Free cookie consent management tool by TermsFeed Policy Generator

source: branches/2886_SymRegGrammarEnumeration/Test/GrammarEnumerationTest.cs @ 15716

Last change on this file since 15716 was 15714, checked in by lkammere, 7 years ago

#2886: Add tree hashing for addition and multiplication.
File size: 27.5 KB
Line 
1using System;
2using System.Collections.Generic;
3using System.Linq;
4using HeuristicLab.Algorithms.DataAnalysis.SymRegGrammarEnumeration;
5using HeuristicLab.Core;
6using HeuristicLab.Data;
7using HeuristicLab.Problems.DataAnalysis;
8using HeuristicLab.Random;
9using Microsoft.VisualStudio.TestTools.UnitTesting;
10
11namespace HeuristicLab.Algorithms.DataAnalysis.MctsSymbolicRegression {
12  //[TestClass()]
13  public class MctsSymbolicRegressionTest {
14    private const int Seed = 1234;
15    private IRandom rand;
16
17    private const double SuccessThreshold = 0.99999;
18
19    [TestInitialize]
20    public void InitTest() {
21      Console.Write("init called... ");
22      rand = new FastRandom(Seed);
23    }
24
25    #region test structure search (no constants)
26    [TestMethod]
27    [TestCategory("Algorithms.DataAnalysis")]
28    [TestProperty("Time", "short")]
29    public void MctsSymbReg_NoConstants_Nguyen1() {
30      // x³ + x² + x
31      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider(Seed);
32      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F1 ")));
33      TestGrammarEnumeration(regProblem);
34    }
35    [TestMethod]
36    [TestCategory("Algorithms.DataAnalysis")]
37    [TestProperty("Time", "short")]
38    public void MctsSymbReg_NoConstants_Nguyen2() {
39      // x^4 + x³ + x² + x
40      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider(Seed);
41      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F2 ")));
42      TestGrammarEnumeration(regProblem);
43    }
44    [TestMethod]
45    [TestCategory("Algorithms.DataAnalysis")]
46    [TestProperty("Time", "short")]
47    public void MctsSymbReg_NoConstants_Nguyen3() {
48      // x^5 + x^4 + x³ + x² + x
49      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider(Seed);
50      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F3 ")));
51      TestGrammarEnumeration(regProblem);
52    }
53    [TestMethod]
54    [TestCategory("Algorithms.DataAnalysis")]
55    [TestProperty("Time", "short")]
56    public void MctsSymbReg_NoConstants_Nguyen4() {
57      // x^6 + x^5 + x^4 + x³ + x² + x
58      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider(Seed);
59      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F4 ")));
60      TestGrammarEnumeration(regProblem);
61    }
62
63    [TestMethod]
64    [TestCategory("Algorithms.DataAnalysis")]
65    [TestProperty("Time", "short")]
66    public void MctsSymbReg_NoConstants_Nguyen7() {
67      // log(x + 1) + log(x² + 1)
68      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider(Seed);
69      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F7 ")));
70      TestGrammarEnumeration(regProblem);
71    }
72
73    [TestMethod]
74    [TestCategory("Algorithms.DataAnalysis")]
75    [TestProperty("Time", "short")]
76    public void MctsSymbReg_NoConstants_Poly10_Part1() {
77      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.VariousInstanceProvider(Seed);
78      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Poly-10")));
79
80      //  Y = X1*X2 + X3*X4 + X5*X6 + X1*X7*X9 + X3*X6*X10
81      //  Y' = X1*X2 + X3*X4 + X5*X6
82      // simplify problem by changing target
83      var ds = ((Dataset)regProblem.Dataset).ToModifiable();
84      var ys = ds.GetDoubleValues("Y").ToArray();
85      var x1 = ds.GetDoubleValues("X1").ToArray();
86      var x2 = ds.GetDoubleValues("X2").ToArray();
87      var x3 = ds.GetDoubleValues("X3").ToArray();
88      var x4 = ds.GetDoubleValues("X4").ToArray();
89      var x5 = ds.GetDoubleValues("X5").ToArray();
90      var x6 = ds.GetDoubleValues("X6").ToArray();
