# The templates files that were here are now in the # templates lib ( ../templates ) file linalg/readme for overview of linalg file linalg/acre.tgz for parallel covariance component , estimation code available website http://csm.pnl.gov/statistics/dll/pml.html contact acre-developers@emsl.pnl.gov size 54347k lib linalg/amd for Minimum degree ordering algorithms for sparse symmetric matrices. , For details on the methods used, see "An approximate minimum degree , ordering algorithm," SIAM J. Matrix Analysis and Applications (to , appear), by Amestoy, Davis, and Duff. Also see TR-94-039, CISE Dept, , Univ. of Florida (available via anonymous ftp to ftp.cis.ufl.edu, or , via the WWW at http://www.cise.ufl.edu/~davis). (Some licensing , restrictions apply). by Timothy A. Davis (University of Florida, email: davis@cise.ufl.edu), , Patrick Amestoy, Iain S. Duff, and John K. Reid alg approximate minimum degree ordering file linalg/amli.tgz for Algebraic Multilevel Iteration (AMLI) method This is a Matlab code and test examples to illustrate the AMLI technique to construct and use preconditioners for linear systems of equations which are of optimal order. The test examples include numerical solution of 2-D second order elliptic problems with continuous and discontinuous coefficients. lang Matlab by Maya Neytcheva, University of Nijmegen contact neytchev@sci.kun.nl file linalg/arpack96.tgz , ARPACK is an implementation of the Implicitly Restarted Arnoldi , Method for computing a few selected eigenvalues and corresponding , eigenvectors of a large (sparse) matrix. It may be used to solve , symmetric, nonsymmetric, standard or generalized eigenvalue problems , and to compute a partial singular value decomposition. A parallel , version is also available (refer to the scalapack index on , netlib for further details). by Dan Sorensen , Rich Lehoucq , and Chao Yang . size 651178 bytes lang Fortran 77 file linalg/axxbc.f for solution of A*X + X*B = C lang fortran by Stephen Nash file linalg/binary-lu for lu decomposition on a binary matrix in binary arithmetic. by Burt Garbow, ANL, 11/86 file linalg/bemw.tgz for 4 implementations of the mixed block elimination method by W. Govaerts and J. D. Pryce lang fortran prec single real, double real, single complex, double complex gams d2a4, d2b4, d2c4 file linalg/bmr.tgz for An MPI implementation of the Broadcast-Multiply-Roll algorithm for , parallel matrix multiplication by Anna Tsao, Supercomputing Research Center, tsao@super.org gams d1b6 encoding compress file linalg/bsmp.f by Bank and Smith for sparse LU made simple gams d2a4, d2b4 file linalg/cg.tgz for A collection of conjugate gradient type iterative methods , for distributed memory architectures, using PVM as the , message passing protocol. Includes demo and X window interface. , (371886 bytes) by Victor Eijkhout lib linalg/colamd for approximate minimum degree column ordering by Stefan I. Larimore, Timothy A. Davis (University of Florida, email: , davis@cise.ufl.edu), John Gilbert, Esmond Ng alg approximate minimum degree column ordering file linalg/conest for estimates the 1-norm of a square, complex matrix A. , reverse communication is used for evaluating matrix-vector product. by N.J. Higham, 1987. file linalg/dcgc for a preconditioned conjugate gradient code written in 'c' using , double precision. by Mark Seager, LLNL, 10/26/86 lang c gams d2b1b, d2b4 file linalg/eiginv for numerical solution of the inverse eigenvalue problem by Burt Garbow, ANL, 11/86 file linalg/gemmw.tgz for a highly portable Level 3 BLAS implementation of Winograd's variant , Strassen's matrix multiplication algorithm by douglas-craig@CS.YALE.EDU ("Craig C. Douglas") Nov 27 1995 encoding tar, gzip gams d1b6 file linalg/goliath.f for Fortran with driver for exact analysis of sparse rectangular rational , linear systems by Peter Alfeld and David Eyre, 1989. file linalg/goliath.tex for user manual for the exact analysis of sparse rectangular rational , linear systems by Peter Alfeld and David Eyre, 