The official SuiteSparse library: a suite of sparse matrix algorithms authored or co-authored by Tim Davis, Texas A&M University

Overview

SuiteSparse: A Suite of Sparse matrix packages at http://suitesparse.com

May 17, 2021. SuiteSparse VERSION 5.10.1

Now includes GraphBLAS, SLIP_LU, and a new interface to the SuiteSparse
Matrix Collection (ssget), via MATLAB and a Java GUI, to
http://sparse.tamu.edu.

Primary author of SuiteSparse (codes and algorithms, excl. METIS): Tim Davis

Code co-authors, in alphabetical order (not including METIS):

Patrick Amestoy, David Bateman, Jinhao Chen.  Yanqing Chen, Iain Duff,
Les Foster, William Hager, Scott Kolodziej, Chris Lourenco, Stefan
Larimore, Erick Moreno-Centeno, Ekanathan Palamadai, Sivasankaran
Rajamanickam, Sanjay Ranka, Wissam Sid-Lakhdar, Nuri Yeralan.

Additional algorithm designers: Esmond Ng and John Gilbert.

Refer to each package for license, copyright, and author information. All codes are authored or co-authored by Timothy A. Davis.


About the BLAS and LAPACK libraries

NOTE: Use of the Intel MKL BLAS is strongly recommended. A recent OpenBLAS can result in severe performance degradation. The reason for this is being investigated, and this may be resolved in the near future. Ignore the comments about OpenBLAS in the various user guides; those are out of date.


SuiteSparse/README

Packages in SuiteSparse, and files in this directory:

GraphBLAS   graph algorithms in the language of linear algebra.
            https://graphblas.org
            A stand-alone package that uses cmake to compile; see
            GraphBLAS/README.txt.  The rest of SuiteSparse still uses
            'make'.  A cmake setup for all of SuiteSparse is in progress.
            author: Tim Davis

SLIP_LU     solves sparse linear systems in exact arithmetic.
            Requires the GNU GMP and MPRF libraries.

AMD         approximate minimum degree ordering.  This is the built-in AMD
            function in MATLAB.
            authors: Tim Davis, Patrick Amestoy, Iain Duff

bin         where the metis-5.1.0 programs are placed when METIS is compiled

BTF         permutation to block triangular form
            authors: Tim Davis, Ekanathan Palamadai

CAMD        constrained approximate minimum degree ordering
            authors: Tim Davis, Patrick Amestoy, Iain Duff, Yanqing Chen

CCOLAMD     constrained column approximate minimum degree ordering
            authors: Tim Davis, Sivasankaran Rajamanickam, Stefan Larimore.
                Algorithm design collaborators: Esmond Ng, John Gilbert
                (for COLAMD)

ChangeLog   a summary of changes to SuiteSparse.  See */Doc/ChangeLog
            for details for each package.

CHOLMOD     sparse Cholesky factorization.  Requires AMD, COLAMD, CCOLAMD,
            the BLAS, and LAPACK.  Optionally uses METIS.  This is chol and
            x=A\b in MATLAB.
            author for all modules: Tim Davis 
            CHOLMOD/Modify module authors: Tim Davis and William W. Hager

COLAMD      column approximate minimum degree ordering.  This is the
            built-in COLAMD function in MATLAB.
            authors (of the code): Tim Davis and Stefan Larimore
            Algorithm design collaborators: Esmond Ng, John Gilbert

Contents.m  a list of contents for 'help SuiteSparse' in MATLAB.

CSparse     a concise sparse matrix package, developed for my
            book, "Direct Methods for Sparse Linear Systems",
            published by SIAM.  Intended primarily for teaching.
            It does have a 'make install' but I recommend using
            CXSparse instead.  In particular, both CSparse and CXSparse
            have the same include filename: cs.h.

            This package is used for the built-in DMPERM in MATLAB.
            author: Tim Davis

CSparse_to_CXSparse
            a Perl script to create CXSparse from CSparse and
            CXSparse_newfiles
            author: David Bateman, Motorola

CXSparse    CSparse Extended.  Includes support for complex matrices
            and both int or long integers.  Use this instead of CSparse
            for production use; it creates a libcsparse.so (or *dylib on
            the Mac) with the same name as CSparse.  It is a superset
            of CSparse.  Any code that links against CSparse should
            also be able to link against CXSparse instead.
            author: Tim Davis, David Bateman

CXSparse_newfiles
            Files unique to CXSparse
            author: Tim Davis, David Bateman

share       'make' places documentation for each package here

include     'make' places user-visible include fomes for each package here

KLU         sparse LU factorization, primarily for circuit simulation.
            Requires AMD, COLAMD, and BTF.  Optionally uses CHOLMOD,
            CAMD, CCOLAMD, and METIS.
            authors: Tim Davis, Ekanathan Palamadai

LDL         a very concise LDL' factorization package
            author: Tim Davis

lib         'make' places shared libraries for each package here

Makefile    to compile all of SuiteSparse
            make            compiles SuiteSparse libraries and runs demos
            make install    compiles SuiteSparse and installs in the
                            current directory (./lib, ./include).
                            Use "sudo make INSTALL=/usr/local" to install
                            in /usr/local/lib and /usr/local/include.
            make uninstall  undoes 'make install'
            make library    compiles SuiteSparse libraries (not demos)
            make distclean  removes all files not in distribution, including
                            ./bin, ./share, ./lib, and ./include.
            make purge      same as 'make distclean'
            make clean      removes all files not in distribution, but
                            keeps compiled libraries and demoes, ./lib,
                            ./share, and ./include.
            make config     displays parameter settings; does not compile

            Each individual package also has each of the above 'make'
            targets.  Doing 'make config' in each package */Lib directory
            displays the exact shared and static library names.

            Things you don't need to do:
            make cx         creates CXSparse from CSparse
            make docs       creates user guides from LaTeX files
            make cov        runs statement coverage tests (Linux only)
            make metis      compiles METIS (also done by 'make')

MATLAB_Tools    various m-files for use in MATLAB
            author: Tim Davis (all parts)
            for spqr_rank: author Les Foster and Tim Davis

            Contents.m      list of contents
            dimacs10        loads matrices for DIMACS10 collection
            Factorize       object-oriented x=A\b for MATLAB
            find_components finds connected components in an image
            GEE             simple Gaussian elimination
            getversion.m    determine MATLAB version
            gipper.m        create MATLAB archive
            hprintf.m       print hyperlinks in command window
            LINFACTOR       predecessor to Factorize package
            MESHND          nested dissection ordering of regular meshes
            pagerankdemo.m  illustrates how PageRank works
            SFMULT          C=S*F where S is sparse and F is full
            shellgui        display a seashell
            sparseinv       sparse inverse subset
            spok            check if a sparse matrix is valid
            spqr_rank       SPQR_RANK package.  MATLAB toolbox for rank
                            deficient sparse matrices: null spaces,
                            reliable factorizations, etc.  With Leslie
                            Foster, San Jose State Univ.
            SSMULT          C=A*B where A and B are both sparse
            SuiteSparseCollection    for the SuiteSparse Matrix Collection
            waitmex         waitbar for use inside a mexFunction

            The SSMULT and SFMULT functions are the basis for the
            built-in C=A*B functions in MATLAB.

Mongoose    graph partitioning.
            authors: Nuri Yeralan, Scott Kolodziej, William Hager, Tim Davis

metis-5.1.0 a modified version of METIS.  See the README.txt files for
            details.
            author: George Karypis; not an integral component of
            SuiteSparse, however.  This is just a copy included with
            SuiteSparse via the open-source license provided by
            George Karypis

RBio        read/write sparse matrices in Rutherford/Boeing format
            author: Tim Davis

README.txt  this file

SPQR        sparse QR factorization.  This the built-in qr and x=A\b in
            MATLAB.
            author of the CPU code: Tim Davis
            author of GPU modules: Tim Davis, Nuri Yeralan,
                Wissam Sid-Lakhdar, Sanjay Ranka

            SPQR/GPUQREngine: GPU support package for SPQR
            (not built into MATLAB, however)
            authors: Tim Davis, Nuri Yeralan, Sanjay Ranka,
                Wissam Sid-Lakhdar

SuiteSparse_config    configuration file for all the above packages.  The
            SuiteSparse_config/SuiteSparse_config.mk is included in the
            Makefile's of all packages.  CSparse and MATLAB_Tools do not
            use SuiteSparse_config.
            author: Tim Davis

SuiteSparse_GPURuntime      GPU support package for SPQR and CHOLMOD
            (not builtin to MATLAB, however).

