source: trunk/EventBenchCore/src/de/ugoe/cs/eventbench/ppm/PredictionByPartialMatch.java @ 6

package de.ugoe.cs.eventbench.ppm;

import java.util.ArrayList;
import java.util.LinkedHashSet;
import java.util.LinkedList;
import java.util.List;
import java.util.Random;
import java.util.Set;

import de.ugoe.cs.eventbench.data.Event;
import de.ugoe.cs.eventbench.markov.IncompleteMemory;
import de.ugoe.cs.util.console.Console;

public class PredictionByPartialMatch {
        
        private String initialSymbol = "GS";
        private String endSymbol = "GE";
        
        private int maxOrder = 3;
        
        private Trie<String> trie;
        
        private Set<String> knownSymbols;
        
        private double probEscape = 0.2d; // TODO getter/setter - steering parameter!
        
        private Random r = new Random(); // TODO should be defined in the constructor
        
        // the training is basically the generation of the trie
        public void train(List<List<Event<?>>> sequences) {
                trie = new Trie<String>();
                knownSymbols = new LinkedHashSet<String>();
                knownSymbols.add(initialSymbol);
                knownSymbols.add(endSymbol);
                
                for(List<Event<?>> sequence : sequences) {
                        List<String> stringSequence = new LinkedList<String>();
                        stringSequence.add(initialSymbol);
                        for( Event<?> event : sequence ) {
                                stringSequence.add(event.getStandardId());
                        }
                        stringSequence.add(endSymbol);
                        
                        trainStringTrie(stringSequence);
                }
        }
        
        private void trainStringTrie(List<String> sequence) {
                knownSymbols = new LinkedHashSet<String>();             
                IncompleteMemory<String> latestActions = new IncompleteMemory<String>(maxOrder);
                int i=0;
                for(String currentAction : sequence) {
                        String currentId = currentAction;
                        latestActions.add(currentId);
                        knownSymbols.add(currentId);
                        i++;
                        if( i>=maxOrder ) {
                                trie.add(latestActions.getLast(maxOrder));
                        }
                }
                int sequenceLength = sequence.size();
                for( int j=maxOrder-1 ; j>0 ; j-- ) {
                        trie.add(sequence.subList(sequenceLength-j, sequenceLength));
                }
        }
        
        // TODO needs to be changed from String to <? extends Event>
        public List<String> randomSequence() {
                List<String> sequence = new LinkedList<String>();
                
                IncompleteMemory<String> context = new IncompleteMemory<String>(maxOrder-1);
                context.add(initialSymbol);
                sequence.add(initialSymbol);
                
                String currentState = initialSymbol;
                
                Console.println(currentState);
                while(!endSymbol.equals(currentState)) {
                        double randVal = r.nextDouble();
                        double probSum = 0.0;
                        List<String> currentContext = context.getLast(maxOrder);
                        for( String symbol : knownSymbols ) {
                                probSum += getProbability(currentContext, symbol);
                                if( probSum>=randVal ) {
                                        context.add(symbol);
                                        currentState = symbol;
                                        sequence.add(currentState);
                                        break;
                                }
                        }
                }
                return sequence;
        }
                
        private double getProbability(List<String> context, String symbol) {
                double result = 0.0d;
                double resultCurrentContex = 0.0d;
                double resultShorterContex = 0.0d;
                
                List<String> contextCopy = new LinkedList<String>(context); // defensive copy

        
                List<String> followers = trie.getFollowingSymbols(contextCopy); // \Sigma'
                int sumCountFollowers = 0; // N(s\sigma')
                for( String follower : followers ) {
                        sumCountFollowers += trie.getCount(contextCopy, follower);
                }
                
                int countSymbol = trie.getCount(contextCopy, symbol); // N(s\sigma)
                if( contextCopy.size()==0 ) {
                        resultCurrentContex = ((double) countSymbol) / sumCountFollowers;
                } else {
                        resultCurrentContex = ((double) countSymbol / sumCountFollowers)*(1-probEscape);
                        contextCopy.remove(0); 
                        double probSuffix = getProbability(contextCopy, symbol);
                        if( followers.size()==0 ) {
                                resultShorterContex = probSuffix;
                        } else {
                                resultShorterContex = probEscape*probSuffix;
                        }
                }
                result = resultCurrentContex+resultShorterContex;
                
                return result;
        }
        
        @Override
        public String toString() {
                return trie.toString();
        }
        
        public void testStuff() {
                // basically an inline unit test without assertions but manual observation
                List<String> list = new ArrayList<String>();
                list.add(initialSymbol);
                list.add("a");
                list.add("b");
                list.add("r");
                list.add("a");
                list.add("c");
                list.add("a");
                list.add("d");
                list.add("a");
                list.add("b");
                list.add("r");
                list.add("a");
                list.add(endSymbol);
                
                PredictionByPartialMatch model = new PredictionByPartialMatch();
                model.trie = new Trie<String>();
                model.trainStringTrie(list);
                model.trie.display();
                Console.println("------------------------");
                model.randomSequence();/*
                Console.println("------------------------");
                model.randomSequence();
                Console.println("------------------------");
                model.randomSequence();
                Console.println("------------------------");*/
                
                List<String> context = new ArrayList<String>();
                String symbol = "a";
                // expected: 5
                Console.traceln(""+model.trie.getCount(context, symbol));
                
                // expected: 0
                context.add("b");
                Console.traceln(""+model.trie.getCount(context, symbol));
                
                // expected: 2
                context.add("r");
                Console.traceln(""+model.trie.getCount(context, symbol));
                
                // exptected: [b, r]
                context = new ArrayList<String>();
                context.add("a");
                context.add("b");
                context.add("r");
                Console.traceln(model.trie.getContextSuffix(context).toString());
                
                // exptected: []
                context = new ArrayList<String>();
                context.add("e");
                Console.traceln(model.trie.getContextSuffix(context).toString());
                
                // exptected: {a, b, c, d, r}
                context = new ArrayList<String>();
                Console.traceln(model.trie.getFollowingSymbols(context).toString());
                
                // exptected: {b, c, d}
                context = new ArrayList<String>();
                context.add("a");
                Console.traceln(model.trie.getFollowingSymbols(context).toString());
                
                // exptected: []
                context = new ArrayList<String>();
                context.add("a");
                context.add("b");
                context.add("r");
                Console.traceln(model.trie.getFollowingSymbols(context).toString());
                
                // exptected: 0.0d
                context = new ArrayList<String>();
                context.add("a");
                Console.traceln(""+model.getProbability(context, "z"));
        }
}