Prefpy

This release represents a minor increment to correct package setup and structure to enable proper publication for the seminal release of PrefPy on PyPI. This commit revision is released on PyPI at https://pypi.python.org/pypi/prefpy. The PrefPy package is now available publically for installation alongside Python 3.4+ with the following command:


pip3 install prefpy


Future development will finalize, consolidate, and generalize the GMM algorithm implementation for any number of alternatives and any moment conditions.

This release represents the consolidation and revision of the EMM algorithm implementation including three different types of configurable iterations dictionary arguments setup: fixed number of iterations for each E-step and M-step, scaling M-step iterations, and total number of MM iterations for given number of E-step iterations.

Future development will finalize, consolidate, and generalize the GMM algorithm implementation for any number of alternatives and any moment conditions.

This release represents the final implementations of the GMM and EMM algorithms as they were to generate the experimental results presented in the ICML '16 camera ready version of the "Learning Mixtures of Plackett-Luce Models" paper.