91      var x7 = ds.GetDoubleValues("X7").ToArray();
92      var x8 = ds.GetDoubleValues("X8").ToArray();
93      var x9 = ds.GetDoubleValues("X9").ToArray();
94      var x10 = ds.GetDoubleValues("X10").ToArray();
95      for (int i = 0; i < ys.Length; i++) {
96        ys[i] -= x1[i] * x7[i] * x9[i];
97        ys[i] -= x3[i] * x6[i] * x10[i];
98      }
99      ds.ReplaceVariable("Y", ys.ToList());
100
101      var modifiedProblemData = new RegressionProblemData(ds, regProblem.AllowedInputVariables, regProblem.TargetVariable);
102
103      TestGrammarEnumeration(modifiedProblemData);
104    }
105
106    [TestMethod]
107    [TestCategory("Algorithms.DataAnalysis")]
108    [TestProperty("Time", "short")]
109    public void MctsSymbReg_NoConstants_Poly10_Part2() {
110      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.VariousInstanceProvider(Seed);
111      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Poly-10")));
112
113      //  Y = X1*X2 + X3*X4 + X5*X6 + X1*X7*X9 + X3*X6*X10
114      //  Y' = X1*X7*X9 + X3*X6*X10
115      // simplify problem by changing target
116      var ds = ((Dataset)regProblem.Dataset).ToModifiable();
117      var ys = ds.GetDoubleValues("Y").ToArray();
118      var x1 = ds.GetDoubleValues("X1").ToArray();
119      var x2 = ds.GetDoubleValues("X2").ToArray();
120      var x3 = ds.GetDoubleValues("X3").ToArray();
121      var x4 = ds.GetDoubleValues("X4").ToArray();
122      var x5 = ds.GetDoubleValues("X5").ToArray();
123      var x6 = ds.GetDoubleValues("X6").ToArray();
124      var x7 = ds.GetDoubleValues("X7").ToArray();
125      var x8 = ds.GetDoubleValues("X8").ToArray();
126      var x9 = ds.GetDoubleValues("X9").ToArray();
127      var x10 = ds.GetDoubleValues("X10").ToArray();
128      for (int i = 0; i < ys.Length; i++) {
129        ys[i] -= x1[i] * x2[i];
130        ys[i] -= x3[i] * x4[i];
131        ys[i] -= x5[i] * x6[i];
132      }
133      ds.ReplaceVariable("Y", ys.ToList());
134
135      var modifiedProblemData = new RegressionProblemData(ds, regProblem.AllowedInputVariables, regProblem.TargetVariable);
136
137      TestGrammarEnumeration(modifiedProblemData);
138    }
139
140    [TestMethod]
141    [TestCategory("Algorithms.DataAnalysis")]
142    [TestProperty("Time", "short")]
143    public void MctsSymbReg_NoConstants_Poly10_Part3() {
144      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.VariousInstanceProvider(Seed);
145      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Poly-10")));
146
147      //  Y = X1*X2 + X3*X4 + X5*X6 + X1*X7*X9 + X3*X6*X10
148      //  Y' = X1*X2 + X1*X7*X9
149      // simplify problem by changing target
150      var ds = ((Dataset)regProblem.Dataset).ToModifiable();
151      var ys = ds.GetDoubleValues("Y").ToArray();
152      var x1 = ds.GetDoubleValues("X1").ToArray();
153      var x2 = ds.GetDoubleValues("X2").ToArray();
154      var x3 = ds.GetDoubleValues("X3").ToArray();
155      var x4 = ds.GetDoubleValues("X4").ToArray();
156      var x5 = ds.GetDoubleValues("X5").ToArray();
157      var x6 = ds.GetDoubleValues("X6").ToArray();
158      var x7 = ds.GetDoubleValues("X7").ToArray();
159      var x8 = ds.GetDoubleValues("X8").ToArray();
160      var x9 = ds.GetDoubleValues("X9").ToArray();
161      var x10 = ds.GetDoubleValues("X10").ToArray();
162      for (int i = 0; i < ys.Length; i++) {
163        ys[i] -= x3[i] * x4[i];
164        ys[i] -= x5[i] * x6[i];
165        ys[i] -= x3[i] * x6[i] * x10[i];
166      }
167      ds.ReplaceVariable("Y", ys.ToList());
168
169      var modifiedProblemData = new RegressionProblemData(ds, regProblem.AllowedInputVariables, regProblem.TargetVariable);
170
171      TestGrammarEnumeration(modifiedProblemData);
172    }
173
174    [TestMethod]
175    [TestCategory("Algorithms.DataAnalysis")]
176    [TestProperty("Time", "short")]
177    public void MctsSymbReg_NoConstants_Poly10_Part4() {
