pyParticleEst - A Python based Particle Estimation framework
Overview
pyParticleEst is a python framework to assist in implementing (Rao-Blackwellized) particle filtering and/or particle smoothing algorithms to solve a user definedestimation problem.
A poster giving an overview of the goal and capabilities of the framework. (from Reglermöte 2014)
A short presentation giving a quick introduction to particle smoothing and filtering (the example code in the presentation was written against an older version of the framework and the API has been improved streamlined since then)
Documentation
The official documentaion can be found on the following link
http://pyparticleest.readthedocs.org
Source code
The project is hosted on github as pyParticleEst and available under the LGPL license.
Examples
In the source code there is a test/ directory which contains a number of examples which have been used to verify the implementation, they can provide a starting point for working with framework.
The file jss_examples.tar.gz contains the example code which reproduces the result for an article submitted to http://www.jstatsoft.org/
Download/Installation
Using PyPI/pip
Many system (such as Ubuntu) can install directy from PyPI using the pip package manager, to do so simply type
pip install pyParticleEst
This will download and install the framework (and any unsatisfied dependencies)
Manual installation
You can download the latest code from github: pyParticleEst
It uses NumPy and SciPy so those packages must be installed on your system.
After downloading the source code you should be able to install it by simply running
python setup.py install
Contact
If you find any problems/bugs feel free to contact me, or ideally make a "pull request" for the github project containing a fix.
Feedback on problems solved using the framework is also welcome, and if used in academic research a reference to the above mentioned technical report is welcomed.