Package Jama

Examples of Jama.Matrix

95
96
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
    IndexSearcher is = new IndexSearcher(dir);
    //IndexSearcher is = lis.getIndexSearcher();
    is.setSimilarity(new MySimilarity());
    ir = is.getIndexReader();
    long time = System.currentTimeMillis();
    Matrix matrix = new Matrix(ir.numDocs(), ir.numDocs());
    int nz = 0;
    for(int i=0;i<ir.numDocs();i++) {
      System.out.println(i);
      //nz++;
      matrix.set(i, i, 1.0);
      Query q = myMoreLikeThisQuery(ir, i);
      //time = System.currentTimeMillis();
      Hits h = is.search(q);
      //System.out.println(System.currentTimeMillis() - time);
      //time = System.currentTimeMillis();
      Iterator iterator = h.iterator();
      while(iterator.hasNext()) {
        Hit hit = (Hit)iterator.next();
        int id = hit.getId();
        float score = hit.getScore();
        //if(score < 0.025) break;
        if(i != id) {
          //matrix.set(i, id, score);
          matrix.set(i, id, matrix.get(i, id) + score/2);
          matrix.set(id, i, matrix.get(id, i) + score/2);
          //nz++;
        }
        //System.out.println(i + "\t" + hit.getId() + "\t" + hit.getScore());
      }
      //System.out.println(System.currentTimeMillis() - time);
      //System.out.println(((double)h.length()) / ir.numDocs());
     
    }
    for(int i=0;i<matrix.getColumnDimension();i++) {
      for(int j=0;j<matrix.getRowDimension();j++) {
        if(matrix.get(i, j) != 0.0) nz++;
      }
    }
    System.out.println(System.currentTimeMillis() - time);
    time = System.currentTimeMillis();
    //matrix.eig();
    System.out.println(System.currentTimeMillis() - time);
    System.out.println(nz);
   
   
   
    int newnz = 0;
    System.out.println(matrix.getColumnDimension() + " " + matrix.getRowDimension() + " " + nz);
    for(int i=0;i<matrix.getColumnDimension();i++) {
      int nonzero = 0;
      for(int j=0;j<matrix.getRowDimension();j++) {
        if(matrix.get(i, j) != 0.0) nonzero++;
      }
      System.out.println(nonzero);
      for(int j=0;j<matrix.getRowDimension();j++) {
        if(matrix.get(i, j) != 0.0) System.out.println(j + " " + matrix.get(i, j));
      }
      newnz += nonzero;
    }
    System.out.println(newnz);
  }
View Full Code Here
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
        svdh.set(fno, i, wv.get(feature));
      }
    }
    System.out.println("Harness ready...");
    svdh.svd(10);
    Matrix lm = svdh.getUt().transpose();
    System.out.println(lm.getRowDimension());
    System.out.println(lm.getColumnDimension());
    System.out.println(termNo);
    System.out.println(featureNo);
    double [] svals = svdh.getS();
    if(true) {
      for(int i=0;i<svals.length;i++) {
        System.out.println(svals[i]);
        for(int j=0;j<lm.getRowDimension();j++) {
          lm.set(j, i, lm.get(j, i) * svals[i]);
        }
      }     
    }

    for(int tn=0;tn<termNo;tn++) {
      System.out.println(terms.get(tn));
      Map<String,Double> cosines = new HashMap<String,Double>();
      for(int i=0;i<terms.size();i++) {
        double termScore = 0.0;
        double otherTermScore = 0.0;
        double product = 0.0;
       
        for(int j=0;j<svals.length;j++) {
          double tVal = lm.get(tn,j);
          double otVal = lm.get(i,j);
          termScore += tVal * tVal;
          otherTermScore += otVal * otVal;
          product += tVal * otVal;
        }
        double cosine = product / (Math.sqrt(termScore) * Math.sqrt(otherTermScore));
 
