org.deeplearning4j.bagofwords.vectorizer

Class BagOfWordsVectorizer.Builder

java.lang.Object

org.deeplearning4j.bagofwords.vectorizer.BagOfWordsVectorizer.Builder

Enclosing class:

BagOfWordsVectorizer

public static class BagOfWordsVectorizer.Builder
extends java.lang.Object

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	isParallel
protected LabelAwareIterator	iterator
protected LabelsSource	labelsSource
protected int	minWordFrequency
protected java.util.Collection<java.lang.String>	stopWords
protected TokenizerFactory	tokenizerFactory
protected VocabCache<VocabWord>	vocabCache

Constructor Summary

Constructors

Constructor and Description
Builder()

Method Summary

Modifier and Type	Method and Description
BagOfWordsVectorizer.Builder	allowParallelTokenization(boolean reallyAllow)
BagOfWordsVectorizer	build()
BagOfWordsVectorizer.Builder	setIterator(DocumentIterator iterator)
BagOfWordsVectorizer.Builder	setIterator(LabelAwareIterator iterator)
BagOfWordsVectorizer.Builder	setIterator(SentenceIterator iterator)
BagOfWordsVectorizer.Builder	setMinWordFrequency(int minWordFrequency)
BagOfWordsVectorizer.Builder	setStopWords(java.util.Collection<java.lang.String> stopWords)
BagOfWordsVectorizer.Builder	setTokenizerFactory(TokenizerFactory tokenizerFactory)
BagOfWordsVectorizer.Builder	setVocab(VocabCache<VocabWord> vocab)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

tokenizerFactory

protected TokenizerFactory tokenizerFactory

iterator

protected LabelAwareIterator iterator

minWordFrequency

protected int minWordFrequency

vocabCache

protected VocabCache<VocabWord> vocabCache

labelsSource

protected LabelsSource labelsSource

stopWords

protected java.util.Collection<java.lang.String> stopWords

isParallel

protected boolean isParallel

Constructor Detail

Builder

public Builder()

Method Detail

allowParallelTokenization

public BagOfWordsVectorizer.Builder allowParallelTokenization(boolean reallyAllow)

setTokenizerFactory

public BagOfWordsVectorizer.Builder setTokenizerFactory(@NonNull
                                                        TokenizerFactory tokenizerFactory)

setIterator

public BagOfWordsVectorizer.Builder setIterator(@NonNull
                                                LabelAwareIterator iterator)

setIterator

public BagOfWordsVectorizer.Builder setIterator(@NonNull
                                                DocumentIterator iterator)

setIterator

public BagOfWordsVectorizer.Builder setIterator(@NonNull
                                                SentenceIterator iterator)

setVocab

public BagOfWordsVectorizer.Builder setVocab(@NonNull
                                             VocabCache<VocabWord> vocab)

setMinWordFrequency

public BagOfWordsVectorizer.Builder setMinWordFrequency(int minWordFrequency)

setStopWords

public BagOfWordsVectorizer.Builder setStopWords(java.util.Collection<java.lang.String> stopWords)

build

public BagOfWordsVectorizer build()