⚠️There's a newer version (1.0.2) of this package.Take me there.

COMIX (version 1.0.1)

Coarsened Mixtures of Hierarchical Skew Kernels

Description

Bayesian fit of a Dirichlet Process Mixture with hierarchical multivariate skew normal kernels and coarsened posteriors. For more information, see Gorsky, Chan and Ma (2024) .

Copy Link

Version

Install

install.packages('COMIX')

Monthly Downloads

251

Version

1.0.1

License

CC0

Maintainer

S. Gorsky

Last Published

November 12th, 2025

Functions in COMIX (1.0.1)

tidyChain

This function creates tidy versions of the stored chain. This object can then be used as input for the other diagnostic functions in this package.

summarizeChain

This function provides post-hoc estimates of the model parameters.

relabelChain

This function relabels the chain to avoid label switching issues.

effectiveSampleSize

This function creates an object that summarizes the effective sample size for the parameters of the model.

calibrateNoDist

This function aligns multiple samples so that their location parameters are equal.

plotHeidelParams

This function creates plots for the Heidelberg-Welch diagnostic and results of test of stationarity for the parameters of the model.

plotGewekeParams

This function creates plots for the Geweke diagnostic and results of test of stationarity for the parameters of the model.

heidelParams

This function creates an object that summarizes the Heidelberg-Welch convergence diagnostic.

plotEffectiveSampleSize

This function creates plots for the effective sample size for the parameters of the model.

gewekeParams

This function creates an object that summarizes the Geweke convergence diagnostic.

comix

This function generates a sample from the posterior of COMIX.

transform_params

Convert between parameterizations of the multivariate skew normal distribution.

plotTracePlots

This function creates trace plots for different parameters of the MCMC chain.

calibrate

This function aligns multiple samples so that their location parameters are equal.

acfParams

The function computes (and by default plots) estimates of the autocovariance or autocorrelation function for the different parameters of the model. This is a wrapper for coda::acf.