FirstOrderConverter.java
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* https://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
/*
* This is not the original file distributed by the Apache Software Foundation
* It has been modified by the Hipparchus project
*/
package org.hipparchus.ode;
/** This class converts second order differential equations to first
* order ones.
*
* <p>This class is a wrapper around a {@link SecondOrderODE} which
* allow to use a {@link ODEIntegrator} to integrate it.</p>
*
* <p>The transformation is done by changing the n dimension state
* vector to a 2n dimension vector, where the first n components are
* the initial state variables and the n last components are their
* first time derivative. The first time derivative of this state
* vector then really contains both the first and second time
* derivative of the initial state vector, which can be handled by the
* underlying second order equations set.</p>
*
* <p>One should be aware that the data is duplicated during the
* transformation process and that for each call to {@link
* #computeDerivatives computeDerivatives}, this wrapper does copy 4n
* scalars : 2n before the call to {@link
* SecondOrderODE#computeSecondDerivatives
* computeSecondDerivatives} in order to dispatch the y state vector
* into z and zDot, and 2n after the call to gather zDot and zDDot
* into yDot. Since the underlying problem by itself perhaps also
* needs to copy data and dispatch the arrays into domain objects,
* this has an impact on both memory and CPU usage. The only way to
* avoid this duplication is to perform the transformation at the
* problem level, i.e. to implement the problem as a first order one
* and then avoid using this class.</p>
*
* @see ODEIntegrator
* @see OrdinaryDifferentialEquation
* @see SecondOrderODE
*/
public class FirstOrderConverter implements OrdinaryDifferentialEquation {
/** Underlying second order equations set. */
private final SecondOrderODE equations;
/** second order problem dimension. */
private final int dimension;
/** Simple constructor.
* Build a converter around a second order equations set.
* @param equations second order equations set to convert
*/
public FirstOrderConverter (final SecondOrderODE equations) {
this.equations = equations;
dimension = equations.getDimension();
}
/** {@inheritDoc}
* <p>The dimension of the first order problem is twice the
* dimension of the underlying second order problem.</p>
* @return dimension of the problem
*/
@Override
public int getDimension() {
return 2 * dimension;
}
/** {@inheritDoc} */
@Override
public double[] computeDerivatives(final double t, final double[] y) {
final double[] yDot = new double[y.length];
// split the state vector in two
final double[] z = new double[dimension];
final double[] zDot = new double[dimension];
System.arraycopy(y, 0, z, 0, dimension);
System.arraycopy(y, dimension, zDot, 0, dimension);
// apply the underlying equations set
final double[] zDDot = equations.computeSecondDerivatives(t, z, zDot);
// build the result state derivative
System.arraycopy(zDot, 0, yDot, 0, dimension);
System.arraycopy(zDDot, 0, yDot, dimension, dimension);
return yDot;
}
}