178      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.VariousInstanceProvider(Seed);
179      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Poly-10")));
180
181      //  Y = X1*X2 + X3*X4 + X5*X6 + X1*X7*X9 + X3*X6*X10
182      //  Y' = X3*X4 + X5*X6 + X3*X6*X10
183      // simplify problem by changing target
184      var ds = ((Dataset)regProblem.Dataset).ToModifiable();
185      var ys = ds.GetDoubleValues("Y").ToArray();
186      var x1 = ds.GetDoubleValues("X1").ToArray();
187      var x2 = ds.GetDoubleValues("X2").ToArray();
188      var x3 = ds.GetDoubleValues("X3").ToArray();
189      var x4 = ds.GetDoubleValues("X4").ToArray();
190      var x5 = ds.GetDoubleValues("X5").ToArray();
191      var x6 = ds.GetDoubleValues("X6").ToArray();
192      var x7 = ds.GetDoubleValues("X7").ToArray();
193      var x8 = ds.GetDoubleValues("X8").ToArray();
194      var x9 = ds.GetDoubleValues("X9").ToArray();
195      var x10 = ds.GetDoubleValues("X10").ToArray();
196      for (int i = 0; i < ys.Length; i++) {
197        ys[i] -= x1[i] * x2[i];
198        ys[i] -= x1[i] * x7[i] * x9[i];
199      }
200      ds.ReplaceVariable("Y", ys.ToList());
201      var modifiedProblemData = new RegressionProblemData(ds, regProblem.AllowedInputVariables, regProblem.TargetVariable);
202
203
204      TestGrammarEnumeration(modifiedProblemData);
205    }
206
207    [TestMethod]
208    [TestCategory("Algorithms.DataAnalysis")]
209    [TestProperty("Time", "short")]
210    public void MctsSymbReg_NoConstants_Poly10_Part5() {
211      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.VariousInstanceProvider(Seed);
212      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Poly-10")));
213
214      //  Y = X1*X2 + X3*X4 + X5*X6 + X1*X7*X9 + X3*X6*X10
215      //  Y' = X1*X2 + X3*X4 + X5*X6 + X1*X7*X9
216      // simplify problem by changing target
217      var ds = ((Dataset)regProblem.Dataset).ToModifiable();
218      var ys = ds.GetDoubleValues("Y").ToArray();
219      var x1 = ds.GetDoubleValues("X1").ToArray();
220      var x2 = ds.GetDoubleValues("X2").ToArray();
221      var x3 = ds.GetDoubleValues("X3").ToArray();
222      var x4 = ds.GetDoubleValues("X4").ToArray();
223      var x5 = ds.GetDoubleValues("X5").ToArray();
224      var x6 = ds.GetDoubleValues("X6").ToArray();
225      var x7 = ds.GetDoubleValues("X7").ToArray();
226      var x8 = ds.GetDoubleValues("X8").ToArray();
227      var x9 = ds.GetDoubleValues("X9").ToArray();
228      var x10 = ds.GetDoubleValues("X10").ToArray();
229      for (int i = 0; i < ys.Length; i++) {
230        ys[i] -= x3[i] * x6[i] * x10[i];
231      }
232      ds.ReplaceVariable("Y", ys.ToList());
233      var modifiedProblemData = new RegressionProblemData(ds, regProblem.AllowedInputVariables, regProblem.TargetVariable);
234
235
236      TestGrammarEnumeration(modifiedProblemData);
237    }
238
239    [TestMethod]
240    [TestCategory("Algorithms.DataAnalysis")]
241    [TestProperty("Time", "short")]
242    public void MctsSymbReg_NoConstants_Poly10_Part6() {
243      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.VariousInstanceProvider(Seed);
244      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Poly-10")));
245
246      //  Y = X1*X2 + X3*X4 + X5*X6 + X1*X7*X9 + X3*X6*X10
247      //  Y' = X1*X2 + X3*X4 + X5*X6 + X3*X6*X10
248      // simplify problem by changing target
249      var ds = ((Dataset)regProblem.Dataset).ToModifiable();
250      var ys = ds.GetDoubleValues("Y").ToArray();
251      var x1 = ds.GetDoubleValues("X1").ToArray();
252      var x2 = ds.GetDoubleValues("X2").ToArray();
253      var x3 = ds.GetDoubleValues("X3").ToArray();
254      var x4 = ds.GetDoubleValues("X4").ToArray();
255      var x5 = ds.GetDoubleValues("X5").ToArray();
256      var x6 = ds.GetDoubleValues("X6").ToArray();
257      var x7 = ds.GetDoubleValues("X7").ToArray();
258      var x8 = ds.GetDoubleValues("X8").ToArray();
259      var x9 = ds.GetDoubleValues("X9").ToArray();
