Abstract

This document describes petsc4py, a Python port to the PETSc libraries.

PETSc (the Portable, Extensible Toolkit for Scientific Computation) is a suite of data structures and routines for the scalable (parallel) solution of scientific applications modeled by partial differential equations. It employs the MPI standard for all message-passing communication.

This package provides an important subset of PETSc functionalities and uses NumPy to efficiently manage input and output of array data.

A good friend of petsc4py is:

Other projects depends on petsc4py:

Overview

PETSc is a suite of data structures and routines for the scalable (parallel) solution of scientific applications modeled by partial differential equations. It employs the MPI standard for all message-passing communication.

PETSc is intended for use in large-scale application projects [petsc-efficient], and several ongoing computational science projects are built around the PETSc libraries. With strict attention to component interoperability, PETSc facilitates the integration of independently developed application modules, which often most naturally employ different coding styles and data structures.

PETSc is easy to use for beginners [petsc-user-ref]. Moreover, its careful design allows advanced users to have detailed control over the solution process. PETSc includes an expanding suite of parallel linear and nonlinear equation solvers that are easily used in application codes written in C, C++, and Fortran. PETSc provides many of the mechanisms needed within parallel application codes, such as simple parallel matrix and vector assembly routines that allow the overlap of communication and computation.

[petsc-user-ref]

S. Balay, S. Abhyankar, M. Adams, J. Brown, P. Brune, K. Buschelman, L. Dalcin, V. Eijkhout, W. Gropp, D. Karpeyev, D. Kaushik, M. Knepley, D. May, L. Curfman McInnes, R. Mills, T. Munson, K. Rupp, P. Sanan, B. Smith, S. Zampini, H. Zhang, and H. Zhang, PETSc Users Manual, ANL-95/11 - Revision 3.10, 2018. http://www.mcs.anl.gov/petsc/petsc-current/docs/manual.pdf

[petsc-efficient]

Satish Balay, Victor Eijkhout, William D. Gropp, Lois Curfman McInnes and Barry F. Smith. Efficient Management of Parallelism in Object Oriented Numerical Software Libraries. Modern Software Tools in Scientific Computing. E. Arge, A. M. Bruaset and H. P. Langtangen, editors. 163–202. Birkhauser Press. 1997.

Components

PETSc is designed with an object-oriented style. Almost all user-visible types are abstract interfaces with implementations that may be chosen at runtime. Those objects are managed through handles to opaque data structures which are created, accessed and destroyed by calling appropriate library routines.

PETSc consists of a variety of components. Each component manipulates a particular family of objects and the operations one would like to perform on these objects. These components provide the functionality required for many parallel solutions of PDEs.

Installation

Using pip or easy_install

You can use pip to install petsc4py and its dependencies (mpi4py is optional but highly recommended):

$ pip install [--user] numpy mpi4py
$ pip install [--user] petsc petsc4py

Alternatively, you can use easy_install (deprecated):

$ easy_install petsc4py

If you already have a working PETSc installation, set environment variables PETSC_DIR and PETSC_ARCH to appropriate values and next use pip:

$ export PETSC_DIR=/path/to/petsc
$ export PETSC_ARCH=arch-linux2-c-opt
$ pip install petsc4py

Using distutils

Requirements

You need to have the following software properly installed in order to build PETSc for Python:

[1]

Unless you have appropiatelly configured and built PETSc without MPI (configure option --with-mpi=0).

[2]

You may need to use a parallelized version of the Python interpreter with some MPI-1 implementations (e.g. MPICH1).

Downloading

The PETSc for Python package is available for download at the project website generously hosted by Bitbucket. You can use curl or wget to get a release tarball.

$ curl -O https://bitbucket.org/petsc/petsc4py/petsc4py-X.Y.Z.tar.gz

$ wget https://bitbucket.org/petsc/petsc4py/petsc4py-X.Y.Z.tar.gz

Building

After unpacking the release tarball:

$ tar -zxf petsc4py-X.Y.Z.tar.gz
$ cd petsc4py-X.Y.Z

the distribution is ready for building.

NOTE:

$ export MACOSX_DEPLOYMENT_TARGET=10.6
$ export SDKROOT=/
$ export ARCHFLAGS='-arch x86_64'

Some environment configuration is needed to inform the location of PETSc. You can set (using setenv, export or what applies to you shell or system) the environment variables PETSC_DIR, and PETSC_ARCH indicating where you have built/installed PETSc:

$ export PETSC_DIR=/usr/local/petsc
$ export PETSC_ARCH=arch-linux2-c-opt

Alternatively, you can edit the file setup.cfg and provide the required information below the [config] section:

[config]
petsc_dir  = /usr/local/petsc
petsc_arch = arch-linux2-c-opt
...

Finally, you can build the distribution by typing:

$ python setup.py build

Installing

After building, the distribution is ready for installation.

If you have root privileges (either by log-in as the root user of by using sudo) and you want to install PETSc for Python in your system for all users, just do:

$ python setup.py install

The previous steps will install the petsc4py package at standard location prefix/lib/pythonX.Y/site-packages.

If you do not have root privileges or you want to install PETSc for Python for your private use, just do:

$ python setup.py install --user

Tutorial

XXX To be written … Any contribution welcome!

Citations

If PETSc for Python been significant to a project that leads to an academic publication, please acknowledge that fact by citing the project.