mallet_model: The model object
In agoldst/dfrtopics: Tools for exploring topic models of text

mallet_model

R Documentation

The model object

Description

A topic model is a complicated beastie, and exploring it requires keeping track of a number of different kinds of data. The mallet_model object strives to encapsulate some of this for you. Package users shouldn't need to invoke the constructor explicitly; to obtain model objects, use either train_model or load_mallet_model. The summary method indicates which elements of the model have been loaded into R's memory.

Usage

mallet_model(
  doc_topics = NULL,
  doc_ids = NULL,
  vocab = NULL,
  top_words = NULL,
  topic_words = NULL,
  params = NULL,
  hyper = NULL,
  metadata = NULL,
  model = NULL,
  ss = NULL,
  instances = NULL
)

## S3 method for class 'mallet_model'
print(x)

## S3 method for class 'mallet_model'
summary(x)

## S3 method for class 'mallet_model_summary'
print(x)

Arguments

`doc_topics`	document-topic matrix
`doc_ids`	vector of document ids corresponding to rows of `doc_topics`
`vocab`	vector of word types corresponding to columns of `topic_words`
`top_words`	data frame with top-ranked words and their weights for each topic
`topic_words`	topic-word matrix (expected to be sparse)
`params`	list of modeling parameters
`hyper`	list of estimated hyperparameters α and β
`metadata`	data frame of metadata
`model`	reference to an RTopicModel object from MALLET
`ss`	final Gibbs sampling state represented as "simplified" `big.matrix`
`instances`	reference to InstanceList (redundant if `model` specified)