystem in interface QRAlgorithm

Specified by:

solveSystem in class AbstractQRAlgorithm

rsolve

public DblMatrix rsolve(DblMatrix Y)

Solve the System R*b = Y for b given Y.