Example usage for org.apache.commons.math3.linear CholeskyDecomposition getSolver

Introduction

In this page you can find the example usage for org.apache.commons.math3.linear CholeskyDecomposition getSolver.

Prototype

public DecompositionSolver getSolver() 

Source Link

Document

Get a solver for finding the A × X = B solution in least square sense.

Usage

From source file:org.ojalgo.benchmark.lab.library.ACM.java

@Override
public MutatingUnaryMatrixOperation<RealMatrix, Array2DRowRealMatrix> getOperationInvert(final int dim,
        final boolean spd) {
    if (spd) {//from   w  w  w. jav a 2s  .  c  o m
        return (matA, result) -> {
            final CholeskyDecomposition chol = new CholeskyDecomposition(matA);
            this.copy(chol.getSolver().getInverse(), result);
        };
    } else {
        return (matA, result) -> {
            final LUDecomposition lu = new LUDecomposition(matA);
            this.copy(lu.getSolver().getInverse(), result);
        };
    }
}

---

From source file:org.rhwlab.variationalbayesian.GaussianMixture.java

public void init(int K) {
    this.K = K;/*from  ww  w . ja v a  2  s.  co m*/
    this.N = new double[K];
    this.m = new RealVector[K];
    inModel = new boolean[K];
    for (int k = 0; k < K; ++k) {
        inModel[k] = true;
    }
    Dln2 = X.getD() * Math.log(2.0);
    Dln2Pi2 = X.getD() * Math.log(2.0 * Math.PI) / 2.0;

    r = new Array2DRowRealMatrix(X.getN(), K);
    if (X instanceof ClusteredDataSource) {
        ClusteredDataSource clSource = (ClusteredDataSource) X;
        for (int n = 0; n < X.getN(); ++n) {
            int cluster = clSource.getCluster(n);
            for (int k = 0; k < K; ++k) {
                if (k == cluster) {
                    r.setEntry(n, k, 1.0);
                } else {
                    r.setEntry(n, k, 0.0);
                }
            }
        }
    } else {
        for (int n = 0; n < X.getN(); ++n) {
            double[] dir = dirichlet(0.001, K);
            for (int k = 0; k < K; ++k) {
                r.setEntry(n, k, dir[k]);
            }
        }
    }
    calculateN();

    if (mu0 == null) {
        mu0 = new ArrayRealVector(X.getD(), 0.0);
    }

    if (W0 == null) {
        W0 = MatrixUtils.createRealIdentityMatrix(X.getD());
        W0 = W0.scalarMultiply(.01);
    }
    CholeskyDecomposition de = new CholeskyDecomposition(W0);
    W0inverse = de.getSolver().getInverse();
    W0det = de.getDeterminant();
    W = new RealMatrix[K];
    Winverse = new RealMatrix[K];
    detW = new double[K];
    detWinv = new double[K];
    /*        
            for (int k=0 ; k<K ; ++k){
    W[k] = W0.copy();
    W[k].scalarMultiply(0.0000000001);
    detW[k] = de.getDeterminant();
            }
     */
    alpha = new double[K];
    beta = new double[K];
    lnLambdaTilde = new double[K];
    nu = new double[K];
    xBar = new RealVector[K];
    for (int k = 0; k < K; ++k) {
        xBar[k] = new ArrayRealVector(X.getD());
        //            nu[k] = nu0 + N[k];
    }
    lnPi = new double[K];
    for (int k = 0; k < K; ++k) {
        //            lnPi[k] = Math.log(N[k]/X.getN());
    }

    S = new RealMatrix[K];
    for (int k = 0; k < K; ++k) {
        S[k] = new Array2DRowRealMatrix(X.getD(), X.getD());
    }

    lnRho = new double[K];
    a0 = new double[K];
    for (int k = 0; k < K; ++k) {
        a0[k] = alpha0;
    }
    C0 = lnC(a0);
    //        alphaBetaNu();
    //        lambdaTilde();
}

---

From source file:org.rhwlab.variationalbayesian.SuperVoxelGaussianMixture.java

public void init(int K) {
    this.K = K;//from w w  w.j  a v a2s  .  co  m
    Dln2 = X.getD() * Math.log(2.0);
    Dln2Pi2 = X.getD() * Math.log(2.0 * Math.PI) / 2.0;

    r = new Array2DRowRealMatrix(X.getN(), K);
    if (mu0 == null) {
        mu0 = new ArrayRealVector(X.getD(), 0.0);
    }
    // initialize the means 

    m = new RealVector[K];
    N = new double[K];
    double Nk = (double) X.getN() / (double) K;
    for (int k = 0; k < K; ++k) {
        m[k] = X.get(k).centroid;
        N[k] = Nk;
    }

    if (W0 == null) {
        W0 = MatrixUtils.createRealIdentityMatrix(X.getD());
        W0 = W0.scalarMultiply(.01);
    }
    CholeskyDecomposition de = new CholeskyDecomposition(W0);
    W0inverse = de.getSolver().getInverse();
    W0det = de.getDeterminant();
    W = new RealMatrix[K];
    detW = new double[K];
    for (int k = 0; k < K; ++k) {
        W[k] = W0.copy();
        detW[k] = de.getDeterminant();
    }

    alpha = new double[K];
    beta = new double[K];
    lnLambdaTilde = new double[K];
    nu = new double[K];
    xBar = new RealVector[K];
    for (int k = 0; k < K; ++k) {
        xBar[k] = new ArrayRealVector(X.getD());
        nu[k] = nu0 + N[k];
    }
    lnPi = new double[K];
    lnPi();
    /*        
            for (k=0 ; k<K ; ++k){
    lnPi[k] = Math.log(N[k]/X.getN());
            }
    */
    S = new RealMatrix[K];
    for (int k = 0; k < K; ++k) {
        S[k] = new Array2DRowRealMatrix(X.getD(), X.getD());
    }

    lnRho = new double[K];
    a0 = new double[K];
    for (int k = 0; k < K; ++k) {
        a0[k] = alpha0;
    }
    C0 = lnC(a0);
    alphaBetaNuLambda();
}

---

• «
• 1
• 2
• »