artofillusion.math
Class SVD

java.lang.Object
  extended by artofillusion.math.SVD

public class SVD
extends java.lang.Object

The SVD class defines methods for solving sets of linear equations by singular value decomposition. It uses classes from the Java Matrix (JAMA) package to factor the matrix. The complete JAMA package, including documentation and source, can be obtained from http://math.nist.gov/javanumerics/jama/


Constructor Summary
SVD()
           
 
Method Summary
static void solve(double[][] a, double[] b)
          Solve a set of M linear equations in N unknows.
static void solve(double[][] a, double[] b, double tol)
           
 
Methods inherited from class java.lang.Object
equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

SVD

public SVD()
Method Detail

solve

public static void solve(double[][] a,
                         double[] b)
Solve a set of M linear equations in N unknows. The parameters are: a: an array of size [M][N] containing the matrix of coefficients. b: an array of length max(M,N) containing the right hand side vector in its first M elements. On exit, the first N elements are overwritten with the solution vector. tol: any singular values smaller than tol*(largest singular value) are set to 0. If tol is omitted, it defaults to 1.0e-8.


solve

public static void solve(double[][] a,
                         double[] b,
                         double tol)


Copyright © 1999-2011 by Peter Eastman.