gen_synth_clda_corpus: Generates a Synthetic c-LDA Corpus

In clintpgeorge/clda: Approximate Inference Algorithms for the Compound Latent Dirichlet Allocation Model

Description

Generates documents using the c-LDA generative process based on a set of predefined values.

Usage

gen_synth_clda_corpus(K, V, J, collection.size, doc.size, alpha.h, gamma.h,
  eta.h)

Arguments

K	number of topics
V	vocabulary size
J	number of collections
collection.size	number of documents in each collection (a list of J elements)
doc.size	number of words in each document
alpha.h	hyperparameter for collection-level Dirichlet sampling
gamma.h	hyperparameter for document-level Dirichlet sampling
eta.h	hyperparameter for topic Diriclet sampling

Details

Last modified on: March 04, 2016

Value

a list of generated corpus and their statistics

See Also

Other corpus: gen_synth_clda_corpus_pi

Examples

## Generates documents with given parameters

J                  <- 2
K                  <- 4
V                  <- 20
alpha.h            <- 2
gamma.h            <- .2
eta.h              <- .25
doc.size           <- 80
collection.size    <- c(40, 40) # number of documents in each collection
ds.name            <- paste("synth-J", J, "-K", K, "-D", D, "-V", V, sep = "")

ds                 <- gen_synth_clda_corpus(K, V, J, collection.size, doc.size, alpha.h, gamma.h, eta.h)

clintpgeorge/clda documentation built on May 13, 2019, 8 p.m.