lda_fgs: Full Gibbs sampler (FGS) for LDA

In clintpgeorge/ldamcmc: Markov chain Monte Carlo Algorithms for the Latent Dirichlet Allocation Model

Description

This a R wrapper function for the C++ implementation of the full Gibbs sampler for LDA—a Markov chain on (z, β, θ).

Usage

lda_fgs(K, V, wid, doc.N, alpha.v, eta, max.iter = 100, burn.in = 0,
  spacing = 1, save.z = 0, save.beta = 0, save.theta = 0, save.lp = 0)

Arguments

K	Number of topics in the corpus
V	Vocabulary size
wid	Vocabulary ids of every word instance in each corpus document (1 X total.N vector). We assume vocabulary id starts with 1
doc.N	Documents' word counts
alpha.v	Hyperparameter vector for θ
eta	Smoothing parameter for the β matrix
max.iter	Maximum number of Gibbs iterations to be performed
burn.in	Burn-in-period for the Gibbs sampler
spacing	Spacing between the stored samples (to reduce correlation)
save.z	if 0 the function does not save z samples
save.beta	if 0 the function does not save β samples
save.theta	if 0 the function does not save θ samples
save.lp	if 0 the function does not save computed log posterior for iterations

Value

the Gibbs sampling output

See Also

Other Gibbs sampling methods: get_grid_neighbors, lda_acgs_hs, lda_acgs, lda_fgs_hs, lda_fgs_st

clintpgeorge/ldamcmc documentation built on Feb. 22, 2020, 12:39 p.m.