1 | #region License Information
|
---|
2 | /* HeuristicLab
|
---|
3 | * Copyright (C) 2002-2015 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
4 | *
|
---|
5 | * This file is part of HeuristicLab.
|
---|
6 | *
|
---|
7 | * HeuristicLab is free software: you can redistribute it and/or modify
|
---|
8 | * it under the terms of the GNU General Public License as published by
|
---|
9 | * the Free Software Foundation, either version 3 of the License, or
|
---|
10 | * (at your option) any later version.
|
---|
11 | *
|
---|
12 | * HeuristicLab is distributed in the hope that it will be useful,
|
---|
13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
15 | * GNU General Public License for more details.
|
---|
16 | *
|
---|
17 | * You should have received a copy of the GNU General Public License
|
---|
18 | * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
|
---|
19 | */
|
---|
20 | #endregion
|
---|
21 |
|
---|
22 | using System;
|
---|
23 | using HeuristicLab.Common;
|
---|
24 | using HeuristicLab.Core;
|
---|
25 | using HeuristicLab.Data;
|
---|
26 | using HeuristicLab.Encodings.RealVectorEncoding;
|
---|
27 | using HeuristicLab.Operators;
|
---|
28 | using HeuristicLab.Parameters;
|
---|
29 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
30 |
|
---|
31 | namespace HeuristicLab.Problems.TestFunctions {
|
---|
32 | /// <summary>
|
---|
33 | /// An operator that improves test functions solutions.
|
---|
34 | /// </summary>
|
---|
35 | /// <remarks>
|
---|
36 | /// It is implemented as described in Laguna, M. and Martí, R. (2003). Scatter Search: Methodology and Implementations in C. Operations Research/Computer Science Interfaces Series, Vol. 24. Springer.<br />
|
---|
37 | /// The operator uses an implementation of the Nelder-Mead method with adaptive parameters as described in Gao, F. and Han, L. (2010). Implementing the Nelder-Mead simplex algorithm with adaptive parameters. Computational Optimization and Applications, Vol. 51. Springer. and conducts relection, expansion, contraction and reduction on the test functions solution.
|
---|
38 | /// </remarks>
|
---|
39 | [Item("SingleObjectiveTestFunctionImprovementOperator", "An operator that improves test functions solutions. It is implemented as described in Laguna, M. and Martí, R. (2003). Scatter Search: Methodology and Implementations in C. Operations Research/Computer Science Interfaces Series, Vol. 24. Springer.")]
|
---|
40 | [StorableClass]
|
---|
41 | public sealed class SingleObjectiveTestFunctionImprovementOperator : SingleSuccessorOperator, ISingleObjectiveTestFunctionImprovementOperator {
|
---|
42 | #region Parameter properties
|
---|
43 | public IValueParameter<DoubleValue> AlphaParameter {
|
---|
44 | get { return (IValueParameter<DoubleValue>)Parameters["Alpha"]; }
|
---|
45 | }
|
---|
46 | public IValueParameter<DoubleValue> BetaParameter {
|
---|
47 | get { return (IValueParameter<DoubleValue>)Parameters["Beta"]; }
|
---|
48 | }
|
---|
49 | public IValueLookupParameter<DoubleMatrix> BoundsParameter {
|
---|
50 | get { return (IValueLookupParameter<DoubleMatrix>)Parameters["Bounds"]; }
|
---|
51 | }
|
---|
52 | public IValueParameter<DoubleValue> DeltaParameter {
|
---|
53 | get { return (IValueParameter<DoubleValue>)Parameters["Delta"]; }
|
---|
54 | }
|
---|
55 | public IValueLookupParameter<ISingleObjectiveTestFunction> TestFunctionParameter {
|
---|
56 | get { return (IValueLookupParameter<ISingleObjectiveTestFunction>)Parameters["TestFunction"]; }
|
---|
57 | }
|
---|
58 | public IValueParameter<DoubleValue> GammaParameter {
|
---|
59 | get { return (IValueParameter<DoubleValue>)Parameters["Gamma"]; }
|
---|
60 | }
|
---|