SuiteSparse_install.m       install SuiteSparse for MATLAB

SuiteSparse_test.m          exhaustive test for SuiteSparse in MATLAB

ssget       MATLAB interface to the SuiteSparse Matrix Collection
            (formerly called the UF Sparse Matrix Collection).
            Includes a UFget function for backward compatibility.
            author: Tim Davis

UMFPACK     sparse LU factorization.  Requires AMD and the BLAS.
            This is the built-in lu and x=A\b in MATLAB.
            author: Tim Davis
            algorithm design collaboration: Iain Duff

Some codes optionally use METIS 5.1.0. This package is located in SuiteSparse in the metis-5.1.0 directory. Its use is optional, so you can remove it before compiling SuiteSparse, if you desire. The use of METIS will improve the ordering quality. METIS has been slightly modified for use in SuiteSparse; see the metis-5.1.0/README.txt file for details. SuiteSparse can use the unmodified METIS 5.1.0, however. To use your own copy of METIS, or a pre-installed copy of METIS use 'make MY_METIS_LIB=-lmymetis' or 'make MY_METIS_LIB=/my/stuff/metis-5.1.0/whereeveritis/libmetis.so MY_METIS_INC=/my/stuff/metis-5.1.0/include'. If you want to use METIS in MATLAB, however, you MUST use the version provided here, in SuiteSparse/metis-5.1.0. The MATLAB interface to METIS required some small changes in METIS itself to get it to work. The original METIS 5.1.0 will segfault MATLAB.

Refer to each package for license, copyright, and author information. All codes are authored or co-authored by Timothy A. Davis. email: [email protected]

Licenses for each package are located in the following files, all in PACKAGENAME/Doc/License.txt:

AMD/Doc/License.txt
BTF/Doc/License.txt
CAMD/Doc/License.txt
CCOLAMD/Doc/License.txt
CHOLMOD/Doc/License.txt
COLAMD/Doc/License.txt
CSparse/Doc/License.txt
CXSparse/Doc/License.txt
GPUQREngine/Doc/License.txt
KLU/Doc/License.txt
LDL/Doc/License.txt
MATLAB_Tools/Doc/License.txt
Mongoose/Doc/License.txt
RBio/Doc/License.txt
SPQR/Doc/License.txt
SuiteSparse_GPURuntime/Doc/License.txt
ssget/Doc/License.txt
UMFPACK/Doc/License.txt
GraphBLAS/Doc/License.txt

These files are also present, but they are simply copies of the above license files for CXSparse and ssget:

CXSparse_newfiles/Doc/License.txt
CSparse/MATLAB/ssget/Doc/License.txt
CXSparse/MATLAB/ssget/Doc/License.txt

METIS 5.0.1 is distributed with SuiteSparse, and is Copyright (c) by George Karypis. Please refer to that package for its License.


QUICK START FOR MATLAB USERS (Linux, Mac, or Windows):

Uncompress the SuiteSparse.zip or SuiteSparse.tar.gz archive file (they contain the same thing). Next, compile the GraphBLAS library (see instructions in GraphBLAS/Doc). Then in the MATLAB Command Window, cd to the SuiteSparse directory and type SuiteSparse_install. All packages will be compiled, and several demos will be run. To run a (long!) exhaustive test, do SuiteSparse_test.


QUICK START FOR THE C/C++ LIBRARIES:

For just GraphBLAS, do this:

cd GraphBLAS/build ; cmake .. ; make ; cd ../Demo ; ./demo 
cd ../build ; sudo make install

For all other packages, type 'make' in this directory. All libraries will be created and copied into SuiteSparse/lib. All include files need by the applications that use SuiteSparse are copied into SuiteSparse/include. All user documenation is copied into SuiteSparse/share/doc.

Be sure to first install all required libraries: BLAS and LAPACK for UMFPACK, CHOLMOD, and SPQR, and GMP and MPFR for SLIP_LU. Be sure to use the latest libraries; SLIP_LU requires MPFR 4.0 for example.

When compiling the libraries, do NOT use the INSTALL=... options for installing. Just do:

make

or to compile just the libraries without running the demos, do:

make library

Any program that uses SuiteSparse can thus use a simpler rule as compared to earlier versions of SuiteSparse. If you add /home/myself/SuiteSparse/lib to your library search patch, you can do the following (for example):

S = /home/myself/SuiteSparse
cc myprogram.c -I$(S)/include -lumfpack -lamd -lcholmod -lsuitesparseconfig -lm

To change the C and C++ compilers, and to compile in parallel use:

AUTOCC=no CC=gcc CX=g++ JOBS=32 make

for example, which changes the compiler to gcc and g++, and runs make with 'make -j32', in parallel with 32 jobs.

Now you can install the libraries, if you wish, in a location other than SuiteSparse/lib, SuiteSparse/include, and SuiteSparse/share/doc, using 'make install INSTALL=...'

Do 'make install' if you want to install the libraries and include files in SuiteSparse/lib and SuiteSparse/include, and the documentation in SuiteSparse/doc/suitesparse-VERSION. This will work on Linux/Unix and the Mac. It should automatically detect if you have the Intel compilers or not, and whether or not you have CUDA. If this fails, see the SuiteSparse_config/SuiteSparse_config.mk file. There are many options that you can either list on the 'make' command line, or you can just edit that file. For example, to compile with your own BLAS:

make BLAS=-lmyblaslibraryhere

NOTE: Use of the Intel MKL BLAS is strongly recommended. The OpenBLAS can result in severe performance degradation, in CHOLMOD in particular.

To list all configuration options (but not compile anything), do:

make config

Any parameter you see in the output of 'make config' with an equal sign can be modified at the 'make' command line.

If you do "make install" by itself, then the packages are all installed in SuiteSparse/lib (libraries), SuiteSparse/include (include .h files), and SuiteSparse/doc/suitesparse-VERSION (documentation). To install in /usr/local, the default location for Linux, do:

make library
sudo make install INSTALL=/usr/local

If you want to install elsewhere, say in /my/path, first ensure that /my/path is in your LD_LIBRARY_PATH. How to do that depends on your system, but in the bash shell, add this to your ~/.bashrc file:

LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/my/path
export LD_LIBRARY_PATH

You may also need to add SuiteSparse/lib to your path. If your copy of SuiteSparse is in /home/me/SuiteSparse, for example, then add this to your ~/.bashrc file:

LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/me/SuiteSparse/lib:/my/path
export LD_LIBRARY_PATH

Then do the following (use "sudo make ..." if needed):

make library
make install INSTALL=/my/path

which puts the files in /my/path/lib, /my/path/include, and /my/path/doc. If you want to selectively put the libraries, include files, and doc files in different locations, do:

make install INSTALL_LIB=/my/libs INSTALL_INCLUDE=/myotherstuff/include INSTALL_DOC=/mydocs

for example. Any term not defined will be set to its default, so if you don't want to install the documentation, but wish to install the libraries and includes in /usr/local/lib and /usr/local/include, do:

make install INSTALL_DOC=/tmp/doc

which copies the documentation to /tmp/doc where you can then remove it later.

Both the static (.a) and shared (.so) libraries are compiled. The lib.a libraries are left in the package Lib folder (AMD/Lib/libamd.a for example). The main exception to this rule is the SuiteSparse_config library, which is in SuiteSparse/libsuiteSparseconfig.a. SuiteSparse_config is required by all packages. The (extremely) optional xerbla library is also an exception, but it is highly unlikely that you need that library.

The 'make uninstall' takes the same command-line arguments.


Step-by-step details:

(1) Use the right BLAS and LAPACK libraries. Determine where your BLAS and LAPACK libraries are. If the default 'make' does not find them, use 'make BLAS=-lmyblaslibraryhere LAPACK=-lmylapackgoeshere'

(2) Install Intel's Threading Building Blocks (TBB). This is optionally used by SuiteSparseQR. Refer to the User Guide in SuiteSparse/SPQR/Doc/spqr_user_guide.pdf for details.

