Implementation of algorithm for testing models of hybrid incompatibility
accumulation on phylogenetic trees. Applicable if you want to test models on a
three species complex and you have branch lengths and counts of the number of
shared/unique incompatibilities (in at least two species pairs).

This implementation was originally written in Python 2.7 and requires NumPy
(http://www.numpy.org/) for scientific computing and SciPy
(http://www.scipy.org/) for its implementation of the Broyden-Fletcher-
Goldfarb-Shanno (BFGS) algorithm.

<evcEqsEPL.py> contains definitions for expectations, variances, covariances as
well as likelihood functions for different models and wrapper functions for
missing data.

<maxlikeEPL.py> contains functions for maximizing the likelihood functions of
different models given observations of incompatibilities between species and
relative divergence times.

<usage.txt> contains the details on how to use this module.