260      var x10 = ds.GetDoubleValues("X10").ToArray();
261      for (int i = 0; i < ys.Length; i++) {
262        ys[i] -= x1[i] * x7[i] * x9[i];
263      }
264      ds.ReplaceVariable("Y", ys.ToList());
265      var modifiedProblemData = new RegressionProblemData(ds, regProblem.AllowedInputVariables, regProblem.TargetVariable);
266
267      TestGrammarEnumeration(modifiedProblemData);
268    }
269
270
271    [TestMethod]
272    [TestCategory("Algorithms.DataAnalysis")]
273    [TestProperty("Time", "long")]
274    public void MctsSymbReg_NoConstants_Poly10_250rows() {
275      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.VariousInstanceProvider(Seed);
276      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Poly-10")));
277      regProblem.TrainingPartition.Start = 0;
278      regProblem.TrainingPartition.End = regProblem.Dataset.Rows;
279      regProblem.TestPartition.Start = 0;
280      regProblem.TestPartition.End = 2;
281      TestGrammarEnumeration(regProblem);
282    }
283
284    [TestMethod]
285    [TestCategory("Algorithms.DataAnalysis")]
286    [TestProperty("Time", "long")]
287    public void MctsSymbReg_NoConstants_Poly10_10000rows() {
288      // as poly-10 but more rows
289      var x1 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
290      var x2 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
291      var x3 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
292      var x4 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
293      var x5 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
294      var x6 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
295      var x7 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
296      var x8 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
297      var x9 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
298      var x10 = Enumerable.Range(0, 10000).Select(_ => rand.NextDouble()).ToList();
299      var ys = new List<double>();
300      for (int i = 0; i < x1.Count; i++) {
301        ys.Add(x1[i] * x2[i] + x3[i] * x4[i] + x5[i] * x6[i] + x1[i] * x7[i] * x9[i] + x3[i] * x6[i] * x10[i]);
302      }
303
304      var ds = new Dataset(new string[] { "a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "y" },
305        new[] { x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, ys });
306
307
308      var problemData = new RegressionProblemData(ds, new string[] { "a", "b", "c", "d", "e", "f", "g", "h", "i", "j" }, "y");
309
310      problemData.TrainingPartition.Start = 0;
311      problemData.TrainingPartition.End = problemData.Dataset.Rows;
312      problemData.TestPartition.Start = 0;
313      problemData.TestPartition.End = 2; // must not be empty
314
315
316      TestGrammarEnumeration(problemData);
317    }
318
319    [TestMethod]
320    [TestCategory("Algorithms.DataAnalysis")]
321    [TestProperty("Time", "short")]
322    public void MctsSymbReg_NoConstants_TwoVars() {
323
324      // y = x1 + x2 + x1*x2 + x1*x2*x2 + x1*x1*x2
325      var x1 = Enumerable.Range(0, 100).Select(_ => rand.NextDouble()).ToList();
326      var x2 = Enumerable.Range(0, 100).Select(_ => rand.NextDouble()).ToList();
327      var ys = x1.Zip(x2, (x1i, x2i) => x1i + x2i + x1i * x2i + x1i * x2i * x2i + x1i * x1i * x2i).ToList();
328
329      var ds = new Dataset(new string[] { "a", "b", "y" }, new[] { x1, x2, ys });
330      var problemData = new RegressionProblemData(ds, new string[] { "a", "b" }, "y");
331
332      TestGrammarEnumeration(problemData);
333    }
334
335    [TestMethod]
336    [TestCategory("Algorithms.DataAnalysis")]
337    [TestProperty("Time", "short")]
338    public void MctsSymbReg_NoConstants_Misleading() {
339
340      // y = a + baaaaa (the effect of the second term should be very small)
341      // the alg will quickly find that a has big effect and will search below a
342      // since we prevent a + a... the algorithm must find the correct expression via a + b...
