Package org.apache.mahout.math.solver

Examples of org.apache.mahout.math.solver.EigenDecomposition

144
145
146
147
148
149
150
151
152
153
154
155
156
157
    }
    startTime(TimingSection.TRIDIAG_DECOMP);

    log.info("Lanczos iteration complete - now to diagonalize the tri-diagonal auxiliary matrix.");
    // at this point, have tridiag all filled out, and basis is all filled out, and orthonormalized
    EigenDecomposition decomp = new EigenDecomposition(triDiag);

    Matrix eigenVects = decomp.getV();
    Vector eigenVals = decomp.getRealEigenvalues();
    endTime(TimingSection.TRIDIAG_DECOMP);
    startTime(TimingSection.FINAL_EIGEN_CREATE);
    for (int row = 0; row < i; row++) {
      Vector realEigen = null;
      // the eigenvectors live as columns of V, in reverse order.  Weird but true.
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
    int desiredRank = 80;
    LanczosState state = new LanczosState(m, desiredRank, initialVector);
    // set initial vector?
    solver.solve(state, desiredRank, true);

    EigenDecomposition decomposition = new EigenDecomposition(m);
    Vector eigenvalues = decomposition.getRealEigenvalues();

    float fractionOfEigensExpectedGood = 0.6f;
    for (int i = 0; i < fractionOfEigensExpectedGood * desiredRank; i++) {
      double s = state.getSingularValue(desiredRank - i - 1);
      double e = eigenvalues.get(eigenvalues.size() - i - 1);
      log.info("{} : L = {}, E = {}", i, s, e);
      assertTrue("Singular value differs from eigenvalue", Math.abs((s-e)/e) < ERROR_TOLERANCE);
      Vector v = state.getRightSingularVector(i);
      Vector v2 = decomposition.getV().viewColumn(eigenvalues.size() - i - 1);
      double error = 1 - Math.abs(v.dot(v2)/(v.norm(2) * v2.norm(2)));
      log.info("error: {}", error);
      assertTrue(i + ": 1 - cosAngle = " + error, error < ERROR_TOLERANCE);
    }
  }
View Full Code Here
515
516
517
518
519
520
521
522
523
524
525
526
527
528
        outerSq.assign(Functions.mult(xisquaredlen));
        bbtSquare.assign(outerSq, Functions.PLUS);

      }

      EigenDecomposition eigen = new EigenDecomposition(bbtSquare);

      Matrix uHat = eigen.getV();
      svalues = eigen.getRealEigenvalues().clone();

      svalues.assign(Functions.SQRT);

      // save/redistribute UHat
      fs.mkdirs(uHatPath);
View Full Code Here
50
51
52
53
54
55
56
57
58
59
60
61
62
63
    Matrix b = new DenseMatrix(a.numRows(), a.numCols());
    b.assign(a);

    assertEquals(0, a.minus(b).aggregate(Functions.PLUS, Functions.ABS), 1.0e-10);

    EigenDecomposition edA = new EigenDecomposition(a);
    EigenDecomposition edB = new EigenDecomposition(b);

    System.out.println(edA.getV());

    assertEquals(0, edA.getV().minus(edB.getV()).aggregate(Functions.PLUS, Functions.ABS), 1.0e-10);
    assertEquals(0, edA.getRealEigenvalues().minus(edA.getRealEigenvalues()).aggregate(Functions.PLUS, Functions.ABS), 1.0e-10);

  }
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
    int desiredRank = 80;
    LanczosState state = new LanczosState(m, desiredRank, initialVector);
    // set initial vector?
    solver.solve(state, desiredRank, true);

    EigenDecomposition decomposition = new EigenDecomposition(m);
    Vector eigenvalues = decomposition.getRealEigenvalues();

    float fractionOfEigensExpectedGood = 0.6f;
    for (int i = 0; i < fractionOfEigensExpectedGood * desiredRank; i++) {
      double s = state.getSingularValue(i);
      double e = eigenvalues.get(i);
      log.info("{} : L = {}, E = {}", i, s, e);
      assertTrue("Singular value differs from eigenvalue", Math.abs((s-e)/e) < ERROR_TOLERANCE);
      Vector v = state.getRightSingularVector(i);
      Vector v2 = decomposition.getV().viewColumn(i);
      double error = 1 - Math.abs(v.dot(v2)/(v.norm(2) * v2.norm(2)));
      log.info("error: {}", error);
      assertTrue(i + ": 1 - cosAngle = " + error, error < ERROR_TOLERANCE);
    }
  }
View Full Code Here
144
145
146
147
148
149
150
151
152
153
154
155
156
157
    }
    startTime(TimingSection.TRIDIAG_DECOMP);

    log.info("Lanczos iteration complete - now to diagonalize the tri-diagonal auxiliary matrix.");
    // at this point, have tridiag all filled out, and basis is all filled out, and orthonormalized
    EigenDecomposition decomp = new EigenDecomposition(triDiag);

