murasyp is Python software for accept/reject statement-based uncertainty modeling.

Or at least, that is what it is going to be. Currently, murasyp consists of code for dealing with simpler imprecise probabilistic models: (closed convex) credal sets and sets of (really) desirable gambles, which can also used as (conditional) lower previsions. It will be extended to support my (Erik Quaeghebeur's) research. It is publically available in case somebody else wishes to test it out; feel free to contact me via my GitHub page or using the contact details on my personal website.

Murasyp can be installed from source. Install Python, download the source with Git by running:

git clone git://github.com/equaeghe/murasyp

and run python setup.py install in the directory in which the repository is cloned. Examples of how to use murasyp can be found in this documentation. You can also browse the source code on GitHub: equaeghe/murasyp. Note that you need Sphinx to generate the documentation and to run the doctests.

Inspiration for starting this project came from Matthias Troffaes's project improb, which is a currently much more developed package for working with imprecise probabilistic models, but with different goals and design choices. I am also indebted to Matthias Troffaes for getting me started with using GitHub and Sphinx.