#region License Information
/* HeuristicLab
* Copyright (C) 2002-2015 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
*
* This file is part of HeuristicLab.
*
* HeuristicLab is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* HeuristicLab is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with HeuristicLab. If not, see .
*/
#endregion
using System;
using System.Linq;
using HeuristicLab.Analysis;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Data;
using HeuristicLab.Operators;
using HeuristicLab.Optimization;
using HeuristicLab.Optimization.Operators;
using HeuristicLab.Parameters;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
namespace HeuristicLab.Algorithms.LocalSearch {
///
/// A local search improvement operator.
///
[Item("LocalSearchImprovementOperator", "A local search improvement operator.")]
[StorableClass("A3DAA4D3-5C87-4E41-8AD4-FA76D803C50D")]
public sealed class LocalSearchImprovementOperator : SingleSuccessorOperator, ILocalImprovementAlgorithmOperator, IStochasticOperator, ISingleObjectiveOperator {
#region IGenericLocalImprovementOperator Properties
public Type ProblemType { get { return typeof(ISingleObjectiveHeuristicOptimizationProblem); } }
public IProblem Problem {
get { return problem; }
set {
if (problem != value) {
if (value != null && !(value is ISingleObjectiveHeuristicOptimizationProblem))
throw new ArgumentException("Only problems of type " + ProblemType.ToString() + " can be assigned.");
if (problem != null) DeregisterProblemEventHandlers();
problem = (ISingleObjectiveHeuristicOptimizationProblem)value;
if (problem != null) RegisterProblemEventHandlers();
UpdateProblem();
}
}
}
#endregion
[Storable]
private ISingleObjectiveHeuristicOptimizationProblem problem;
[Storable]
private LocalSearchMainLoop loop;
[Storable]
private BestAverageWorstQualityAnalyzer qualityAnalyzer;
#region Parameter Properties
public IConstrainedValueParameter MoveGeneratorParameter {
get { return (IConstrainedValueParameter)Parameters["MoveGenerator"]; }
}
public IConstrainedValueParameter MoveMakerParameter {
get { return (IConstrainedValueParameter)Parameters["MoveMaker"]; }
}
public IConstrainedValueParameter MoveEvaluatorParameter {
get { return (IConstrainedValueParameter)Parameters["MoveEvaluator"]; }
}
public IValueLookupParameter SampleSizeParameter {
get { return (IValueLookupParameter)Parameters["SampleSize"]; }
}
public ValueParameter AnalyzerParameter {
get { return (ValueParameter)Parameters["Analyzer"]; }
}
public ScopeTreeLookupParameter QualityParameter {
get { return (ScopeTreeLookupParameter)Parameters["Quality"]; }
}
public ILookupParameter RandomParameter {
get { return (ILookupParameter)Parameters["Random"]; }
}
#region ILocalImprovementOperator Parameters
public IValueLookupParameter MaximumIterationsParameter {
get { return (IValueLookupParameter)Parameters["MaximumIterations"]; }
}
public ILookupParameter EvaluatedSolutionsParameter {
get { return (ILookupParameter)Parameters["EvaluatedSolutions"]; }
}
public ILookupParameter ResultsParameter {
get { return (ILookupParameter)Parameters["Results"]; }
}
#endregion
#endregion
#region Properties
public IMoveGenerator MoveGenerator {
get { return MoveGeneratorParameter.Value; }
set { MoveGeneratorParameter.Value = value; }
}
public IMoveMaker MoveMaker {
get { return MoveMakerParameter.Value; }
set { MoveMakerParameter.Value = value; }
}
public ISingleObjectiveMoveEvaluator MoveEvaluator {
get { return MoveEvaluatorParameter.Value; }
set { MoveEvaluatorParameter.Value = value; }
}
public MultiAnalyzer Analyzer {
get { return AnalyzerParameter.Value; }
set { AnalyzerParameter.Value = value; }
}
#endregion
[StorableConstructor]
private LocalSearchImprovementOperator(bool deserializing) : base(deserializing) { }
private LocalSearchImprovementOperator(LocalSearchImprovementOperator original, Cloner cloner)
: base(original, cloner) {
this.loop = cloner.Clone(original.loop);
this.qualityAnalyzer = cloner.Clone(original.qualityAnalyzer);
this.problem = cloner.Clone(original.problem);
RegisterEventHandlers();
}
public LocalSearchImprovementOperator()
: base() {
Parameters.Add(new ConstrainedValueParameter("MoveGenerator", "The operator used to generate moves to the neighborhood of the current solution."));
Parameters.Add(new ConstrainedValueParameter("MoveMaker", "The operator used to perform a move."));
