Example usage for org.apache.commons.math3.fitting.leastsquares MultivariateJacobianFunction MultivariateJacobianFunction

Introduction

In this page you can find the example usage for org.apache.commons.math3.fitting.leastsquares MultivariateJacobianFunction MultivariateJacobianFunction.

Prototype

MultivariateJacobianFunction

Source Link

Usage

From source file:org.orekit.propagation.conversion.AbstractPropagatorConverter.java

/** Find the propagator that minimize the mean square error for a sample of {@link SpacecraftState states}.
 * @param initial initial estimation parameters (position, velocity, free parameters)
 * @return fitted parameters//from  w  ww.  jav  a 2  s . com
 * @exception OrekitException if propagator cannot be adapted
 * @exception MaxCountExceededException if maximal number of iterations is exceeded
 */
private double[] fit(final double[] initial) throws OrekitException, MaxCountExceededException {

    final MultivariateVectorFunction f = getObjectiveFunction();
    final MultivariateMatrixFunction jac = getObjectiveFunctionJacobian();
    final MultivariateJacobianFunction fJac = new MultivariateJacobianFunction() {
        /** {@inheritDoc} */
        @Override
        public Pair<RealVector, RealMatrix> value(final RealVector point) {
            final double[] p = point.toArray();
            return new Pair<RealVector, RealMatrix>(MatrixUtils.createRealVector(f.value(p)),
                    MatrixUtils.createRealMatrix(jac.value(p)));
        }
    };
    final LeastSquaresProblem problem = new LeastSquaresBuilder().maxIterations(maxIterations)
            .maxEvaluations(Integer.MAX_VALUE).model(fJac).target(target).weight(new DiagonalMatrix(weight))
            .start(initial).checker(checker).build();
    optimum = optimizer.optimize(problem);
    return optimum.getPoint().toArray();
}

---

From source file:uk.ac.diamond.scisoft.analysis.optimize.ApacheOptimizer.java

public MultivariateJacobianFunction createJacobianFunction() {
    final int size = coords[0].getSize();
    final AFunction afn;
    final CoordinatesIterator it;
    final DoubleDataset vd, pvd;
    final double[][] dm = new double[size][n];
    if (function instanceof AFunction) {
        afn = (AFunction) function;// w  w w  . j av a  2  s . c  o m
        it = afn.getIterator(coords);
        vd = (DoubleDataset) DatasetFactory.zeros(coords[0].getShapeRef(), Dataset.FLOAT64);
        pvd = vd.clone();
    } else {
        afn = null;
        it = null;
        vd = null;
        pvd = null;
    }

    MultivariateJacobianFunction f = new MultivariateJacobianFunction() {
        @SuppressWarnings("null")
        @Override
        public Pair<RealVector, RealMatrix> value(RealVector point) {
            IMonitor monitor = function.getMonitor();
            if (monitor != null && monitor.isCancelled()) {
                throw new IllegalMonitorStateException("Monitor cancelled");
            }

            if (point instanceof ArrayRealVector) {
                setParameterValues(((ArrayRealVector) point).getDataRef());
            } else {
                setParameterValues(point.toArray());
            }
            final double[] dv;
            if (afn != null) {
                dv = vd.getData();
                AFunction afn = (AFunction) function;
                afn.fillWithValues(vd, it);
                double[] pd = pvd.getData();
                for (int i = 0; i < n; i++) { // assuming number of parameters is less than number of coordinates
                    IParameter p = params.get(i);
                    afn.fillWithPartialDerivativeValues(p, pvd, it);
                    for (int j = 0; j < size; j++) {
                        dm[j][i] = pd[j];
                    }
                }
            } else {
                dv = calculateValues().getData();
                for (int i = 0; i < n; i++) {
                    IParameter p = params.get(i);
                    DoubleDataset dp = (DoubleDataset) DatasetUtils
                            .cast(function.calculatePartialDerivativeValues(p, coords), Dataset.FLOAT64);
                    double[] pd = dp.getData();
                    for (int j = 0; j < size; j++) {
                        dm[j][i] = pd[j];
                    }
                }
            }
            return new Pair<RealVector, RealMatrix>(new ArrayRealVector(dv, false),
                    new Array2DRowRealMatrix(dm, false));
        }
    };

    return f;
}

---