met 0.1.5.0

Multinomial Exact Tests

met.py is a Python module that allows you to define a pair of multinomial distributions (conceptually ‘control’ and ‘test’ or ‘reference’ and ‘site’ distributions) and then compute one- and two-sided p values to test whether the ‘site’ distribution is equivalent to the ‘reference’ distribution. The likelihood of all possible ‘reference’ distributions can be evaluated and the distribution of p values can be expressed in terms of the likelihood of the observed ‘reference’ distribution.

The met module defines one class, one exception, and several functions. The class (Multinom) defines objects that represent specific combinations of multinomial data for site and reference conditions. Methods of Multinom objects allow one-sided and two-sided exact tests to be performed. Attributes of Multinom objects allow access to additional information generated by the exact tests, such as the number of different rearrangements of site data that were found to be more extreme than the reference data.

Functions in the met module allow p values to be calculated for a variety of reference probability distributions that might have produced the observed reference data, producing a likelihood estimate for each p value.