Grammar.java

package edu.berkeley.nlp.PCFGLA;

import edu.berkeley.nlp.PCFGLA.smoothing.*;
import edu.berkeley.nlp.math.SloppyMath;
import edu.berkeley.nlp.syntax.StateSet;
import edu.berkeley.nlp.syntax.Tree;
import edu.berkeley.nlp.util.*;
import edu.berkeley.nlp.util.PriorityQueue;

import java.io.IOException;
import java.io.PrintWriter;
import java.io.Writer;
import java.util.*;

import opennlp.maxent.GIS;
import opennlp.maxent.GISModel;

/**
 * Simple implementation of a PCFG grammar, offering the ability to look up
 * rules by their child symbols. Rule probability estimates are just relative
 * frequency estimates off of training trees.
 */
public class Grammar implements java.io.Serializable {

	/**
	 * @author leon
	 * 
	 */
	public static enum RandomInitializationType {
		INITIALIZE_WITH_SMALL_RANDOMIZATION, INITIALIZE_LIKE_MMT
		// initialize like in the Matzuyaki, Miyao, and Tsujii paper
	}

	public static class RuleNotFoundException extends Exception {
		private static final long serialVersionUID = 2L;
	}

	public int finalLevel;

	public boolean[] isGrammarTag;
	public boolean useEntropicPrior = false;

	private List<BinaryRule>[] binaryRulesWithParent;
	private List<BinaryRule>[] binaryRulesWithLC;
	private List<BinaryRule>[] binaryRulesWithRC;
	private BinaryRule[][] splitRulesWithLC;
	private BinaryRule[][] splitRulesWithRC;
	private BinaryRule[][] splitRulesWithP;
	public List<UnaryRule>[] unaryRulesWithParent;
	public List<UnaryRule>[] unaryRulesWithC;
	private List<UnaryRule>[] sumProductClosedUnaryRulesWithParent;

	/** the number of states */
	public short numStates;

	/** the number of substates per state */
	public short[] numSubStates;

	// private List<Rule> allRules;

	public Map<BinaryRule, BinaryRule> binaryRuleMap;
	BinaryRule bSearchRule;
	public Map<UnaryRule, UnaryRule> unaryRuleMap;
	UnaryRule uSearchRule;

	UnaryCounterTable unaryRuleCounter = null;

	BinaryCounterTable binaryRuleCounter = null;

	CounterMap<Integer, Integer> symbolCounter = new CounterMap<Integer, Integer>();

	private static final long serialVersionUID = 1L;

	protected Numberer tagNumberer;

	public List<UnaryRule>[] closedSumRulesWithParent = null;
	public List<UnaryRule>[] closedSumRulesWithChild = null;

	public List<UnaryRule>[] closedViterbiRulesWithParent = null;
	public List<UnaryRule>[] closedViterbiRulesWithChild = null;

	public UnaryRule[][] closedSumRulesWithP = null;
	public UnaryRule[][] closedSumRulesWithC = null;

	public UnaryRule[][] closedViterbiRulesWithP = null;
	public UnaryRule[][] closedViterbiRulesWithC = null;

	private Map bestSumRulesUnderMax = null;
	private Map bestViterbiRulesUnderMax = null;
	public double threshold;

	public Smoother smoother = null;

	/**
	 * A policy giving what state to go to next, starting from a given state,
	 * going to a given state. This array is indexed by the start state, the end
	 * state, the start substate, and the end substate.
	 */
	private int[][] closedViterbiPaths = null;
	private int[][] closedSumPaths = null;

	public boolean findClosedPaths;

	/**
	 * If we are in logarithm mode, then this grammar's scores are all given as
	 * logarithms. The default is to have a score plus a scale factor.
	 */
	boolean logarithmMode;

	public Tree<Short>[] splitTrees;

	public void clearUnaryIntermediates() {
		ArrayUtil.fill(closedSumPaths, 0);
		ArrayUtil.fill(closedViterbiPaths, 0);
	}

	public void addBinary(BinaryRule br) {
		// System.out.println("BG adding rule " + br);
		binaryRulesWithParent[br.parentState].add(br);
		binaryRulesWithLC[br.leftChildState].add(br);
		binaryRulesWithRC[br.rightChildState].add(br);
		// allRules.add(br);
		binaryRuleMap.put(br, br);
	}

	public void addUnary(UnaryRule ur) {
		// System.out.println(" UG adding rule " + ur);
		// closeRulesUnderMax(ur);
		if (!unaryRulesWithParent[ur.parentState].contains(ur)) {
			unaryRulesWithParent[ur.parentState].add(ur);
			unaryRulesWithC[ur.childState].add(ur);
			// allRules.add(ur);
			unaryRuleMap.put(ur, ur);
		}
	}

	public Numberer getTagNumberer() {
		return tagNumberer;
	}

	// @SuppressWarnings("unchecked")
	// public List<BinaryRule> getBinaryRulesByParent(int state) {
	// if (state >= binaryRulesWithParent.length) {
	// return Collections.EMPTY_LIST;
	// }
	// return binaryRulesWithParent[state];
	// }
	//	
	@SuppressWarnings("unchecked")
	public List<UnaryRule> getUnaryRulesByParent(int state) {
		if (state >= unaryRulesWithParent.length) {
			return Collections.EMPTY_LIST;
		}
		return unaryRulesWithParent[state];
	}