343      // however b has a small effect so the branch might not be identified as relevant
344      var @as = Enumerable.Range(0, 100).Select(_ => rand.NextDouble()).ToList();
345      var bs = Enumerable.Range(0, 100).Select(_ => rand.NextDouble()).ToList();
346      var cs = Enumerable.Range(0, 100).Select(_ => rand.NextDouble() * 1.0e-3).ToList();
347      var ds = Enumerable.Range(0, 100).Select(_ => rand.NextDouble()).ToList();
348      var es = Enumerable.Range(0, 100).Select(_ => rand.NextDouble()).ToList();
349      var ys = new double[@as.Count];
350      for (int i = 0; i < ys.Length; i++)
351        ys[i] = @as[i] + bs[i] + @as[i] * bs[i] * cs[i];
352
353      var dataset = new Dataset(new string[] { "a", "b", "c", "d", "e", "y" }, new[] { @as, bs, cs, ds, es, ys.ToList() });
354
355      var problemData = new RegressionProblemData(dataset, new string[] { "a", "b", "c", "d", "e" }, "y");
356
357      TestGrammarEnumeration(problemData);
358    }
359    #endregion
360
361    #region restricted structure but including numeric constants
362
363    [TestMethod]
364    [TestCategory("Algorithms.DataAnalysis")]
365    [TestProperty("Time", "short")]
366    public void MctsSymbRegKeijzer7() {
367      // ln(x)
368      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.KeijzerInstanceProvider(Seed);
369      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Keijzer 7 f(")));
370      // some Keijzer problem instances have very large test partitions (here we are not concerened about test performance)
371      if (regProblem.TestPartition.End - regProblem.TestPartition.Start > 1000) regProblem.TestPartition.End = regProblem.TestPartition.Start + 1000;
372      TestGrammarEnumeration(regProblem);
373    }
374
375   
376    [TestMethod]
377    [TestCategory("Algorithms.DataAnalysis")]
378    [TestProperty("Time", "short")]
379    public void MctsSymbRegBenchmarkNguyen5() {
380      // sin(x²)cos(x) - 1
381      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider();
382      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F5 ")));
383      TestGrammarEnumeration(regProblem);
384    }
385
386    [TestMethod]
387    [TestCategory("Algorithms.DataAnalysis")]
388    [TestProperty("Time", "short")]
389    public void MctsSymbRegBenchmarkNguyen6() {
390      // sin(x) + sin(x + x²)
391      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider();
392      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F6 ")));
393      TestGrammarEnumeration(regProblem);
394    }
395   
396    [TestMethod]
397    [TestCategory("Algorithms.DataAnalysis")]
398    [TestProperty("Time", "short")]
399    public void MctsSymbRegBenchmarkNguyen7() {
400      //  log(x + 1) + log(x² + 1)
401      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider(Seed);
402      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F7 ")));
403      TestGrammarEnumeration(regProblem);
404    }
405    [TestMethod]
406    [TestCategory("Algorithms.DataAnalysis")]
407    [TestProperty("Time", "short")]
408    public void MctsSymbRegBenchmarkNguyen8() {
409      // Sqrt(x)
410      // = x ^ 0.5
411      // = exp(0.5 * log(x))
412      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider(Seed);
413      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F8 ")));
414      TestGrammarEnumeration(regProblem);
415    }
416   
417    // [TestMethod]
418    [TestCategory("Algorithms.DataAnalysis")]
419    [TestProperty("Time", "short")]
420    public void MctsSymbRegBenchmarkNguyen9() {
421      //  sin(x) + sin(y²)
422      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider();
423      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F9 ")));
424      TestGrammarEnumeration(regProblem);
425    }
426
427    // [TestMethod]
428    [TestCategory("Algorithms.DataAnalysis")]
429    [TestProperty("Time", "short")]
430    public void MctsSymbRegBenchmarkNguyen10() {
431      // 2sin(x)cos(y)
432      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider();
