gnu.trove.iterator.TObjectIntIterator Java Examples

private void buildAttributeModels(int datasetId, List<EntityProfile> profiles) {
    attrNameIndex = new TObjectIntHashMap<>();
    profiles.forEach((profile) -> {
        profile.getAttributes().forEach((attribute) -> {
            attrNameIndex.putIfAbsent(attribute.getName(), attrNameIndex.size() + 1);
        });
    });

    int currentAttributes = attrNameIndex.size();
    attributeModels[datasetId] = new ITextModel[currentAttributes];
    final TObjectIntIterator<String> it = attrNameIndex.iterator();
    while (it.hasNext()) {
        it.advance();
        attributeModels[datasetId][it.value() - 1] = RepresentationModel.getModel(datasetId, repModel, simMetric, it.key());
    }

    profiles.forEach((profile) -> {
        profile.getAttributes().forEach((attribute) -> {
            updateModel(datasetId, attribute);
        });
    });

    for (int i = 0; i < currentAttributes; i++) {
        attributeModels[datasetId][i].finalizeModel();
    }
}
 

protected float getSigmaSimilarity(TokenNGramsWithGlobalWeights oModel) {
    float totalTerms2 = oModel.getNoOfTotalTerms();
    final TObjectIntMap<String> itemVector2 = oModel.getItemsFrequency();

    float numerator = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        int frequency2 = itemVector2.get(iterator.key());
        if (0 < frequency2) {
            numerator += iterator.value() / noOfTotalTerms * getIdfWeight(iterator.key())
                       + frequency2 / totalTerms2 * oModel.getIdfWeight(iterator.key());
        }
    }

    final Set<String> allKeys = new HashSet<>(itemsFrequency.keySet());
    allKeys.addAll(itemVector2.keySet());
    float denominator = 0.0f;
    denominator = allKeys.stream().map((key) -> itemsFrequency.get(key) / noOfTotalTerms  * getIdfWeight(key) + 
            itemVector2.get(key) / totalTerms2 * oModel.getIdfWeight(key)).reduce(denominator, (accumulator, _item) -> accumulator + _item);

    return (float)(numerator / denominator);
}
 

protected float getTfIdfCosineSimilarity(TokenNGramsWithGlobalWeights oModel) {
    float totalTerms2 = oModel.getNoOfTotalTerms();
    final TObjectIntMap<String> itemVector2 = oModel.getItemsFrequency();

    float numerator = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        int frequency2 = itemVector2.get(iterator.key());
        if (0 < frequency2) {
            numerator += (iterator.value() / noOfTotalTerms) * getIdfWeight(iterator.key())
                       * (frequency2 / totalTerms2) * oModel.getIdfWeight(iterator.key());
        }
    }

    float denominator = getVectorMagnitude() * oModel.getVectorMagnitude();
    return (float)(numerator / denominator);
}
 

protected float getTfIdfGeneralizedJaccardSimilarity(CharacterNGramsWithGlobalWeights oModel) {
    float totalTerms2 = oModel.getNoOfTotalTerms();
    final TObjectIntMap<String> itemVector2 = oModel.getItemsFrequency();

    float numerator = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        int frequency2 = itemVector2.get(iterator.key());
        if (0 < frequency2) {
            numerator += Math.min(iterator.value() / noOfTotalTerms * getIdfWeight(iterator.key()),
                                  frequency2 / totalTerms2 * oModel.getIdfWeight(iterator.key()));
        }
    }

    final Set<String> allKeys = new HashSet<>(itemsFrequency.keySet());
    allKeys.addAll(itemVector2.keySet());
    float denominator = 0.0f;
    denominator = allKeys.stream().map((key) -> Math.max(itemsFrequency.get(key) / noOfTotalTerms  * getIdfWeight(key),
            itemVector2.get(key) / totalTerms2 * oModel.getIdfWeight(key))).reduce(denominator, (accumulator, _item) -> accumulator + _item);

    return (float)(numerator / denominator);
}
 

protected float getTfIdfCosineSimilarity(CharacterNGramsWithGlobalWeights oModel) {
    float totalTerms2 = oModel.getNoOfTotalTerms();
    final TObjectIntMap<String> itemVector2 = oModel.getItemsFrequency();

    float numerator = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        int frequency2 = itemVector2.get(iterator.key());
        if (0 < frequency2) {
            numerator += (iterator.value() / noOfTotalTerms) * getIdfWeight(iterator.key())
                       * (frequency2 / totalTerms2) * oModel.getIdfWeight(iterator.key());
        }
    }