	@SuppressWarnings("unchecked")
	public List<UnaryRule>[] getSumProductClosedUnaryRulesByParent() {
		return sumProductClosedUnaryRulesWithParent;
	}

	@SuppressWarnings("unchecked")
	public List<BinaryRule> getBinaryRulesByLeftChild(int state) {
		// System.out.println("getBinaryRulesByLeftChild not supported anymore.");
		// return null;
		if (state >= binaryRulesWithLC.length) {
			return Collections.EMPTY_LIST;
		}
		return binaryRulesWithLC[state];
	}

	@SuppressWarnings("unchecked")
	public List<BinaryRule> getBinaryRulesByRightChild(int state) {
		// System.out.println("getBinaryRulesByRightChild not supported anymore.");
		// return null;
		if (state >= binaryRulesWithRC.length) {
			return Collections.EMPTY_LIST;
		}
		return binaryRulesWithRC[state];
	}

	@SuppressWarnings("unchecked")
	public List<UnaryRule> getUnaryRulesByChild(int state) {
		// System.out.println("getUnaryRulesByChild not supported anymore.");
		// return null;
		if (state >= unaryRulesWithC.length) {
			return Collections.EMPTY_LIST;
		}
		return unaryRulesWithC[state];
	}

	public String toString_old() {
		/*
		 * StringBuilder sb = new StringBuilder(); List<String> ruleStrings =
		 * new ArrayList<String>(); for (int state = 0; state < numStates;
		 * state++) { List<BinaryRule> leftRules =
		 * getBinaryRulesByLeftChild(state); for (BinaryRule r : leftRules) {
		 * ruleStrings.add(r.toString()); } } for (int state = 0; state <
		 * numStates; state++) { UnaryRule[] unaries =
		 * getClosedViterbiUnaryRulesByChild(state); for (int r = 0; r <
		 * unaries.length; r++) { UnaryRule ur = unaries[r];
		 * ruleStrings.add(ur.toString()); } } for (String ruleString :
		 * CollectionUtils.sort(ruleStrings)) { sb.append(ruleString);
		 * sb.append("\n"); }
		 */
		return null;// sb.toString();
	}

	public void writeData(Writer w) throws IOException {
		finalLevel = (short) (Math.log(numSubStates[1]) / Math.log(2));
		PrintWriter out = new PrintWriter(w);
		for (int state = 0; state < numStates; state++) {
			BinaryRule[] parentRules = this.splitRulesWithP(state);
			for (int i = 0; i < parentRules.length; i++) {
				BinaryRule r = parentRules[i];
				out.print(r.toString());
			}
		}
		for (int state = 0; state < numStates; state++) {
			UnaryRule[] unaries = this
					.getClosedViterbiUnaryRulesByParent(state);
			for (int r = 0; r < unaries.length; r++) {
				UnaryRule ur = unaries[r];
				out.print(ur.toString());
			}
		}
		out.flush();
	}

	public String toString() {
		// splitRules();
		StringBuilder sb = new StringBuilder();
		List<String> ruleStrings = new ArrayList<String>();
		for (int state = 0; state < numStates; state++) {
			BinaryRule[] parentRules = this.splitRulesWithP(state);
			for (int i = 0; i < parentRules.length; i++) {
				BinaryRule r = parentRules[i];
				ruleStrings.add(r.toString());
			}
		}
		for (int state = 0; state < numStates; state++) {
			UnaryRule[] unaries = this.getClosedSumUnaryRulesByParent(state);
			// this.getClosedSumUnaryRulesByParent(state);//
			for (int r = 0; r < unaries.length; r++) {
				UnaryRule ur = unaries[r];
				ruleStrings.add(ur.toString());
			}
			// UnaryRule[] unaries2 =
			// this.getClosedViterbiUnaryRulesByParent(state);
			// for (int r = 0; r < unaries2.length; r++) {
			// UnaryRule ur = unaries2[r];
			// ruleStrings.add(ur.toString());
			// }
		}
		for (String ruleString : CollectionUtils.sort(ruleStrings)) {
			sb.append(ruleString);
			// sb.append("\n");
		}
		return sb.toString();
	}

	public int getNumberOfRules() {
		int nRules = 0;
		for (int state = 0; state < numStates; state++) {
			BinaryRule[] parentRules = this.splitRulesWithP(state);
			for (int i = 0; i < parentRules.length; i++) {
				BinaryRule bRule = parentRules[i];
				double[][][] scores = bRule.getScores2();
				for (int j = 0; j < scores.length; j++) {
					for (int k = 0; k < scores[j].length; k++) {
						if (scores[j][k] != null) {
							nRules += scores[j][k].length;
						}
					}
				}
			}
			UnaryRule[] unaries = this.getClosedSumUnaryRulesByParent(state);
			for (int r = 0; r < unaries.length; r++) {
				UnaryRule uRule = unaries[r];
				// List<UnaryRule> unaries = this.getUnaryRulesByParent(state);
				// for (UnaryRule uRule : unaries){
				if (uRule.childState == uRule.parentState)
					continue;
				double[][] scores = uRule.getScores2();
				for (int j = 0; j < scores.length; j++) {
					if (scores[j] != null) {
						nRules += scores[j].length;
					}
				}
			}
		}
		return nRules;
	}

