1 | using System;
|
---|
2 | using System.Collections.Generic;
|
---|
3 | using System.Linq;
|
---|
4 | using HeuristicLab.Common;
|
---|
5 | using HeuristicLab.Random;
|
---|
6 |
|
---|
7 | namespace HeuristicLab.Problems.Instances.DataAnalysis {
|
---|
8 | public class Feynman13 : FeynmanDescriptor {
|
---|
9 | private readonly int testSamples;
|
---|
10 | private readonly int trainingSamples;
|
---|
11 |
|
---|
12 | public Feynman13() : this((int) DateTime.Now.Ticks, 10000, 10000, null) { }
|
---|
13 |
|
---|
14 | public Feynman13(int seed) {
|
---|
15 | Seed = seed;
|
---|
16 | trainingSamples = 10000;
|
---|
17 | testSamples = 10000;
|
---|
18 | noiseRatio = null;
|
---|
19 | }
|
---|
20 |
|
---|
21 | public Feynman13(int seed, int trainingSamples, int testSamples, double? noiseRatio) {
|
---|
22 | Seed = seed;
|
---|
23 | this.trainingSamples = trainingSamples;
|
---|
24 | this.testSamples = testSamples;
|
---|
25 | this.noiseRatio = noiseRatio;
|
---|
26 | }
|
---|
27 |
|
---|
28 | public override string Name {
|
---|
29 | get {
|
---|
30 | return string.Format("I.13.4 1/2*m*(v**2+u**2+w**2) | {0} samples | noise ({1})", trainingSamples,
|
---|
31 | noiseRatio == null ? "no noise" : noiseRatio.ToString());
|
---|
32 | }
|
---|
33 | }
|
---|
34 |
|
---|
35 | protected override string TargetVariable { get { return noiseRatio == null ? "K" : "K_noise"; } }
|
---|
36 |
|
---|
37 | protected override string[] VariableNames {
|
---|
38 | get { return new[] {"m", "v", "u", "w", noiseRatio == null ? "K" : "K_noise"}; }
|
---|
39 | }
|
---|
40 |
|
---|
41 | protected override string[] AllowedInputVariables { get { return new[] {"m", "v", "u", "w"}; } }
|
---|
42 |
|
---|
43 | public int Seed { get; private set; }
|
---|
44 |
|
---|
45 | protected override int TrainingPartitionStart { get { return 0; } }
|
---|
46 | protected override int TrainingPartitionEnd { get { return trainingSamples; } }
|
---|
47 | protected override int TestPartitionStart { get { return trainingSamples; } }
|
---|
48 | protected override int TestPartitionEnd { get { return trainingSamples + testSamples; } }
|
---|
49 |
|
---|
50 | protected override List<List<double>> GenerateValues() {
|
---|
51 | var rand = new MersenneTwister((uint) Seed);
|
---|
52 |
|
---|
53 | var data = new List<List<double>>();
|
---|
54 | var m = ValueGenerator.GenerateUniformDistributedValues(rand.Next(), TestPartitionEnd, 1, 5).ToList();
|
---|
55 | var v = ValueGenerator.GenerateUniformDistributedValues(rand.Next(), TestPartitionEnd, 1, 5).ToList();
|
---|
56 | var u = ValueGenerator.GenerateUniformDistributedValues(rand.Next(), TestPartitionEnd, 1, 5).ToList();
|
---|
57 | var w = ValueGenerator.GenerateUniformDistributedValues(rand.Next(), TestPartitionEnd, 1, 5).ToList();
|
---|
58 |
|
---|
59 | var K = new List<double>();
|
---|
60 |
|
---|
61 | data.Add(m);
|
---|
62 | data.Add(v);
|
---|
63 | data.Add(u);
|
---|
64 | data.Add(w);
|
---|
65 | data.Add(K);
|
---|
66 |
|
---|
67 | for (var i = 0; i < m.Count; i++) {
|
---|
68 | var res = 1.0 / 2 * m[i] * (Math.Pow(v[i], 2) + Math.Pow(u[i], 2) + Math.Pow(w[i], 2));
|
---|
69 | K.Add(res);
|
---|
70 | }
|
---|
71 |
|
---|
72 | if (noiseRatio != null) {
|
---|
73 | var K_noise = new List<double>();
|
---|
74 | var sigma_noise = (double) noiseRatio * K.StandardDeviationPop();
|
---|
75 | K_noise.AddRange(K.Select(md => md + NormalDistributedRandom.NextDouble(rand, 0, sigma_noise)));
|
---|
76 | data.Remove(K);
|
---|
77 | data.Add(K_noise);
|
---|
78 | }
|
---|
79 |
|
---|
80 | return data;
|
---|
81 | }
|
---|
82 | }
|
---|
83 | } |
---|