jvx.numeric
Class PnBiconjugateGradient

java.lang.Object
  jvx.numeric.PnBiconjugateGradient

public class PnBiconjugateGradient

extends java.lang.Object

Solve sparse linear equation systems using preconditioned biconjugate gradient method.

Constructor Summary

PnBiconjugateGradient()
Constructor

Method Summary

double	getActualEstimatedError() Get estimated error of last linbcg call.
int	getActualNumOfTakenIterations() Get number of iterations used in last linbcg call.
double	getConvergenceTest() There are four different methods measure the convergence of the solution of the linear equation used in linbcg.
int	getDimension() Get dimension.
double	getEPS() Set EPS used during solving of linear equations.
int	getMaxNumIterations() Get maximum number of allowed iterations.
double	getTolerance() Get desired convergence tolerance of linbcg.
boolean	isUsingPrecondioning() Check if precondioning is enabled.
double	linbcg(PnSparseMatrix smat, PdVector x, PdVector b) Method solves a sparse linear equation system A*x=b using a preconditioned biconjugate gradient method.
static void	preconditionJacobiLeft(PnSparseMatrix smat, PdVector b) Perform a diagonal (Jacobi) preconditioning from the left to the a linear equation system smat.x=b.
static void	preconditionJacobiLeft(PnSparseMatrix smat, PdVector b, int dim) Perform a diagonal (Jacobi) preconditioning from the left to the a linear equation system smat.x=b.
void	setConvergenceTest(int aTest) There are four different methods measure the convergence of the solution of the linear equation used in linbcg.
void	setDimension(int dim) Set dimension of surface matrix is based on.
void	setEPS(double eps) Set EPS used during solving of linear equations.
void	setMaxNumIterations(int itmax) Set maximum number of allowed iterations.
void	setTolerance(double tol) Set desired convergence tolerance of linbcg.
void	usePreconditioning(boolean flag) Enable/Disable preconditioning before solving the linear equation system.

Methods inherited from class java.lang.Object

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

Constructor Detail

PnBiconjugateGradient

public PnBiconjugateGradient()

Constructor

Method Detail

getActualEstimatedError

public double getActualEstimatedError()

Get estimated error of last linbcg call.

getActualNumOfTakenIterations

public int getActualNumOfTakenIterations()

Get number of iterations used in last linbcg call.

setMaxNumIterations

public void setMaxNumIterations(int itmax)

Set maximum number of allowed iterations. Default value is 1000.

getMaxNumIterations

public int getMaxNumIterations()

Get maximum number of allowed iterations.

setTolerance

public void setTolerance(double tol)

Set desired convergence tolerance of linbcg.

getTolerance

public double getTolerance()

Get desired convergence tolerance of linbcg.

setDimension

public void setDimension(int dim)

Set dimension of surface matrix is based on. Coordinate vectors must have length equal dim*(dim of sparseMatrix).

getDimension

public int getDimension()

Get dimension.

setEPS

public void setEPS(double eps)

Set EPS used during solving of linear equations.

getEPS

public double getEPS()

Set EPS used during solving of linear equations. Default value is PuMath.EPS.

setConvergenceTest

public void setConvergenceTest(int aTest)

There are four different methods measure the convergence of the solution of the linear equation used in linbcg. The default is 1.

1. |A*x-b|/|b| 2. |Diag(A)^-1*x-b|/|Diag(A)^-1*b| With Diag(A) denoting the diagonal part of A.
3. This routine uses its own estimate of error in x, and requires its magnitude, divided by the magnitude of x, to be less than tol.
4. Same as 3., except that the largest (in absolute value) component of the error and the largest component of x are used instead of the vector magnitude.

getConvergenceTest

public double getConvergenceTest()

There are four different methods measure the convergence of the solution of the linear equation used in linbcg. Get the index of the method actually used.

See Also:

setConvergenceTest(int)

usePreconditioning

public void usePreconditioning(boolean flag)

Enable/Disable preconditioning before solving the linear equation system. Method uses a Jacobi preconditioning. If the preconditioning is enabled, the method linbcg modifies the given matrix and vector.

isUsingPrecondioning

public boolean isUsingPrecondioning()

Check if precondioning is enabled.

linbcg

public double linbcg(PnSparseMatrix smat,
                     PdVector x,
                     PdVector b)

Method solves a sparse linear equation system A*x=b using a preconditioned biconjugate gradient method. Preconditioning is done by inverting the diagonal part of A.

Returns:

estimated error of solution or -1 at error.

preconditionJacobiLeft

public static void preconditionJacobiLeft(PnSparseMatrix smat,
                                          PdVector b,
                                          int dim)

Perform a diagonal (Jacobi) preconditioning from the left to the a linear equation system smat.x=b. Setting the parameter dim larger than 1, allow to handle more than one equation system with the same matrix in one run. The first equation system uses the 0-th, dim-th, 2*dim-th, 3*dim-th,... entries of b, the second the 1-th, (1+dim)-th, (1+2*dim)-th, ... and so on.

preconditionJacobiLeft

public static void preconditionJacobiLeft(PnSparseMatrix smat,
                                          PdVector b)

Perform a diagonal (Jacobi) preconditioning from the left to the a linear equation system smat.x=b.