1 | #region License Information
|
---|
2 | /* HeuristicLab
|
---|
3 | * Copyright (C) 2002-2018 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.Collections.Generic;
|
---|
23 | using System.Linq;
|
---|
24 | using System.Text;
|
---|
25 | using HeuristicLab.Analysis;
|
---|
26 | using HeuristicLab.Common;
|
---|
27 | using HeuristicLab.Core;
|
---|
28 | using HeuristicLab.Data;
|
---|
29 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
30 | using HeuristicLab.Optimization;
|
---|
31 | using HeuristicLab.Parameters;
|
---|
32 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
33 | using static HeuristicLab.Problems.DataAnalysis.Symbolic.SymbolicExpressionHashExtensions;
|
---|
34 |
|
---|
35 | namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Analyzers {
|
---|
36 | [Item("SymbolicDataAnalysisBuildingBlockAnalyzer", "An analyzer that uses tree hashing to identify the most common subtrees (building blocks) in the population")]
|
---|
37 | [StorableClass]
|
---|
38 | public sealed class SymbolicDataAnalysisBuildingBlockAnalyzer : SymbolicDataAnalysisAnalyzer {
|
---|
39 | private const string BuildingBlocksResultName = "BuildingBlocks";
|
---|
40 | private const string MinimumSubtreeLengthParameterName = "MinimumSubtreeLength";
|
---|
41 | private const string SimplifyTreesParameterName = "SimplifyTrees";
|
---|
42 |
|
---|
43 | private Dictionary<int, DataRow> hashToRow = new Dictionary<int, DataRow>();
|
---|
44 |
|
---|
45 | #region parameters
|
---|
46 | public IValueLookupParameter<IntValue> MinimumSubtreeLengthParameter {
|
---|
47 | get { return (IValueLookupParameter<IntValue>)Parameters[MinimumSubtreeLengthParameterName]; }
|
---|
48 | }
|
---|
49 |
|
---|
50 | public IValueLookupParameter<BoolValue> SimplifyTreesParameter {
|
---|
51 | get { return (IValueLookupParameter<BoolValue>)Parameters[SimplifyTreesParameterName]; }
|
---|
52 | }
|
---|
53 | #endregion
|
---|
54 |
|
---|
55 | #region parameter properties
|
---|
56 | public IntValue MinimumSubtreeLength {
|
---|
57 | get { return MinimumSubtreeLengthParameter.ActualValue; }
|
---|
58 | }
|
---|
59 |
|
---|
60 | public BoolValue SimplifyTrees {
|
---|
61 | get { return SimplifyTreesParameter.ActualValue; }
|
---|
62 | }
|
---|
63 | #endregion
|
---|
64 |
|
---|
65 | public override void InitializeState() {
|
---|
66 | base.InitializeState();
|
---|
67 |
|
---|
68 | hashToRow = new Dictionary<int, DataRow>();
|
---|
69 | }
|
---|
70 |
|
---|
71 | [StorableHook(HookType.AfterDeserialization)]
|
---|
72 | private void AfterDeserialization() {
|
---|
73 | if (!Parameters.ContainsKey(SimplifyTreesParameterName)) {
|
---|
74 | Parameters.Add(new ValueLookupParameter<BoolValue>(SimplifyTreesParameterName, new BoolValue(false)));
|
---|
75 | }
|
---|
76 | }
|
---|
77 |
|
---|
78 | public SymbolicDataAnalysisBuildingBlockAnalyzer() {
|
---|
79 | Parameters.Add(new ValueLookupParameter<IntValue>(MinimumSubtreeLengthParameterName, new IntValue(3)));
|
---|
80 | Parameters.Add(new ValueLookupParameter<BoolValue>(SimplifyTreesParameterName, new BoolValue(false)));
|
---|
81 | }
|
---|
82 |
|
---|
83 | private SymbolicDataAnalysisBuildingBlockAnalyzer(SymbolicDataAnalysisBuildingBlockAnalyzer original, Cloner cloner) : base(original, cloner) {
|
---|
84 | }
|
---|
85 |
|
---|
86 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
87 | return new SymbolicDataAnalysisBuildingBlockAnalyzer(this, cloner);
|
---|
88 | }
|
---|
89 |
|
---|
90 | [StorableConstructor]
|
---|
91 | private SymbolicDataAnalysisBuildingBlockAnalyzer(bool deserializing) : base(deserializing) { }
|
---|
92 |
|
---|
93 | public override IOperation Apply() {
|
---|
94 | DataTable dt;
|
---|
95 |
|
---|