    float denominator = getVectorMagnitude() * oModel.getVectorMagnitude();
    return (float)(numerator / denominator);
}
 

protected float getSigmaSimilarity(CharacterNGramsWithGlobalWeights oModel) {
    float totalTerms2 = oModel.getNoOfTotalTerms();
    final TObjectIntMap<String> itemVector2 = oModel.getItemsFrequency();

    float numerator = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        int frequency2 = itemVector2.get(iterator.key());
        if (0 < frequency2) {
            numerator += iterator.value() / noOfTotalTerms * getIdfWeight(iterator.key())
                       + frequency2 / totalTerms2 * oModel.getIdfWeight(iterator.key());
        }
    }

    final Set<String> allKeys = new HashSet<>(itemsFrequency.keySet());
    allKeys.addAll(itemVector2.keySet());
    float denominator = 0.0f;
    denominator = allKeys.stream().map((key) -> itemsFrequency.get(key) / noOfTotalTerms  * getIdfWeight(key) + 
            itemVector2.get(key) / totalTerms2 * oModel.getIdfWeight(key)).reduce(denominator, (accumulator, _item) -> accumulator + _item);

    return (float)(numerator / denominator);
}
 

protected float getTfIdfGeneralizedJaccardSimilarity(TokenNGramsWithGlobalWeights oModel) {
    float totalTerms2 = oModel.getNoOfTotalTerms();
    final TObjectIntMap<String> itemVector2 = oModel.getItemsFrequency();

    float numerator = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        int frequency2 = itemVector2.get(iterator.key());
        if (0 < frequency2) {
            numerator += Math.min(iterator.value() / noOfTotalTerms * getIdfWeight(iterator.key()),
                                  frequency2 / totalTerms2 * oModel.getIdfWeight(iterator.key()));
        }
    }

    final Set<String> allKeys = new HashSet<>(itemsFrequency.keySet());
    allKeys.addAll(itemVector2.keySet());
    float denominator = 0.0f;
    denominator = allKeys.stream().map((key) -> Math.max(itemsFrequency.get(key) / noOfTotalTerms  * getIdfWeight(key),
            itemVector2.get(key) / totalTerms2 * oModel.getIdfWeight(key))).reduce(denominator, (accumulator, _item) -> accumulator + _item);

    return (float)(numerator / denominator);
}
 

protected float getTfCosineSimilarity(BagModel oModel) {
    float totalTerms2 = oModel.getNoOfTotalTerms();

    TObjectIntMap<String> itemVector1 = itemsFrequency;
    TObjectIntMap<String> itemVector2 = oModel.getItemsFrequency();
    if (itemVector2.size() < itemVector1.size()) {
        itemVector1 = oModel.getItemsFrequency();
        itemVector2 = itemsFrequency;
    }

    float numerator = 0.0f;
    for (TObjectIntIterator<String> iterator = itemVector1.iterator(); iterator.hasNext();) {
        iterator.advance();
        numerator += iterator.value() * itemVector2.get(iterator.key()) / noOfTotalTerms / totalTerms2;
    }

    float denominator = getVectorMagnitude() * oModel.getVectorMagnitude();
    return (float)(numerator / denominator);
}
 

@Override
public float getEntropy(boolean normalized) {
    float entropy = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        float p_i = (iterator.value() / noOfTotalTerms);
        entropy -= (p_i * (Math.log10(p_i) / Math.log10(2.0d)));
    }
    
    if (normalized) {
        float maxEntropy = (float) Math.log10(noOfTotalTerms) / (float) Math.log10(2.0f);
        return entropy / maxEntropy;
    } 
        
    return entropy;
}
 

protected float getEnhancedJaccardSimilarity(BagModel oModel) {
    TObjectIntMap<String> itemVector1 = itemsFrequency;
    TObjectIntMap<String> itemVector2 = oModel.getItemsFrequency();
    if (itemVector2.size() < itemVector1.size()) {
        itemVector1 = oModel.getItemsFrequency();
        itemVector2 = itemsFrequency;
    }

    float numerator = 0.0f;
    for (TObjectIntIterator<String> iterator = itemVector1.iterator(); iterator.hasNext();) {
        iterator.advance();
        numerator += Math.min(iterator.value(), itemVector2.get(iterator.key()));
    }

    float denominator = noOfTotalTerms + oModel.getNoOfTotalTerms() - numerator;
    return numerator / (float)denominator;
}
 