	public void printUnaryRules() {
		// System.out.println("BY PARENT");
		for (int state1 = 0; state1 < numStates; state1++) {
			List<UnaryRule> unaries = this.getUnaryRulesByParent(state1);
			for (UnaryRule uRule : unaries) {
				UnaryRule uRule2 = (UnaryRule) unaryRuleMap.get(uRule);
				if (!uRule.getScores2().equals(uRule2.getScores2()))
					System.out.print("BY PARENT:\n" + uRule + "" + uRule2
							+ "\n");
			}
		}
		// System.out.println("VITERBI CLOSED");
		for (int state1 = 0; state1 < numStates; state1++) {
			UnaryRule[] unaries = this
					.getClosedViterbiUnaryRulesByParent(state1);
			for (int r = 0; r < unaries.length; r++) {
				UnaryRule uRule = unaries[r];
				// System.out.print(uRule);
				UnaryRule uRule2 = (UnaryRule) unaryRuleMap.get(uRule);
				if (unariesAreNotEqual(uRule, uRule2))
					System.out.print("VITERBI CLOSED:\n" + uRule + "" + uRule2
							+ "\n");
			}
		}

		/*
		 * System.out.println("FROM RULE MAP"); for (UnaryRule uRule :
		 * unaryRuleMap.keySet()){ System.out.print(uRule); }
		 */

		// System.out.println("AND NOW THE BINARIES");
		// System.out.println("BY PARENT");
		for (int state1 = 0; state1 < numStates; state1++) {
			BinaryRule[] parentRules = this.splitRulesWithP(state1);
			for (int i = 0; i < parentRules.length; i++) {
				BinaryRule bRule = parentRules[i];
				BinaryRule bRule2 = (BinaryRule) binaryRuleMap.get(bRule);
				if (!bRule.getScores2().equals(bRule2.getScores2()))
					System.out.print("BINARY: " + bRule + "" + bRule2 + "\n");
			}
		}
		/*
		 * System.out.println("FROM RULE MAP"); for (BinaryRule bRule :
		 * binaryRuleMap.keySet()){ System.out.print(bRule); }
		 */

	}

	public boolean unariesAreNotEqual(UnaryRule u1, UnaryRule u2) {
		// two cases:
		// 1. u2 is null and u1 is a selfRule
		if (u2 == null) {
			return false;
			/*
			 * double[][] s1 = u1.getScores2(); for (int i=0; i<s1.length; i++){
			 * if (s1[i][i] != 1.0) return true; }
			 */
		} else { // compare all entries
			double[][] s1 = u1.getScores2();
			double[][] s2 = u2.getScores2();
			for (int i = 0; i < s1.length; i++) {
				if (s1[i] == null || s2[i] == null)
					continue;
				for (int j = 0; j < s1[i].length; j++) {
					if (s1[i][j] != s2[i][j])
						return true;
				}
			}
		}
		return false;
	}

	@SuppressWarnings("unchecked")
	public void init() {
		binaryRuleMap = new HashMap<BinaryRule, BinaryRule>();
		unaryRuleMap = new HashMap<UnaryRule, UnaryRule>();
		// allRules = new ArrayList<Rule>();
		bestSumRulesUnderMax = new HashMap();
		bestViterbiRulesUnderMax = new HashMap();
		binaryRulesWithParent = new List[numStates];
		binaryRulesWithLC = new List[numStates];
		binaryRulesWithRC = new List[numStates];
		unaryRulesWithParent = new List[numStates];
		unaryRulesWithC = new List[numStates];
		closedSumRulesWithParent = new List[numStates];
		closedSumRulesWithChild = new List[numStates];
		closedViterbiRulesWithParent = new List[numStates];
		closedViterbiRulesWithChild = new List[numStates];
		isGrammarTag = new boolean[numStates];

		// if (findClosedPaths) {
		closedViterbiPaths = new int[numStates][numStates];
		// }
		closedSumPaths = new int[numStates][numStates];

		for (short s = 0; s < numStates; s++) {
			binaryRulesWithParent[s] = new ArrayList<BinaryRule>();
			binaryRulesWithLC[s] = new ArrayList<BinaryRule>();
			binaryRulesWithRC[s] = new ArrayList<BinaryRule>();
			unaryRulesWithParent[s] = new ArrayList<UnaryRule>();
			unaryRulesWithC[s] = new ArrayList<UnaryRule>();
			closedSumRulesWithParent[s] = new ArrayList<UnaryRule>();
			closedSumRulesWithChild[s] = new ArrayList<UnaryRule>();
			closedViterbiRulesWithParent[s] = new ArrayList<UnaryRule>();
			closedViterbiRulesWithChild[s] = new ArrayList<UnaryRule>();

			double[][] scores = new double[numSubStates[s]][numSubStates[s]];
			for (int i = 0; i < scores.length; i++) {
				scores[i][i] = 1;
			}
			UnaryRule selfR = new UnaryRule(s, s, scores);
			// relaxSumRule(selfR);
			relaxViterbiRule(selfR);
		}
	}