Parameters.Add(new ConstrainedValueParameter("MoveEvaluator", "The operator used to evaluate a move."));
Parameters.Add(new ValueLookupParameter("MaximumIterations", "The maximum number of generations which should be processed.", new IntValue(150)));
Parameters.Add(new ValueLookupParameter("SampleSize", "Number of moves that MultiMoveGenerators should create. This is ignored for Exhaustive- and SingleMoveGenerators.", new IntValue(300)));
Parameters.Add(new LookupParameter("EvaluatedSolutions", "The number of evaluated moves."));
Parameters.Add(new ValueParameter("Analyzer", "The operator used to analyze the solution.", new MultiAnalyzer()));
Parameters.Add(new LookupParameter("Results", "The name of the collection where the results are stored."));
Parameters.Add(new ScopeTreeLookupParameter("Quality", "The quality/fitness value of a solution."));
Parameters.Add(new LookupParameter("Random", "The random number generator to use."));
loop = new LocalSearchMainLoop();
((ResultsCollector)((SingleSuccessorOperator)loop.OperatorGraph.InitialOperator).Successor).CollectedValues.Remove(loop.BestLocalQualityParameter.Name);
ParameterizeLSMainLoop();
qualityAnalyzer = new BestAverageWorstQualityAnalyzer();
Analyzer.Operators.Add(qualityAnalyzer);
RegisterEventHandlers();
}
public override IDeepCloneable Clone(Cloner cloner) {
return new LocalSearchImprovementOperator(this, cloner);
}
[StorableHook(HookType.AfterDeserialization)]
private void AfterDeserialization() {
RegisterEventHandlers();
}
#region Event Handler Registration
private void RegisterEventHandlers() {
MoveGeneratorParameter.ValueChanged += new EventHandler(MoveGeneratorParameter_ValueChanged);
if (problem != null)
RegisterProblemEventHandlers();
}
private void RegisterProblemEventHandlers() {
problem.Reset += new EventHandler(problem_Reset);
problem.OperatorsChanged += new EventHandler(problem_OperatorsChanged);
}
private void DeregisterProblemEventHandlers() {
problem.Reset -= new EventHandler(problem_Reset);
problem.OperatorsChanged -= new EventHandler(problem_OperatorsChanged);
}
#endregion
#region Event Handlers
private void MoveGeneratorParameter_ValueChanged(object sender, EventArgs e) {
ChooseMoveOperators();
ParameterizeLSMainLoop();
}
private void problem_Reset(object sender, EventArgs e) {
UpdateProblem();
}
private void problem_OperatorsChanged(object sender, EventArgs e) {
UpdateProblem();
}
#endregion
#region Parameterize and Update Methods
private void UpdateProblem() {
UpdateMoveOperators();
ChooseMoveOperators();
ParameterizeMoveGenerators();
ParameterizeLSMainLoop();
ParameterizeAnalyzers();
UpdateAnalyzers();
}
private void ParameterizeLSMainLoop() {
loop.AnalyzerParameter.ActualName = AnalyzerParameter.Name;
loop.BestLocalQualityParameter.ActualName = QualityParameter.Name;
loop.EvaluatedMovesParameter.ActualName = EvaluatedSolutionsParameter.Name;
loop.IterationsParameter.ActualName = "LocalIterations";
loop.MaximumIterationsParameter.ActualName = MaximumIterationsParameter.Name;
loop.MoveEvaluatorParameter.ActualName = MoveEvaluatorParameter.Name;
loop.MoveGeneratorParameter.ActualName = MoveGeneratorParameter.Name;
loop.MoveMakerParameter.ActualName = MoveMakerParameter.Name;
loop.QualityParameter.ActualName = QualityParameter.Name;
loop.RandomParameter.ActualName = RandomParameter.Name;
loop.ResultsParameter.ActualName = ResultsParameter.Name;
if (problem != null) {
loop.BestKnownQualityParameter.ActualName = problem.BestKnownQualityParameter.Name;
loop.MaximizationParameter.ActualName = problem.MaximizationParameter.Name;
}
if (MoveEvaluator != null) {
loop.MoveQualityParameter.ActualName = MoveEvaluator.MoveQualityParameter.ActualName;
}
}
private void ParameterizeAnalyzers() {
qualityAnalyzer.ResultsParameter.ActualName = ResultsParameter.Name;
if (problem != null) {
qualityAnalyzer.MaximizationParameter.ActualName = problem.MaximizationParameter.Name;
if (MoveEvaluator != null)
qualityAnalyzer.QualityParameter.ActualName = MoveEvaluator.MoveQualityParameter.ActualName;
qualityAnalyzer.BestKnownQualityParameter.ActualName = problem.BestKnownQualityParameter.Name;
}
}
private bool IsSubclassOfGeneric(Type generic, Type toCheck) {
while (toCheck != typeof(object)) {
var cur = toCheck.IsGenericType ? toCheck.GetGenericTypeDefinition() : toCheck;
