private void Initialise(Func <TInput, Func <double, double, ROutput> > func, TInput input, Func <TInput, ROutput, double> error, IProgress <KeyValuePair <int, Result> > progress, CancellationToken token, int count)
{
//function to optimise
Func <double, double, double> fc = (a, b) => error(input, func(input)(a, b));
// wrap-up in class
var of = new UserFunction2D(fc);
// build default genetic algorithm class
var w = AccordWrapFactory.BuildDefaultMinimisation(of);
if (progress == null & token == default(CancellationToken))
{
w.RunMultipleEpochs(count);
}
else
{
w.RunMultipleEpochs(count, progress);
}
}
// public static BackgroundWorkerObservable<ChromosomeWrapper> Create<TInput, ROutput>(Func<double, double, ROutput> func, TInput input, Func<TInput, ROutput, double> error, int count = 10)
// { //function to optimise
// Func<double, double, double> fc = (a, b) => error(input,func(a, b));
// // wrap-up in class
// var of = new UserFunction2D(fc);
// var main = BuildGeneticProgram(of);
// return new BackgroundWorkerObservable<ChromosomeWrapper>(main, Observable.Repeat((bool?)true, 1), Observable.Repeat(count, 1), Observable.Repeat(0, 1));
// }
// //public static BackgroundWorkerObservable<Accord.Genetic.IChromosome> Create2<TInput, ROutput>(Func<double, double, ROutput> func, TInput input, Func<TInput, ROutput, double> error, int count = 10)
// //{
// // var main = BuildFunction2(func, input, error);
// // return new BackgroundWorkerObservable<Accord.Genetic.IChromosome>(main, null, Observable.Repeat((bool?)true, 1), Observable.Repeat(count, 1), Observable.Repeat(0, 1));
// //}
// public static BackgroundWorkerObservableQueue<ChromosomeWrapper> Create<TInput, ROutput>(IObservable<Tuple<TInput, Func<double, double, ROutput>>> funcinput, Func<TInput, ROutput, double> error, int count = 10)
// {
// var fc2 = funcinput.Select(_ =>
// {
// //function to optimise
// Func<double, double, double> fc = (a, b) => error(_.Item1,_.Item2(a, b));
// // wrap-up in class
// var of = new UserFunction2D(fc);
// var main = BuildGeneticProgram(of);
// return main;
// });
// return new BackgroundWorkerObservableQueue<ChromosomeWrapper>(fc2, Observable.Repeat((bool?)true, 1), Observable.Repeat(count, 1), Observable.Repeat(0, 1), Observable.Repeat(1, 1));
// }
//public static BackgroundWorkerObservableQueue<ChromosomeWrapper> Create2<TInput, ROutput>(IObservable<Tuple<TInput, Func<double, double, ROutput>>> funcinput, Func<TInput, ROutput, double> error, int count = 10)
//{
// var fc2 = funcinput.Select(_ =>
// {
// //function to optimise
// //function to optimise
// Func<double, double, double> fc = (a, b) => error(_.Item1, _.Item2(a, b));
// // wrap-up in class
// var of = new UserFunction2D(fc);
// return BuildFunction2<TInput, ROutput>(of);
// });
// var c =new ChromosomeWrapper { Chromosome= new BinaryChromosome(64) };
// return new BackgroundWorkerObservableQueue<ChromosomeWrapper>(fc2, c,Observable.Repeat((bool?)true, 1), Observable.Repeat(count, 1), Observable.Repeat(0, 1), Observable.Repeat(1, 1));
//}
private static Func <int, ChromosomeWrapper> BuildGeneticProgram(OptimizationFunction2D of)
{
// build default genetic algorithm class to minimise output
var w = AccordWrapFactory.BuildDefaultMinimisation(of);
//cyclical method to run
Func <int, ChromosomeWrapper> main = (i) =>
{
w.Population.RunEpoch();
// var _ = w.Evaluate();
//var score = _.Output[0, 1];
//double[] xy = _.BestSolution.Split(' ').Select(d => double.Parse(d)).ToArray();
w.Population.RunEpoch();
var x = w.Evaluate();
return(new ChromosomeWrapper {
Chromosome = w.Population.BestChromosome, Iteration = i, Score = x.Output[0, 1], Input = x.BestSolution.Split(' ').Select(d => double.Parse(d)).ToArray()
});
};
return(main);
}
