1 | using System;
|
---|
2 | using System.Collections.Generic;
|
---|
3 | using System.Diagnostics;
|
---|
4 | using System.Linq;
|
---|
5 | using System.Text;
|
---|
6 | using HeuristicLab.Algorithms.Bandits;
|
---|
7 | using HeuristicLab.Common;
|
---|
8 | using HeuristicLab.Problems.GrammaticalOptimization;
|
---|
9 |
|
---|
10 | namespace HeuristicLab.Algorithms.GrammaticalOptimization {
|
---|
11 | public class AlternativesSampler {
|
---|
12 | public event Action<string, double> FoundNewBestSolution;
|
---|
13 | public event Action<string, double> SolutionEvaluated;
|
---|
14 |
|
---|
15 | private readonly int maxLen;
|
---|
16 | private readonly Random random;
|
---|
17 | private readonly IProblem problem;
|
---|
18 |
|
---|
19 | public AlternativesSampler(IProblem problem, int maxLen) {
|
---|
20 | this.problem = problem;
|
---|
21 | this.maxLen = maxLen;
|
---|
22 | this.random = new Random(31415);
|
---|
23 | }
|
---|
24 |
|
---|
25 | public void Run(int maxIterations) {
|
---|
26 | double bestQuality = double.MinValue;
|
---|
27 | InitPolicies(problem.Grammar);
|
---|
28 | for (int i = 0; i < maxIterations; i++) {
|
---|
29 | var sentence = SampleSentence(problem.Grammar);
|
---|
30 | var quality = problem.Evaluate(sentence) / problem.GetBestKnownQuality(maxLen);
|
---|
31 | DistributeReward(quality);
|
---|
32 |
|
---|
33 | RaiseSolutionEvaluated(sentence, quality);
|
---|
34 |
|
---|
35 | if (quality > bestQuality) {
|
---|
36 | bestQuality = quality;
|
---|
37 | RaiseFoundNewBestSolution(sentence, quality);
|
---|
38 | }
|
---|
39 | }
|
---|
40 | }
|
---|
41 |
|
---|
42 |
|
---|
43 | private Dictionary<char, IPolicy> ntPolicy;
|
---|
44 | private List<Tuple<char, int>> updateChain;
|
---|
45 |
|
---|
46 | private void InitPolicies(IGrammar grammar) {
|
---|
47 | this.ntPolicy = new Dictionary<char, IPolicy>();
|
---|
48 | this.updateChain = new List<Tuple<char, int>>();
|
---|
49 | foreach (var nt in grammar.NonTerminalSymbols) {
|
---|
50 | ntPolicy.Add(nt, new EpsGreedyPolicy(random, grammar.GetAlternatives(nt).Count(), 0.1));
|
---|
51 | }
|
---|
52 | }
|
---|
53 |
|
---|
54 | private string SampleSentence(IGrammar grammar) {
|
---|
55 | updateChain.Clear();
|
---|
56 | return CompleteSentence(grammar, grammar.SentenceSymbol.ToString());
|
---|
57 | }
|
---|
58 |
|
---|
59 | public string CompleteSentence(IGrammar g, string phrase) {
|
---|
60 | if (phrase.Length > maxLen) throw new ArgumentException();
|
---|
61 | if (g.MinPhraseLength(phrase) > maxLen) throw new ArgumentException();
|
---|
62 | bool done = phrase.All(g.IsTerminal); // terminal phrase means we are done
|
---|
63 | while (!done) {
|
---|
64 | int ntIdx; char nt;
|
---|
65 | Grammar.FindFirstNonTerminal(g, phrase, out nt, out ntIdx);
|
---|
66 |
|
---|
67 | int maxLenOfReplacement = maxLen - (phrase.Length - 1); // replacing aAb with maxLen 4 means we can only use alternatives with a minPhraseLen <= 2
|
---|
68 | Debug.Assert(maxLenOfReplacement > 0);
|
---|
69 |
|
---|
70 | var alts = g.GetAlternatives(nt);
|
---|
71 | string selectedAlt;
|
---|
72 | // if the choice is restricted then one of the allowed alternatives is selected randomly
|
---|
73 | if (alts.Any(alt => g.MinPhraseLength(alt) > maxLenOfReplacement)) {
|
---|
74 | var allowedAlts = alts.Where(alt => g.MinPhraseLength(alt) <= maxLenOfReplacement);
|
---|
75 | Debug.Assert(allowedAlts.Any());
|
---|
76 | // replace nt with random alternative
|
---|
77 | selectedAlt = allowedAlts.SelectRandom(random);
|
---|
78 | } else {
|
---|
79 | // all alts are allowed => select using bandit policy
|
---|
80 | var selectedAltIdx = ntPolicy[nt].SelectAction();
|
---|
81 | selectedAlt = alts.ElementAt(selectedAltIdx);
|
---|
82 | updateChain.Add(Tuple.Create(nt, selectedAltIdx));
|
---|
83 | }
|
---|
84 |
|
---|
85 | // replace nt with alt
|
---|
86 | phrase = phrase.Remove(ntIdx, 1);
|
---|
87 | phrase = phrase.Insert(ntIdx, selectedAlt);
|
---|
88 |
|
---|
89 | done = phrase.All(g.IsTerminal); // terminal phrase means we are done
|
---|
90 | }
|
---|
91 | return phrase;
|
---|
92 | }
|
---|
93 |
|
---|
94 |
|
---|
95 | private void DistributeReward(double reward) {
|
---|
96 | foreach (var e in updateChain) {
|
---|
97 | var nt = e.Item1;
|
---|
98 | var action = e.Item2;
|
---|
99 | ntPolicy[nt].UpdateReward(action, reward);
|
---|
100 | }
|
---|
101 | }
|
---|
102 |
|
---|
103 | private void RaiseSolutionEvaluated(string sentence, double quality) {
|
---|
104 | var handler = SolutionEvaluated;
|
---|
105 | if (handler != null) handler(sentence, quality);
|
---|
106 | }
|
---|
107 | private void RaiseFoundNewBestSolution(string sentence, double quality) {
|
---|
108 | var handler = FoundNewBestSolution;
|
---|
109 | if (handler != null) handler(sentence, quality);
|
---|
110 | }
|
---|
111 | }
|
---|
112 | }
|
---|