1989. file linalg/guptri for This package of routines contains robust software with error , bounds for computing the generalized Schur decomposition of an , arbitrary pencil A - zB (regular or singular). The decomposition , (guptri - generalized upper triangular form) is a generalization , of the Schur canonical form of A -zI to matrix pencils and reveals , the Kronecker structure of a singular pencil. , More information of the package is placed in README, , where you also can find references to papers describing software, , algorithms and error bounds used in the package. The package is , developed by Jim Demmel and Bo Kagstrom (adresses in README). gams d4b4 lang fortran file linalg/hssxev for - An out-of-core symmetric eigensolve r for large , dense problems. It uses block householder reductions to reduce the full , dense matrix to banded form. The banded form is then reduced to , tridiagonal form and all eigenvalues are computed. Specified , eigenvectors are computed using inverse iteration with the band , matrix and then back transformed to orginal form. by Roger Grimes, Boeing Computer Services, Nov 1987 gams d4a1 file linalg/ibmblas3 for A Fortran implementation of the Level 3 BLAS optimized for the , IBM 3090. Bo Kagstrom bokg@cs.umu.se Tue Oct 30 10:37:01 1990 gams d1b file linalg/iccg for nonsymmetric sparse solver using implicite normal equations by Dongarra, Leaf, and Minkoff. lang fortran gams d2a4 file linalg/iccg-doc by Dongarra, Leaf, and Minkoff. file linalg/iccg-paper lang troff by Dongarra, Leaf, and Minkoff. file linalg/ilu.tgz lang C and Fortran 77 by John R. Gilbert , Sivan Toledo for Pivoting incomplete LU factorization. Designed to be used as a , preconditioner for nonsymmetric sparse iterative solvers and , provided with an interface to PETSc. The code supports , drop tolerances, threashold pivoting, and fill ratios. , The code is based on the Gilbert-Peierls factorization code. file linalg/l3abdsol for almost block diagonal linear systems by Cyphers and Paprzycki ref SMU Software Report 92-3 prec single or double lang Fortran 77 gams d2a2 file linalg/lalqmr for A package implementing the Freund, Gutknecht, and Nachtigal version of , the look-ahead Lanczos algorithm. Includes driver code to compute , eigenvalues of matrices, as well as a linear systems solver using the , quasi-minimal residual method. , Roland Freund and Noel Nachtigal, RIACS by Noel M. Nachtigal Tue Jan 14 16:20:35 1992 gams d2a1, d4a2 file linalg/laspack.tgz for LASPack is a package for solving large sparse systems of linear ' equations like those which arise from discretization of partial ' differential equations. , It contains classical as well as selected state-of-the-art algorithms , which are commonly used for large sparse systems such as CG-like methods , for non-symmetric systems (CGN, GMRES, BiCG, QMR, CGS, and BiCGStab) and , multilevel methods such as multigrid and conjugate gradient method , preconditioned by multigrid and BPX preconditioners. , LASPack is written in ANSI C and is thus largely portable. , Postscript and HTML version of the reference manual are included. by Tomas Skalicky lang c gams d2a4 file linalg/linpackc++ for partial interface to LINPACK routines for C++ , single values are passed by reference, arrays by pointer gams d2a1, d3a1 file linalg/linpack.b for limbo version of LINPACK by Nick Zhang date 11/13/98 lang limbo gams d2a1, d3a1 file linalg/lsqr for finds a solution x to the following problems: , 1. Unsymmetric equations -- solve A*x = b , 2. Linear least squares -- solve A*x = b , in the least-squares sense , 3. Damped least squares -- solve ( A )*x = ( b ) , damp*I ) ( 0 , in the least-squares sense , where A is a matrix with m rows and n columns, b is an , m-vector, and damp is a scalar. (All quantities are real.) , The matrix A is intended to be large and sparse. It is accessed , by means of subroutine calls of the form , CALL APROD ( mode,m,n,x,y,LENIW,LENRW,IW,RW ) lang fortran gams d2a4, d9a1 file linalg/metis-4.0.tar.gz for