relabel_kl_log: Relabel membership vector by minimizing KL-algorithm...
In baruuum/relabelKL: Relabeling MCMC Output

relabel_kl_log

R Documentation

Relabel membership vector by minimizing KL-algorithm (log-scale)

Description

Relabels the membership vectors of a mixed membership model or mixed membership stochastic blockmodel, using the KL-algorithm proposed by Stephens (2000).

Usage

relabel_kl_log(
  lphi,
  renormalize = FALSE,
  maxit = 100L,
  nthreads = 0L,
  verbose = TRUE
)

Arguments

`lphi`	cube of length `S`, each element of which is a matrix of dimension N times K, where N is the number of individuals and K is the number of extreme types (or classes).
`renormalize`	if `true`, renormalizes the rows of each slice of `lphi`
`maxit`	the number of maximum iterations to run, defaults to 100.
`nthreads`	number of threads to use in parallel processing
`verbose`	if TRUE, number of iterations and corresponding KL-distance values are printed.

Details

OpenMP is enabled if available. Due to overhead, the inner-most loop is parallelized only if the number of latent classes/types is larger than 3.

Value

The function returns a Rcpp::List of three elements. 1) A cube, permuted, of the same dimensions as phi but with the labels permuted; 2) perms is a S times K matrix containing the permutations necessary to produce permuted from phi (i.e., the mapping from phi to permuted); and 3) the number of iterations run.