Extract meaningful information from the von Mises-Fisher mixture model

Source: R/movMF_info.R

Given a mixture of von Mises-Fisher distributions, this function computes several related quantities of the data on the unit hypersphere with respect to the specified model.

movMF_info(data, means, concentrations, weights)

Arguments

data

an \((n\times d)\) data matrix.

means

an \((k\times d)\) matrix of means.

concentrations

a vector of length \(k\) of concentration parameters.

weights

a vector of length \(k\) of mixing weights.

Value

a named list containing

densities

a vector of length \(n\) of the densities of the data points.

clustering

a vector of length \(n\) of the hard clustering results.

loglkd

the log-likelihood of the data.

AIC

the Akaike information criterion.

BIC

the Bayesian information criterion.