METIS is a set of programs for partitioning graphs, partitioning finite , element meshes, and for producing fill reducing orderings for sparse , matrices. size 508,601K file linalg/optimqr-1.0.tgz for OptimQR is a software package that generates (Fortran) QR solvers , for sparse systems. OptimQR is given the structure of the , system, and will then apply a heuristic B&B search to the , row/column ordering problem, in order to create a near optimal , QR solver. OptimQR is good for solving a large number of , systems with a fixed sparsity pattern. Typical system dimensions , are below 100x100, as the optimization problem becomes very time , consuming as the dimension increases. size 73k date 5/6/1999 by Jakob Østergaard contact Jakob Østergaard url http://ostenfeld.dk/~jakob/OptimQR/ gams d file linalg/pcg.tgz by Graham F. Carey, carey@cfdlab.ae.utexas.edu (512) 471-4207 contact pcg-support@c3.lanl.gov for The PCG package is a software system for solving systems , of linear equations by means of preconditioned conjugate , gradient-type iterative methods on a variety of computer , architectures. The software is designed to give high , performance with nearly identical user interface across different , scalar, vector and parallel platforms as well as across different , programming models such as shared memory, data parallel, and , message passing programming interfaces. , , This distribution includes FORTRAN source code for , uni-processor (serial) workstations and Intel iPSC860 , parallel computers only. Contact Graham F. Carey for , other supported architectures. file linalg/psblas1.0.tgz by PSBLAS Team - University of Roma "Tor Vergata" for A Fortran 90 toolkit to implement parallel iterative solvers for sparse ; linear systems. This library is intended for the development of ; iterative solvers based on Krylov subspace methods on distributed memory ; computers. It is implemented in mixed Fortran 90/Fortran 77 and C, and ; requires the BLACS message passing interface. See README for info on ; current development. lib linalg/qmr for iterative solution of linear systems by Roland Freund and Noel Nachtigal alg QMR (with lookahead, no-lookahead, transpose-free, and other variants) gams d2a1, d2b1, d2c1, d4a1, d4a2, d4a4 file linalg/qmrpack.tgz for iterative solution of linear systems by Roland Freund and Noel Nachtigal alg QMR (with lookahead, no-lookahead, transpose-free, and other variants) size 277 kilobytes # Since this is a large, binary file you cannot use email to get it. gams d2a1, d2b1, d2c1, d4a1, d4a2, d4a4 file linalg/sgefac for a 'c' implementation of the LINPACK routines sgefa and sgesl , which do LU decomposition with partial pivoting (single precision). by Mark Seager, LLNL, 10/26/86 gams d2a1 lang c lib linalg/spooles lang c contact Cleve Ashcraft for SPOOLES is a library for solving sparse linear systems of equations. file linalg/sonest for estimates the 1-norm of a square, real matrix A. , reverse communication is used for evaluating matrix-vector product. by N.J. Higham, 1987. file linalg/sparsdyn for SparseDynamics is a sparse matrix algorithm development tool that , runs on a Sun workstation under suntools. It allows the developer to , observe the dynamic changes in a sparse matrix as it is operated on by , a sparse matrix algorithm. In its current state, it only demonstrates , an early sequential version of the D2 algorithm. SparseDynamics is a , prototype package, so no documentation is provided on how to connect , it to a different sparse matrix algorithm. It has been tested on a Sun , 3/60 with 4 megabytes, a SPARCsystem 330, and a SPARCstation 1. For , more information, refer to "A nondeterministic parallel algorithm for , unsymmetric sparse LU factorization," T. A. Davis, and P.