com.numericalmethod.suanshu.algebra.linear.matrix.doubles.matrixtype.sparse.solver.iterative.preconditioner

Interface Preconditioner

All Known Implementing Classes:

IdentityPreconditioner, JacobiPreconditioner, SSORPreconditioner

public interface Preconditioner

Preconditioning reduces the condition number of the coefficient matrix of a linear system to accelerate the convergence when the system is solved by an iterative method.

See Also:

Wikipedia: Preconditioner

Method Summary

Modifier and Type	Method and Description
Vector	solve(Vector x) Solve Mv = x, where M is the preconditioner matrix.
Vector	transposeSolve(Vector x) Solve Mtv = x, where M is the preconditioner matrix.

Method Detail

solve

Vector solve(Vector x)

Solve Mv = x, where M is the preconditioner matrix. In effect, this method returns v as the product M^-1x.

Parameters:

x - a vector

Returns:

M^-1x

transposeSolve

Vector transposeSolve(Vector x)

Solve M^tv = x, where M is the preconditioner matrix. In effect, this method returns v as the product M^-tx. This method is usually called by a solver which solves a dual system A^tx^* = b^*. BiconjugateGradientSolver is an example of such a solver.

Parameters:

x - a vector

Returns:

M^-tx