public static void Main(string[] args)
{
/*
* The context of this tutorial is that you have an existing time stepping simulation
* model.
*/
IClonableObjectiveEvaluator <IHyperCube <double> > evaluator;
IEvolutionEngine <IHyperCube <double> > uniformRandomSampling;
var simulation = SimulationFactory.CreateAwbmSimulation();
var data = DataHandling.GetSampleClimate();
SimulationFactory.SetSampleSimulation(simulation, data);
int from = simulation.GetStart();
int to = simulation.GetEnd();
evaluator = BuildUrsEvaluator(simulation, data.Runoff, from, to);
var paramSpace = CreateFeasibleAwbmParameterSpace(simulation);
uniformRandomSampling = new UniformRandomSampling <IHyperCube <double> >(evaluator, new BasicRngFactory(0), paramSpace, 3000);
var ursResults = uniformRandomSampling.Evolve();
Console.WriteLine(MetaheuristicsHelper.GetHumanReadable(ursResults));
}
static void Main(string[] args)
{
/*
* This program primarily demonstrates how to wrap and calibrate a model written
* in a native language. While technological aspects of what it demonstrates
* are not inherently limited nor specific to optimisation, this is
* a use case often encountered (e.g. to calibrate existing models written in
* C++ , C, or Fortran)
*/
IClonableObjectiveEvaluator <IHyperCube <double> > evaluator;
IEvolutionEngine <IHyperCube <double> > uniformRandomSampling;
NativeModelWrapper.AwbmWrapper.PermitMultiThreading = true;
using (var simulation = new AwbmWrapper())
{
var data = DataHandling.GetSampleClimate();
SimulationFactory.SetSampleSimulation(simulation, data);
int from = simulation.GetStart();
int to = simulation.GetEnd();
evaluator = AWBM_URS.MainClass.BuildUrsEvaluator(simulation, data.Runoff, from, to);
var paramSpace = AWBM_URS.MainClass.CreateFeasibleAwbmParameterSpace(simulation);
uniformRandomSampling = new UniformRandomSampling <IHyperCube <double> >(evaluator, new BasicRngFactory(0), paramSpace, 3000);
var ursResults = uniformRandomSampling.Evolve();
Console.WriteLine(MetaheuristicsHelper.GetHumanReadable(ursResults));
}
}