Package gnu.trove

Examples of gnu.trove.TIntArrayList

143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
    }
  }
 
  public void computeExpectations(ArrayList<SumLattice> lattices) {
    double[][] gammas;   
    TIntArrayList cache = new TIntArrayList();
    for (int i = 0; i < lattices.size(); i++) {
      if (lattices.get(i) == null) { continue; }
      SumLattice lattice = lattices.get(i);
      FeatureVectorSequence fvs = (FeatureVectorSequence)lattice.getInput();
      gammas = lattice.getGammas();
      for (int ip = 0; ip < fvs.size(); ++ip) {
        cache.resetQuick();
        FeatureVector fv = fvs.getFeatureVector(ip);
        int fi;
        for (int loc = 0; loc < fv.numLocations(); loc++) {
          fi = fv.indexAtLocation(loc);
          // binary constraint features
          if (constraints.containsKey(fi)) {
            cache.add(fi);
          }
        }
        if (constraints.containsKey(fv.getAlphabet().size())) {
          cache.add(fv.getAlphabet().size());
        }
        for (int s = 0; s < map.getNumStates(); ++s) {
          int li = map.getLabelIndex(s);
          if (li != StateLabelMap.START_LABEL) {
            double gammaProb = Math.exp(gammas[ip+1][s]);
            for (int j = 0; j < cache.size(); j++) {
              constraints.get(cache.getQuick(j)).incrementExpectation(li,gammaProb);
            }
          }
        }
      }
    }
View Full Code Here
40
41
42
43
44
45
46
  public TwoLabelGEConstraints() {
    this.constraintsList = new ArrayList<TwoLabelGEConstraint>();
    this.constraintsMap = new TIntIntHashMap();
    this.map = null;
    this.cache = new TIntArrayList();
  }
View Full Code Here
47
48
49
50
51
52
53
 
  protected TwoLabelGEConstraints(ArrayList<TwoLabelGEConstraint> constraintsList, TIntIntHashMap constraintsMap, StateLabelMap map) {
    this.constraintsList = constraintsList;
    this.constraintsMap = constraintsMap;
    this.map = map;
    this.cache = new TIntArrayList();
  }
View Full Code Here
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
    }
  }
 
  public void computeExpectations(ArrayList<SumLattice> lattices) {
    double[][][] xis;
    TIntArrayList cache = new TIntArrayList();
    for (int i = 0; i < lattices.size(); i++) {
      if (lattices.get(i) == null) { continue; }
      FeatureVectorSequence fvs = (FeatureVectorSequence)lattices.get(i).getInput();
      SumLattice lattice = lattices.get(i);
      xis = lattice.getXis();
      for (int ip = 1; ip < fvs.size(); ++ip) {
        cache.resetQuick();
        FeatureVector fv = fvs.getFeatureVector(ip);
        int fi;
        for (int loc = 0; loc < fv.numLocations(); loc++) {
          fi = fv.indexAtLocation(loc);
          // binary constraint features
          if (constraintsMap.containsKey(fi)) {
            cache.add(constraintsMap.get(fi));
          }
        }
        for (int prev = 0; prev < map.getNumStates(); ++prev) {
          int liPrev = map.getLabelIndex(prev);
          if (liPrev != StateLabelMap.START_LABEL) {
            for (int curr = 0; curr < map.getNumStates(); ++curr) {
              int liCurr = map.getLabelIndex(curr);
              if (liCurr != StateLabelMap.START_LABEL) {
                double prob = Math.exp(xis[ip][prev][curr]);
                for (int j = 0; j < cache.size(); j++) {
                  constraintsList.get(cache.getQuick(j)).expectation[liPrev][liCurr] += prob;
                }
              }
            }
          }
        }
View Full Code Here
36
37
38
39
40
41
42
43
44
45
public class PostingUtil {

  /** Get an array of all the ids in a given IterablePosting stream */
  public static int[] getIds(final IterablePosting ip) throws IOException
  {
    final TIntArrayList ids = new TIntArrayList();
    while(ip.next() != IterablePosting.EOL)
      ids.add(ip.getId());
    return ids.toNativeArray();
  }
View Full Code Here
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
    final boolean loadTagInformation = fieldCount > 0;
   
    final int[][] documentTerms = new int[fieldCount+4][];
    for(int i=0;i<fieldCount+3;i++)
      documentTerms[i] = new int[df];
    final TIntArrayList blockids = new TIntArrayList(df); //ideally we'd have TF here
 
    if (loadTagInformation) { //if there are tag information to process
      documentTerms[0][0] = file.readGamma() - 1;
      documentTerms[1][0] = file.readUnary();
      for(int fi=0;fi < fieldCount;fi++)
        documentTerms[2+fi][0] = file.readUnary() -1;
      int blockfreq = documentTerms[fieldCount+2][0] = file.readUnary() - DocumentBlockCountDelta;
      int tmpBlocks[] = new int[blockfreq];
      int previousBlockId = -1;
      for(int j=0;j<blockfreq;j++)
      {
        tmpBlocks[j] = previousBlockId = file.readGamma() + previousBlockId;
      }
      blockids.add(tmpBlocks);
     