(3) Determine what other command line options you need for 'make'. All options can be set at the 'make' command line without the need to edit this file. Browse that file to see what options you can control. If you choose different options and wish to recompile, be sure to do 'make distclean' in this directory first, to remove all files not in the original distribution.

(4) Type "make" in this directory. All packages will be be compiled. METIS 5.1.0 will be compiled if you have it (note that METIS require CMake to build it). Several demos will be run. To compile just the libraries, without running any demos, use "make library". The libraries will appear in /Lib/.so.* (*.dylib for the Mac). Include files, as needed by user programs that use CHOLMOD, AMD, CAMD, COLAMD, CCOLAMD, BTF, KLU, UMFPACK, LDL, etc. are in /Include/.h. The include files required by user programs are then copied into SuiteSparse/include, and the compiled libraries are copied into SuiteSparse/lib. Documentation is copied into SuiteSparse/doc. The GraphBLAS libraries are created by cmake and placed in GraphBLAS/build. NOTE: on Linux, you may see some errors when you compile METIS ('make: *** No rule to make target 'w'.). You can safely ignore those errors.

(6) To install, type "make install". This will place copies of all libraries in SuiteSparse/lib, and all include files in SuiteSparse/include, and all documentation in SuiteSparse/doc/suitesparse-VERSION. You can change the install location by "make install INSTALL=/my/path" which puts the libraries in /my/path/lib, the include files in /my/path/include, and documentation in /my/path/doc. These directories are created if they do not already exist.

(7) To uninstall, type "make uninstall", which reverses "make install" by removing the SuiteSparse libraries, include files, and documentation from the place they were installed. If you pass INSTALL_***= options to 'make install', you must pass the same to 'make uninstall'.

Issues
  • How to compile against MKL

    How to compile against MKL

    Hi, This s really a question rather than a bug, but I have a hard time understanding the documentation of how to compile suitesparse with MKL

    The ReadMe it says you can do

    make BLAS=-lmyblaslibraryhere
    
    

    But what is -lmyblaslibraryhere ??

    My MKL libs are in /opt/intel/oneapi/mkl/latest/lib/intel64, which is not seen by the dynamic linker, and what is myblas? libmkt-rt? or liblmkl_rt or -L/opt/intel/oneapi/mkl/latest/lib/intel64 -lmkl? ??

    so exactly what I should put there?

    The ReadMe said If I do the above I don't have to edit SuiteSparse_config.mk. But I figure it would be easier for me if I just edit it, but the ReadMe doesn't provide any info of how to do that if I want to. There is a stackoverflow thread but the config file looks different, there, must be from an old version of SuiteSparse.

    So, taking a cue from line 190 and 191 I edited he block between line 98 and 112 so it becomes

    ifeq ($(TCOV),yes)
            # Each package has a */Tcov directory for extensive testing, including
            # statement coverage.  The Tcov tests require Linux and gcc, and use
            # the vanilla BLAS.  For those tests, the packages use 'make TCOV=yes',
            # which overrides the following settings:
            MKLROOT = /opt/intel/oneapi/mkl/latest
            AUTOCC = no
            CC = gcc
            CXX = g++
            BLAS = -lmkl_intel_lp64 -lmkl_core -lmkl_intel_thread -liomp5 -lpthread -lm
            LAPACK = -lmkl_intel_lp64 -lmkl_core -lmkl_intel_thread -liomp5 -lpthread -lm
            CFLAGS += --coverage
            OPTIMIZATION = -g
            LDFLAGS += --coverage
        endif
    

    But when I do make config I still see

    BLAS library:             BLAS=            -lblas
    LAPACK library:           LAPACK=          -llapack
    

    So what is the right way to do it?

    Please help. Thanks!

    Edited: So if I do

    export MKLROOT=/opt/intel/oneapi/mkl/latest/
    

    make config produces (without editing SuiteSparse_config.mk)

    BLAS library:             BLAS=            -lmkl_intel_lp64 -lmkl_core -lmkl_intel_thread -liomp5 -lpthread -lm
    LAPACK library:           LAPACK=          
    

    But make failed because ld cannot find these libraries

    usr/bin/ld: cannot find -lmkl_intel_lp64
    /usr/bin/ld: cannot find -lmkl_core
    /usr/bin/ld: cannot find -lmkl_intel_thread
    /usr/bin/ld: cannot find -liomp5
    
    opened by beew 14
  • How to put maximum (in row) value on diagonal?

    How to put maximum (in row) value on diagonal?

    Is it possible to get R matrix with maximum (in row) value on diagonal? Small example: A= {{1 0 0},{1 1e-5 5}} spqr(A)->R R= {{1e-5 1 5}, {0,1,0}} So we get 1e-5 on diagonal, but there is 5 in this row. If we drop this row by tolerance we lose some valuable row. My actual matrixes A, R, in attachment. Problem is in rows 1318, 1328, 1337, 1347.. Matrices.zip

    opened by KukushkinAleksei 13
  • New paper that yields 12-24x faster runtime for BFS

    New paper that yields 12-24x faster runtime for BFS

    Is your feature request related to a problem? Please describe. Not a problem; just came across a recent paper (April 2021) that claims to dramatically increase the performance of BFS over current implementations of GraphBLAS. We're going to be using RedisGraph's BFS functionality, and RedisGraph uses SuiteSparse.

    Describe the solution you'd like Potentially implement the paper in question and thereby increase BFS performance.

    Describe alternatives you've considered N/A

    Additional context While I'm a software engineer, I'm not at all a specialist in BLAS or C programming, so if this paper is not implementable as written, or if it's too much of a refactoring effort, I understand. Just wanted to bring this to your attention if it's helpful!

    Also, let me know if this is a feature discussion that belongs upstream (e.g. the GraphBLAS C API). Again, not being a specialist in this, I'm not sure whether the paper suggests an overhaul in current GraphBLAS implementations or in the GraphBLAS API itself.

    opened by alexandergunnarson 9
  • GraphBLAS/Source/Generated/GB_AxB__any_first_bool.c:278:5: error: this 'if' clause does not guard... [-Werror=misleading-indentation]

    GraphBLAS/Source/Generated/GB_AxB__any_first_bool.c:278:5: error: this 'if' clause does not guard... [-Werror=misleading-indentation]

    Using latest GCC 11, I can see the following error:

    /home/abuild/rpmbuild/BUILD/SuiteSparse-5.9.0/GraphBLAS/Source/Template/GB_bitmap_AxB_saxpy_A_bitmap_B_sparse_template.c: In function 'GB_Asaxpy3B__any_first_bool':
    /home/abuild/rpmbuild/BUILD/SuiteSparse-5.9.0/GraphBLAS/Source/Generated/GB_AxB__any_first_bool.c:278:5: error: this 'if' clause does not guard... [-Werror=misleading-indentation]
      278 |     if (exists && !cb) cx = (ax) ; cb |= exists
          |     ^~
    /home/abuild/rpmbuild/BUILD/SuiteSparse-5.9.0/GraphBLAS/Source/Template/GB_bitmap_AxB_saxpy_A_bitmap_B_sparse_template.c:327:33: note: in expansion of macro 'GB_BITMAP_MULTADD'
      327 |                                 GB_BITMAP_MULTADD (
          |                                 ^~~~~~~~~~~~~~~~~
    /home/abuild/rpmbuild/BUILD/SuiteSparse-5.9.0/GraphBLAS/Source/Template/GB_bitmap_AxB_saxpy_A_bitmap_B_sparse_template.c:328:37: note: ...this statement, but the latter is misleadingly indented as if it were guarded by the 'if'
      328 |                                     Hf [pH+ii], Hx [pH+ii],
          |                                     ^~
    /home/abuild/rpmbuild/BUILD/SuiteSparse-5.9.0/GraphBLAS/Source/Generated/GB_AxB__any_first_bool.c:278:36: note: in definition of macro 'GB_BITMAP_MULTADD'
      278 |     if (exists && !cb) cx = (ax) ; cb |= exists
          |                                    ^~
    

    I see the project intentionally enables it with:

