public class FirstOrderConverter extends Object implements FirstOrderDifferentialEquations
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.
One should be aware that the data is duplicated during the
transformation process and that for each call to
computeDerivatives, this wrapper does copy 4n
scalars : 2n before the call to
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.
|Constructor and Description|
|Modifier and Type||Method and Description|
Get the current time derivative of the state vector.
Get the dimension of the problem.
public int getDimension()
The dimension of the first order problem is twice the dimension of the underlying second order problem.
public void computeDerivatives(double t, double y, double yDot)
Copyright © 2003–2014 The Apache Software Foundation. All rights reserved.