lda

Type Members

1. class CVB0LDA extends LDA[SoftAssignmentModelState, CVB0LDADocument, (String, Array[Array[Double]])] with SoftAssignmentModel[LDAModelParams, LDADocumentParams, CVB0LDADocument]

   CVB0 learning and inference model for vanilla LDA.

2. class CVB0LDADocument extends LDADocument[(String, Array[Array[Double]])] with SoftAssignmentDocument

   CVB0LDA document.

3. class GibbsLDA extends LDA[HardAssignmentModelState, GibbsLDADocument, (String, Array[Int])] with HardAssignmentModel[LDAModelParams, LDADocumentParams, GibbsLDADocument]

   Collapsed Gibbs sampler for LDA learning and inference.

4. class GibbsLDADocument extends LDADocument[(String, Array[Int])] with HardAssignmentDocument

   Holds a bag of observations and an assignment for each observation.

5. trait LDA [ModelState, Doc <: LDADocument[DocState], DocState] extends TopicModel[LDAModelParams, ModelState, LDADocumentParams, Doc, DocState] with ClosedTopicSet with DirichletTermSmoothing with DirichletTopicSmoothing

   LDA models are a TopicModel on a fixed set of k topics with dirichlet term and topic smoothing.

6. trait LDACompanion [Model <: LDA[ModelState, Doc, DocState], ModelState, Doc <: LDADocument[DocState], DocState] extends TopicModelCompanion[LDAModelParams, Model, ModelState, LDADocumentParams, Doc, DocState] with LocalUpdateModelCompanion[LDAModelParams, Model, ModelState, LDADocumentParams, Doc, DocState] with LogProbabilityEstimateModelCompanion[Model, Doc]

   LDA model companion - LDA models are all local-update models.

7. trait LDADataset [Item] extends LazyIterable[LDADocumentParams] with Dataset[Item, String, LDADataset[Item]] with TermDataset[Item]

   An LDA dataset is a collection of items; each item has an ID (retrieved with the getID function) and a sequence of terms (retrieved with the getTerms function).

8. trait LDADocument [DocState] extends Stateful[DocState]

   LDA documents contain a set of word observations and a distribution over topics.

9. case class LDADocumentParams (id: String, terms: Array[Int]) extends Product with Serializable

   Parameters describing a single LDA document

10. case class LDAModelParams (numTopics: Int, numTerms: Int, topicSmoothing: DirichletParams, termSmoothing: DirichletParams, termIndex: Option[Index[String]], tokenizer: Option[Tokenizer]) extends Product with Serializable

    Parameters describing LDA models.