    GraphBLAS/Demo/Include/graphblas_demos.h:#pragma GCC diagnostic error "-Wmisleading-indentation"
    GraphBLAS/Source/GB_warnings.h:#pragma GCC diagnostic error "-Wmisleading-indentation"
    
    opened by marxin 9
  • Installation problem with MATLAB in Windows 10

    Installation problem with MATLAB in Windows 10

    I'm trying to install SuiteSparse 5.7.2 on MATLAB R2020a using Windows 10 and Microsoft Visual C++ 2019. However, the following issue appers:

    Installing SuiteSparse for MATLAB version 9.8.0.1323502 (R2020a)
    
    Compiling UMFPACK for MATLAB Version 9.8.0.1323502 (R2020a)
    with 64-bit BLAS
    with CHOLMOD, CAMD, CCOLAMD, and METIS
    ............................b64.c
    B:\Francisco\Documents\MATLAB\SuiteSparse-5.7.2\metis-5.1.0\GKlib\GKlib.h(68): fatal error C1083: No se puede abrir el archivo incluir: 'regex.h': No such file or directory
    
    
    UMFPACK not installed
    
    Compiling CHOLMOD with METIS 5.1.0 for MATLAB Version 9.8.0.1323502 (R2020a)
    with 64-bit BLAS
    
    .b64.c
    B:\Francisco\Documents\MATLAB\SuiteSparse-5.7.2\metis-5.1.0\GKlib\GKlib.h(68): fatal error C1083: No se puede abrir el archivo incluir: 'regex.h': No such file or directory
    
    
    CHOLMOD not installed
    
    ....
    
    opened by fjramireg 9
  • Usage of deprecated and removed tbb::task_scheduler_init and task API (Intel oneTBB)

    Usage of deprecated and removed tbb::task_scheduler_init and task API (Intel oneTBB)

    tbb::task_scheduler_init has been deprecated since Intel TBB 4.3 Update 5 and in the new Intel oneTBB 2021.x releases it is finally completely gone.

    According to documentation,

    Using task_scheduler_init is optional in Intel® Threading Building Blocks (Intel® TBB) 2.2. By default, Intel TBB 2.2 automatically creates a task scheduler the first time that a thread uses task scheduling services and destroys it when the last such thread exits.

    and Intel does not recommend specifying the number of threads anywhere:

    The reason for not specifying the number of threads in production code is that in a large software project, there is no way for various components to know how many threads would be optimal for other threads. Hardware threads are a shared global resource. It is best to leave the decision of how many threads to use to the task scheduler.

    Please remove the usage of tbb::task_scheduler_init and the tbb/task_scheduler_init.h header, or at least detect their existence in the build system if you want to preserve the behavior with older TBB versions.

    https://github.com/DrTimothyAldenDavis/SuiteSparse/blob/1869379f464f0f8dac471edb4e6d010b2b0e639d/SPQR/Source/spqr_parallel.cpp#L91-L92


    Actually the whole tbb::task API is gone. task_group is suggested as a replacement.

    opened by unrelentingtech 8
  • Question about Intel MKL BLAS

    Question about Intel MKL BLAS

    Hi, I'm trying to set up SuiteSparse to be used in MATLAB, following these quick start instructions.

    A note in the README reads "use of the Intel MKL BLAS is strongly recommended." If I understand correctly, the CMakeLists.txt file for compiling GraphBLAS must reflect this preference for Intel MKL BLAS. However, I am a bit confused with lines 119-147 of the config file. https://github.com/DrTimothyAldenDavis/SuiteSparse/blob/9d97e0a2fb3d03fac66cd45d3a84338b0a97f368/GraphBLAS/CMakeLists.txt#L119-L147

    If these lines are commented out, how will Intel MKL BLAS be used? I was expecting something like this (from the FindBLAS documentation):

    set(BLA_VENDOR Intel10_64lp)
    find_package(BLAS)
    
    opened by christian-cahig 6
  • 5.10.1: build fails against tbb 2021.4.0

    5.10.1: build fails against tbb 2021.4.0

    Describe the bug Looks like build fails when in build env is tbb 2021.4.0

    To Reproduce Install tbb 2021.4.0 and build.

    Expected behavior Build should not fail.

    Screenshots

    /home/tkloczko/rpmbuild/BUILD/suitesparse-5.10.1/SuiteSparse-5.10.1
    + cd SPQR
    + /usr/bin/make -O -j12 V=1 VERBOSE=1 -C Lib 'CFLAGS=-O2 -g -grecord-gcc-switches -pipe -Wall -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -Wp,-D_GLIBCXX_ASSERTIONS -specs=/usr/lib/rpm/redhat/redhat-hardened-cc1 -fstack-protector-strong -specs=/usr/lib/rpm/redhat/redhat-annobin-cc1 -m64 -mtune=generic -fasynchronous-unwind-tables -fstack-clash-protection -fcf-protection -fdata-sections -ffunction-sections -flto=auto -flto-partition=none -I/usr/include/metis -DHAVE_TBB -DNPARTITION' TBB=-ltbb BLAS=-lflexiblas LIBRARY_SUFFIX=
    /usr/bin/make install INSTALL=/home/tkloczko/rpmbuild/BUILD/suitesparse-5.10.1/SuiteSparse-5.10.1
    make[1]: Entering directory '/home/tkloczko/rpmbuild/BUILD/suitesparse-5.10.1/SuiteSparse-5.10.1/SPQR/Lib'
    /usr/bin/g++ -O2 -g -grecord-gcc-switches -pipe -Wall -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -Wp,-D_GLIBCXX_ASSERTIONS -specs=/usr/lib/rpm/redhat/redhat-hardened-cc1 -fstack-protector-strong -specs=/usr/lib/rpm/redhat/redhat-annobin-cc1 -m64 -mtune=generic -fasynchronous-unwind-tables -fstack-clash-protection -fcf-protection -fdata-sections -ffunction-sections -flto=auto -flto-partition=none -I/usr/include/metis -DHAVE_TBB -DNPARTITION   -O3 -fexceptions -fPIC -fopenmp   -DNPARTITION -I../../CHOLMOD/Include -I../../SuiteSparse_config -I../Include -c ../Source/spqr_parallel.cpp
    ../Source/spqr_parallel.cpp:10:10: fatal error: tbb/task_scheduler_init.h: No such file or directory
       10 | #include <tbb/task_scheduler_init.h>
          |          ^~~~~~~~~~~~~~~~~~~~~~~~~~~
    compilation terminated.
    make[1]: *** [Makefile:126: spqr_parallel.o] Error 1
    make[1]: Leaving directory '/home/tkloczko/rpmbuild/BUILD/suitesparse-5.10.1/SuiteSparse-5.10.1/SPQR/Lib'
    
    [[email protected] SPECS]$ rpm -ql tbb-devel | grep include/tbb
    /usr/include/tbb
    /usr/include/tbb/blocked_range.h
    /usr/include/tbb/blocked_range2d.h
    /usr/include/tbb/blocked_range3d.h
    /usr/include/tbb/blocked_rangeNd.h
    /usr/include/tbb/cache_aligned_allocator.h
    /usr/include/tbb/collaborative_call_once.h
    /usr/include/tbb/combinable.h
    /usr/include/tbb/concurrent_hash_map.h
    /usr/include/tbb/concurrent_lru_cache.h
    /usr/include/tbb/concurrent_map.h
    /usr/include/tbb/concurrent_priority_queue.h
    /usr/include/tbb/concurrent_queue.h
    /usr/include/tbb/concurrent_set.h
    /usr/include/tbb/concurrent_unordered_map.h
    /usr/include/tbb/concurrent_unordered_set.h
    /usr/include/tbb/concurrent_vector.h
    /usr/include/tbb/enumerable_thread_specific.h
    /usr/include/tbb/flow_graph.h
    /usr/include/tbb/flow_graph_abstractions.h
    /usr/include/tbb/global_control.h
    /usr/include/tbb/info.h
    /usr/include/tbb/memory_pool.h
    /usr/include/tbb/null_mutex.h
    /usr/include/tbb/null_rw_mutex.h
    /usr/include/tbb/parallel_for.h
    /usr/include/tbb/parallel_for_each.h
    /usr/include/tbb/parallel_invoke.h
    /usr/include/tbb/parallel_pipeline.h
    /usr/include/tbb/parallel_reduce.h
    /usr/include/tbb/parallel_scan.h
    /usr/include/tbb/parallel_sort.h
    /usr/include/tbb/partitioner.h
    /usr/include/tbb/profiling.h
    /usr/include/tbb/queuing_mutex.h
    /usr/include/tbb/queuing_rw_mutex.h
    /usr/include/tbb/scalable_allocator.h
    /usr/include/tbb/spin_mutex.h
    /usr/include/tbb/spin_rw_mutex.h
    /usr/include/tbb/task.h
    /usr/include/tbb/task_arena.h
    /usr/include/tbb/task_group.h
    /usr/include/tbb/task_scheduler_observer.h
    /usr/include/tbb/tbb.h
    /usr/include/tbb/tbb_allocator.h
    /usr/include/tbb/tbbmalloc_proxy.h
    /usr/include/tbb/tick_count.h
    /usr/include/tbb/version.h
    [[email protected] SPECS]$ rpm -ql tbb-devel | grep task_scheduler_init.h
    [[email protected] SPECS]$
    