View Full Code Here
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
     * @param data данные для анализа
     */
    public static void singularDecomposition(SSAData data) {
        double inclosureMatrix[][] = data.getInclosureMatrix();
        double transp[][] = transpositionMatrix(inclosureMatrix);
        Matrix S = new Matrix(inclosureMatrix).times(new Matrix(transp));
        //int d = new Matrix(inclosureMatrix).rank(); //ранг матрицы вложений
        EigenvalueDecomposition decomposition = new EigenvalueDecomposition(S);
        Matrix eigenvalue = decomposition.getD();   //матрица с собственными значениями
        Matrix eigenvec = decomposition.getV();     //матрица собственных векторов
        List<Double> eigenvalueList = new ArrayList<Double>();
        //формируем набор собственных значений, стоящих на диагонали
        for (int i = 0; i < eigenvalue.getRowDimension(); i++) {
            for (int j = 0; j < eigenvalue.getRowDimension(); j++) {
                if (i == j) {
                    eigenvalueList.add(eigenvalue.get(i, j));
                }
            }
        }
        Comparator comparator = Collections.reverseOrder();
        /*
         * собственные значения должны быть в убывающем порядке, поэтому
         * сортируем их в обратном порядке (изначально значения в возрастающем
         * порядке)
         */
        Collections.sort(eigenvalueList, comparator);
        data.setEigenValueList(eigenvalueList);
        double sumValueList = 0;
        List<Double> percentList;
        List<Double> accruePercentList;
        for (int i = 0; i < data.getEigenValueList().size(); i++) {
            sumValueList = sumValueList + data.getEigenValueList().get(i);
        }
        //формирование процентов собственных чисел и накопленных процентов
        percentList = new ArrayList<Double>();
        accruePercentList = new ArrayList<Double>();
        double accruePercent = 0;
        for (int i = 0; i < data.getEigenValueList().size(); i++) {
            percentList.add(data.getEigenValueList().get(i) / sumValueList * 100);
            accruePercent += percentList.get(i);
            accruePercentList.add(accruePercent);
        }
        data.setAccruePercentList(accruePercentList);
        data.setPercentList(percentList);

