smile.math.matrix

Interface IMatrix

All Known Implementing Classes:

BandMatrix, Matrix, SparseMatrix

public interface IMatrix

An abstract interface of matrix. The most important method is the matrix vector multiplication, which is the only operation needed in many iterative matrix algorithms, e.g. biconjugate gradient method for solving linear equations and power iteration and Lanczos algorithm for eigen decomposition, which are usually very effecient for very large and sparse matrices.

Author:

Haifeng Li

Method Summary

Methods

Modifier and Type	Method and Description
void	asolve(double[] b, double[] x) Solve Ad * x = b for the preconditioner matrix Ad.
void	atx(double[] x, double[] y) y = A' * x
void	atxpy(double[] x, double[] y) y = A' * x + y
void	atxpy(double[] x, double[] y, double b) y = A' * x + b * y
void	ax(double[] x, double[] y) y = A * x
void	axpy(double[] x, double[] y) y = A * x + y
void	axpy(double[] x, double[] y, double b) y = A * x + b * y
double	get(int i, int j) Returns the entry value at row i and column j.
int	ncols() Returns the number of columns.
int	nrows() Returns the number of rows.
void	set(int i, int j, double x) Set the entry value at row i and column j.

Method Detail

nrows

int nrows()

Returns the number of rows.

ncols

int ncols()

Returns the number of columns.

get

double get(int i,
         int j)

Returns the entry value at row i and column j.

set

void set(int i,
       int j,
       double x)

Set the entry value at row i and column j.

ax

void ax(double[] x,
      double[] y)

y = A * x

axpy

void axpy(double[] x,
        double[] y)

y = A * x + y

axpy

void axpy(double[] x,
        double[] y,
        double b)

y = A * x + b * y

atx

void atx(double[] x,
       double[] y)

y = A' * x

atxpy

void atxpy(double[] x,
         double[] y)

y = A' * x + y

atxpy

void atxpy(double[] x,
         double[] y,
         double b)

y = A' * x + b * y

asolve

void asolve(double[] b,
          double[] x)

Solve A_d * x = b for the preconditioner matrix A_d. The preconditioner matrix A_d is close to A and should be easy to solve for linear systems. This method is useful for preconditioned conjugate gradient method. The preconditioner matrix could be as simple as the trivial diagonal part of A in some cases.