      for (int i = 1; i < df; i++) {         
        documentTerms[0][i]  = file.readGamma() + documentTerms[0][i - 1];
        documentTerms[1][i]  = file.readUnary();
        for(int fi=0;fi < fieldCount;fi++)
          documentTerms[2+fi][i] = file.readUnary() -1;
       
        blockfreq = documentTerms[2+fieldCount][i] = file.readUnary() - DocumentBlockCountDelta;
        tmpBlocks = new int[blockfreq];
        previousBlockId = -1;
        for(int j=0;j<blockfreq;j++)
        {
          tmpBlocks[j] = previousBlockId = file.readGamma() + previousBlockId;
        }
        blockids.add(tmpBlocks);
      }
    } else { //no tag information to process         
     
      documentTerms[0][0] = file.readGamma() - 1;
      documentTerms[1][0] = file.readUnary();
     
      int blockfreq = documentTerms[2][0] = file.readUnary() - DocumentBlockCountDelta;
      int tmpBlocks[] = new int[blockfreq];
      int previousBlockId = -1;
      for(int j=0;j<blockfreq;j++)
      {
        tmpBlocks[j] = previousBlockId = file.readGamma() + previousBlockId;
      }
      blockids.add(tmpBlocks);
     
      for (int i = 1; i < df; i++) {         
        documentTerms[0][i]  = file.readGamma() + documentTerms[0][i - 1];
        documentTerms[1][i]  = file.readUnary();

        blockfreq = documentTerms[2][i] = file.readUnary() - DocumentBlockCountDelta;
        tmpBlocks = new int[blockfreq];
        previousBlockId = -1;
        for(int j=0;j<blockfreq;j++)
        {
          tmpBlocks[j] = previousBlockId = file.readGamma() + previousBlockId;
        }
        blockids.add(tmpBlocks);
      }
    }
    documentTerms[documentTerms.length -1] = blockids.toNativeArray();
    return documentTerms;
  }
View Full Code Here
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
    final boolean loadTagInformation = fieldCount > 0;
   
    final int[][] documentTerms = new int[4+fieldCount][];
    for(int i=0;i<fieldCount+3;i++)
      documentTerms[i] = new int[df];
    final TIntArrayList blockids = new TIntArrayList(df); //ideally we'd have TF here

    try{
      final BitIn file = this.file[pointer.getFileNumber()].readReset(startOffset, startBitOffset);
 
      if (loadTagInformation) { //if there are tag information to process
        //documentTerms[2] = new int[df];
        documentTerms[0][0] = file.readGamma() - 1;       
        documentTerms[1][0] = file.readUnary();
        for(int fi=0;fi < fieldCount;fi++)
          documentTerms[2+fi][0] = file.readUnary() -1;
        int blockfreq = documentTerms[2+fieldCount][0] = file.readUnary() - DocumentBlockCountDelta;
        int tmpBlocks[] = new int[blockfreq];
        int previousBlockId = -1;
        for(int j=0;j<blockfreq;j++)
        {
          tmpBlocks[j] = previousBlockId = file.readGamma() + previousBlockId;
        }
        blockids.add(tmpBlocks);
       
        for (int i = 1; i < df; i++) {         
          documentTerms[0][i]  = file.readGamma() + documentTerms[0][i - 1];
          documentTerms[1][i]  = file.readUnary();
          for(int fi=0;fi < fieldCount;fi++)
            documentTerms[2+fi][0] = file.readUnary() -1;
          blockfreq = documentTerms[2+fieldCount][i] = file.readUnary() - DocumentBlockCountDelta;
          tmpBlocks = new int[blockfreq];
          previousBlockId = -1;
          for(int j=0;j<blockfreq;j++)
          {
            tmpBlocks[j] = previousBlockId = file.readGamma() + previousBlockId;
          }
          blockids.add(tmpBlocks);
        }
      } else { //no tag information to process         
       
        documentTerms[0][0] = file.readGamma() - 1;
        documentTerms[1][0] = file.readUnary();
       
        int blockfreq = documentTerms[2][0] = file.readUnary() - DocumentBlockCountDelta;
        int tmpBlocks[] = new int[blockfreq];
        int previousBlockId = -1;
        for(int j=0;j<blockfreq;j++)
        {
          tmpBlocks[j] = previousBlockId = file.readGamma() + previousBlockId;
        }
        blockids.add(tmpBlocks);
       
        for (int i = 1; i < df; i++) {         
          documentTerms[0][i]  = file.readGamma() + documentTerms[0][i - 1];
          documentTerms[1][i]  = file.readUnary();

