Skip to content

videlec/pplpy

 
 

Repository files navigation

PPL Python wrapper

This Python package provides a wrapper to the C++ Parma Polyhedra Library (PPL).

The whole package started as a fork of a tiny part of the Sage software.

How it works

The names of objects and methods are the same as in the library:

>>> import ppl
>>> x = ppl.Variable(0)
>>> y = ppl.Variable(1)
>>> z = ppl.Variable(2)
>>> cs = ppl.Constraint_System()
>>> cs.insert(x >= 0)
>>> cs.insert(y >= 0)
>>> cs.insert(z >= 0)
>>> cs.insert(x + y + z == 1)
>>> poly = ppl.C_Polyhedron(cs)
>>> poly.minimized_generators()
Generator_System {point(1/1, 0/1, 0/1), point(0/1, 1/1, 0/1), point(0/1, 0/1, 1/1)}

The available objects and functions from the ppl Python module are:

  • Variable, Variables_Set, Linear_Expression (defined in ppl.linear_algebra)
  • MIP_Problem (defined in ppl.mip_problem)
  • C_Polyhedron, NNC_Polyhedron (defined in ppl.polyhedron)
  • Generator, Generator_System, Poly_Gen_Relation, point, closure_point, ray, line (defined in ppl.generator)
  • Constraint, Constraint_System, Poly_Con_Relation, inequality, equation, strict_inequality (defined in ppl.constraint)

Installation

The project is available at Python Package Index and can be installed with pip:

$ pip install pplpy

Note that if you have gmp and ppl installed in a non standard directory (e.g. you use brew on MacOSX) then you need to set appropriately the variables CFLAGS before calling pip. For example:

$ export CFLAGS="-I/path/to/gmp/include/ -L/path/to/gmp/lib/ -I/path/to/ppl/include/ -L/path/to/ppl/lib $CFLAGS"
$ pip install pplpy

Using from Cython

All Python classes from pplpy are extension types and can be used with Cython. Each extension type carries an attribute thisptr that holds a pointer to the corresponding C++ object from ppl.

A complete example is provided with the files tests/testpplpy.pyx and tests/setup.py.

Source

You can find the latest version of the source code on gitlab: https://gitlab.com/videlec/pplpy

Documentation

An online version of the documentation is available at http://www.labri.fr/perso/vdelecro/pplpy/latest/

Compiling the html documentation requires make and sphinx. Before building the documentation, you need to install the pplpy package (sphinx uses Python introspection). The documentation source code is contained in the repository docs where there is a standard Makefile with a target html. Running make html in the docs repository builds the documentation inside docs/build/html. For more configuration options, run make help.

License

pplpy is distributed under the terms of the GNU General Public License (GPL) published by the Free Software Foundation; either version 3 of the License, or (at your option) any later version. See http://www.gnu.org/licenses/.

Requirements

On Debian/Ubuntu systems these can be installed with:

$ sudo apt-get install libgmp-dev libmpfr-dev libmpc-dev libppl-dev cython
$ pip install cysignals --user
$ pip install gmpy2 --pre --user

The pip optional option --user allows to install python packages for a single user with no need for administrator rights. The two pip install commands might be replaced by sudo pip install PKG (not recommended). On recent Debian/Ubuntu systems, cysignals is also available as a package under the name python-cysignals for Python 2 and python3-cysignals for Python 3.

About

Python wrapper to the PPL library

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Cython 95.4%
  • Python 4.3%
  • C++ 0.3%