/**
 * PUBLIC: Returns the preferred variant of the term. This is a form
 * of the term which actually occurred in the classified content.
 */
public String getPreferredName() {
  if (variants.isEmpty())
    return getStem();
  Variant maxKey = null;
  int maxValue = -1;
  TObjectIntIterator<Variant> iter = variants.iterator();
  while (iter.hasNext()) {
    iter.advance();
    int thisValue = iter.value();
    Variant thisKey = iter.key();
    // select variant with most occurrences, or lowest lexical value if equal for predictability
    if ((thisValue > maxValue) ||
        ((thisValue == maxValue) && (thisKey.getValue().compareTo(maxKey.getValue()) < 0))) {
      maxValue = thisValue;
      maxKey = thisKey;
    }
  }
  return maxKey.getValue();
}
 

protected void merge(Term other) {
  if (other == this) return;
  
  this.score = this.score + other.score;    
  this.totalOccurrences = this.totalOccurrences + other.totalOccurrences;

  TObjectIntIterator<Variant> iter = other.variants.iterator();
  while (iter.hasNext()) {
    iter.advance();
    Variant key = iter.key();
    int value = iter.value();
    if (this.variants.containsKey(key))
      this.variants.adjustValue(key, value);
    else
      this.variants.put(key, value);
    key.replaceTerm(this);
  }
}
 

public static Entities getAidaEntitiesForKBEntities(Set<KBIdentifiedEntity> entities) throws EntityLinkingDataAccessException {
  TObjectIntHashMap<KBIdentifiedEntity> kbEntities = DataAccess.getInternalIdsForKBEntities(entities);
  Entities aidaEntities = new Entities();
  for (TObjectIntIterator<KBIdentifiedEntity> itr = kbEntities.iterator(); itr.hasNext(); ) {
    itr.advance();
    aidaEntities.add(new Entity(itr.key(), itr.value()));
  }
  return aidaEntities;
}
 

@Override
protected float getVectorMagnitude() {
    float magnitude = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        magnitude += Math.pow(iterator.value() * getIdfWeight(iterator.key()) / noOfTotalTerms, 2.0);
    }

    return (float) Math.sqrt(magnitude);
}
 

public boolean updatePositions(NetworkState<TPos> state){
    if(!railLinkHolds.isEmpty()) {
        TObjectIntIterator<TPos> iterator = railLinkHolds.iterator();
        while(iterator.hasNext()) {
            iterator.advance();
            if(iterator.value() == 1) {
                iterator.remove();
            } else {
                iterator.setValue(iterator.value() - 1);
            }
        }
    }
    return false;
}
 

/**
 * 得到索引集合
 * @return
 */
public Set<Integer> getIndexSet() {
	Set<Integer> set = new HashSet<Integer>();
	for (TObjectIntIterator<String> it = data.iterator(); it.hasNext();) {
		it.advance();
		set.add(it.value());
	}
	return set;
}
 

public Map<String,Integer> toMap() {
	HashMap<String,Integer> map = new HashMap<String,Integer>();
	for (TObjectIntIterator<String> it = data.iterator(); it.hasNext();) {
		it.advance();
		map.put(it.key(), it.value());
	}
	return map;
}
 

/**
 * 得到标签集合
 */
public Set<String> toSet() {
	Set<String> set = new HashSet<String>();
	for (TObjectIntIterator<String> it = data.iterator(); it.hasNext();) {
		it.advance();
		set.add(it.key());
	}
	return set;
}
 

/**
 * 将标签集合输出为字符串
 */
public String toString() {
	StringBuffer sb = new StringBuffer();
	for (TObjectIntIterator<String> it = data.iterator(); it.hasNext();) {
		it.advance();
		sb.append(it.key());
		sb.append(",");
	}
	return sb.toString();
}
 