96 | if (!ResultCollection.ContainsKey(BuildingBlocksResultName)) {
|
---|
97 | dt = new DataTable(BuildingBlocksResultName);
|
---|
98 | ResultCollection.Add(new Result(BuildingBlocksResultName, dt));
|
---|
99 | } else {
|
---|
100 | dt = (DataTable)ResultCollection[BuildingBlocksResultName].Value;
|
---|
101 | }
|
---|
102 |
|
---|
103 | var minLength = MinimumSubtreeLength.Value - 1; // -1 because the HashNode.Size property returns the size without current node (-1)
|
---|
104 | var simplify = SimplifyTrees.Value;
|
---|
105 |
|
---|
106 | var expressions = new Dictionary<int, string>();
|
---|
107 | var expressionCounts = new Dictionary<int, int>();
|
---|
108 |
|
---|
109 | int totalCount = 0; // total number of examined subtrees
|
---|
110 |
|
---|
111 | // count hashes
|
---|
112 | foreach (var tree in SymbolicExpressionTree) {
|
---|
113 | var hashNodes = tree.Root.GetSubtree(0).GetSubtree(0).MakeNodes();
|
---|
114 | var simplified = simplify ? hashNodes.Simplify() : hashNodes.Sort();
|
---|
115 |
|
---|
116 | for (int i = 0; i < simplified.Length; i++) {
|
---|
117 | HashNode<ISymbolicExpressionTreeNode> s = simplified[i];
|
---|
118 | if (s.IsChild || s.Size < minLength) {
|
---|
119 | continue;
|
---|
120 | }
|
---|
121 | ++totalCount;
|
---|
122 | var hash = s.CalculatedHashValue;
|
---|
123 | if (expressions.ContainsKey(hash)) {
|
---|
124 | expressionCounts[hash]++;
|
---|
125 | continue;
|
---|
126 | }
|
---|
127 |
|
---|
128 | var sb = new StringBuilder();
|
---|
129 | for (int j = i - s.Size; j < i; ++j) {
|
---|
130 | sb.Append(GetLabel(simplified[j].Data)).Append(" ");
|
---|
131 | }
|
---|
132 | sb.Append(GetLabel(simplified[i].Data));
|
---|
133 | expressions[hash] = sb.ToString();
|
---|
134 | expressionCounts[hash] = 1;
|
---|
135 | }
|
---|
136 | }
|
---|
137 |
|
---|
138 | // fill in values for existing rows
|
---|
139 | foreach (var t in hashToRow) {
|
---|
140 | var hash = t.Key;
|
---|
141 | var row = t.Value;
|
---|
142 |
|
---|
143 | expressionCounts.TryGetValue(hash, out int count);
|
---|
144 | row.Values.Add(count);
|
---|
145 | }
|
---|
146 |
|
---|
147 | var nValues = dt.Rows.Any() ? dt.Rows.Max(x => x.Values.Count) : 0;
|
---|
148 |
|
---|
149 | // check if we have new rows
|
---|
150 | foreach (var t in expressionCounts.OrderByDescending(x => x.Value).Take(10)) {
|
---|
151 | var hash = t.Key;
|
---|
152 | var count = t.Value;
|
---|
153 | var label = expressions[hash];
|
---|
154 |
|
---|
155 | if (hashToRow.ContainsKey(hash)) {
|
---|
156 | continue;
|
---|
157 | }
|
---|
158 | var row = new DataRow(label) { VisualProperties = { StartIndexZero = true } };
|
---|
159 | if (nValues > 0) {
|
---|
160 | row.Values.AddRange(Enumerable.Repeat<double>(0, nValues - 1)); // pad with zeroes
|
---|
161 | }
|
---|
162 | row.Values.Add(count);
|
---|
163 | dt.Rows.Add(row);
|
---|
164 | hashToRow[hash] = row;
|
---|
165 | }
|
---|
166 |
|
---|
167 | return base.Apply();
|
---|
168 | }
|
---|
169 |
|
---|
170 | private static string GetLabel(ISymbolicExpressionTreeNode node) {
|
---|
171 | if (node is ConstantTreeNode constant) {
|
---|
172 | return "C";
|
---|
173 | }
|
---|
174 | if (node is VariableTreeNode variable) {
|
---|
175 | return variable.VariableName;
|
---|
176 | }
|
---|
177 | if (node.Symbol is Addition) {
|
---|
178 | return "+";
|
---|
179 | }
|
---|
180 | if (node.Symbol is Subtraction) {
|
---|
181 | return "-";
|
---|
182 | }
|
---|
183 | if (node.Symbol is Multiplication) {
|
---|
184 | return "*";
|
---|
185 | }
|
---|
186 | if (node.Symbol is Division) {
|
---|
187 | return "/";
|
---|
188 | }
|
---|
189 | return node.Symbol.ToString();
|
---|
190 | }
|
---|
191 | }
|
---|
192 | }
|
---|