    Desktop (please complete the following information):

    • OS: [e.g. iOS]
    • compiler [e.g. gcc 7.4, icc 19.0]
    • BLAS and LAPACK library, if applicable
    • Version [e.g. 22]

    Additional context Add any other context about the problem here.

    opened by kloczek 6
  • Invalid read in GB_wait with hypersparse matrix

    Invalid read in GB_wait with hypersparse matrix

    valgrind is complaining about an invalid memory read during GrB_Matrix_assign of a hyper sparse matrix. Later I'm dying with corrupted GrB objects. I'm confident they are related, though I have not done you the courtesy of making a minimal reproducer.

    ==51679== Invalid read of size 8 ==51679== at 0x20086D17: GB_wait (GB_wait.c:274) ==51679== by 0x2005917C: GB_block (GB_block.c:33) ==51679== by 0x200689F9: GB_subassigner (GB_subassigner.c:1441) ==51679== by 0x20055982: GB_assign (GB_assign.c:634) ==51679== by 0x2003D9A6: GrB_Matrix_assign (GrB_Matrix_assign.c:46)

    ==51679== Address 0xed6bce8 is 8 bytes before a block of size 8 alloc'd ==51679== at 0x442E01F: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) ==51679== by 0x200531AE: GB_Global_malloc_function (GB_Global.c:345) ==51679== by 0x2005DA69: GB_malloc_memory (GB_malloc_memory.c:125) ==51679== by 0x2005FA1D: GB_new (GB_new.c:191) ==51679== by 0x201CC1E0: GB_builder (GB_builder.c:787) ==51679== by 0x20086A7F: GB_wait (GB_wait.c:160) ==51679== by 0x2005917C: GB_block (GB_block.c:33) ==51679== by 0x200689F9: GB_subassigner (GB_subassigner.c:1441) ==51679== by 0x20055982: GB_assign (GB_assign.c:634) ==51679== by 0x2003D9A6: GrB_Matrix_assign (GrB_Matrix_assign.c:46)

    Heres the relevant slice of GB_wait.c. It looks to me like we're reading Ah[-1] when kA==0, and glancing at GB_new it does not seem to allocate A->h such that A->h[-1] is valid memory. So I declare you are OffByOne. Forgive me if I am mistaken.

        if (A->is_hyper)
        { 
            // find tjfirst in A->h 
            int64_t pright = A->nvec - 1 ;
            bool found ;
            int64_t *GB_RESTRICT Ah = A->h ;
            GB_SPLIT_BINARY_SEARCH (tjfirst, Ah, kA, pright, found) ;
            // Ah [0 ... kA-1] excludes vector tjfirst.  The list
            // Ah [kA ... A->nvec-1] includes tjfirst.
            ASSERT (kA >= 0 && kA <= A->nvec) ;
            ASSERT (GB_IMPLIES (kA > 0 && kA < A->nvec, Ah [kA-1] < tjfirst)) ;
            ASSERT (GB_IMPLIES (found, Ah [kA] == tjfirst)) ;
            anz0 = A->p [kA] ;
            jlast = Ah [kA-1] ; // ### THIS IS WHERE VALGRIND HATES THE READ ###
        }
        else
        { 
            kA = tjfirst ;
            anz0 = A->p [tjfirst] ;
            jlast = tjfirst - 1 ;
        }
    

    And the contents of my A struct (not from the same run that valgrind was watching so ignore memory addresses).

    {magic = 32199698172899170, type = 0x44176ac0 <GB_opaque_GrB_FP64>, 
      type_size = 8, hyper_ratio = 0.0625, plen = 1, vlen = 100, vdim = 100, 
      nvec = 1, nvec_nonempty = 1, h = 0x2aaac4001b30, p = 0x2aaac4000fe0, 
      i = 0x2aaac4003180, x = 0x2aaac40041a0, nzmax = 1, hfirst = 0, 
      Pending = 0x0, nzombies = 0, AxB_method_used = GxB_DEFAULT, 
      queue_next = 0x0, queue_prev = 0x0, enqueued = false, p_shallow = false, 
      h_shallow = false, i_shallow = false, x_shallow = false, is_hyper = true, 
      is_csc = false, is_slice = false}
    
    opened by jbachan 6
  • Allow re-use of factorization on multiple threads

    Allow re-use of factorization on multiple threads

    I'm using CHOLMOD (via Haskell, but I don't think that's important here) and I'd like to analyze a matrix structure once and then re-factor on multiple threads. I tried this, using copy_factor to make a factor for each thread, and it failed with various seg-faults and other complaints from which I concluded/guessed that there must be some workspace in Common that is being re-used (?).

    So for now I've "solved" this by re-analyzing once per cpu and keeping a concurrent queue of common/factor pairs which get re-used by the various threads. But that is clearly a waste of work.

    Is there any way to analyze just once and then make the rest of this work?

    Thanks!

    opened by adamConnerSax 6
  • UMFPACK does not work for large matrices

    UMFPACK does not work for large matrices

    Hi all,

    I use SuiteSparse in my C# solver, I have already tested libraries CSParse, Cholmod, UMFPACK and all are running well on test examples. I have taken the dll’s at http://wo80.bplaced.net/math/packages.html The problem is UMFPACK, which works only for small and middle-size matrices, for a matrix with 1.38 million dofs I get:

    00:01:11.88, Error CSparse.Interop.SuiteSparse.Umfpack.UmfpackException: ERROR: out of memory at CSparse.Interop.SuiteSparse.Umfpack.UmfpackContext1.Factorize() in C:\D_A_T_A\PROJEKTY\A_femCalc\jobreader\CSparse.Interop\Interop\SuiteSparse\Umfpack\UmfpackContext.cs:line 60 at CSparse.Interop.SuiteSparse.Umfpack.UmfpackContext1.Solve(UmfpackSolve sys, T[] input, T[] result) in C:\D_A_T_A\PROJEKTY\A_femCalc\jobreader\CSparse.Interop\Interop\SuiteSparse\Umfpack\UmfpackContext.cs:line 100

    Other two solvers CSparse, Cholmod work well (I have 128GB RAM anyway). The UMFPACK manual says that:

    Insufficient memory to perform the symbolic analysis. If the analysis requires more than 2GB of memory and you are using the 32-bit ("int") version of UMFPACK, then you are guaranteed to run out of memory. Try using the 64-bit version of UMFPACK.

    which exactly looks like the problem, which I experience. So my question, where can I get 64bit (long) dll version of UMFPACK? I work in Windows 10 64bit.

    thank you very much

    Vita

    opened by vstembera 5
  • SuiteSparse compilation error

    SuiteSparse compilation error

    On a Rocky8 system, gcc 8.5.0 with LAPACK 3.10.1, the compile goes to 100% and then at the end gives:

    ranlib libsliplu.a SLIP_gmp.o: In function SLIP_mpfr_get_q': SLIP_gmp.c:(.text+0x28bc): undefined reference tompfr_get_q' collect2: error: ld returned 1 exit status

    opened by susanchacko 3
  • SPQR documentation: compatibility fix for TeX Live 2022

    SPQR documentation: compatibility fix for TeX Live 2022

    The \htmladdnormallink macro has been removed from hyperref as of TeX Live 2022. It was simply an alias for \href (with arguments swapped), so use the latter instead.

    opened by svillemot 0
  • One-line queries of the SuiteSparse matrix collection (discussion, not an issue)

    One-line queries of the SuiteSparse matrix collection (discussion, not an issue)

    Hello Prof. Davis,

    Fwiw, here's a one-page gist that uses python pandas for simple queries like

    '(posdef == 1)  &  (1000 <= rows <= 20000)'
    '(symm == 1)  &  (posdef == 0)' 
    

    It just queries to a subset csv -- it doesn't download like ssgetpy.

