next up previous
Next: Sparse Matrix Storage Up: ITPACKV 2D User's Guide Previous: Introduction

   
Applicability

The ITPACKV package is designed to solve linear systems of the form Au=b where A is a symmetric and positive definite (or slightly nonsymmetric) system of N linear equations, b is the right hand side, and u is the desired solution vector. The basic solution methods provided are the Jacobi, SOR, Symmetric SOR, and Reduced System methods. All methods except SOR are accelerated by either Conjugate Gradient or Chebyshev acceleration. When using the Reduced System methods, it is required that the system be reordered into a red-black system, that is, a system of the form

\begin{displaymath}\left[\begin{array}{cc} D_R & H \\ K & D_B \end{array} \right], \end{displaymath}

where DR and DB are diagonal matrices. A switch to compute, if possible, the red-black indexing, permute the linear system, and permute associated vectors is provided.