	/**
	 * Construct the optimal grammar from a list of Tree<StateSet>. This
	 * assumes that trainTrees has all the inside/outside probabilities correct.
	 * It performs the M step of EM--it finds the best grammar given the
	 * sufficient statistics (i.e. i/o probabilities).
	 * 
	 * @param trainTrees
	 *            A set of StateSet trees which have had their inside/outside
	 *            probabilities calculated.
	 * @param dummy
	 *            A dummy parameter because otherwise java objects that this
	 *            method shares the signature of Grammar(List of Tree of
	 *            Strings).
	 */
	/*
	 * comment out unused constructor public Grammar(List<Tree<StateSet>>
	 * trainTrees, Grammar old_grammar) { this.tagNumberer =
	 * Numberer.getGlobalNumberer("tags"); unaryRuleCounter = new
	 * Counter<UnaryRule>(); binaryRuleCounter = new Counter<BinaryRule>();
	 * symbolCounter = new CounterMap<Integer,Integer>(); numStates =
	 * tagNumberer.total(); numSubStates = old_grammar.numSubStates; init();
	 * 
	 * for (Tree<StateSet> trainTree : trainTrees) {
	 * tallyStateSetTree(trainTree, old_grammar); } for (UnaryRule unaryRule :
	 * unaryRuleCounter.keySet()) { double unaryProbability =
	 * unaryRuleCounter.getCount(unaryRule) /
	 * symbolCounter.getCount(unaryRule.getParentState
	 * (),unaryRule.getParentSubState());
	 * unaryRule.setScore(Math.log(unaryProbability)); addUnary(unaryRule); }
	 * for (BinaryRule binaryRule : binaryRuleCounter.keySet()) { double
	 * binaryProbability = binaryRuleCounter.getCount(binaryRule) /
	 * symbolCounter
	 * .getCount(binaryRule.getParentState(),binaryRule.getParentSubState());
	 * binaryRule.setScore(Math.log(binaryProbability)); addBinary(binaryRule);
	 * } }
	 */

	/**
	 * This constructor generates a grammar with the rule probabilities read as
	 * though there were no substates, but with a bit of randomness added. This
	 * is the way we should initialize the EM algorithm.
	 * 
	 * @param trainTrees
	 *            The training trees, which don't need to have their
	 *            inside-outside probabilities calculated correctly.
	 * @param randomness
	 *            The size of the region to be uniformly sampled from in adding
	 *            extra random weight to the rules.
	 */
	/*
	 * comment out unused constructor public Grammar(List<Tree<StateSet>>
	 * trainTrees, int[] nSubStates, int maxN, double randomness) {
	 * this.tagNumberer = Numberer.getGlobalNumberer("tags"); unaryRuleCounter =
	 * new Counter<UnaryRule>(); binaryRuleCounter = new Counter<BinaryRule>();
	 * symbolCounter = new CounterMap<Integer, Integer>(); numStates =
	 * tagNumberer.total(); numSubStates = nSubStates; maxNumSubStates = maxN;
	 * init();
	 * 
	 * //tally trees as though there were no subsymbols for (Tree<StateSet>
	 * trainTree : trainTrees) { tallyUninitializedStateSetTree(trainTree); }
	 * //add randomness Random random = new Random(); for (UnaryRule unaryRule :
	 * unaryRuleCounter.keySet()) { double r = random.nextDouble()*randomness;
	 * unaryRuleCounter.incrementCount(unaryRule,r); } for (BinaryRule
	 * binaryRule : binaryRuleCounter.keySet()) { double r =
	 * random.nextDouble()*randomness;
	 * binaryRuleCounter.incrementCount(binaryRule,r); } //re-tally the parent
	 * counts because adding the randomness ruined them symbolCounter = new
	 * CounterMap<Integer, Integer>(); for (UnaryRule unaryRule :
	 * unaryRuleCounter.keySet()) { symbolCounter.incrementCount(
	 * unaryRule.getParentState(), unaryRule.getParentSubState(),
	 * unaryRuleCounter.getCount(unaryRule)); } for (BinaryRule binaryRule :
	 * binaryRuleCounter.keySet()) {
	 * symbolCounter.incrementCount(binaryRule.getParentState
	 * (),binaryRule.getParentSubState(),
	 * binaryRuleCounter.getCount(binaryRule)); } //set the scores of all the
	 * rules based on these counts for (UnaryRule unaryRule :
	 * unaryRuleCounter.keySet()) { double unaryProbability =
	 * unaryRuleCounter.getCount(unaryRule) /
	 * symbolCounter.getCount(unaryRule.getParentState(),
	 * unaryRule.getParentSubState());
	 * unaryRule.setScore(Math.log(unaryProbability)); addUnary(unaryRule); }
	 * for (BinaryRule binaryRule : binaryRuleCounter.keySet()) { double
	 * binaryProbability = binaryRuleCounter.getCount(binaryRule) /
	 * symbolCounter
	 * .getCount(binaryRule.getParentState(),binaryRule.getParentSubState());
	 * binaryRule.setScore(Math.log(binaryProbability)); addBinary(binaryRule);
	 * } }
	 */

