ARPACK Users' Guide: Solution of Large-Scale Eigenvalue Problems with Implicitly Restarted Arnoldi Methods

Abstract

List of figures List of tables Preface 1. Introduction to ARPACK. Important features Getting started Reverse communication interface Availability Installation Documentation Dependence on LAPACK and BLAS Expected performance P_ARPACK Contributed additions Trouble shooting and problems 2. Getting started with ARPACK. Directory structure and contents Getting started An example for a symmetric Eigenvalue problem 3. General use of ARPACK. Naming conventions, Precisions, and types Shift and invert spectral transformation mode Reverse communication structure for shift-Invert Using the computational modes Computational modes for real symmetric problems Postprocessing for Eigenvectors using dseupd Computational modes for real nonsymmetric problems Postprocessing for Eigenvectors Using dneupd Computational modes for complex problems Postprocessing for Eigenvectors Using zneupd 4. The implicitly restarted Arnoldi method: structure of the Eigenvalue problem Krylov subspaces and projection methods The Arnoldi factorization Restarting the Arnoldi method The generalized Eigenvalue problem Stopping Criterion 5. Computational routines. ARPACK subroutines LAPACK routines used by ARPACK BLAS routines used by ARPACK Appendix A. Templates and driver routines Symmetric drivers Real Nonsymmetric drivers Complex drivers Band drivers The singular value decomposition Appendix B. Tracking the progress of ARPACK. Obtaining trace output Check-pointing ARPACK Appendix C. The XYaupd ARPACK Routines. DSAUPD DNAUPD ZNAUPD Bibliography Index.