|
Class Summary |
| LDADocumentTopicMapper |
|
| LDADriver |
Estimates an LDA model from a corpus of documents, which are SparseVectors of word counts. |
| LDAInference |
Class for performing infererence on a document, which involves computing (an approximation to)
p(word|topic) for each word and topic, and a prior distribution p(topic) for each topic. |
| LDAInference.InferredDocument |
An estimate of the probabilities for each document. |
| LDAReducer |
A very simple reducer which simply logSums the input doubles and outputs a new double for sufficient
statistics, and sums log likelihoods. |
| LDASampler |
Takes in a Matrix of topic distributions (such as generated by LDADriver,
CVB0Driver or
InMemoryCollapsedVariationalBayes0, and constructs
a set of samplers over this distribution, which may be sampled from by providing a distribution
over topics, and a number of samples desired |
| LDAState |
|
| LDAWordTopicMapper |
Runs inference on the input documents (which are sparse vectors of word counts) and outputs the sufficient
statistics for the word-topic assignments. |