	/**
	 * Rather than calling some all-in-one constructor that takes a list of
	 * trees as training data, you call Grammar() to create an empty grammar,
	 * call tallyTree() repeatedly to include all the training data, then call
	 * optimize() to take it into account.
	 * 
	 * @param oldGrammar
	 *            This is the previous grammar. We use this to copy the split
	 *            trees that record how each state is split recursively. These
	 *            parameters are intialized if oldGrammar is null.
	 */
	@SuppressWarnings("unchecked")
	public Grammar(short[] nSubStates, boolean findClosedPaths,
			Smoother smoother, Grammar oldGrammar, double thresh) {
		this.tagNumberer = Numberer.getGlobalNumberer("tags");
		this.findClosedPaths = findClosedPaths;
		this.smoother = smoother;
		this.threshold = thresh;
		unaryRuleCounter = new UnaryCounterTable(nSubStates);
		binaryRuleCounter = new BinaryCounterTable(nSubStates);
		symbolCounter = new CounterMap<Integer, Integer>();
		numStates = (short) tagNumberer.total();
		numSubStates = nSubStates;
		bSearchRule = new BinaryRule((short) 0, (short) 0, (short) 0);
		uSearchRule = new UnaryRule((short) 0, (short) 0);
		logarithmMode = false;
		if (oldGrammar != null) {
			splitTrees = oldGrammar.splitTrees;
		} else {
			splitTrees = new Tree[numStates];
			boolean hasAnySplits = false;
			for (int tag = 0; !hasAnySplits && tag < numStates; tag++) {
				hasAnySplits = hasAnySplits || numSubStates[tag] > 1;
			}
			for (int tag = 0; tag < numStates; tag++) {
				ArrayList<Tree<Short>> children = new ArrayList<Tree<Short>>(
						numSubStates[tag]);
				if (hasAnySplits) {
					for (short substate = 0; substate < numSubStates[tag]; substate++) {
						children.add(substate, new Tree<Short>(substate));
					}
				}
				splitTrees[tag] = new Tree<Short>((short) 0, children);
			}
		}
		init();
	}

	public void setSmoother(Smoother smoother) {
		this.smoother = smoother;
	}

	public static double generateMMTRandomNumber(Random r) {
		double f = r.nextDouble();
		f = f * 2 - 1;
		f = f * Math.log(3);
		return Math.exp(f);
	}

	public void optimize(double randomness) {
		// System.out.print("Optimizing Grammar...");
		init();
		// checkNumberOfSubstates();
		if (randomness > 0.0) {
			Random random = GrammarTrainer.RANDOM;
			// switch (randomInitializationType ) {
			// case INITIALIZE_WITH_SMALL_RANDOMIZATION:
			// add randomness
			for (UnaryRule unaryRule : unaryRuleCounter.keySet()) {
				double[][] unaryCounts = unaryRuleCounter.getCount(unaryRule);
				for (int i = 0; i < unaryCounts.length; i++) {
					if (unaryCounts[i] == null)
						unaryCounts[i] = new double[numSubStates[unaryRule
								.getParentState()]];
					for (int j = 0; j < unaryCounts[i].length; j++) {
						double r = random.nextDouble() * randomness;
						unaryCounts[i][j] += r;
					}
				}
				unaryRuleCounter.setCount(unaryRule, unaryCounts);
			}
			for (BinaryRule binaryRule : binaryRuleCounter.keySet()) {
				double[][][] binaryCounts = binaryRuleCounter
						.getCount(binaryRule);
				for (int i = 0; i < binaryCounts.length; i++) {
					for (int j = 0; j < binaryCounts[i].length; j++) {
						if (binaryCounts[i][j] == null)
							binaryCounts[i][j] = new double[numSubStates[binaryRule
									.getParentState()]];
						for (int k = 0; k < binaryCounts[i][j].length; k++) {
							double r = random.nextDouble() * randomness;
							binaryCounts[i][j][k] += r;
						}
					}
				}
				binaryRuleCounter.setCount(binaryRule, binaryCounts);
			}
			// break;
			// case INITIALIZE_LIKE_MMT:
			// //multiply by a random factor
			// for (UnaryRule unaryRule : unaryRuleCounter.keySet()) {
			// double[][] unaryCounts = unaryRuleCounter.getCount(unaryRule);
			// for (int i = 0; i < unaryCounts.length; i++) {
			// if (unaryCounts[i]==null)
			// continue;
			// for (int j = 0; j < unaryCounts[i].length; j++) {
			// double r = generateMMTRandomNumber(random);
			// unaryCounts[i][j] *= r;
			// }
			// }
			// unaryRuleCounter.setCount(unaryRule, unaryCounts);
			// }
			// for (BinaryRule binaryRule : binaryRuleCounter.keySet()) {
			// double[][][] binaryCounts =
			// binaryRuleCounter.getCount(binaryRule);
			// for (int i = 0; i < binaryCounts.length; i++) {
			// for (int j = 0; j < binaryCounts[i].length; j++) {
			// if (binaryCounts[i][j]==null)
			// continue;
			// for (int k = 0; k < binaryCounts[i][j].length; k++) {
			// double r = generateMMTRandomNumber(random);
			// binaryCounts[i][j][k] *= r;
			// }
			// }
			// }
			// binaryRuleCounter.setCount(binaryRule, binaryCounts);
			// }
			// break;
			// }
		}