    Bytheway, there are quite a few duplicate id s in ssstats.csv ?

    (This is not really an issue -- move it to Discussions ?)

    cheers -- denis

    opened by denis-bz 0
  • Question about compressed-column representation

    Question about compressed-column representation

    I hope that this is not an inappropriate place to ask this question.

    When using the CXSparse library, and using a compressed-column representation, is it guaranteed that explicit (non-null) matrix elements are stored in order of increasing row index (for each column)?

    opened by szhorvat 1
  • Possible bug in the -lblas library on Ubuntu 20.04.

    Possible bug in the -lblas library on Ubuntu 20.04.

    Describe the bug Incorrect numerical results for CHOLMOD and SPQR on larger problems when using -lblas on Ubuntu 20.04.

    To Reproduce I have not yet replicated the bug.

    Expected behavior Low residuals.

    Screenshots

    Desktop (please complete the following information):

    • OS: Unbuntu 20.04.
    • compiler: any?
    • BLAS and LAPACK library, if applicable: -lblas

    Additional context Workaround: use the Intel MKL instead, which fixes this problem completely. I suspect a bug in the BLAS, or perhaps an interface issue (32 bit vs 64 bit integer parameters).

    bug 
    opened by DrTimothyAldenDavis 0
Releases(v5.12.0)
  • v5.12.0(Apr 10, 2022)

  • v5.11.0(Mar 15, 2022)

  • v5.10.1(May 24, 2021)

    May 17, 2021, SuiteSparse 5.10.1

    * CUDA: remove sm_30 from SuiteSparse_config.mk
    * GraphBLAS v5.0.5: minor bug fix
    * minor changes to Makefiles
    
    Source code(tar.gz)
    Source code(zip)
  • v5.10.0(May 17, 2021)

  • v5.9.0(Mar 3, 2021)

    GraphBLAS upgraded from v3.3 to v4.0.3 with many new features, and increased performance. Betweeness Centrality about 2x faster, and now faster than the GAP benchmark for larger matrices. BFS about 5x faster. For the very latest versions of GraphBLAS, see https://github.com/DrTimothyAldenDavis/GraphBLAS where stable releases are more frequent. Those updates are added to this SuiteSparse meta-package on a slower release cycle.

    Source code(tar.gz)
    Source code(zip)
  • v5.8.1(Jul 14, 2020)

  • v5.8.0(Jul 3, 2020)

    SuiteSparse 5.8.0, July 3, 2020:

    * SLIP_LU v1.0.1 added: for solving Ax=b exactly.  Requires
        the GNU GMP and MPRF libraries.
    * GraphBLAS v3.3.1: see the GraphBLAS/Doc/Changlog
    * replaced UFget with ssget: affects nearly all packages:
        UMFPACK, KLU, CHOLMOD, CXSparse/CSparse, etc,
        but their version numbers are left unchanged since it affects
        the MATLAB tests only, not the compiled libraries.
    * ssget v2.2.0: better URL redirects
    * updates to SuiteSparse build system
    
    Source code(tar.gz)
    Source code(zip)
  • v5.8.0-draft(Jul 1, 2020)

  • v5.7.2(Apr 8, 2020)

  • v5.7.1(Feb 20, 2020)

    Feb 20, 2020, SuiteSparse 5.7.1

    * SuiteSparse_config: update version number
    * Makefile: fixed install issue with README.txt
    

    Feb 20, 2020, SuiteSparse 5.7.0

    * GraphBLAS 3.2.0: better performance, new ANY and PAIR operators,
        structural mask, GrB_DESC_* from 1.3 C API Specification.
    * CHOLMOD 3.0.14: minor update to cholmod_check to print a matrix
    * added: CONTRIBUTIING.md, CODE_OF_CONDUCT.md, README.md.
    
    Source code(tar.gz)
    Source code(zip)
  • v5.7.0(Feb 20, 2020)

    Feb 20, 2020, SuiteSparse 5.7.0

    * GraphBLAS 3.2.0: better performance, new ANY and PAIR operators,
        structural mask, GrB_DESC_* from 1.3 C API Specification.
    * CHOLMOD 3.0.14: minor update to cholmod_check to print a matrix
    * added: CONTRIBUTIING.md, CODE_OF_CONDUCT.md, README.md.
    
    Source code(tar.gz)
    Source code(zip)
  • v5.6.0(Oct 21, 2019)

    SuiteSparse v5.6.0, Oct 21, 2019, with GraphBLAS v3.1.1.

    Release notes for GraphBLAS, since v2.3.5 in SuiteSparse v5.5.0:

    GraphBLAS Version 3.1.1, Oct 21, 2019

    * minor edits: user guide and comments in code
    

    GraphBLAS Version 3.1.0, Oct 2, 2019

    * added MATLAB interface: GraphBLAS/GraphBLAS is new.  In Source/, added
        global pointer to printf for MATLAB mexPrintf, pointer to
        mexMakeMemoryPersistent for Sauna workspace.  Changed how GraphBLAS
        objects are printed with GxB_print.  Changed how duplicate indices are
        handled in assign and extract, to match the MATLAB stadard.  Added
        helper functions for MATLAB (GB_matlab_helper.[ch]).
        Code size: @GrB is 9.7KLOC, test/ is 4.5KLOC).
    * bug counter added to this ChangeLog: to count # of bugs that appeared
        in formal releases that affect production code.  The count excludes
        bug fixes for test code, bugs in the demo codes, bugs introduced
        in beta versions that were fixed before any formal release, and
        bugs prior to version 1.0.  Code size of Source/, Include/ excluding
        Source/Generated, and Config/*.m4, is 42,659 lines (not including the
        new MATLAB interface).  13 bugs / 42K lines is a bug rate of 0.3 bugs
        per KLOC, much lower than most commercial software, but higher than
        UMFPACK, CHOLMOD, etc (with about 0.1 bug per KLOC).  GraphBLAS is a
        much more complex library, from the external view, than solving Ax=b.
        If UMFPACK has a bug, then Ax-b is typically large; there is no
        'residual' to check for GraphBLAS.  Also, in GraphBLAS, the test suite
        has about the same size as the main library (32K lines in Test/ and
        Tcov/).  For UMFPACK, etc, the test suite is always about 1/3 the size
        of the library itself.
    * 'make dox': for doxygen removed (not really that useful)
    * (13) bug fix to GB_reduce_to_vector: to avoid integer divide-by-zero for
        a matrix with n=0 columns.
    * (12) bug fix to GB_accum_mask: when C+=T if C has no entries except
        pending tuples
    * (11) bug fix to GB_resize: when pending tuples exist and vdim is growing
        from vdim <= 1 to vdim > 1, GB_WAIT(A) is required first.
    * (10) bug fix to GB_subref_phase1: "int nI" parameter should be int64_t.
    

    GraphBLAS Version 3.0.1, July 26, 2019

    * version number: Three changes to the user-visible API are not
        backward-compatible with V2.x: the added parameters to GxB_init and
        GxB_SelectOp_new, and the change in the type of the Thunk argument for
        GxB_select.  Thus, the SO version of SuiteSparse:GraphBLAS is now 3, no
        longer 2.  This change only affects SuiteSparse:GraphBLAS GxB_*
        extenstions, not any GrB_* functions or definitions.
    * added GxB_Scalar: acts like a GrB_Vector of length 1.
    * OpenMP parallelism: added nthreads and chunk parameters to GxB_set/get.
    * added parameter to GxB_init: bool user_malloc_is_thread_safe,
        for the MATLAB mexFunction interface, or any other malloc library that
        might not be thread-safe.  mxMalloc is not thread-safe.
        This change is not backward compatible with Version 2.x.
    * changed thunk parameter of GxB_select:  was (void *), now GxB_Scalar.
        This change is not backward compatible with Version 2.x.
    * added parameter to GxB_SelectOp_new: to specify the type of the Thunk.
        This change is not backward compatible with Version 2.x.
    * added options to GxB_get: determine if a matrix is hypersparse or not,
        global library, API information, nthreads, and chunk.
    * added options to GxB_set: nthreads, and chunk.
    * new operators and semirings: RDIV (f(x,y)=y/x) and RMINUS (y-x)
        binary operators.
    