433      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F10 ")));
434      TestGrammarEnumeration(regProblem);
435    }
436   
437    [TestMethod]
438    [TestCategory("Algorithms.DataAnalysis")]
439    [TestProperty("Time", "short")]
440    public void MctsSymbRegBenchmarkNguyen11() {
441      // x ^ y  , x > 0, y > 0   
442      // = exp(y * log(x))
443      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider(Seed);
444      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F11 ")));
445      TestGrammarEnumeration(regProblem);
446    }
447    [TestMethod]
448    [TestCategory("Algorithms.DataAnalysis")]
449    [TestProperty("Time", "short")]
450    public void MctsSymbRegBenchmarkNguyen12() {
451      // x^4 - x³ + y²/2 - y
452      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.NguyenInstanceProvider(Seed);
453      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("F12 ")));
454      TestGrammarEnumeration(regProblem);
455    }
456
457    #endregion
458
459    #region keijzer
460    [TestMethod]
461    [TestCategory("Algorithms.DataAnalysis")]
462    [TestProperty("Time", "long")]
463    public void MctsSymbRegBenchmarkKeijzer5() {
464      // (30 * x * z) / ((x - 10)  * y²)
465      // = 30 x z / (xy² - y²)
466      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.KeijzerInstanceProvider(Seed);
467      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Keijzer 5 f(")));
468      // some Keijzer problem instances have very large test partitions (here we are not concerened about test performance)
469      if (regProblem.TestPartition.End - regProblem.TestPartition.Start > 1000) regProblem.TestPartition.End = regProblem.TestPartition.Start + 1000;
470      TestGrammarEnumeration(regProblem);
471    }
472
473    [TestMethod]
474    [TestCategory("Algorithms.DataAnalysis")]
475    [TestProperty("Time", "short")]
476    public void MctsSymbRegBenchmarkKeijzer6() {
477      // Keijzer 6 f(x) = Sum(1 / i) From 1 to X  , x \in [0..120]
478      // we can only approximate this
479      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.KeijzerInstanceProvider(Seed);
480      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Keijzer 6 f(")));
481      // some Keijzer problem instances have very large test partitions (here we are not concerened about test performance)
482      if (regProblem.TestPartition.End - regProblem.TestPartition.Start > 1000) regProblem.TestPartition.End = regProblem.TestPartition.Start + 1000;
483      TestGrammarEnumeration(regProblem);
484    }
485
486    [TestMethod]
487    [TestCategory("Algorithms.DataAnalysis")]
488    [TestProperty("Time", "short")]
489    public void MctsSymbRegBenchmarkKeijzer8() {
490      // sqrt(x)
491      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.KeijzerInstanceProvider(Seed);
492      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Keijzer 8 f(")));
493      // some Keijzer problem instances have very large test partitions (here we are not concerened about test performance)
494      if (regProblem.TestPartition.End - regProblem.TestPartition.Start > 1000) regProblem.TestPartition.End = regProblem.TestPartition.Start + 1000;
495      TestGrammarEnumeration(regProblem);
496    }
497
498    [TestMethod]
499    [TestCategory("Algorithms.DataAnalysis")]
500    [TestProperty("Time", "short")]
501    public void MctsSymbRegBenchmarkKeijzer9() {
502      // arcsinh(x)  i.e. ln(x + sqrt(x² + 1))
503      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.KeijzerInstanceProvider(Seed);
504      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Keijzer 9 f(")));
505      // some Keijzer problem instances have very large test partitions (here we are not concerened about test performance)
506      if (regProblem.TestPartition.End - regProblem.TestPartition.Start > 1000) regProblem.TestPartition.End = regProblem.TestPartition.Start + 1000;
507      TestGrammarEnumeration(regProblem);
508    }
509
510    [TestMethod]