		// smooth
		// if (useEntropicPrior) {
		// System.out.println("\nGrammar uses entropic prior!");
		// normalizeWithEntropicPrior();
		// }
		
		//SSIE
		overwriteWithMaxent();
		
		normalize();
		smooth(false); // this also adds the rules to the proper arrays
		// System.out.println("done.");
	}

	public void removeUnlikelyRules(double thresh, double power) {
		// System.out.print("Removing everything below "+thresh+" and rasiing rules to the "
		// +power+"th power... ");
		if (isLogarithmMode())
			power = Math.log(power);
		int total = 0, removed = 0;
		for (int state = 0; state < numStates; state++) {
			for (int r = 0; r < splitRulesWithP[state].length; r++) {
				BinaryRule rule = splitRulesWithP[state][r];
				for (int lC = 0; lC < rule.scores.length; lC++) {
					for (int rC = 0; rC < rule.scores[lC].length; rC++) {
						if (rule.scores[lC][rC] == null)
							continue;
						boolean isNull = true;
						for (int p = 0; p < rule.scores[lC][rC].length; p++) {
							total++;
							if (rule.scores[lC][rC][p] < thresh) {
								// System.out.print(".");
								rule.scores[lC][rC][p] = 0;
								removed++;
							} else {
								if (power != 1)
									rule.scores[lC][rC][p] = Math.pow(
											rule.scores[lC][rC][p], power);
								isNull = false;
							}
						}
						if (isNull)
							rule.scores[lC][rC] = null;
					}
				}
				splitRulesWithP[state][r] = rule;
			}
			for (UnaryRule rule : unaryRulesWithParent[state]) {
				for (int c = 0; c < rule.scores.length; c++) {
					if (rule.scores[c] == null)
						continue;
					boolean isNull = true;
					for (int p = 0; p < rule.scores[c].length; p++) {
						total++;
						if (rule.scores[c][p] <= thresh) {
							removed++;
							rule.scores[c][p] = 0;
						} else {
							if (power != 1)
								rule.scores[c][p] = Math.pow(rule.scores[c][p],
										power);
							isNull = false;
						}
					}
					if (isNull)
						rule.scores[c] = null;
				}
			}
		}
		// System.out.print("done.\nRemoved "+removed+" out of "+total+" rules.\n");
	}

	public void smooth(boolean noNormalize) {
		smoother.smooth(unaryRuleCounter, binaryRuleCounter);
		if (!noNormalize)
			normalize();

		// if (threshold>0){
		// removeUnlikelyRules(threshold);
		// normalize();
		// }

		// compress and add the rules
		for (UnaryRule unaryRule : unaryRuleCounter.keySet()) {
			double[][] unaryCounts = unaryRuleCounter.getCount(unaryRule);
			for (int i = 0; i < unaryCounts.length; i++) {
				if (unaryCounts[i] == null)
					continue;
				/**
				 * allZero records if all probabilities are 0. If so, we want to
				 * null out the matrix element.
				 */
				double allZero = 0;
				int j = 0;
				while (allZero == 0 && j < unaryCounts[i].length) {
					allZero += unaryCounts[i][j++];
				}
				if (allZero == 0) {
					unaryCounts[i] = null;
				}
			}
			unaryRule.setScores2(unaryCounts);
			addUnary(unaryRule);
		}
		computePairsOfUnaries();
		for (BinaryRule binaryRule : binaryRuleCounter.keySet()) {
			double[][][] binaryCounts = binaryRuleCounter.getCount(binaryRule);
			for (int i = 0; i < binaryCounts.length; i++) {
				for (int j = 0; j < binaryCounts[i].length; j++) {
					if (binaryCounts[i][j] == null)
						continue;
					/**
					 * allZero records if all probabilities are 0. If so, we
					 * want to null out the matrix element.
					 */
					double allZero = 0;
					int k = 0;
					while (allZero == 0 && k < binaryCounts[i][j].length) {
						allZero += binaryCounts[i][j][k++];
					}
					if (allZero == 0) {
						binaryCounts[i][j] = null;
					}
				}
			}
			binaryRule.setScores2(binaryCounts);
			addBinary(binaryRule);
		}
		// Reset all counters:
		unaryRuleCounter = new UnaryCounterTable(numSubStates);
		binaryRuleCounter = new BinaryCounterTable(numSubStates);
		symbolCounter = new CounterMap<Integer, Integer>();
		/*
		 * // tally usage of closed unary rule paths if (findClosedPaths) { int
		 * maxSize = numStates * numStates; int size = 0; for (int i=0;
		 * i<numStates; i++) { for (int j=0; j<numStates; j++) { if
		 * (closedViterbiPaths[i][j]!=null) size++; } }
		 * System.out.println("Closed viterbi unary path data structure covers "
		 * + size + " / " + maxSize + " = " + (((double) size) / maxSize) +
		 * " state pairs"); }
		 */
		// checkNumberOfSubstates();

