Free cookie consent management tool by TermsFeed Policy Generator

source: branches/2974_Constants_Optimization/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Importer/SymbolicExpressionImporter.cs @ 16676

Last change on this file since 16676 was 16676, checked in by gkronber, 5 years ago

#2974: merged r16478:16672 from trunk:HeuristicLab.Problems.DataAnalysis.Symbolic to branch:HeuristicLab.Problems.DataAnalysis.Symbolic
File size: 11.5 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2019 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
22using System;
23using System.Collections.Generic;
24using System.Diagnostics;
25using System.Linq;
26using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
27
28namespace HeuristicLab.Problems.DataAnalysis.Symbolic {
29  public class SymbolicExpressionImporter {
30    private const string VARSTART = "VAR";
31    private const string LAGGEDVARSTART = "LAGVARIABLE";
32    private const string INTEGRALSTART = "INTEG";
33    private const string DEFUNSTART = "DEFUN";
34    private const string ARGSTART = "ARG";
35    private const string INVOKESTART = "CALL";
36    private const string TIMELAGSTART = "LAG";
37    private Dictionary<string, Symbol> knownSymbols = new Dictionary<string, Symbol>()
38      {
39        {"+", new Addition()},
40        {"/", new Division()},
41        {"*", new Multiplication()},
42        {"-", new Subtraction()},
43        {"ABS", new Absolute() },
44        {"EXP", new Exponential()},
45        {"LOG", new Logarithm()},
46        {"POW", new Power()},
47        {"ROOT", new Root()},
48        {"SQR", new Square()},
49        {"SQRT", new SquareRoot()},
50        {"CUBE", new Cube()},
51        {"CUBEROOT", new CubeRoot()},
52        {"SIN",new Sine()},
53        {"COS", new Cosine()},
54        {"TAN", new Tangent()},
55        {"TANH", new HyperbolicTangent ()},
56        {"AIRYA", new AiryA()},
57        {"AIRYB", new AiryB()},
58        {"BESSEL", new Bessel()},
59        {"COSINT", new CosineIntegral()},
60        {"SININT", new SineIntegral()},
61        {"HYPCOSINT", new HyperbolicCosineIntegral()},
62        {"HYPSININT", new HyperbolicSineIntegral()},
63        {"FRESNELSININT", new FresnelSineIntegral()},
64        {"FRESNELCOSINT", new FresnelCosineIntegral()},
65        {"NORM", new Norm()},
66        {"ERF", new Erf()},
67        {"GAMMA", new Gamma()},
68        {"PSI", new Psi()},
69        {"DAWSON", new Dawson()},
70        {"EXPINT", new ExponentialIntegralEi()},
71        {"AQ", new AnalyticQuotient() },
72        {"MEAN", new Average()},
73        {"IF", new IfThenElse()},
74        {">", new GreaterThan()},
75        {"<", new LessThan()},
76        {"AND", new And()},
77        {"OR", new Or()},
78        {"NOT", new Not()},
79        {"XOR", new Xor()},
80        {"DIFF", new Derivative()},
81        {"PROG", new ProgramRootSymbol()},
82        {"MAIN", new StartSymbol()},
83        {"FACTOR", new FactorVariable() },
84        {"BINFACTOR", new BinaryFactorVariable()}
85      };
86
87    Constant constant = new Constant();
88    Variable variable = new Variable();
89    LaggedVariable laggedVariable = new LaggedVariable();
90    Defun defun = new Defun();
91    TimeLag timeLag = new TimeLag();
92    Integral integral = new Integral();
93    FactorVariable factorVar = new FactorVariable();
94    BinaryFactorVariable binFactorVar = new BinaryFactorVariable();
95
96    ProgramRootSymbol programRootSymbol = new ProgramRootSymbol();
97    StartSymbol startSymbol = new StartSymbol();
98
99    public ISymbolicExpressionTree Import(string str) {
100      str = str.Replace("(", " ( ").Replace(")", " ) ");
101      ISymbolicExpressionTreeNode root = programRootSymbol.CreateTreeNode();
102      ISymbolicExpressionTreeNode start = startSymbol.CreateTreeNode();
103      ISymbolicExpressionTreeNode mainBranch = ParseSexp(new Queue<Token>(GetTokenStream(str)));
104      if (mainBranch.Symbol is ProgramRootSymbol) {