        int size = eigenvec.getColumnDimension();
        Matrix V[] = new Matrix[size];
        Matrix U[] = new Matrix[size];
        Matrix X[] = new Matrix[size]; //элементарные матрицы сингулярного разложения
        ArrayList listSeries = new ArrayList();
        for (int j = 0; j < eigenvec.getColumnDimension(); j++) {
            double uVec[][] = new double[size][1];
            ArrayList series = new ArrayList();
            for (int k = 0; k < eigenvec.getRowDimension(); k++) {
                /*
                 * векторы должны соответствовать собственным числа (!), поэтому
                 * начинаем с последнего собственного вектора
                 */
                uVec[k][0] = eigenvec.get(k, eigenvec.getColumnDimension() - j - 1);
                series.add(uVec[k][0]);
            }
            listSeries.add(series);
            U[j] = new Matrix(uVec);
            V[j] = new Matrix(transp).times(U[j]);
        }
        data.setEigenVectors(listSeries);
        for (int i = 0; i < V.length; i++) {
            for (int j = 0; j < V[i].getRowDimension(); j++) {
                for (int k = 0; k < V[i].getColumnDimension(); k++) {
View Full Code Here
130
131
132
133
134
135
136
137
138
139
140
     *
     * @param model модель JList (список групп)
     * @param data данные для анализа
     */
    public static void grouping(DefaultListModel model, SSAData data) {
        Matrix grouX[] = new Matrix[model.getSize()];
        for (int i = 0; i < model.getSize(); i++) {
            GroupListObject obj = (GroupListObject) model.get(i);
            for (int j = 0; j < obj.getGroups().size(); j++) {
                UnselectListObject unselect = (UnselectListObject) obj.getGroups().get(j);
                if (j == 0) {
View Full Code Here
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
    public static void averagedCovariance(SSAData data) {
        double avg;
        double K = data.getTimeSeries().size() - data.getL() + 1; //количество векторов вложения
        List<Double> covarianceList = new ArrayList<Double>();
        double transp[][] = transpositionMatrix(data.getInclosureMatrix());
        Matrix S = new Matrix(data.getInclosureMatrix()).times(new Matrix(transp));
        S = S.times(1.0 / K); //ковариационная матрица
        int size = S.getColumnDimension();
        int N = size + size - 1;
        int n;
        for (int k = 0; k < N; k++) {
            if ((k % 2) == 0) {
                if (k >= 0 && k < size) {
                    avg = 0;
                    n = 0;
                    for (int m = 0; m <= k; m++) {
                        avg += S.get(m, size - 1 - (k - m));
                        n++;
                    }
                    avg = avg / (n);
                    covarianceList.add(avg);
                }
                if (k >= size && k < N) {
                    avg = 0;
                    n = 0;
                    for (int m = k - size + 1; m <= N - size; m++) {
                        avg += S.get(m, size - 1 - (k - m));
                        n++;
                    }
                    avg = avg / (n);
                    covarianceList.add(avg);
                }
View Full Code Here
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
        } else {
          p0 = new double[stations.size()];
          p0[0] = 1; // Assumes that all jobs enters in first station
        }
        try {
          Matrix b = new Matrix(p0, 1);
          Matrix P = new Matrix(buildProbabilityMatrix(stations, input, classKeys.get(cl), res));
          // V = (P-eye(3))' \ (-b') where \ is "mldivide"
          Matrix V = P.minus(Matrix.identity(stations.size(), stations.size())).solveTranspose(b.uminus());
          vis = V.getColumnPackedCopy();
        } catch (Exception e) {
          // Matrix is singular
          res.add("Cannot compute correctly visits for " + output.getClassNames()[cl] + " as" + " network thopology was badly specified");
        }

      } else {
        // Closed class, system is indefinded, so sets visits to reference station to
        // 1 and builds a smaller P matrix
        stations = new Vector<Object>(input.getStationKeysNoSourceSink());
        // Finds reference station
        Object refStat = input.getClassRefStation(classKeys.get(cl));
        if (refStat == null) {
          refStat = stations.get(0);
          res.add("Reference station for " + output.getClassNames()[cl] + " was " + "not set. " + input.getStationName(refStat)
              + " was chosen.");
        }

        // Sets visits to reference station (if allowed) to 1
        if (stationKeys.contains(refStat)) {
          visits[stationKeys.lastIndexOf(refStat)][cl] = 1;
        }
        // Removes reference station from stations vector and computes p0
        stations.remove(refStat);
        double[] p0 = getRoutingProbability(refStat, classKeys.get(cl), input, stations, res);

        try {
          Matrix b = new Matrix(p0, 1);
          Matrix P = new Matrix(buildProbabilityMatrix(stations, input, classKeys.get(cl), res));
          // V = (P-eye(3))' \ (-b') where \ is "mldivide"
          Matrix V = P.minus(Matrix.identity(stations.size(), stations.size())).solveTranspose(b.uminus());
          vis = V.getColumnPackedCopy();
        } catch (Exception e) {
          // Matrix is singular
          res.add("Cannot compute correctly visits for " + output.getClassNames()[cl] + " as" + " network thopology was badly specified");
        }
      }
View Full Code Here
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
  /**
       * computes the variance of the extimator of the variance
       */
  protected double calcExtimatorVar(int K, int polyOrder) {
    Matrix X = new Matrix(K, polyOrder + 1);
    Matrix Xt;
    Matrix square;
    for (int i = 0; i < K; i++) {
      for (int j = 0; j <= polyOrder; j++) {
        X.set(i, j, (Math.pow((4 * ((double) i + 1) - 1) / (2 * (double) K), j)));
      }
    }
    Xt = X.transpose();
    square = Xt.times(X);
    square = square.inverse();
    //System.out.println("calcConst "+square.get(0,0));
    return (.645 * square.get(0, 0));
  }
View Full Code Here
57
58
59
60
61
62
63
64
65
66
67
68
69
70
        { v2x, 0, 0, -v1x, 0, 0, 0, 0, 0 }, { 0, v2y, 0, 0, -v1y, 0, 0, 0, 0 }, { 0, 0, v2z, 0, 0, -v1z, 0, 0, 0 },
        { 0, 0, 0, v3x, 0, 0, -v2x, 0, 0 }, { 0, 0, 0, 0, v3y, 0, 0, -v2y, 0 }, { 0, 0, 0, 0, 0, v3z, 0, 0, -v2z } };

