boyle / meagre-crowd Goto Github PK

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

symmetric solver

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

http://www.cise.ufl.edu/research/sparse/cholmod/

CHOLMOD is a set of ANSI C routines for sparse Cholesky factorization and
update/downdate. A MATLAB interface is provided.

license: GNU LGPL/GPL

dependencies: BLAS, LAPACK, METIS

Note: symmetric positive definite matrices ONLY, does include various helper functions including basic math operations, and file IO (matrix-market)

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

symmetric solver

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

http://www.hsl.rl.ac.uk/hsl_matlab/ma57/ma57.html

The HSL Fortran 77 package MA57 implements a multifrontal algorithm for the solution of one or more sparse symmetric systems A X = B. An L D L' factorization of A is computed, where the matrix L is a unit lower triangular matrix and D is a block diagonal matrix with blocks of order 1 or, if necessary, 2. To limit the fill-in during the factorization process the matrix A is reordered using either the Approximate Minimum Degree (AMD) algorithm or, if METIS software is available, the Nested Dissection algorithm. Before the factorization starts, scaling of the matrix A is performed using the MC64 package from HSL 2007.

dependencies: METIS, AMD, MC64

license: Free academic use for personal research and teaching, needs agreements, etc.

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from http://web.eecs.utk.edu/~dongarra/etemplates/node388.html#tab:SparseDirectSolver

Multifrontal symmetric

license: free for academic use

superseeded by MA57

dependencies: mc19

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

When a solver can only handle symmetric/symmetric posdef, etc this should be checked and error reported if it will fail.

What is the appropriate posdef test?

(note: cholmod will give a bad answer if not posdef, probably an error code available from cholmod to indicate this?)

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

dependencies: pord, scalapack (lapack, blas, blacs (mpi)), mpi, ptscotch (scotch, mpi), parmetis (metis, mpi)

license: Public domain (?)

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

dependencies: none (single .so)

license: academic/commercial (pardiso.lic - node/user locked)

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

Hi,

When I try to test it, I got error of missing BeBOP library:
checking for library containing load_sparse_matrix... no
I found BeBOP is Berkeley Benchmarking and OPtimization, and got libbebop_util and libsparse_matrix_converter, and set LD_LIBRARY_PATH.
But it still report missing BeBOP library.
Would you please give some help ?

Best wishes,
Jiandong Cheng

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

-- oblio is c++ and its a bit of a pain to integrate into a C program as a result

license: CeCILL-C license. CeCILL-C is fully compatible with the FSF's LGPL

http://hips.gforge.inria.fr/doc/license.html

Hierarchical Iterative Parallel Solver

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

If you loop more than once no timing info is produced.

$ ./meagre-crowd -i tests/test.mm -b tests/rhs1.mm -s umfpack -t -r 100

$ ./meagre-crowd -i tests/test.mm -b tests/rhs1.mm -s umfpack -t
analyze, factorize, evaluate, total (ms)
0.079, 0.086, 0.008, 0.173

If you add more timing requests (-tttt) this used to give you more info on initialization, etc. now it gives you no info at all. Should really remove this "feature" from the documentation and just make it binary on/off behaviour. Multiple -t should be the same as one -t.

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

http://acts.nersc.gov/hypre/

iterative and direct solvers for sparse systems, FEMs, etc

includes algebraic multigrid solvers, krylov, etc...

does not support COMPLEX values

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

What are the dependencies?

What is the license?

Is there anything preventing getting this solver integrated into meagre-crowd?

from x. Li "Direct Solvers for Sparse Matrices" Mar 2010 - list of solvers

dependencies: metis, cholmod (lapack, blas, colamd, ccolamd, camd, amd)

license: GNU GPL

UMFPACK is a set of routines for solving unsymmetric sparse linear systems, Ax=b, using the Unsymmetric MultiFrontal method. Written in ANSI/ISO C, with a MATLAB (Version 6.0 and later) interface. Appears as a built-in routine (for lu, backslash, and forward slash) in MATLAB. Includes a MATLAB interface, a C-callable interface, and a Fortran-callable interface. Note that "UMFPACK" is pronounced in two syllables, "Umph Pack". It is not "You Em Ef Pack".