Package SPARSELIB++

• List of Modules
• Components
  • Documentation from ITL
  • Paper from ITL
  • Source from ITL

Package : SPARSELIB++

 
C++ class library for sparse matrix computation. Includes matrix classes in
several storage formats (e.g. compressed row, compressed column, coordinate),
and basic management utilities. Uses Sparse BLAS Toolkit for kernel operations
(e.g. matrix-vector multiply) to enhance portability and performance on wide
range of architectures. Includes common preconditioners for iterative
solvers. Can be built with nearly any C++ matrix/vector class; shipped with MV++.
 
Type         : Subprogram package
 
Portability   : Portable
 
Availability : Public domain
 
Language     : C++
 
Reference    : SparseLib++ Home Page
 
Developer    : R. Pozo, K. Remington and A. Lumsdaime
 
Distributor  : R. Pozo (pozo@cam.nist.gov), via anonymous ftp from
               math.nist.gov. or via the SparseLib++ Home Page
 
On sites     : (1) ITL
 

Implementation data for ITL

 
ITL     : Unix Workstation Network, National Institute of Standards and
          Technology (NIST), Gaithersburg, MD. Available to NIST staff.
 
Version : 1.4
Support : Unsupported
 
Access available only to NIST staff on internal Unix systems. They may access this
package provided the /itl tree is cross-mounted.
 
   Source        : Anonymous ftp from math.nist.gov in /pub/pozo/src
   Documentation : Anonymous ftp from math.nist.gov in /pub/pozo/docs
   Paper         : Anonymous ftp from math.nist.gov in
                   /pub/pozo/papers/sparse.ps.Z