    Matrix eigenVects = decomp.getV();
    Vector eigenVals = decomp.getRealEigenvalues();
    endTime(TimingSection.TRIDIAG_DECOMP);
    startTime(TimingSection.FINAL_EIGEN_CREATE);
    for (int row = 0; row < i; row++) {
      Vector realEigen = null;
View Full Code Here
136
137
138
139
140
141
142
143
144
145
146
147
148
149
    }
    startTime(TimingSection.TRIDIAG_DECOMP);

    log.info("Lanczos iteration complete - now to diagonalize the tri-diagonal auxiliary matrix.");
    // at this point, have tridiag all filled out, and basis is all filled out, and orthonormalized
    EigenDecomposition decomp = new EigenDecomposition(triDiag);

    Matrix eigenVects = decomp.getV();
    Vector eigenVals = decomp.getRealEigenvalues();
    endTime(TimingSection.TRIDIAG_DECOMP);
    startTime(TimingSection.FINAL_EIGEN_CREATE);
    for (int row = 0; row < i; row++) {
      Vector realEigen = null;
      // the eigenvectors live as columns of V, in reverse order.  Weird but true.
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
    int desiredRank = 80;
    LanczosState state = new LanczosState(m, desiredRank, initialVector);
    // set initial vector?
    solver.solve(state, desiredRank, true);

    EigenDecomposition decomposition = new EigenDecomposition(m);
    Vector eigenvalues = decomposition.getRealEigenvalues();

    float fractionOfEigensExpectedGood = 0.6f;
    for (int i = 0; i < fractionOfEigensExpectedGood * desiredRank; i++) {
      double s = state.getSingularValue(desiredRank - i - 1);
      double e = eigenvalues.get(eigenvalues.size() - i - 1);
      log.info("{} : L = {}, E = {}", new Object[] {i, s, e});
      assertTrue("Singular value differs from eigenvalue", Math.abs((s-e)/e) < ERROR_TOLERANCE);
      Vector v = state.getRightSingularVector(i);
      Vector v2 = decomposition.getV().viewColumn(eigenvalues.size() - i - 1);
      double error = 1 - Math.abs(v.dot(v2)/(v.norm(2) * v2.norm(2)));
      log.info("error: {}", error);
      assertTrue(i + ": 1 - cosAngle = " + error, error < ERROR_TOLERANCE);
    }
  }
View Full Code Here
144
145
146
147
148
149
150
151
152
153
154
155
156
157
    }
    startTime(TimingSection.TRIDIAG_DECOMP);

    log.info("Lanczos iteration complete - now to diagonalize the tri-diagonal auxiliary matrix.");
    // at this point, have tridiag all filled out, and basis is all filled out, and orthonormalized
    EigenDecomposition decomp = new EigenDecomposition(triDiag);

    Matrix eigenVects = decomp.getV();
    Vector eigenVals = decomp.getRealEigenvalues();
    endTime(TimingSection.TRIDIAG_DECOMP);
    startTime(TimingSection.FINAL_EIGEN_CREATE);
    for (int row = 0; row < i; row++) {
      Vector realEigen = null;
View Full Code Here
TOP

Related Classes of org.apache.mahout.math.solver.EigenDecomposition

  • org.apache.mahout.math.decomposer.lanczos.LanczosSolver
  • org.apache.mahout.math.decomposer.lanczos.TestLanczosSolver
  • org.apache.mahout.math.DenseMatrix
  • org.apache.mahout.math.DenseSymmetricTest
  • org.apache.mahout.math.DenseVector
  • org.apache.mahout.math.hadoop.stochasticsvd.SSVDSolver
  • org.apache.mahout.math.Matrix
  • org.apache.mahout.math.Vector
    • 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.