-C. Yew, , SIAM Journal on Matrix Analysis and Applications (July 1990). gams d2a4, s3 file linalg/symmlq for is designed to solve the system of linear equations , A*x = b , where A is an n*n symmetric matrix and b is a given vector. , The matrix A is not required to be positive definite. , (If A is known to be definite, the method of conjugate gradients , may be used -- it will require about the same number of iterations , as SYMMLQ but slightly less work per iteration.) , The matrix A is intended to be large and sparse. It is accessed , by means of a subroutine call of the form , CALL APROD( n,x,y ) , which must return the product y = A*x for any given vector x. gams d2b4 lib templates for "one-liner" drivers of common numerical algorithms , Also, codes from Templates book. file linalg/testmats # A collection of 44 parametrized test matrices, in the form of MATLAB # M-files. The matrices are mostly square, dense, nonrandom, and of # arbitrary dimension. The collection includes matrices with known # inverses or known eigenvalues; ill-conditioned or rank deficient # matrices; and symmetric, positive definite, orthogonal, defective, # involutary, and totally positive matrices. In addition, there are # some further M-files of interest for viewing and modifying the test # matrices. # The M-files are provided in the form of a Unix shar file. # By Nick Higham, July 4 1989. file linalg/tricyclic.f title tricyc for vectorized in-place tridiagonal solution by Dodson and Levin ref SIMAX 13:4 1246-1254 (1992) size 11 kilobytes prec real age stable gams D2a2a rel ok # No pivoting file linalg/underwood for A version of block lanczos based on Richard Underwood's work. file linalg/umfpack3.2.tar.gz title Unsymmetric-pattern Multifrontal Package (UMFPACK Version 3.2) lang ANSI C, MATLAB m-file by Tim Davis, University of Florida (davis@cise.ufl.edu) gams d2a4, d1b3, d1b9 prec double (with "int" integers), double (with "long" integers) for An ANSI C code for sparse LU factorization is presented that combines , a left-looking column pre-ordering strategy with a right-looking , unsymmetric-pattern multifrontal numerical factorization. , The pre-ordering and symbolic analysis phase computes an upper bound on , fill-in, work, and memory usage during the subsequent numerical , factorization. , User-callable routines are provided for ordering and analyzing a sparse , matrix, computing the numerical factorization, solving a system with , the LU factors, transposing and permuting a sparse matrix, and , converting between sparse matrix representations. The simple user , interface shields the user from the details of the complex sparse , factorization data structures by returning simple handles to opaque , objects. Additional user-callable routines are provided for printing , and extracting the contents of these opaque objects. An even simpler , way to use the package is through its MATLAB interface. , See also http://www.cise.ufl.edu/research/sparse/umfpack . file linalg/umfpack2.2.tgz title Unsymmetric-pattern Multifrontal Package (UMFPACK Version 2.2) lang ANSI Fortran 77 by Tim Davis, University of Florida (davis@cis.ufl.edu or na.tdavis@na-net.ornl.gov). Joint work with Iain Duff. gams d2a4 prec single, double, complex, complex double for UMFPACK Version 2.2 is a package for solving systems of sparse , linear systems, Ax=b, where A is sparse and can be unsymmetric. , It is written in ANSI Fortran 77. There are options for , choosing a good pivot order, factorizing a subsequent matrix , with the same pivot order and nonzero pattern as a previously , factorized matrix, and solving systems of linear equations with , the factors (with A, L, or U; or with their transposes in the , single/double precision versions). Iterative refinement, with , sparse backward error estimates, can be performed. Single and , double precision, complex, and complex double precision , (complex*16) routines are available. (Note that complex*16 is , not ANSI Fortran-77, but is a common extension to it). file linalg/oblio.tgz for sparse symmetric linear systems lang C++ by Florin Dobrian and Alex Pothen alg mininum degree, multifrontal # intended as a good example of object oriented design file linalg/acre.tgz for restricted maximum likelihood estimation by Joel Malard .