    double[][] arrayb = { { 1 }, { 1 }, { 1 }, { 0 }, { 0 }, { 0 }, { 0 }, { 0 }, { 0 }, };

    Matrix A = new Matrix(arraya);
    Matrix b = new Matrix(arrayb);
    Matrix x = A.solve(b);

    Sector3D out = new Sector3D(x.get(0, 0), x.get(1, 0), x.get(2, 0), x.get(3, 0), x.get(4, 0), x.get(5, 0), x.get(6, 0), x.get(7, 0), x.get(8,
        0), 3, v1, v2, v3);

    return out;
  }
View Full Code Here
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
        { 0, 0, v2z, 0, 0, -v1z }, { 0, 0, 0, 0, 0, 1 } };

    double[][] arrayb = { { 1 }, { 1 }, { 0 }, { 0 }, { 0 }, { 0 } };
    //t double[][] arrayb ={{1,1,1,0,0,0,0,0,0}};

    Matrix A = new Matrix(arraya);
    Matrix b = new Matrix(arrayb);
    Matrix x = A.solve(b);

    BetaVertex vf1 = new BetaVertex(x.get(0, 0), x.get(1, 0), x.get(2, 0));
    BetaVertex vf2 = new BetaVertex(x.get(3, 0), x.get(4, 0), x.get(5, 0));

    //BetaVertex v3 = new BetaVertex();
    //BetaVertex v4 = new BetaVertex();

    Sector3D out = new Sector3D(vf1, vf2, v1, v2, 2, st1, st2);
View Full Code Here
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
        { 0, 0, v2z, 0, 0, -v1z }, { 0, 0, 0, 1, 0, 0 } };

    double[][] arrayb = { { 1 }, { 1 }, { 0 }, { 0 }, { 0 }, { 0 } };
    //t double[][] arrayb ={{1,1,1,0,0,0,0,0,0}};

    Matrix A = new Matrix(arraya);
    Matrix b = new Matrix(arrayb);
    Matrix x = A.solve(b);

    BetaVertex vf1 = new BetaVertex(x.get(0, 0), x.get(1, 0), x.get(2, 0));
    BetaVertex vf2 = new BetaVertex(x.get(3, 0), x.get(4, 0), x.get(5, 0));

    //BetaVertex v3 = new BetaVertex();
    //BetaVertex v4 = new BetaVertex();

    Sector3D out = new Sector3D(vf1, vf2, v1, v2, 2, st1, st2);
View Full Code Here
TOP

Related Classes of Jama.Matrix

  • br.edu.utfpr.cm.JGitMinerWeb.util.matrix.DoubleDenseMatrix
  • ca.nengo.dynamics.impl.CanonicalModel
  • ca.nengo.dynamics.impl.ImpulseIntegral
  • ca.nengo.math.impl.PolynomialCurveFitter
  • ca.nengo.math.impl.WeightedCostApproximator
  • ca.nengo.math.impl.WeightedCostApproximatorTest
  • ca.nengo.model.nef.impl.AdditiveGaussianExpressModel
  • ca.nengo.model.nef.impl.DecodableEnsembleImpl
  • caterpillarssa.SpectrumAnalysis
  • com.heatonresearch.aifh.regression.TrainLeastSquares
    • 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.