summaryrefslogtreecommitdiff
path: root/academic/suitesparse/README
blob: d4f385acca77bb5bde8ee32372fbd15186b4ae9a (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
suitesparse is a collection of libraries for computations involving sparse
matrices.  The package includes the following libraries:

AMD: approximate minimum degree ordering.
BTF: permutation to block triangular form (beta).
CAMD: constrained approximate minimum degree ordering.
COLAMD: column approximate minimum degree ordering.
CCOLAMD: constrained column approximate minimum degree ordering.
CHOLMOD: sparse Cholesky factorization.
CXSparse: CSparse extended: complex matrix, int and long int support.
KLU: sparse LU factorization, primarily for circuit simulation.
LDL: a simple LDL^t factorization.
SQPR: a multithread, multifrontal, rank-revealing sparse QR
      factorization method.
UMFPACK: sparse LU factorization.
SuiteSparse_config:  configuration file for all the above packages.
RBio: read/write files in Rutherford/Boeing format.

Suitesparse is an optional dependency for the octave package.

NOTES:

* CHOLMOD is now compiled to use the METIS library. It was not
used previously because the licensing terms of earlier METIS
versions excluded it from Octave in some circumstances. Recent
versions of METIS now use the Apache License Version 2.0 and
should cause no problems. However if it is still desired to
build CHOLMOD without METIS, set the USEMETIS environment
variable to "no" when running the SlackBuild e.g.
    USEMETIS=no sh suitesparse.SlackBuild

* CXSparse is a superset of CSparse, and the two share common header
names, so it does not make sense to build both. CXSparse is built
by default.

* SlackBuilds.org also has all the parts of this suite as individual
builds - DO NOT try and install the individual parts and the suite;
pick one or the other.