61 | public IValueLookupParameter<IntValue> ImprovementAttemptsParameter {
|
---|
62 | get { return (IValueLookupParameter<IntValue>)Parameters["ImprovementAttempts"]; }
|
---|
63 | }
|
---|
64 | public IValueLookupParameter<IItem> SolutionParameter {
|
---|
65 | get { return (IValueLookupParameter<IItem>)Parameters["Solution"]; }
|
---|
66 | }
|
---|
67 | #endregion
|
---|
68 |
|
---|
69 | #region Properties
|
---|
70 | private DoubleValue Alpha {
|
---|
71 | get { return AlphaParameter.Value; }
|
---|
72 | }
|
---|
73 | private DoubleValue Beta {
|
---|
74 | get { return BetaParameter.Value; }
|
---|
75 | }
|
---|
76 | private DoubleValue Delta {
|
---|
77 | get { return DeltaParameter.Value; }
|
---|
78 | }
|
---|
79 | private DoubleValue Gamma {
|
---|
80 | get { return GammaParameter.Value; }
|
---|
81 | }
|
---|
82 | #endregion
|
---|
83 |
|
---|
84 | [StorableConstructor]
|
---|
85 | private SingleObjectiveTestFunctionImprovementOperator(bool deserializing) : base(deserializing) { }
|
---|
86 | private SingleObjectiveTestFunctionImprovementOperator(SingleObjectiveTestFunctionImprovementOperator original, Cloner cloner) : base(original, cloner) { }
|
---|
87 | public SingleObjectiveTestFunctionImprovementOperator()
|
---|
88 | : base() {
|
---|
89 | #region Create parameters
|
---|
90 | Parameters.Add(new ValueParameter<DoubleValue>("Alpha", new DoubleValue(1.0)));
|
---|
91 | Parameters.Add(new ValueParameter<DoubleValue>("Beta", new DoubleValue(2.0)));
|
---|
92 | Parameters.Add(new ValueParameter<DoubleValue>("Delta", new DoubleValue(0.5)));
|
---|
93 | Parameters.Add(new ValueParameter<DoubleValue>("Gamma", new DoubleValue(0.5)));
|
---|
94 | Parameters.Add(new ValueLookupParameter<ISingleObjectiveTestFunction>("TestFunction", "The operator used to evaluate solutions."));
|
---|
95 | Parameters.Add(new ValueLookupParameter<DoubleMatrix>("Bounds", "The lower and upper bounds in each dimension."));
|
---|
96 | Parameters.Add(new ValueLookupParameter<IntValue>("ImprovementAttempts", "The number of improvement attempts the operator should perform.", new IntValue(100)));
|
---|
97 | Parameters.Add(new ValueLookupParameter<IItem>("Solution", "The solution to be improved. This parameter is used for name translation only.")); // TODO: Problematic, this cannot be wired! IImprovementOperators need to be generic
|
---|
98 | #endregion
|
---|
99 | }
|
---|
100 |
|
---|
101 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
102 | return new SingleObjectiveTestFunctionImprovementOperator(this, cloner);
|
---|
103 | }
|
---|
104 |
|
---|
105 | public override IOperation Apply() {
|
---|
106 | RealVector bestSol = ExecutionContext.Scope.Variables[SolutionParameter.ActualName].Value as RealVector;
|
---|
107 | if (bestSol == null)
|
---|
108 | throw new ArgumentException("Cannot improve solution because it has the wrong type.");
|
---|
109 |
|
---|
110 | var bounds = BoundsParameter.ActualValue;
|
---|
111 | var function = TestFunctionParameter.ActualValue;
|
---|
112 | var maxIterations = ImprovementAttemptsParameter.ActualValue.Value;
|
---|
113 |
|
---|
114 | double bestSolQuality = function.Evaluate(bestSol);
|
---|
115 |
|
---|
116 | // create perturbed solutions
|
---|
117 | RealVector[] simplex = new RealVector[bestSol.Length];
|
---|
118 | for (int i = 0; i < simplex.Length; i++) {
|
---|
119 | simplex[i] = bestSol.Clone() as RealVector;
|
---|
120 | simplex[i][i] += 0.1 * (bounds[0, 1] - bounds[0, 0]);
|
---|
121 | if (simplex[i][i] > bounds[0, 1]) simplex[i][i] = bounds[0, 1];
|
---|
122 | if (simplex[i][i] < bounds[0, 0]) simplex[i][i] = bounds[0, 0];
|
---|
123 | }
|
---|
124 |
|
---|
125 | // improve solutions
|