511    [TestCategory("Algorithms.DataAnalysis")]
512    [TestProperty("Time", "short")]
513    public void MctsSymbRegBenchmarkKeijzer11() {
514      // xy + sin( (x-1) (y-1) )
515      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.KeijzerInstanceProvider();
516      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Keijzer 11 f(")));
517      // some Keijzer problem instances have very large test partitions (here we are not concerened about test performance)
518      if (regProblem.TestPartition.End - regProblem.TestPartition.Start > 1000) regProblem.TestPartition.End = regProblem.TestPartition.Start + 1000;
519      TestGrammarEnumeration(regProblem);
520    }
521
522    [TestMethod]
523    [TestCategory("Algorithms.DataAnalysis")]
524    [TestProperty("Time", "short")]
525    public void MctsSymbRegBenchmarkKeijzer12() {
526      // x^4 - x³ + y² / 2 - y,  same as Nguyen 12             
527      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.KeijzerInstanceProvider(Seed);
528      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Keijzer 12 f(")));
529      // some Keijzer problem instances have very large test partitions (here we are not concerened about test performance)
530      if (regProblem.TestPartition.End - regProblem.TestPartition.Start > 1000) regProblem.TestPartition.End = regProblem.TestPartition.Start + 1000;
531      TestGrammarEnumeration(regProblem);
532    }
533
534    [TestMethod]
535    [TestCategory("Algorithms.DataAnalysis")]
536    [TestProperty("Time", "short")]
537    public void MctsSymbRegBenchmarkKeijzer14() {
538      // 8 / (2 + x² + y²)
539      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.KeijzerInstanceProvider(Seed);
540      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Keijzer 14 f(")));
541      // some Keijzer problem instances have very large test partitions (here we are not concerened about test performance)
542      if (regProblem.TestPartition.End - regProblem.TestPartition.Start > 1000) regProblem.TestPartition.End = regProblem.TestPartition.Start + 1000;
543      TestGrammarEnumeration(regProblem);
544    }
545
546    [TestMethod]
547    [TestCategory("Algorithms.DataAnalysis")]
548    [TestProperty("Time", "short")]
549    public void MctsSymbRegBenchmarkKeijzer15() {
550      // x³ / 5 + y³ / 2 - y - x
551      var provider = new HeuristicLab.Problems.Instances.DataAnalysis.KeijzerInstanceProvider(Seed);
552      var regProblem = provider.LoadData(provider.GetDataDescriptors().Single(x => x.Name.Contains("Keijzer 15 f(")));
553      // some Keijzer problem instances have very large test partitions (here we are not concerened about test performance)
554      if (regProblem.TestPartition.End - regProblem.TestPartition.Start > 1000) regProblem.TestPartition.End = regProblem.TestPartition.Start + 1000;
555      TestGrammarEnumeration(regProblem);
556    }
557    #endregion
558
559    private void TestGrammarEnumeration(IRegressionProblemData problemData)
560    {
561      // Configure algorithm and problem
562      var alg = new GrammarEnumerationAlgorithm();
563      var problem = new RegressionProblem();
564      alg.Problem = problem;
565      problem.ProblemData = problemData;
566     
567      // Run
568      alg.Start();
569
570      // Evaluate results
571      var eps = 1.0 - SuccessThreshold;
572
573      // Check if algorithm terminated correctly
574      Assert.IsTrue(alg.Results.ContainsKey("Best solution quality (training)"), "No training solution returned!");
575      Assert.IsTrue(alg.Results.ContainsKey("Best solution quality (test)"), "No test solution returned!");
576
577      // Check resultss
578      Assert.AreEqual(1.0, ((DoubleValue)alg.Results["Best solution quality (training)"].Value).Value, eps, "Training quality too low!");
579      Assert.AreEqual(1.0, ((DoubleValue)alg.Results["Best solution quality (test)"].Value).Value, eps, "Test quality too low!");
580    }
581  }
582}
Note: See TracBrowser for help on using the repository browser.