public double[] Run(double[] Xobdata, double[] obdatay, double[] para //,double noise
)
{
Parameters.Verbosity = VerbosityLevels.OnlyCritical;
//Debug.Listeners.Add(new TextWriterTraceListener(Console.Out));
var optMethod = new NelderMead();
optMethod.Add(new GoldenSection(1e-8, 0.001));
optMethod.Add(new DeltaFConvergence(1e-7));
var opt = new OneDObjectiveFunc(Xobdata, obdatay, para);
optMethod.Add(opt);
double[] xStar;
var xInit = para;
try
{
var fstar_opt = optMethod.Run(out xStar, xInit);
var output = new double[4] {
xStar[0], xStar[1], xStar[2], fstar_opt
};
return(output);
}
catch
{
var output = new double[4] {
100, 1000, 1000, 100000
};
return(output);
}
// return xStar;
}
public double[] Run(double[,] Xobdata, double[] obdatay, double[] para)
{
Parameters.Verbosity = VerbosityLevels.OnlyCritical;
// this next line is to set the Debug statements from OOOT to the Console.
//Debug.Listeners.Add(new TextWriterTraceListener(Console.Out));
// var optMethod = new GradientBasedOptimization();
// var Y = ThreeDinput.GetfVactor(obdatay);
// var optMethod = new GradientBasedOptimization();
// var optMethod = new GeneralizedReducedGradientActiveSet();
// var optMethod = new HillClimbing();
var optMethod = new NelderMead();
// optMethod.Add(new StochasticNeighborGenerator);
// var optMethod = new NelderMead();
// optMethod.Add(new CyclicCoordinates());
// optMethod.Add(new CyclicCoordinates());
optMethod.Add(new GoldenSection(1e-7, 0.01));
// optMethod.Add(new DeltaXConvergence(1e-10));
optMethod.Add(new DeltaFConvergence(1e-6));
// optMethod.Add(new MaxSpanInPopulationConvergence(1e-3));
//optMethod.Add(new inequalityWithConstant())
var opt = new OneDObjectiveFunc(Xobdata, obdatay, para);
optMethod.Add(opt);
// optMethod.Add(new Inequality(opt, Xobdata, obdatay));
// optMethod.Add(new OptimizationToolbox.greaterThanConstant { constant = 0.0, index = 0 });
//optMethod.Add(new OptimizationToolbox.greaterThanConstant { constant = 0.0, index = 1 });
//optMethod.Add(new OptimizationToolbox.greaterThanConstant { constant = 0.0, index = 2 });
//optMethod.Add(new OptimizationToolbox.lessThanConstant() { constant = 0.30, index = 2 });
//optMethod.Add(new OptimizationToolbox.squaredExteriorPenalty(optMethod, 1.0));
// var p = new double[2] { para[0], para[1] };
double[] xStar;
var xInit = para;
try
{
var fstar_opt = optMethod.Run(out xStar, xInit);
var output = new double[5] {
xStar[0], xStar[1], xStar[2], xStar[3], fstar_opt
};
// var output = new double[4] { xStar[0], xStar[1], xStar[2], fstar_opt };//1d
return(output);
}
catch
{
// var output = new double[4] { 1000, 1000, 1000, 100000000 };//1d
var output = new double[5] {
100, 1000, 1000, 1000, 100000000
};
return(output);
}
// return xStar;
}
public Inequality(OneDObjectiveFunc opt, double[] Xobdata, double[] obdatay)
{
this.Xobdata = Xobdata;
this.obdatay = obdatay;
Opt = opt;
}