cabo40 / pyrichlet

Data analysis library for Python using "Dirichlet Process"-based models.

Project description

Pyrichlet is a package for doing density estimation and clustering using Gaussian mixtures with BNP weighting models

Installation

With pip:

pip install pyrichlet

For a specific version:

pip install pyrichlet==0.0.4

Usage

This is a quick guide. For a more detailed usage see https://pyrichlet.readthedocs.io/en/latest/index.html.

The mixture models that this package implements are

• DirichletDistributionMixture
• DirichletProcessMixture
• PitmanYorMixture
• GeometricProcessMixture
• BetaInBetaMixture
• BetaInDirichletMixture
• BetaBernoulliMixture
• BetaBinomialMixture

They can be fitted for an array or dataframe using a Gibbs sampler or variational Bayes methods,

from pyrichlet import mixture_models

mm = mixture_models.DirichletProcessMixture()
y = [1, 2, 3, 4]
mm.fit_gibbs(y)

mm.fit_variational(y, n_groups=2)

and use the fitted class to do density estimation

x = 2.5
f_x = mm.gibbs_eap_density(x)
f_x = mm.var_eap_density(x)

or clustering

mm.var_map_cluster()
mm.gibbs_map_cluster()
mm.gibbs_eap_spectral_consensus_cluster()