Class RungeKuttaIntegrator

    • Constructor Detail

      • RungeKuttaIntegrator

        protected RungeKuttaIntegrator​(String name,
                                       double[] c,
                                       double[][] a,
                                       double[] b,
                                       org.apache.commons.math4.legacy.ode.nonstiff.RungeKuttaStepInterpolator prototype,
                                       double step)
        Simple constructor. Build a Runge-Kutta integrator with the given step. The default step handler does nothing.
        name - name of the method
        c - time steps from Butcher array (without the first zero)
        a - internal weights from Butcher array (without the first empty row)
        b - propagation weights for the high order method from Butcher array
        prototype - prototype of the step interpolator to use
        step - integration step
    • Method Detail

      • singleStep

        public double[] singleStep​(FirstOrderDifferentialEquations equations,
                                   double t0,
                                   double[] y0,
                                   double t)
        Fast computation of a single step of ODE integration.

        This method is intended for the limited use case of very fast computation of only one step without using any of the rich features of general integrators that may take some time to set up (i.e. no step handlers, no events handlers, no additional states, no interpolators, no error control, no evaluations count, no sanity checks ...). It handles the strict minimum of computation, so it can be embedded in outer loops.

        This method is not used at all by the integrate(ExpandableStatefulODE, double) method. It also completely ignores the step set at construction time, and uses only a single step to go from t0 to t.

        As this method does not use any of the state-dependent features of the integrator, it should be reasonably thread-safe if and only if the provided differential equations are themselves thread-safe.

        equations - differential equations to integrate
        t0 - initial time
        y0 - initial value of the state vector at t0
        t - target time for the integration (can be set to a value smaller than t0 for backward integration)
        state vector at t