The following installation guidelines are mainly for experienced users working on a Linux environment. It is highly recommended to use docker or conda, for quick installation and for package and dependency updates. You can find simple installation instructions for gospl in this document.

Installation via PyPI

gospl can be installed via pip from PyPI.

pip install gospl

The installation however requires additional dependencies and is likely to failed if some initial packages are not already available. When using pip, it is recommended to use virtualenv. For a guide on how to use it, readers are invited to look at visit the Python documentation.

Installing dependencies

On linux, you will need to install MPI, PETSc and Hdf5 via the your Linux distribution’s package manager.

For example, the commands in this table will install MPICH from your distribution.

Distribution

Status

Download / Repository Link

Install method

Debian

stable

official Debian repository

sudo apt-get install mpich

Debian & Ubuntu

unstable (latest packages)

NeuroDebian

sudo apt-get install mpich

Ubuntu

stable

official Ubuntu repository

sudo apt-get install mpich

OpenSuse

stable

OpenSuse Repository

zypper in mpich

Fedora

stable

official Fedora repository

dnf install mpich

Centos/RHEL

stable

EPEL repository

yum install mpich

Using the apt-get command. This will likely be done by:

sudo apt-get update
sudo apt-get install mpich libmpich-dev libhdf5-mpich-dev
sudo apt-get install libblas-dev liblapack-dev
sudo apt-get install libscotchparmetis-dev libmetis-dev

If you have PETSc already installed, you can point to your PETSC_DIR and PETSC_ARCH or you can unset these paths before installing the petsc4py module via pip:

unset PETSC_DIR
unset PETSC_ARCH

Installing PETSc and gospl

Then the remaining library are installed locally by running:

pip install --upgrade pip setuptools wheel
pip install petsc --user
pip install petsc4py --user
pip install gospl --user