		// Romain: added the computation for the sum-product closure
		// TODO: fix the code and add this back in
		// sumProductClosedUnaryRulesWithParent =
		// sumProductUnaryClosure(unaryRulesWithParent);

	}

	public void clearCounts() {
		unaryRuleCounter = new UnaryCounterTable(numSubStates);
		binaryRuleCounter = new BinaryCounterTable(numSubStates);
		symbolCounter = new CounterMap<Integer, Integer>();

	}

	/**
	 * Normalize the unary & binary probabilities so that they sum to 1 for each
	 * parent. The binaryRuleCounter and unaryRuleCounter are assumed to contain
	 * probabilities, NOT log probabilities!
	 */
	public void normalize() {
		// tally the parent counts
		tallyParentCounts();
		// turn the rule scores into fractions
		for (UnaryRule unaryRule : unaryRuleCounter.keySet()) {
			double[][] unaryCounts = unaryRuleCounter.getCount(unaryRule);
			int parentState = unaryRule.getParentState();
			int nParentSubStates = numSubStates[parentState];
			int nChildStates = numSubStates[unaryRule.childState];
			double[] parentCount = new double[nParentSubStates];
			for (int i = 0; i < nParentSubStates; i++) {
				parentCount[i] = symbolCounter.getCount(parentState, i);
			}
			boolean allZero = true;
			for (int j = 0; j < nChildStates; j++) {
				if (unaryCounts[j] == null)
					continue;
				for (int i = 0; i < nParentSubStates; i++) {
					if (parentCount[i] != 0) {
						double nVal = (unaryCounts[j][i] / parentCount[i]);
						if (nVal < threshold
								|| SloppyMath.isVeryDangerous(nVal))
							nVal = 0;
						unaryCounts[j][i] = nVal;
					}
					allZero = allZero && (unaryCounts[j][i] == 0);
				}
			}
			if (allZero) {
				System.out.println("Maybe an underflow? Rule: " + unaryRule
						+ "\n" + ArrayUtil.toString(unaryCounts));
			}
			unaryRuleCounter.setCount(unaryRule, unaryCounts);
		}
		for (BinaryRule binaryRule : binaryRuleCounter.keySet()) {
			double[][][] binaryCounts = binaryRuleCounter.getCount(binaryRule);
			int parentState = binaryRule.parentState;
			int nParentSubStates = numSubStates[parentState];
			double[] parentCount = new double[nParentSubStates];
			for (int i = 0; i < nParentSubStates; i++) {
				parentCount[i] = symbolCounter.getCount(parentState, i);
			}
			for (int j = 0; j < binaryCounts.length; j++) {
				for (int k = 0; k < binaryCounts[j].length; k++) {
					if (binaryCounts[j][k] == null)
						continue;
					for (int i = 0; i < nParentSubStates; i++) {
						if (parentCount[i] != 0) {
							double nVal = (binaryCounts[j][k][i] / parentCount[i]);
							if (nVal < threshold
									|| SloppyMath.isVeryDangerous(nVal))
								nVal = 0;
							binaryCounts[j][k][i] = nVal;
						}
					}
				}
			}
			binaryRuleCounter.setCount(binaryRule, binaryCounts);
		}
	}