105        // when a root symbol was parsed => use main branch as root
106        root = mainBranch;
107      } else {
108        // only a main branch was given => insert the main branch into the default tree template
109        root.AddSubtree(start);
110        start.AddSubtree(mainBranch);
111      }
112      return new SymbolicExpressionTree(root);
113    }
114
115    private IEnumerable<Token> GetTokenStream(string str) {
116      return
117             from strToken in str.Split(new string[] { " " }, StringSplitOptions.RemoveEmptyEntries).AsEnumerable()
118             let t = Token.Parse(strToken)
119             where t != null
120             select t;
121    }
122
123    private ISymbolicExpressionTreeNode ParseSexp(Queue<Token> tokens) {
124      if (tokens.Peek().Symbol == TokenSymbol.LPAR) {
125        ISymbolicExpressionTreeNode tree;
126        Expect(Token.LPAR, tokens);
127        if (tokens.Peek().StringValue.StartsWith(VARSTART)) {
128          tree = ParseVariable(tokens);
129        } else if (tokens.Peek().StringValue.StartsWith(LAGGEDVARSTART)) {
130          tree = ParseLaggedVariable(tokens);
131        } else if (tokens.Peek().StringValue.StartsWith(TIMELAGSTART)) {
132          tree = ParseTimeLag(tokens);
133          tree.AddSubtree(ParseSexp(tokens));
134        } else if (tokens.Peek().StringValue.StartsWith(INTEGRALSTART)) {
135          tree = ParseIntegral(tokens);
136          tree.AddSubtree(ParseSexp(tokens));
137        } else if (tokens.Peek().StringValue.StartsWith(DEFUNSTART)) {
138          tree = ParseDefun(tokens);
139          while (!tokens.Peek().Equals(Token.RPAR)) {
140            tree.AddSubtree(ParseSexp(tokens));
141          }
142        } else if (tokens.Peek().StringValue.StartsWith(ARGSTART)) {
143          tree = ParseArgument(tokens);
144        } else if (tokens.Peek().StringValue.StartsWith(INVOKESTART)) {
145          tree = ParseInvoke(tokens);
146          while (!tokens.Peek().Equals(Token.RPAR)) {
147            tree.AddSubtree(ParseSexp(tokens));
148          }
149        } else if (tokens.Peek().StringValue.StartsWith("FACTOR")) {
150          tree = ParseFactor(tokens);
151        } else if (tokens.Peek().StringValue.StartsWith("BINFACTOR")) {
152          tree = ParseBinaryFactor(tokens);
153        } else {
154          Token curToken = tokens.Dequeue();
155          tree = CreateTree(curToken);
156          while (!tokens.Peek().Equals(Token.RPAR)) {
157            tree.AddSubtree(ParseSexp(tokens));
158          }
159        }
160        Expect(Token.RPAR, tokens);
161        return tree;
162      } else if (tokens.Peek().Symbol == TokenSymbol.NUMBER) {
163        ConstantTreeNode t = (ConstantTreeNode)constant.CreateTreeNode();
164        t.Value = tokens.Dequeue().DoubleValue;
165        return t;
166      } else throw new FormatException("Expected function or constant symbol");
167    }
168
169    private ISymbolicExpressionTreeNode ParseInvoke(Queue<Token> tokens) {
170      Token invokeTok = tokens.Dequeue();
171      Debug.Assert(invokeTok.StringValue == "CALL");
172      InvokeFunction invokeSym = new InvokeFunction(tokens.Dequeue().StringValue);
173      ISymbolicExpressionTreeNode invokeNode = invokeSym.CreateTreeNode();
174      return invokeNode;
175    }
176
177    private ISymbolicExpressionTreeNode ParseArgument(Queue<Token> tokens) {
178      Token argTok = tokens.Dequeue();
179      Debug.Assert(argTok.StringValue == "ARG");
180      Argument argument = new Argument((int)tokens.Dequeue().DoubleValue);
181      ISymbolicExpressionTreeNode argNode = argument.CreateTreeNode();
182      return argNode;
183    }
184
185    private ISymbolicExpressionTreeNode ParseDefun(Queue<Token> tokens) {
186      Token defTok = tokens.Dequeue();
187      Debug.Assert(defTok.StringValue == "DEFUN");
188      DefunTreeNode t = (DefunTreeNode)defun.CreateTreeNode();
189      t.FunctionName = tokens.Dequeue().StringValue;
190      return t;
191    }