    Source code(tar.gz)
    Source code(zip)
  • v5.5.0(Oct 21, 2019)

    SuiteSparse v5.5.0, Oct 20, 2019.

    This release includes v2.3.5 of GraphBLAS, for the Collected Algorithms of the ACM.

    Release notes for SuiteSparse v5.5.0:

    * GraphBLAS 2.3.5: Collected Algorithm of the ACM
    * UMFPACK 5.7.9: fix for compiling in MATLAB R2018b; BLAS library
    * SPQR, CHOLMOD: fix to *_make.m for compiling in MATLAB; same version
    * KLU: fix to Tcov/Makefile; no change to version number
    * CXSparse 3.2.0: version was incorrect in CXSparse/Include/cs.h;
        the corresponding CSparse v3.2.0 had the correct version information
        in its cs.h include file.
    * ssget and MATLAB_Tools/SuiteSparseCollection: update to sparse.tamu.edu
    * Mongoose 2.0.4: update to sparse.tamu.edu
    
    Source code(tar.gz)
    Source code(zip)
  • v5.4.0(Oct 19, 2019)

    SuiteSparse v5.4.0, Dec 28, 2018.

    * GraphBLAS 2.2.2: many upgrades and new features, a few bug fixes
    * CHOLMOD 3.0.13: fix to cholmod_core.h (for latest CUDA)
    * SPQR 2.0.9: fix to SuiteSparseQR.hpp (for latest CUDA)
    * UMFPACK 5.7.8: minor change to umf_analyze.h (not a bug, but the
        parameter names in the *.h did not match the *.c.
    * ssget: new matrices
    * Mongoose 2.0.3: simpler cmake
    * SuiteSparse_config: added JOBS option for parallel make, also added to
        GraphBLAS, CHOLMOD, SPQR, UMFPACK, Mongoose, and metis-5.1.0
    
    Source code(tar.gz)
    Source code(zip)
  • v5.3.0(Oct 19, 2019)

    SuiteSparse v5.3.0, July 5, 2018

    * GraphBLAS 2.0.3: bug fix to GxB_resize, better cmake script
    * new package: Mongoose (version 2.0.2)
    * fixed metis gk_arch.h for Windows
    * UMFPACK 5.7.7: modified comments in umfpack*symbolic.h
    * added contributor license for all of SuiteSparse
    * updated and renamed MATLAB_Tools/UFcollection to SuiteSparseCollection
    
    Source code(tar.gz)
    Source code(zip)
  • v5.2.0(Oct 19, 2019)

    SuiteSparse v5.2.0, Mar 15, 2018

    * GraphBLAS 2.0.1: bug fix to GxB_kron
    * SuiteSparse_config: corrected back to SO_VERSION 5
    
    * GraphBLAS 2.0.0: with changes to API to conform to the latest
        specification.  The SO_VERSION of GraphBLAS must change,
        as a result, since this affects both the ABI and API interface.
    * CHOLMOD 3.1.12: bug fix (no change to the CHOLMOD ABI or API)
    * KLU 1.3.9: minor edit, not a bug fix, but code is more clear now
    
    Source code(tar.gz)
    Source code(zip)
  • v5.1.2(Oct 19, 2019)

    SuiteSparse v5.1.2, Dec 28, 2017.

    * improved build process for GraphBLAS
    * minor change to CSparse/Lib/Makefile, no change in CSparse version
    

    5.1.1:

    * GraphBLAS added to top-level SuiteSparse/Makefile
    * GraphBLAS 1.1.1: bug fix to *assign, split AxB for faster compile,
        added memory usage statistics, AxB performance improvment
    * minor update to [AMD CAMD KLU]/Doc/Makefile's, no change to
        version numbers of AMD, CAMD, or KLU
    
    Source code(tar.gz)
    Source code(zip)
  • v5.1.0(Oct 19, 2019)

  • v5.0.0(Oct 19, 2019)

  • v4.5.6(Oct 19, 2019)

    SuiteSparse v4.5.6, Oct 3, 2017

    * changed COLAMD, CAMD, and CCOLAMD to BSD 3-clause,
        to match AMD.  No other changes; version numbers of
        packages unchanged.
    
    Source code(tar.gz)
    Source code(zip)
  • v4.5.5(Oct 19, 2019)

    SuiteSparse v4.5.5, Apr 17, 2017.

    * minor fix to SuiteSparse/Makefile for 'make install'
    

    4.5.4:

    * minor update to SPQR for ACM TOMS submission
    
    Source code(tar.gz)
    Source code(zip)
  • v4.5.3(Oct 19, 2019)

    SuiteSparse v4.5.3, May 4, 2016.

    * minor changes to Makefiles
    

    4.5.2:

    * licensing simplified (no other change); refer to PACKAGE/Doc/License.txt
        for the license for each package.
    

    4.5.1:

    * update to Makefiles.  Version 4.5.0 is broken on the Mac.
        That version also compiles *.so libraries on Linux with
        underlinked dependencies to other libraries in SuiteSparse.
        For example, AMD requires SuiteSparse_config.  The links to
        required libraries are now explicitly included in each library,
        in SuiteSparse 4.5.1.
    * minor change to CHOLMOD/Check/cholmod_write.c, when compiling with
        options that disable specific modules
    
    Source code(tar.gz)
    Source code(zip)
  • v4.5.0(Oct 19, 2019)

    SuiteSparse v4.5.0, Jan 30, 2016.

    * better Makefiles for creating and installing shared libraries
    * CHOLMOD now uses METIS 5.1.0, which is distributed with SuiteSparse
    * fix for MATLAB R2015b, which changed how it creates empty matrices,
        as compared to prior versions of MATLAB.  This change in MATLAB
        breaks many of the mexFunctions in prior versions of SuiteSparse.
        If you use MATLAB R2015b, you must upgrade to SuiteSparse 4.5.0
        or later.
    
    Source code(tar.gz)
    Source code(zip)
  • v4.4.7(Oct 19, 2019)

    SuiteSparse v4.4.7, Jan 1, 2016.

    * note that this minor update fails on the Mac, so its
        listed on my web page as a 'beta' release.
    * Improved the Makefiles of all packages.  They now create *.so
        shared libraries (*.dylib on the Mac).  Also, there is now
        only a single SuiteSparse_config.mk file.  It now determines
        your system automatically, and whether or not you have METIS
        and CUDA.  It also automatically detects if you have the Intel
        compiler or not, and uses it if it finds it.  There should be
        no need to edit this file for most cases, but you may need to
        for your particular system.  With this release, there are almost
        no changes to the source code, except for the VERSION numbers
        defined in the various include *.h files for each package.
    
    Source code(tar.gz)
    Source code(zip)
  • v4.4.6(Oct 19, 2019)

    SuiteSparse v4.4.6, Aug 2015.

    * SPQR: changed default tol when A has infs, from inf to realmax (~1e308)
    

    4.4.5:

    * CHOLMOD 3.0.6:
        - minor fix to CHOLMOD (disabling modules did not work as expected)
        - added MATLAB interface for row add/delete (lurowmod)
    * KLU 1.3.3: Fix for klu_dump.c (debugging case only)
    * UFcollection:  added additional stats for matrix collection
    * AMD: changed the default license.  See AMD/Doc/License.txt for details.
    
    Source code(tar.gz)
    Source code(zip)
  • v4.4.4(Oct 19, 2019)

    SuiteSparse v4.4.4, Mar 24, 2015.

    * CHOLMOD version number corrected.  In 4.4.3, the CHOLMOD_SUBSUB_VERSION
        string was left at '4' (it should have been '5', for CHOLMOD 3.0.5).
        This version of SuiteSparse corrects this glitch.
    * Minor changes to comments in SuiteSparse_config.
    * SPQR version 2.0.1 released (minor update to documentation)
    

    4.4.3:

    * CHOLMOD 3.0.5: minor bug fix to MatrixOps/cholmod_symmetry
    
    Source code(tar.gz)
    Source code(zip)
  • v4.4.2(Oct 19, 2019)

    SuiteSparse v4.4.2, Jan 7, 2015.

