static void Main(string[] args)
{
#region Variables
//string variablesFileName = @"C:/Users/Ivan/Repositiries/AquatoxBasedOptimization/JupyterNotebooks/variables.xlsx";
string variablesFileName = @"C:/Users/ivanry/Documents/AquatoxBasedOptimization/JupyterNotebooks/variables_20.xlsx";
AquatoxVariablesFileReader variablesReader = new AquatoxVariablesFileReader();
Dictionary <string, AquatoxParameterToTune> modelVariables = variablesReader.ReadParameters(variablesFileName);
#endregion Variables
#region Generating
double defaultMinDelta = 10;
double defaultMaxDelta = 10;
List <(double From, double To)> generatingBoundaries = modelVariables.Select(pair => pair.Value.MakeGenerationBoundaries(defaultMinDelta, defaultMaxDelta)).ToList();
#endregion Generating
#region Constrains
double softConstrainWight = 1000;
HardAndSoftConstrain[] constrains = modelVariables.Select(pair => pair.Value.MakeConstrain(softConstrainWight)).ToArray();
HardAndSoftConstrainerParameters hardAndSoftConstrainerParameters = new HardAndSoftConstrainerParameters(constrains);
HardAndSoftConstrainer constrainer = new HardAndSoftConstrainer(hardAndSoftConstrainerParameters);
#endregion Constrains
#region Model
IOutputVariablesReader outputVariablesReader = new OutputVariablesReaderFromExcel();
Dictionary <string, int> variablesAndIndices = outputVariablesReader.Read();
IAquatoxOutputFileProcessor outputFileProcessor = new AquatoxOutputFileProcessor(variablesAndIndices);
AquatoxModelParameters modelParameters = new AquatoxModelParameters();
//modelParameters.InputParameters = new Dictionary<string, string>() { { "par1", "_param1_" }, { "par2", "_param2_" }, { "par3", "_param3_" }, { "par4", "_param4_" }, { "par5", "_param5_" } };
modelParameters.InputParameters = Enumerable.Range(1, modelVariables.Count).ToDictionary(i => $"par{i}", i => $"_param{i}_");
AquatoxModel model = new AquatoxModel(outputFileProcessor);
model.SetParameters(modelParameters);
#endregion Model
int dimension = modelParameters.InputParameters.Count;
#region Problem
PredefinedDistanceCalculator distanceCalculator = new PredefinedDistanceCalculator();
OutputObservationsReaderFromExcel outputObservationsReader = new OutputObservationsReaderFromExcel();
var observations = outputObservationsReader.ReadOutputVariableObservations();
AquatoxParametersTuningProblem tuningProblem = new AquatoxParametersTuningProblem(dimension, constrainer);
tuningProblem.SetDistanceCalculator(distanceCalculator);
tuningProblem.SetObservations(observations);
tuningProblem.SetModel(model);
#endregion Problem
#region Algorithm
BestAlternativeHistoryMaker bestAltHistMaker = new BestAlternativeHistoryMaker();
IterationValuesHistoryMaker iterationHistoryMaker = new IterationValuesHistoryMaker();
DifferentialEvolutionParameters differentialEvolutionParameters = new DifferentialEvolutionParameters();
differentialEvolutionParameters.CrossoverProbability = 0.5;
differentialEvolutionParameters.DifferentialWeight = 1;
//differentialEvolutionParameters.GenerationParameters.GenerationFrom = Enumerable.Repeat(0.0, dimension).ToArray();
//differentialEvolutionParameters.GenerationParameters.GenerationTo = Enumerable.Repeat(10.0, dimension).ToArray();
differentialEvolutionParameters.GenerationParameters.GenerationFrom = generatingBoundaries.Select(pair => pair.From).ToArray();
differentialEvolutionParameters.GenerationParameters.GenerationTo = generatingBoundaries.Select(pair => pair.To).ToArray();
differentialEvolutionParameters.GenerationParameters.GenerationType = Optimization.EvolutionaryAlgorithms.PopulationGenerationType.Uniform;
differentialEvolutionParameters.Iterations = 20;
differentialEvolutionParameters.Size = 20;
ParallelDifferentialEvolution differentialEvolutionParallel = new ParallelDifferentialEvolution();
differentialEvolutionParallel.SetParameters(differentialEvolutionParameters);
differentialEvolutionParallel.SetProblem(tuningProblem);
differentialEvolutionParallel.AddStatsFollowers(new List <IAlgorithmIterationFollower> {
bestAltHistMaker, iterationHistoryMaker
});
differentialEvolutionParallel.Evaluate();
bestAltHistMaker.SaveToFile();
iterationHistoryMaker.SaveToFile();
#endregion Algorithm
var initialValueBag = tuningProblem.CalculateCriterion(new RealVectorAlternatives(new double[][] { modelVariables.Select(pair => pair.Value.InitialValue).ToArray(), differentialEvolutionParallel.BestSolution }));
var initialValueFitness = initialValueBag.Values.ToArray()[0].Value;
File.WriteAllText("initial.txt", initialValueFitness.ToString());
Console.WriteLine("End!");
Console.Read();
}
public void SetDistanceCalculator(PredefinedDistanceCalculator distanceCalculator)
{
_distanceCalculator = distanceCalculator;
}
