public EvolutionaryAlgorithmBuilderContinuousMO(
Population population,
DecisionSpace decisionSpace,
HyperParameterManager hyperParameters,
IParentSelectionOperator parentSelector,
IRecombinationOperator recombinationOperator,
IMutationOperator mutationOperator,
IReinsertionOperator reinsertionOperator)
{
this.decisionSpace = decisionSpace;
HyperParameters.AddFromExistingHyperParameterSet(hyperParameters);
this.population = population;
this.parentSelector = parentSelector;
this.recombinationOperator = recombinationOperator;
this.mutationOperator = mutationOperator;
this.reinsertionOperator = reinsertionOperator;
}
/// <summary>
/// Creates an Evolutionary Algorithm.
/// </summary>
/// <param name="initialPopulation">The initial population (can be empty).</param>
/// <param name="fitnessCalculator">A <see cref="FitnessCalculatorSingleObjective"/>. <see cref="Optimiser"/></param>
/// <param name="initialIndividualGenerator">Creates new decision vectors to build the first population. <seealso cref="Base.Conversion.IModel"/></param>
/// <param name="parentSelector">The <see cref="IParentSelectionOperator"/> to use.</param>
/// <param name="recombinationOperator">The <see cref="IRecombinationOperator"/> to use.</param>
/// <param name="mutationOperator">The <see cref="IMutationOperator"/> to use.</param>
/// <param name="reinsertionOperator">The <see cref="IReinsertionOperator"/> to use.</param>
/// <param name="hyperParameters">The <see cref="HyperParameterManager"/> object with relevant settings.</param>
public EvolutionaryAlgorithm(
Population initialPopulation,
IFitnessCalculator fitnessCalculator,
Func <DecisionVector> initialIndividualGenerator,
IParentSelectionOperator parentSelector,
IRecombinationOperator recombinationOperator,
IMutationOperator mutationOperator,
IReinsertionOperator reinsertionOperator,
HyperParameterManager hyperParameters)
: base(initialPopulation, fitnessCalculator)
{
this.initialIndividualGenerator = initialIndividualGenerator;
this.parentSelector = parentSelector;
numberOfParents = hyperParameters.GetHyperParameterValue <int>(EvolutionaryAlgorithmHyperParameters.Number_Of_Parents);
this.recombinationOperator = recombinationOperator;
this.mutationOperator = mutationOperator;
this.reinsertionOperator = reinsertionOperator;
}
public static OptimiserBuilder GetBuilder(DecisionSpace problemSpace,
AvailableOperators.ParentSelector parentSelection,
AvailableOperators.RecombinationOperator recombination,
AvailableOperators.MutationOperators mutation,
AvailableOperators.ReinsertionOperators reinsertion,
int?populationSize = null)
{
var hyps = EvolutionaryAlgorithmHyperParameters.GetDefaultHyperParameters();
if (populationSize != null)
{
hyps.UpdateHyperParameterValue(
EvolutionaryAlgorithmHyperParameters.Population_Size, populationSize);
}
var population = new Population(
hyps.GetHyperParameterValue <int>(EvolutionaryAlgorithmHyperParameters.Population_Size));
IParentSelectionOperator parentSelector;
switch (parentSelection)
{
case AvailableOperators.ParentSelector.Tournament:
parentSelector = new ParentSelectionTournament(
20,
false);
break;
case AvailableOperators.ParentSelector.Greedy:
parentSelector = new ParentSelectionGreedy();
break;
case AvailableOperators.ParentSelector.Random:
parentSelector = new ParentSelectionRandom();
break;
case AvailableOperators.ParentSelector.Roulette:
parentSelector = new ParentSelectionRoulette(
false);
break;
default:
throw new NotImplementedException();
}
IRecombinationOperator recombinationOperator;
hyps.UpdateHyperParameterValue(EvolutionaryAlgorithmHyperParameters.Number_Of_Parents, 2);
switch (recombination)
{
case AvailableOperators.RecombinationOperator.MultiPoint:
recombinationOperator = new CrossoverMultiPoint(
2);
break;
case AvailableOperators.RecombinationOperator.ArithmeticTwoParentWeighted:
recombinationOperator = new CrossoverArithmeticWeighted(
false,
0.5);
break;
case AvailableOperators.RecombinationOperator.Sbx:
recombinationOperator = new CrossoverSimulatedBinary(
2);
break;
case AvailableOperators.RecombinationOperator.SbxSa2:
recombinationOperator = new CrossoverSimulatedBinarySelfAdaptive2(
population, 2);
break;
case AvailableOperators.RecombinationOperator.ArithmeticMultiParent:
recombinationOperator = new CrossoverArithmeticMultiParent();
hyps.UpdateHyperParameterValue(
EvolutionaryAlgorithmHyperParameters.Number_Of_Parents,
4);
break;
case AvailableOperators.RecombinationOperator.Uniform:
recombinationOperator = new CrossoverUniform(
0.5);
break;
case AvailableOperators.RecombinationOperator.Pcx:
recombinationOperator = new RecombinationParentCentric(
0.1,
0.1);
hyps.UpdateHyperParameterValue(
EvolutionaryAlgorithmHyperParameters.Number_Of_Parents,
6);
break;
default:
throw new NotImplementedException();
}
IMutationOperator mutationOperator;
switch (mutation)
{
case AvailableOperators.MutationOperators.AddRandomNumber:
mutationOperator = new MutationAddRandomNumber(
0.1,
0.1,
1);
break;
case AvailableOperators.MutationOperators.RandomSwap:
mutationOperator = new MutationRandomSwap(
0.1);
break;
case AvailableOperators.MutationOperators.ReplaceWithRandomNumber:
mutationOperator = new MutationReplaceWithRandomNumber(
0.1,
1);
break;
case AvailableOperators.MutationOperators.None:
mutationOperator = new MutationReplaceWithRandomNumber(
0, 1);
break;
default:
throw new NotImplementedException();
}
IReinsertionOperator reinsertionOperator;
switch (reinsertion)
{
case AvailableOperators.ReinsertionOperators.ReplaceWorst:
reinsertionOperator = new ReinsertionReplaceWorst();
break;
case AvailableOperators.ReinsertionOperators.ReplaceRandom:
reinsertionOperator = new ReinsertionReplaceRandom();
break;
default:
throw new NotImplementedException();
}
return(new EvolutionaryAlgorithmBuilderContinuousSO(population, problemSpace, hyps,
parentSelector, recombinationOperator, mutationOperator, reinsertionOperator));
}