    * CHOLMOD 3.0.4: serious bug fix in supernodal factorization,
        introduced in CHOLMOD 3.0.0 (SuiteSparse 4.3.0).  Can cause segfault,
        and has no user workaround.
    
    Source code(tar.gz)
    Source code(zip)
  • v4.4.1(Oct 19, 2019)

    SuiteSparse v4.4.1, Oct 23, 2014.

    Minor update:  two bug fixes (affecting Windows only)
    
    * CHOLMOD 3.0.3:
        minor update to CHOLMOD (non-ANSI C usage in one *.c file, affects
        Windows only)
    * KLU 1.3.2:
        minor fix to MATLAB install; no change to C except version nubmer
    
    Source code(tar.gz)
    Source code(zip)
  • v4.4.0(Oct 19, 2019)

    SuiteSparse v4.4.0, Oct 10, 2014.

    MAJOR UPDATE:  new GPU-acceleration for SPQR
    
    * AMD 2.4.1:
        minor fix to MATLAB install; no change to C except version nubmer
    * BTF 1.2.1:
        minor fix to MATLAB install; no change to C except version nubmer
    * CAMD 2.4.1:
        minor fix to MATLAB install; no change to C except version nubmer
    * CCOLAMD 2.9.1:
        minor fix to MATLAB install; no change to C except version nubmer
    * CHOLMOD 3.0.2:
        update to accomodate GPU-accelerated SPQR
        added CHOLMOD/Include/cholmod_function.h
    * COLAMD 2.9.1:
        minor fix to MATLAB install; no change to C except version nubmer
    * CSparse 3.1.4:
        minor fix to MATLAB install; no change to C except version nubmer
    * CXSparse 3.1.4:
        minor fix to MATLAB install; no change to C except version nubmer
    * GPUQREngine 1.0.0:
        FIRST RELEASE.  Used by SPQR 2.0.0
    * KLU 1.3.1:
        minor fix to MATLAB install; no change to C except version nubmer
        update to KLU/Tcov/Makefile
    * LDL 2.2.1:
        minor fix to MATLAB install; no change to C except version nubmer
    * RBio 2.2.1:
        minor fix to MATLAB install; no change to C except version nubmer
    * SPQR 2.0.0:
        MAJOR UPDATE.  added GPU support.  Up to 11x faster than on CPU
    * SuiteSparse_GPURuntime 1.0.0:
        FIRST RELEASE.  Used by SPQR 2.0.0
    * UMFPACK 5.7.1:
        minor fix to MATLAB install; no change to C except version nubmer
    * MATLAB_Tools:
        modified SSMULT/ssmult_install.m.  No change to C code
    
    Source code(tar.gz)
    Source code(zip)
  • v4.3.1(Oct 19, 2019)

    SuiteSparse v4.3.1, July 18, 2014.

    Minor update:  added cholmod_rowfac_mask2 function to CHOLMOD
    
    * CHOLMOD 3.0.1:
        added cholmod_rowfac_mask2 function.  Minor fix to build process
    * SPQR 1.3.3:
        minor fix to build process
    
    Source code(tar.gz)
    Source code(zip)
Owner
Tim Davis
Faculty member at Texas A&M University, Department of Computer Science and Engineering. You can also reach me at [email protected]
Tim Davis
An FPGA accelerator for general-purpose Sparse-Matrix Dense-Matrix Multiplication (SpMM).

Sextans Sextans is an accelerator for general-purpose Sparse-Matrix Dense-Matrix Multiplication (SpMM). One exciting feature is that we only need to p

linghao.song 23 Jul 22, 2022
very simple matrix library, int addition to a recreation of DCM attitude estimation in the form of matrix of C.

matrix very simple matrix library, int addition to a recreation of DCM attitude estimation in the form of matrix of C. the lib include matrix inversio

null 4 Mar 21, 2022
matrix-effect This is a dumb matrix effect type thing

matrix-effect This is a dumb matrix effect type thing. It's only like one source file which should compile... Define __POSIX or __WIN though, for posi

null 41 Jun 8, 2022
This is official repository of the course Industrial Informatics LT, Year 2021/22, at University of Modena and Reggio Emilia, held at Fondazione Universitaria di Mantova

Industrial informatics LT - Mantova - 2021/22 This is official repository of the course Industrial Informatics LT, Year 2020/21, at University of Mode

High-Performance Real-Time Lab 4 Jun 27, 2022
The QPEP-Enhanced Direct Sparse Odometry (DSO) with Loop Closure

QPEP-DSO: Quadratic Pose Estimation Problems (QPEP) Enhanced Direct Sparse Odometry with Loop Closure Code Arch The codes are motivated by DSO (https:

Jin Wu 8 Jun 23, 2022
Fast, hierarchical, sparse Voxel Grid

Treexy is a library that implements a compact hierarchical data structure that can store and manipulate volumetric data, discretized on a three-dimens

Davide Faconti 271 Aug 6, 2022
Implementation of Monocular Direct Sparse Localization in a Prior 3D Surfel Map (DSL)

Implementation of Monocular Direct Sparse Localization in a Prior 3D Surfel Map (DSL)

Haoyang Ye 76 Jul 20, 2022
Fast, hierarchical, sparse Voxel Grid

Treexy is a library that implements a compact hierarchical data structure that can store and manipulate volumetric data, discretized on a three-dimens

Davide Faconti 274 Aug 7, 2022
OpenVDB - Sparse volume data structure and tools

OpenVDB AX Nano Houdini Website | Discussion Forum | Documentation OpenVDB is an open source C++ library comprising a novel hierarchical data structur

Academy Software Foundation 1.8k Aug 6, 2022
CPU Sparse Voxel Octree Implementation

Sparse Voxel Octrees This project provides a multithreaded, CPU Sparse Voxel Octree implementation in C++, capable of raytracing large datasets in rea

Benedikt Bitterli 609 Aug 6, 2022
the ares multi-system console emulation suite

ares ares is a multi-system emulator that began development on October 14th, 2004. It is a descendent of higan and bsnes, and focuses on accuracy and

higan 315 Aug 3, 2022
Test suite for evaluating C code on 42 Basecamp

Joãonette Test suite based on Moulinette, made by "Iniciativa João" ⛺️ Setup Clone this repo ( I suggest cloning to the root and calling it joaonette

Lucas Kuhn 13 Aug 2, 2022
Violent Fungus is a command and control (C2) software suite, providing red teams post-exploitation persistence and other juicy stuff.

Violent Fungus is a command and control (C2) software suite, providing red teams post-exploitation persistence and other juicy stuff.

Chris Humphries 33 Jul 3, 2022
the ares multi-system console emulation suite

ares is a multi-system emulator that began development on October 14th, 2004. It is a descendent of higan and bsnes, and focuses on accuracy and prese

ares 315 Jul 30, 2022
A combined suite of utilities for exporting images to retro formats.

ImageTools A combined suite of utilities for exporting images to retro formats. It was developed for use on Windows but might compile on other systems

David Walters 4 Dec 6, 2021
AWS Ambit Scenario Designer for Unreal Engine 4 (Ambit) is a suite of tools to streamline content creation at scale for autonomous vehicle and robotics simulation applications.

AWS Ambit Scenario Designer for Unreal Engine 4 Welcome to AWS Ambit Scenario Designer for Unreal Engine 4 (Ambit), a suite of tools to streamline 3D

AWS Samples 47 Jul 24, 2022
Suite of C++ libraries for radio astronomy data processing

Casacore A suite of C++ libraries for radio astronomy data processing. Installation Debian / Ubuntu Casacore is now part of Debian and Ubuntu, use apt

null 93 Jun 16, 2022
Blender is the free and open source 3D creation suite

Blender is the free and open source 3D creation suite. It supports the entirety of the 3D pipeline-modeling, rigging, animation, simulation, rendering, compositing, motion tracking and video editing.

Blender 6.3k Aug 11, 2022
The FLIP Fluids addon is a tool that helps you set up, run, and render high quality liquid fluid effects all within Blender, the free and open source 3D creation suite.

FLIP Fluids The FLIP Fluids addon is a tool that helps you set up, run, and render liquid simulation effects. Our custom built fluid engine is based a

Ryan Guy 1.3k Aug 4, 2022