if (generic == cur) {
return true;
}
toCheck = toCheck.BaseType;
}
return false;
}
private void UpdateAnalyzers() {
Analyzer.Operators.Clear();
if (problem != null) {
foreach (IAnalyzer analyzer in problem.Operators.OfType()) {
if (!IsSubclassOfGeneric(typeof(AlleleFrequencyAnalyzer<>), analyzer.GetType()) &&
!(analyzer is PopulationSimilarityAnalyzer)) {
IAnalyzer clone = analyzer.Clone() as IAnalyzer;
foreach (IScopeTreeLookupParameter param in clone.Parameters.OfType())
param.Depth = 0;
Analyzer.Operators.Add(clone, false);
}
}
}
Analyzer.Operators.Add(qualityAnalyzer, false);
}
private void UpdateMoveOperators() {
IMoveGenerator oldMoveGenerator = MoveGenerator;
IMoveMaker oldMoveMaker = MoveMaker;
ISingleObjectiveMoveEvaluator oldMoveEvaluator = MoveEvaluator;
ClearMoveParameters();
if (problem != null) {
foreach (IMultiMoveGenerator generator in problem.Operators.OfType().OrderBy(x => x.Name))
MoveGeneratorParameter.ValidValues.Add(generator);
foreach (IExhaustiveMoveGenerator generator in problem.Operators.OfType().OrderBy(x => x.Name))
MoveGeneratorParameter.ValidValues.Add(generator);
if (oldMoveGenerator != null) {
IMoveGenerator newMoveGenerator = MoveGeneratorParameter.ValidValues.FirstOrDefault(x => x.GetType() == oldMoveGenerator.GetType());
if (newMoveGenerator != null) MoveGenerator = newMoveGenerator;
}
ChooseMoveOperators(oldMoveMaker, oldMoveEvaluator);
}
}
private void ChooseMoveOperators(IMoveMaker oldMoveMaker = null, ISingleObjectiveMoveEvaluator oldMoveEvaluator = null) {
if (oldMoveMaker == null) oldMoveMaker = MoveMaker;
if (oldMoveEvaluator == null) oldMoveEvaluator = MoveEvaluator;
MoveMakerParameter.ValidValues.Clear();
MoveEvaluatorParameter.ValidValues.Clear();
if (MoveGenerator != null && Problem != null) {
IMoveGenerator generator = MoveGeneratorParameter.Value;
foreach (IMoveMaker moveMaker in MoveHelper.GetCompatibleMoveMakers(generator, Problem.Operators.OfType()).OrderBy(x => x.Name))
MoveMakerParameter.ValidValues.Add(moveMaker);
foreach (ISingleObjectiveMoveEvaluator moveEvaluator in MoveHelper.GetCompatibleSingleObjectiveMoveEvaluators(generator, Problem.Operators.OfType()).OrderBy(x => x.Name))
MoveEvaluatorParameter.ValidValues.Add(moveEvaluator);
if (oldMoveMaker != null) {
IMoveMaker mm = MoveMakerParameter.ValidValues.FirstOrDefault(x => x.GetType() == oldMoveMaker.GetType());
if (mm != null) MoveMaker = mm;
}
if (oldMoveEvaluator != null) {
ISingleObjectiveMoveEvaluator me = MoveEvaluatorParameter.ValidValues.FirstOrDefault(x => x.GetType() == oldMoveEvaluator.GetType());
if (me != null) MoveEvaluator = me;
}
}
}
private void ClearMoveParameters() {
MoveGeneratorParameter.ValidValues.Clear();
MoveMakerParameter.ValidValues.Clear();
MoveEvaluatorParameter.ValidValues.Clear();
}
private void ParameterizeMoveGenerators() {
if (problem != null) {
foreach (IMultiMoveGenerator generator in problem.Operators.OfType())
generator.SampleSizeParameter.ActualName = SampleSizeParameter.Name;
}
}
#endregion
public override IOperation Apply() {
IScope currentScope = ExecutionContext.Scope;
Scope localScope = new Scope();
Scope individual = new Scope();
foreach (IVariable var in currentScope.Variables)
individual.Variables.Add(var); // add reference to variable otherwise the analyzer fails (it's looking down the tree)
localScope.SubScopes.Add(individual);
currentScope.SubScopes.Add(localScope);
int index = currentScope.SubScopes.Count - 1;
SubScopesProcessor processor = new SubScopesProcessor();
SubScopesRemover remover = new SubScopesRemover();
remover.RemoveAllSubScopes = false;
remover.SubScopeIndexParameter.Value = new IntValue(index);
if (index > 0) {
EmptyOperator eo = new EmptyOperator();
for (int i = 0; i < index - 1; i++) {
processor.Operators.Add(eo);
}
}
VariableCreator variableCreator = new VariableCreator();
variableCreator.CollectedValues.Add(new ValueParameter(loop.IterationsParameter.ActualName, new IntValue(0)));
variableCreator.CollectedValues.Add(new ValueParameter(loop.BestLocalQualityParameter.ActualName, new DoubleValue(0)));
variableCreator.Successor = loop;
processor.Operators.Add(variableCreator);
processor.Successor = remover;
OperationCollection next = new OperationCollection(base.Apply());
next.Insert(0, ExecutionContext.CreateChildOperation(processor));
return next;
}
}
}