úÎ+N%ä&      !"#$%None&'()*+,-./0123456789&()*+,-&'()*+,-None:featureSeq f z/ extracts features at each position of z G following current position, using f to combine features into bigrams. ; features f z1 extracts features at the current position of z, + using f to combine features into bigrams. :;<=:;:;<=NoneSentence is a vector of tokens. FieldA is a part of a word token, such as word form, lemma or POS tag TokenF is a representation of a word, which consists of a number of fields.  parse text# returns a lazy list of sentences.     None #Container for the Word Class model > LDA model learn options xs- runs the LDA Gibbs sampler for word classes  with options on sentences xs", and returns the resulting model - together with progressive class assignments ?AConvert a stream of sentences into a stream of batches ready for  sampling. @!Extract features from a sentence A)Convert text strings into symbols (ints) ! summary m4 returns a textual summary of word classes found in  model m BinterpWordClasses m lambda doc# gives the class probabilities for  word type in context doc according to evolving model m. It 7 interpolates the prior word type probability with the 1 context-conditioned probabilities using alpha: 2 P(d,w) = lambda * P(z|d) + (1-lambda) * P(z|d,w) #wordTypeClasses m/ returns a Map from word types to unnormalized ! distributions over word classes $ label m s' returns for each word in sentences s, - unnormalized probabilities of word classes. % predict m s3 returns for each word in sentence s, unnormalized 4 probabilities of words given predicted word class. CpredictDoc n m ws+ returns unnormalized probabilities of top n , most probable document ids given the model m and words ws. The 1 candidate document ids are taken from the model m. The weights 2 are computed according to the following formula: F P(d|{w}) " £_z[n(d,z)+a £_{w in ws}(n(w,z)+b)/(£_{w in V} n(w,z)+b)] H DEFGHIJKLMNOPQRS T>UVWXYZ[\  ? batch size  no. repeats feature indices stream of sentences stream of batches @A!"B#$%]C^_`abcdefghi"  !"#$%' !"#$%   3 DEFGHIJKLMNOPQRS T>UVWXYZ[\  ?@A!"B#$%]C^_`abcdefghij      !"#$%&'()*+,--./012345645745845945:45;45<45=45>45?45@45ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnop colada-0.5.5Colada.WordClass NLP.CoNLL NLP.SymbolsColada.Featuresswift-lda-0.4.1 NLP.SwiftLDA topicWords topicDocstopics wordTopics docTopics FinalizedSentenceFieldTokenparseOptions WordClass featureTableldaModeloptions wordTypeTablealphasum batchSizebetaexponentfeatIds initPassesinitSizelambdapassespre progressiverepeatsseedtopicNumtopndefaultOptionslearnsummary summarizewordTypeClasseslabelpredictSymbolstoAtomAtoAtomB fromAtomA fromAtomB evalSymbols runSymbolsmonad-atom-0.4.1Control.Monad.AtommappingemptyevalAtom evalAtomTrunAtomrunAtomT AtomTableAtomTAtomfromAtom maybeToAtomtoAtom featureSeqfeaturesatindex _ldaModel prepareData featurize symbolizeinterpWordClasses predictDoc_featIds _topicNum _alphasum_beta_passes_repeats _batchSize_seed_topn _initSize _initPasses _exponent _progressive_pre_lambda_wordTypeTable _featureTable_options$fSerializeOptionsBatchRepeatSentSymb prepareSentextractFeaturescombinecompress decompressfst3countst$fSerializeWordClass$fSerializeAtomTable$fSerializeVector$fSerializeText$fSerializeFinalized