Package cc.mallet.grmm.inference

Examples of cc.mallet.grmm.inference.Inferencer

139
140
141
142
143
144
145
146
147
148
149
150
    testing.addThruPipe (testSource);

    ACRF.Template[] tmpls = parseModelFile (modelFile.value);
    ACRFEvaluator eval = createEvaluator (evalOption.value);

    Inferencer inf = createInferencer (inferencerOption.value);
    Inferencer maxInf = createInferencer (maxInferencerOption.value);

    ACRF acrf = new ACRF (pipe, tmpls);
    acrf.setInferencer (inf);
    acrf.setViterbiInferencer (maxInf);
View Full Code Here
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
  public void testBadVariable ()
  {
    FactorGraph fg = createBoltzmannChain (5);
    Assignment assn = fg.sampleContinuousVars (new Randoms (23423));
    FactorGraph sliced = (FactorGraph) fg.slice (assn);
    Inferencer bp = new TRP ();
    bp.computeMarginals (sliced);

    try {
      bp.lookupMarginal (new Variable (2));
      fail ("Expected exception");
    } catch (IllegalArgumentException e) {
      // expected
      System.out.println ("OK: As expected, got exception "+e);
    }
View Full Code Here
155
156
157
158
159
160
161
162
163
164
165
166
    ACRFExtractorTrainer trainer = createTrainer (trainerOption.value);
    ACRFEvaluator eval = createEvaluator (evalOption.value);
    ExtractionEvaluator extractionEval = createExtractionEvaluator (extractionEvalOption.value);

    Inferencer inf = createInferencer (inferencerOption.value);
    Inferencer maxInf = createInferencer (maxInferencerOption.value);

    trainer.setPipes (tokPipe, new TokenSequence2FeatureVectorSequence ())
            .setDataSource (trainSource, testSource)
            .setEvaluator (eval)
            .setTemplates (tmpls)
View Full Code Here
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
  // Reports true joint likelihood of estimated parameters on the training set.
  public static void reportTrainingLikelihood (ACRF acrf, InstanceList trainingList)
  {
    double total = 0;
    Inferencer inf = acrf.getInferencer ();
    for (int i = 0; i < trainingList.size (); i++) {
      Instance inst = trainingList.get (i);
      ACRF.UnrolledGraph unrolled = acrf.unroll (inst);
      inf.computeMarginals (unrolled);
      double lik = inf.lookupLogJoint (unrolled.getAssignment ());
      total += lik;
      logger.info ("...instance "+i+" likelihood = "+lik);
    }
    logger.info ("Unregularized joint likelihood = "+total);
  }
View Full Code Here
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
        if (unrolled.factors ().size () == 0) {
          System.err.println ("WARNING: FactorGraph for instance " + instance.getName () + " : no factors.");
          continue;
        }

        Inferencer inf = acrf.getInferencer ();
        inf.computeMarginals (unrolled);

        Assignment target = unrolled.getAssignment ();
        for (Iterator it = unrolled.unrolledVarSetIterator (); it.hasNext ();) {
          ACRF.UnrolledVarSet vs = (ACRF.UnrolledVarSet) it.next ();
          Factor marg = inf.lookupMarginal (vs);
          for (AssignmentIterator assnIt = vs.assignmentIterator (); assnIt.hasNext (); assnIt.advance ()) {
            if (marg.value (assnIt) > wrongWrongThreshold) {
              Assignment assn = assnIt.assignment ();
              for (int vi = 0; vi < vs.size (); vi++) {
                Variable var = vs.get (vi);
View Full Code Here
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
      mdl.addFactor (v1, v2, ptlarr);
    }

    // STEP 2: Compute marginals

    Inferencer inf = new JunctionTreeInferencer ();
    inf.computeMarginals (mdl);

    // STEP 3: Collect the results
    //   We'll just print them out

    for (int varnum = 0; varnum < allVars.length; varnum++) {
      Variable var = allVars[varnum];
      Factor ptl = inf.lookupMarginal (var);
      for (AssignmentIterator it = ptl.assignmentIterator (); it.hasNext (); it.advance()) {
        int outcome = it.indexOfCurrentAssn ();
        System.out.println (var+"  "+outcome+"   "+ptl.value (it));
      }
      System.out.println ();
View Full Code Here
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
      mdl.addFactor (v1, v2, ptlarr);
    }

    // STEP 2: Compute marginals

    Inferencer inf = new JunctionTreeInferencer ();
    inf.computeMarginals (mdl);

    // STEP 3: Collect the results
    //   We'll just print them out

    for (int varnum = 0; varnum < allVars.length; varnum++) {
      Variable var = allVars[varnum];
      Factor ptl = inf.lookupMarginal (var);
      for (AssignmentIterator it = ptl.assignmentIterator (); it.hasNext ();) {
        int outcome = it.indexOfCurrentAssn ();
        System.out.println (var+"  "+outcome+"   "+ptl.value (it));
      }
      System.out.println ();
View Full Code Here
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
  public void testBadVariable ()
  {
    FactorGraph fg = createBoltzmannChain (5);
    Assignment assn = fg.sampleContinuousVars (new Randoms (23423));
    FactorGraph sliced = (FactorGraph) fg.slice (assn);
    Inferencer bp = new TRP ();
    bp.computeMarginals (sliced);

    try {
      bp.lookupMarginal (new Variable (2));
      fail ("Expected exception");
    } catch (IllegalArgumentException e) {
      // expected
      System.out.println ("OK: As expected, got exception "+e);
    }
View Full Code Here
113
114
115
116
117
118
119
120
121
122
123
124
    testing.addThruPipe (testSource);

    ACRF.Template[] tmpls = parseModelFile (modelFile.value);
    ACRFEvaluator eval = createEvaluator (evalOption.value);

    Inferencer inf = createInferencer (inferencerOption.value);
    Inferencer maxInf = createInferencer (maxInferencerOption.value);

    ACRF acrf = new ACRF (pipe, tmpls);
    acrf.setInferencer (inf);
    acrf.setViterbiInferencer (maxInf);
View Full Code Here
TOP

Related Classes of cc.mallet.grmm.inference.Inferencer

  • cc.mallet.grmm.examples.SimpleGraphExample
  • cc.mallet.grmm.learning.extract.AcrfExtractorTui
  • cc.mallet.grmm.learning.GenericAcrfTui
  • cc.mallet.grmm.learning.PwplACRFTrainer
  • cc.mallet.grmm.learning.PwplACRFTrainer$Maxable
  • cc.mallet.grmm.test.TestAbstractBeliefPropagation
    • 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.