lda_fgs_ppc: LDA: Full Gibbs Sampler with Posterior Predictive Value

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

Description

Implements the Full Gibbs sampler for the LDA model—a Markov chain on (β, θ, z). The log posterior predictive value is based on Zhe Chen (2015)

Usage

lda_fgs_ppc(num_topics, vocab_size, docs_tf, alpha_h, eta_h, max_iter, burn_in,
  spacing, verbose)

Arguments

num_topics	Number of topics in the corpus
vocab_size	Vocabulary size
docs_tf	A list of corpus documents read from the Blei corpus using read_docs (term indices starts with 0)
alpha_h	Hyperparameter for θ sampling
eta_h	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)
verbose	from 0, 1, 2

Value

The Markov chain output as a list of

lppv	log posterior predictive values of each document
lppc	averge of log posterior predictive values

See Also

Other MCMC: lda_acgs_st, lda_cgs_em_perplexity, lda_cgs_em, lda_cgs_perplexity, lda_fgs_BF_perplexity, lda_fgs_perplexity, lda_fgs_st_perplexity

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