Examples of cc.mallet.types.FeatureVectorSequence$Iterator

cc.mallet.types.FeatureVectorSequence
An iterator over the FeatureVectors in the sequence.

    logger.info("Gathering constraints...");
    assert (constraints.structureMatches(crf.parameters));
    constraints.zero();


    for (Instance instance : ilist) {
      FeatureVectorSequence input = (FeatureVectorSequence) instance.getData();
      FeatureSequence output = (FeatureSequence) instance.getTarget();
      double instanceWeight = ilist.getInstanceWeight(instance);
      Transducer.Incrementor incrementor =
        instanceWeight == 1.0 ? constraints.new Incrementor()
      : constraints.new WeightedIncrementor(instanceWeight);

View Full Code Here

    double value = 0;
    double unlabeledWeight, labeledWeight, weight;
    for (int ii = batchAssignments[0]; ii < batchAssignments[1]; ii++) {
      Instance instance = trainingSet.get(ii);
      double instanceWeight = trainingSet.getInstanceWeight(instance);
      FeatureVectorSequence input = (FeatureVectorSequence) instance.getData();
      FeatureSequence output = (FeatureSequence) instance.getTarget();


      labeledWeight = new SumLatticeDefault (this.crf, input, output, null).getTotalWeight();
      if (Double.isInfinite (labeledWeight)) {
        ++numInfLabeledWeight;

View Full Code Here

  public void computeExpectations() {
    expectations.zero();


    // now, update the expectations due to each instance for entropy reg.
    for (int ii = 0; ii < data.size(); ii++) {
      FeatureVectorSequence input = (FeatureVectorSequence) data.get(ii).getData();
      SumLattice lattice = new SumLatticeDefault(crf,input, true);


      // udpate the expectations
      EntropyLattice entropyLattice = new EntropyLattice(
          input, lattice.getGammas(), lattice.getXis(), crf,

View Full Code Here

    // Reset constraints[] to zero before we fill them again
    assert (constraints.structureMatches(crf.parameters));
    constraints.zero();


    for (Instance instance : ilist) {
      FeatureVectorSequence input = (FeatureVectorSequence) instance.getData();
      FeatureSequence output = (FeatureSequence) instance.getTarget();
      double instanceWeight = ilist.getInstanceWeight(instance);
      //System.out.println ("Constraint-gathering on instance "+i+" of "+ilist.size());
      Transducer.Incrementor incrementor = instanceWeight == 1.0 ? constraints.new Incrementor() : constraints.new WeightedIncrementor(instanceWeight);
      new SumLatticeDefault (this.crf, input, output, incrementor);

View Full Code Here

    // Calculate the value of each instance, and also fill in expectations
    double unlabeledWeight, labeledWeight, weight;
    for (int ii = 0; ii < trainingSet.size(); ii++) {
      Instance instance = trainingSet.get(ii);
      double instanceWeight = trainingSet.getInstanceWeight(instance);
      FeatureVectorSequence input = (FeatureVectorSequence) instance.getData();
      FeatureSequence output = (FeatureSequence) instance.getTarget();
      labeledWeight = new SumLatticeDefault (this.crf, input, output, (Transducer.Incrementor)null).getTotalWeight();
      String instanceName = instance.getName() == null ? "instance#"+ii : instance.getName().toString();
      //System.out.println ("labeledWeight = "+labeledWeight);
      if (Double.isInfinite (labeledWeight)) {

View Full Code Here

          featureIndicesArr[index] = featureIndices.get(index);
        }
        fvs[l] = featureInductionOption.value ? new AugmentableFeatureVector(features, featureIndicesArr, null, featureIndicesArr.length) : 
          new FeatureVector(features, featureIndicesArr);
      }
      carrier.setData(new FeatureVectorSequence(fvs));
      if (isTargetProcessing()) {
        carrier.setTarget(target);
      }
      else {
        carrier.setTarget(new LabelSequence(getTargetAlphabet()));

View Full Code Here

          featureIndicesArr[index] = featureIndices.get(index);
        }
         fvs[l] = featureInductionOption.value ? new AugmentableFeatureVector(features, featureIndicesArr, null, featureIndicesArr.length) : 
          new FeatureVector(features, featureIndicesArr);
      }
      carrier.setData(new FeatureVectorSequence(fvs));
      if (isTargetProcessing())
        carrier.setTarget(target);
      else
        carrier.setTarget(new LabelSequence(getTargetAlphabet()));
      return carrier;

View Full Code Here

        for (int i = start; i < end; ++i) {
          // skip if the instance doesn't have any constraints
          if (!constraintBits.get(i)) {
            continue;
          }
          FeatureVectorSequence fvs =
              (FeatureVectorSequence) data.get(i).getData();
          SumLattice lattice = lattices.get(i);
          assert(lattice != null)
              : "Lattice is null:: " + i + ", size: " + lattices.size();
          new GELattice(

View Full Code Here

      BitSet constraintBits = geCriteria.getConstraintBits();
      for (int ii = start; ii < end; ii++) {
        if (!constraintBits.get(ii)) {
          continue;
        }
        FeatureVectorSequence fvs = (FeatureVectorSequence)data.get(ii).getData();
        lattices.put(ii, new SumLatticeDefault(crf,fvs,true));
      }
      return null;
    }

View Full Code Here

    Set<Integer> indices = constraints.keySet();
    // true if at least on feature constraint is present anywhere in the
    // *instance*
    boolean featurePresent = false;
    for (int i = start; i < end; ++i) {
      FeatureVectorSequence fvs =
          (FeatureVectorSequence) ilist.get(i).getData();
      featurePresent = false;
      for (int ip = 0; ip < fvs.size(); ++ip) {
        FeatureVector fv = fvs.getFeatureVector(ip);
        // set flag and bit if any constraint is present
        for (int index : indices) {
          if (fv.value(index) > 0.0) {
            featurePresent = true;
            break;

View Full Code Here

0 1 2 3 4

TOP

Related Classes of cc.mallet.types.FeatureVectorSequence$Iterator

ar.glyphsets.implicitgeometry.Indexed

cc.mallet.classify.RankMaxEnt

cc.mallet.classify.RankMaxEntTrainer$MaximizableTrainer

cc.mallet.fst.CRF

cc.mallet.fst.CRFOptimizableByBatchLabelLikelihood

cc.mallet.fst.CRFOptimizableByLabelLikelihood

cc.mallet.fst.CRFTrainerByStochasticGradient

cc.mallet.fst.MEMMTrainer

cc.mallet.fst.semi_supervised.constraints.OneLabelGEConstraints

cc.mallet.fst.semi_supervised.constraints.OneLabelL2RangeGEConstraints

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.