@Override
protected float getVectorMagnitude() {
    float magnitude = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        magnitude += Math.pow(iterator.value() * getIdfWeight(iterator.key()) / noOfTotalTerms, 2.0);
    }

    return (float) Math.sqrt(magnitude);
}
 

protected float getVectorMagnitude() {
    float magnitude = 0.0f;
    for (TObjectIntIterator<String> iterator = itemsFrequency.iterator(); iterator.hasNext();) {
        iterator.advance();
        magnitude += Math.pow(iterator.value() / noOfTotalTerms, 2.0);
    }

    return (float) Math.sqrt(magnitude);
}
 

public static <T> void writeTObjectIntMapToFile(File file, TObjectIntHashMap<T> map) throws IOException {
  BufferedWriter writer = getBufferedUTF8Writer(file);
  for (TObjectIntIterator<T> itr = map.iterator(); itr.hasNext(); ) {
    itr.advance();
    writer.write(itr.key() + "\t" + itr.value());
    writer.newLine();
  }
  writer.flush();
  writer.close();
}
 

public static TIntObjectHashMap<String> getAllWordIds() throws EntityLinkingDataAccessException {
  TObjectIntHashMap<String> wordIds = DataAccess.getAllWordIds();
  TIntObjectHashMap<String> idWords = new TIntObjectHashMap<>(wordIds.size());
  for (TObjectIntIterator<String> itr = wordIds.iterator(); itr.hasNext(); ) {
    itr.advance();
    idWords.put(itr.value(), itr.key());
  }
  return idWords;
}
 

public static TIntObjectHashMap<String> getAllWordIds() throws EntityLinkingDataAccessException {
  TObjectIntHashMap<String> wordIds = DataAccess.getAllWordIds();
  TIntObjectHashMap<String> idWords = new TIntObjectHashMap<String>(wordIds.size());
  for (TObjectIntIterator<String> itr = wordIds.iterator(); itr.hasNext(); ) {
    itr.advance();
    idWords.put(itr.value(), itr.key());
  }
  return idWords;
}
 

public static TIntObjectHashMap<String> getAllWordIds() throws EntityLinkingDataAccessException {
  TObjectIntHashMap<String> wordIds = DataAccess.getAllWordIds();
  TIntObjectHashMap<String> idWords = new TIntObjectHashMap<String>(wordIds.size());
  for (TObjectIntIterator<String> itr = wordIds.iterator(); itr.hasNext(); ) {
    itr.advance();
    idWords.put(itr.value(), itr.key());
  }
  return idWords;
}
 

@Override
public Entities getAllEntities(boolean isNamedEntity) throws EntityLinkingDataAccessException {
  TObjectIntHashMap<KBIdentifiedEntity> entityIds = getAllEntityIds(isNamedEntity);
  Entities entities = new Entities();
  for (TObjectIntIterator<KBIdentifiedEntity> itr = entityIds.iterator();
      itr.hasNext(); ) {
    itr.advance();
    entities.add(new Entity(itr.key(), itr.value()));
  }
  return entities;
}
 

@Override public Entities getAllEntities() throws EntityLinkingDataAccessException {
  TObjectIntHashMap<KBIdentifiedEntity> entityIds = getAllEntityIds();
  Entities entities = new Entities();
  for (TObjectIntIterator<KBIdentifiedEntity> itr = entityIds.iterator(); itr.hasNext(); ) {
    itr.advance();
    entities.add(new Entity(itr.key(), itr.value()));
  }
  return entities;
}
 

public static Map<KBIdentifiedEntity, EntityMetaData> getEntitiesMetaData(Set<KBIdentifiedEntity> entities)
    throws EntityLinkingDataAccessException {
  TObjectIntHashMap<KBIdentifiedEntity> ids = getInternalIdsForKBEntities(entities);
  TIntObjectHashMap<EntityMetaData> metadata = getEntitiesMetaData(ids.values());
  Map<KBIdentifiedEntity, EntityMetaData> result = new HashMap<KBIdentifiedEntity, EntityMetaData>();
  for (TObjectIntIterator<KBIdentifiedEntity> itr = ids.iterator(); itr.hasNext(); ) {
    itr.advance();
    int id = itr.value();
    result.put(itr.key(), metadata.get(id));
  }
  return result;
}
 

public static Entities getEntities(Set<KBIdentifiedEntity> kbEntities) throws EntityLinkingDataAccessException {
  TObjectIntHashMap<KBIdentifiedEntity> ids = DataAccess.getInternalIdsForKBEntities(kbEntities);
  Entities entities = new Entities();
  for (TObjectIntIterator<KBIdentifiedEntity> itr = ids.iterator(); itr.hasNext(); ) {
    itr.advance();
    entities.add(new Entity(itr.key(), itr.value()));
  }
  return entities;
}
 

public TObjectIntIterator<String> iterator()	{
	return data.iterator();
}