---|
126 | for (int i = 0; i < maxIterations; i++) {
|
---|
127 | // order according to their objective function value
|
---|
128 | Array.Sort(simplex, (x, y) => function.Evaluate(x).CompareTo(function.Evaluate(y)));
|
---|
129 |
|
---|
130 | // calculate centroid
|
---|
131 | RealVector centroid = new RealVector(bestSol.Length);
|
---|
132 | foreach (var vector in simplex)
|
---|
133 | for (int j = 0; j < centroid.Length; j++)
|
---|
134 | centroid[j] += vector[j];
|
---|
135 | for (int j = 0; j < centroid.Length; j++)
|
---|
136 | centroid[j] /= simplex.Length;
|
---|
137 |
|
---|
138 | // reflection
|
---|
139 | RealVector reflectionPoint = new RealVector(bestSol.Length);
|
---|
140 | for (int j = 0; j < reflectionPoint.Length; j++)
|
---|
141 | reflectionPoint[j] = centroid[j] + Alpha.Value * (centroid[j] - simplex[simplex.Length - 1][j]);
|
---|
142 | double reflectionPointQuality = function.Evaluate(reflectionPoint);
|
---|
143 | if (function.Evaluate(simplex[0]) <= reflectionPointQuality
|
---|
144 | && reflectionPointQuality < function.Evaluate(simplex[simplex.Length - 2]))
|
---|
145 | simplex[simplex.Length - 1] = reflectionPoint;
|
---|
146 |
|
---|
147 | // expansion
|
---|
148 | if (reflectionPointQuality < function.Evaluate(simplex[0])) {
|
---|
149 | RealVector expansionPoint = new RealVector(bestSol.Length);
|
---|
150 | for (int j = 0; j < expansionPoint.Length; j++)
|
---|
151 | expansionPoint[j] = centroid[j] + Beta.Value * (reflectionPoint[j] - centroid[j]);
|
---|
152 | simplex[simplex.Length - 1] = function.Evaluate(expansionPoint) < reflectionPointQuality ? expansionPoint : reflectionPoint;
|
---|
153 | }
|
---|
154 |
|
---|
155 | // contraction
|
---|
156 | if (function.Evaluate(simplex[simplex.Length - 2]) <= reflectionPointQuality
|
---|
157 | && reflectionPointQuality < function.Evaluate(simplex[simplex.Length - 1])) {
|
---|
158 | RealVector outsideContractionPoint = new RealVector(bestSol.Length);
|
---|
159 | for (int j = 0; j < outsideContractionPoint.Length; j++)
|
---|
160 | outsideContractionPoint[j] = centroid[j] + Gamma.Value * (reflectionPoint[j] - centroid[j]);
|
---|
161 | if (function.Evaluate(outsideContractionPoint) <= reflectionPointQuality) {
|
---|
162 | simplex[simplex.Length - 1] = outsideContractionPoint;
|
---|
163 | if (function.Evaluate(reflectionPoint) >= function.Evaluate(simplex[simplex.Length - 1])) {
|
---|
164 | RealVector insideContractionPoint = new RealVector(bestSol.Length);
|
---|
165 | for (int j = 0; j < insideContractionPoint.Length; j++)
|
---|
166 | insideContractionPoint[j] = centroid[j] - Gamma.Value * (reflectionPoint[j] - centroid[j]);
|
---|
167 | if (function.Evaluate(insideContractionPoint) < function.Evaluate(simplex[simplex.Length - 1])) simplex[simplex.Length - 1] = insideContractionPoint;
|
---|
168 | }
|
---|
169 | }
|
---|
170 | }
|
---|
171 |
|
---|
172 | // reduction
|
---|
173 | for (int j = 1; j < simplex.Length; j++)
|
---|
174 | for (int k = 0; k < simplex[j].Length; k++)
|
---|
175 | simplex[j][k] = simplex[0][k] + Delta.Value * (simplex[j][k] - simplex[0][k]);
|
---|
176 | }
|
---|
177 |
|
---|
178 | for (int i = 0; i < simplex[0].Length; i++) {
|
---|
179 | if (simplex[0][i] > bounds[0, 1]) simplex[0][i] = bounds[0, 1];
|
---|
180 | if (simplex[0][i] < bounds[0, 0]) simplex[0][i] = bounds[0, 0];
|
---|
181 | }
|
---|
182 |
|
---|
183 | ExecutionContext.Scope.Variables[SolutionParameter.ActualName].Value = simplex[0];
|
---|
184 | ExecutionContext.Scope.Variables.Add(new Variable("LocalEvaluatedSolutions", new IntValue(maxIterations)));
|
---|
185 |
|
---|
186 | return base.Apply();
|
---|
187 | }
|
---|
188 | }
|
---|
189 | }
|
---|