          blockfreq = documentTerms[2][i] = file.readUnary() - DocumentBlockCountDelta;
          tmpBlocks = new int[blockfreq];
          previousBlockId = -1;
          for(int j=0;j<blockfreq;j++)
          {
            tmpBlocks[j] = previousBlockId = file.readGamma() + previousBlockId;
          }
          blockids.add(tmpBlocks);
        }
      }
      documentTerms[documentTerms.length-1] = blockids.toNativeArray();
      return documentTerms;
    } catch (IOException ioe) {
      logger.error("Problem reading block inverted index", ioe);
      return null;
    }
View Full Code Here
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
    {
      filters[i].new_query(this, srq, results);
    }
   
    int doccount = -1;//doccount is zero-based, so 0 means 1 document
    TIntArrayList docatnumbers = new TIntArrayList();//list of resultset index numbers to keep
    byte docstatus; int thisDocId;
    int[] docids = results.getDocids();
    //int elitesetsize = results.getExactResultSize();
    //the exact result size is the total number of
    //documents that would be retrieved if we
    //didn't do any cropping
    int elitesetsize = results.getResultSize();
    for(int thisdoc = 0; thisdoc < elitesetsize; thisdoc++)
    {
      //run this document through all the filters
      docstatus = PostFilter.FILTER_OK;
      thisDocId = docids[thisdoc];
      //run this doc through the filters
      for(int i=0;i<filters_length; i++)
      {
        if ( ( docstatus = filters[i].filter(this, srq, results, thisdoc, thisDocId) )
          == PostFilter.FILTER_REMOVE
        )
          break;
          //break if the document has to be removed
      }
      //if it's not being removed, then
      if (docstatus != PostFilter.FILTER_REMOVE) //TODO this should always be true
      {
        //success, another real document
        doccount++;
        //check if it's in our results "WINDOW"
        if (doccount >= Start)
        {
          if (doccount <= End)
          {  //add to the list of documents to keep
            docatnumbers.add(thisdoc);
            //System.out.println("Keeping @"+thisdoc);
          }
          else
          {
            //we've now got enough results, break
            break;
          }
        }
      }
      else
      {
        //System.out.println("Removed");
      }
    }
    //since doccount is zero-based, we add one so that it
    //corresponds to the real number of documents.
    doccount++;
    rq.setNumberOfDocumentsAfterFiltering(doccount);
    if (docatnumbers.size() < docids.length)
    {
      //result set is definently shorter, replace with new one
      rq.setResultSet( results.getResultSet(docatnumbers.toNativeArray()));
      rq.getResultSet().setExactResultSize(results.getExactResultSize());
    }
  }
           
        catch(Exception e){}
View Full Code Here
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
  protected TIntArrayList[] createPointerForTerm(LexiconEntry le)
  {
    TIntArrayList[] tmpArray = new TIntArrayList[4+fieldCount];
    final int tmpNT = le.getDocumentFrequency();
    for(int i = 0; i < fieldCount+3; i++)
      tmpArray[i] = new TIntArrayList(tmpNT);
    if (le instanceof BlockEntryStatistics)
    {
      tmpArray[fieldCount+3] = new TIntArrayList(((BlockEntryStatistics)le).getBlockCount());
    }
    else
    {
      tmpArray[fieldCount+3] = new TIntArrayList(le.getFrequency());
    }
    return tmpArray;
  }
View Full Code Here
402
403
404
405
406
407
408
409
  protected TIntArrayList[] createPointerForTerm(LexiconEntry le)
  {
    TIntArrayList[] tmpArray = new TIntArrayList[2 + fieldCount];
    final int tmpNT = le.getDocumentFrequency();
    for(int i = 0; i < fieldCount+2; i++)
      tmpArray[i] = new TIntArrayList(tmpNT);
    return tmpArray;
  }
View Full Code Here
TOP

Related Classes of gnu.trove.TIntArrayList

  • cc.mallet.classify.constraints.ge.MaxEntFLGEConstraints
  • cc.mallet.classify.constraints.ge.MaxEntRangeL2FLGEConstraints
  • cc.mallet.classify.constraints.pr.MaxEntFLPRConstraints
  • cc.mallet.cluster.GreedyAgglomerativeByDensity
  • cc.mallet.cluster.tui.Text2Clusterings
  • cc.mallet.fst.semi_supervised.constraints.OneLabelGEConstraints
  • cc.mallet.fst.semi_supervised.constraints.OneLabelL2RangeGEConstraints
  • cc.mallet.fst.semi_supervised.constraints.TwoLabelGEConstraints
  • cc.mallet.fst.semi_supervised.pr.constraints.OneLabelL2IndPRConstraints
  • cc.mallet.fst.semi_supervised.pr.constraints.OneLabelL2PRConstraints
    • Copyright © 2018 www.massapicom. All rights reserved.
    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.