192
193    private ISymbolicExpressionTreeNode ParseTimeLag(Queue<Token> tokens) {
194      Token varTok = tokens.Dequeue();
195      Debug.Assert(varTok.StringValue == "LAG");
196      LaggedTreeNode t = (LaggedTreeNode)timeLag.CreateTreeNode();
197      t.Lag = (int)tokens.Dequeue().DoubleValue;
198      return t;
199    }
200
201    private ISymbolicExpressionTreeNode ParseIntegral(Queue<Token> tokens) {
202      Token varTok = tokens.Dequeue();
203      Debug.Assert(varTok.StringValue == "INTEGRAL");
204      LaggedTreeNode t = (LaggedTreeNode)integral.CreateTreeNode();
205      t.Lag = (int)tokens.Dequeue().DoubleValue;
206      return t;
207    }
208
209    private ISymbolicExpressionTreeNode ParseVariable(Queue<Token> tokens) {
210      Token varTok = tokens.Dequeue();
211      Debug.Assert(varTok.StringValue == "VARIABLE");
212      VariableTreeNode t = (VariableTreeNode)variable.CreateTreeNode();
213      t.Weight = tokens.Dequeue().DoubleValue;
214      t.VariableName = tokens.Dequeue().StringValue;
215      return t;
216    }
217
218    private ISymbolicExpressionTreeNode ParseFactor(Queue<Token> tokens) {
219      Token tok = tokens.Dequeue();
220      Debug.Assert(tok.StringValue == "FACTOR");
221      FactorVariableTreeNode t = (FactorVariableTreeNode)(new FactorVariable()).CreateTreeNode(); // create a new symbol each time on purpose
222      var varNameTok = tokens.Dequeue();
223      Debug.Assert(tok.Symbol == TokenSymbol.SYMB);
224      t.VariableName = varNameTok.StringValue;
225
226      var weights = new List<double>();
227      while (tokens.Peek().Symbol == TokenSymbol.NUMBER) {
228        weights.Add(tokens.Dequeue().DoubleValue);
229      }
230
231      t.Weights = weights.ToArray();
232
233      // create a set of (virtual) values to match the number of weights
234      t.Symbol.VariableNames = new string[] { t.VariableName };
235      t.Symbol.VariableValues = new[]
236      { new KeyValuePair<string, Dictionary<string,int>>(
237        t.VariableName,
238        weights.Select((_, i) => Tuple.Create(_,i)).ToDictionary(tup=>"X" + tup.Item2, tup=>tup.Item2)) };
239      return t;
240    }
241
242    private ISymbolicExpressionTreeNode ParseBinaryFactor(Queue<Token> tokens) {
243      Token tok = tokens.Dequeue();
244      Debug.Assert(tok.StringValue == "BINFACTOR");
245      var t = (BinaryFactorVariableTreeNode)binFactorVar.CreateTreeNode();
246      var varNameTok = tokens.Dequeue();
247      Debug.Assert(varNameTok.Symbol == TokenSymbol.SYMB);
248      t.VariableName = varNameTok.StringValue;
249
250      var varValTok = tokens.Dequeue();
251      Debug.Assert(varValTok.Symbol == TokenSymbol.SYMB);
252      t.VariableValue = varValTok.StringValue;
253
254      var weightTok = tokens.Dequeue();
255      Debug.Assert(weightTok.Symbol == TokenSymbol.NUMBER);
256      t.Weight = weightTok.DoubleValue;
257
258      return t;
259    }
260
261
262    private ISymbolicExpressionTreeNode ParseLaggedVariable(Queue<Token> tokens) {
263      Token varTok = tokens.Dequeue();
264      Debug.Assert(varTok.StringValue == "LAGVARIABLE");
265      LaggedVariableTreeNode t = (LaggedVariableTreeNode)laggedVariable.CreateTreeNode();
266      t.Weight = tokens.Dequeue().DoubleValue;
267      t.VariableName = tokens.Dequeue().StringValue;
268      t.Lag = (int)tokens.Dequeue().DoubleValue;
269      return t;
270    }
271
272    private ISymbolicExpressionTreeNode CreateTree(Token token) {
273      if (token.Symbol != TokenSymbol.SYMB) throw new FormatException("Expected function symbol, but got: " + token.StringValue);
274      return knownSymbols[token.StringValue].CreateTreeNode();
275    }
276
277    private void Expect(Token token, Queue<Token> tokens) {
278      Token cur = tokens.Dequeue();
279      if (!token.Equals(cur)) throw new FormatException("Expected: " + token.StringValue + ", but got: " + cur.StringValue);
280    }
281  }
282}
Note: See TracBrowser for help on using the repository browser.