mboct-numerical-pkg^©

mboct-numerical-pkg belongs to a suite of packages which can be used for pre- and postprocessing of MBDyn models (https://www.mbdyn.org) with GNU-Octave (http://www.gnu.org/software/octave/). This package contains interfaces to several well known numerical solvers.

List of features

• Solve large scale sparse symmetric linear systems of equations using PaStiX (https://gitlab.inria.fr/solverstack/pastix), MUMPS (https://github.com/group-gu/mumps) or UMFPACK (http://faculty.cse.tamu.edu/davis/suitesparse.html).
• Solve large scale generalized symmetric eigenvalue problems with ARPACK (https://www.caam.rice.edu/software/ARPACK/).
• Compute fill in reducing node orderings with METIS (http://glaros.dtc.umn.edu/gkhome/metis/metis/overview).
• Compute sparse matrix vector products for symmetric sparse matrices where only the upper or lower triangular part of the matrix is stored.
• Compute selected eigenvalues, and optionally, eigenvectors of a real generalized symmetric-definite banded eigenproblem using LAPACK (http://www.netlib.org/lapack/).
• Computes all eigenvalues and, optionally, eigenvectors of a real symmetric matrix in packed storage using LAPACK.

Copyright^© 2019-2021

Reinhard

Installation

The following code is an example how mboct-numerical-pkg can be installed on an Ubuntu system:

sudo apt-get install octave liboctave-dev libsuitesparse-dev libarpack2-dev libmumps-seq-dev libmetis-dev libmkl-full-dev

git clone -b master https://github.com/octave-user/mboct-numerical-pkg.git

make -C mboct-numerical-pkg install_local

PaStiX installation

• Follow the instructions on (https://gitlab.inria.fr/solverstack/pastix) to install PaStiX (optional but recommended).

MUMPS installation

• Follow the instructions on (https://github.com/group-gu/mumps) to install MUMPS (optional).

GNU Octave installation

• Follow the instructions on (http://www.gnu.org/software/octave/) to install GNU Octave.
• Make sure, that mkoctfile is installed.
  mkoctfile --version

GNU Octave package installation:

• Install the following packages from github.
  for pkg in numerical; do
git clone https://github.com/octave-user/mboct-${pkg}-pkg.git && make -C mboct-${pkg}-pkg install_local
done

Usage

• Run Octave.
  octave
• At the Octave prompt load the package.
  pkg load mboct-numerical-pkg
• At the Octave prompt execute the tests.
  test numerical_tests