	// public void normalizeWithEntropicPrior(){
	// for (int iter=1; iter<=6; iter++){
	// tallyParentCounts();
	// // turn the rule scores into fractions
	// for (UnaryRule unaryRule : unaryRuleCounter.keySet()) {
	// double[][] unaryCounts = unaryRuleCounter.getCount(unaryRule);
	// int parentState = unaryRule.getParentState();
	// int nParentSubStates = numSubStates[parentState];
	// double[] parentCount = new double[nParentSubStates];
	// for (int i = 0; i < nParentSubStates; i++) {
	// parentCount[i] = symbolCounter.getCount(parentState, i);
	// }
	// double[][] theta = new double[unaryCounts.length][];
	// if (iter==1){
	// // initialize the thetas
	// for (int j = 0; j < unaryCounts.length; j++) {
	// if (unaryCounts[j]==null) continue;
	// theta[j] = new double[nParentSubStates];
	// for (int i = 0; i < nParentSubStates; i++) {
	// double val = unaryCounts[j][i];
	// theta[j][i]=Math.pow(val,1-(1/parentCount[i]));
	// }
	// }
	// unaryRule.setScores2(unaryCounts);
	// unaryRuleCounter.setCount(unaryRule,theta);
	// }
	// // compute lambdas
	// else{
	// theta = unaryCounts;
	// unaryCounts = unaryRule.getScores2();
	// for (int j = 0; j < unaryCounts.length; j++) {
	// if (unaryCounts[j]==null) continue;
	// for (int i = 0; i < nParentSubStates; i++) {
	// theta[j][i] = unaryCounts[j][i]/parentCount[i];
	// }
	// }
	// for (int j = 0; j < unaryCounts.length; j++) {
	// if (unaryCounts[j]==null) continue;
	// for (int i = 0; i < nParentSubStates; i++) {
	// if (unaryCounts[j][i]==0) {
	// theta[j][i]=0;
	// continue;
	// }
	// double val = theta[j][i];
	// double lambda = -((unaryCounts[j][i]/val) + Math.log(val) + 1);
	// // compute thetas
	// val = -1.0*unaryCounts[j][i];
	// theta[j][i]= val/SloppyMath.lambert(val,(1+lambda));
	// // if (SloppyMath.isDangerous(theta[j][i]))
	// //
	// System.out.println("Maybe an underflow: count "+val+" lambda "+lambda+" theta "
	// +theta[j][i] + " div "+SloppyMath.lambert(val,(1+lambda)));
	// }
	// }
	// unaryRuleCounter.setCount(unaryRule,theta);
	// }
	// }
	// for (BinaryRule binaryRule : binaryRuleCounter.keySet()) {
	// double[][][] binaryCounts = binaryRuleCounter.getCount(binaryRule);
	// int parentState = binaryRule.parentState;
	// int nParentSubStates = numSubStates[parentState];
	// double[] parentCount = new double[nParentSubStates];
	// for (int i = 0; i < nParentSubStates; i++) {
	// parentCount[i] = symbolCounter.getCount(parentState, i);
	// }
	// double[][][] theta = new
	// double[binaryCounts.length][binaryCounts[0].length][];
	// if (iter==1){
	// // initialize the thetas
	// for (int j = 0; j < binaryCounts.length; j++) {
	// for (int k = 0; k < binaryCounts[j].length; k++) {
	// if (binaryCounts[j][k]==null) continue;
	// theta[j][k] = new double[nParentSubStates];
	// for (int i = 0; i < nParentSubStates; i++) {
	// double val = binaryCounts[j][k][i];
	// theta[j][k][i]=Math.pow(val,1-(1/parentCount[i]));
	// }
	// }
	// }
	// binaryRule.setScores2(binaryCounts);
	// binaryRuleCounter.setCount(binaryRule,theta);
	// }
	// else{
	// theta = binaryCounts;
	// binaryCounts = binaryRule.getScores2();
	// for (int j = 0; j < binaryCounts.length; j++) {
	// for (int k = 0; k < binaryCounts[j].length; k++) {
	// if (binaryCounts[j][k]==null) continue;
	// for (int i = 0; i < nParentSubStates; i++) {
	// theta[j][k][i] = binaryCounts[j][k][i]/parentCount[i];
	// }
	// }
	// }
	// binaryCounts = binaryRule.getScores2();
	// for (int j = 0; j < binaryCounts.length; j++) {
	// for (int k = 0; k < binaryCounts[j].length; k++) {
	// if (binaryCounts[j][k]==null) continue;
	// for (int i = 0; i < nParentSubStates; i++) {
	// if (binaryCounts[j][k][i]==0) {
	// theta[j][k][i]=0;
	// continue;
	// }
	// double val = theta[j][k][i];
	// double lambda = -((binaryCounts[j][k][i]/val) + Math.log(val) + 1);
	// // compute thetas
	// val = -1.0*binaryCounts[j][k][i];
	// theta[j][k][i]= val/SloppyMath.lambert(val,(1+lambda));
	// if (SloppyMath.isDangerous(theta[j][k][i]))
	// System.out.println("Maybe an underflow: count "+val+" lambda "+lambda+" theta "
	// +theta[j][k][i] + " div "+SloppyMath.lambert(val,(1+lambda)));
	// if (val==0) theta[j][k][i]=0;
	// }
	// }
	// }
	// binaryRuleCounter.setCount(binaryRule,theta);
	// }
	// }
	// }
	// }

	/*
	 * Check number of substates
	 */
	public void checkNumberOfSubstates() {
		for (UnaryRule unaryRule : unaryRuleCounter.keySet()) {
			double[][] unaryCounts = unaryRuleCounter.getCount(unaryRule);
			int nParentSubStates = numSubStates[unaryRule.parentState];
			int nChildSubStates = numSubStates[unaryRule.childState];
			if (unaryCounts.length != nChildSubStates) {
				System.out.println("Unary Rule " + unaryRule + " should have "
						+ nChildSubStates + " childsubstates.");
			}
			if (unaryCounts[0] != null
					&& unaryCounts[0].length != nParentSubStates) {
				System.out.println("Unary Rule " + unaryRule + " should have "
						+ nParentSubStates + " parentsubstates.");
			}
		}
		for (BinaryRule binaryRule : binaryRuleCounter.keySet()) {
			double[][][] binaryCounts = binaryRuleCounter.getCount(binaryRule);
			int nParentSubStates = numSubStates[binaryRule.parentState];
			int nLeftChildSubStates = numSubStates[binaryRule.leftChildState];
			int nRightChildSubStates = numSubStates[binaryRule.rightChildState];
			if (binaryCounts.length != nLeftChildSubStates) {
				System.out.println("Unary Rule " + binaryRule + " should have "
						+ nLeftChildSubStates + " left childsubstates.");
			}
			if (binaryCounts[0].length != nRightChildSubStates) {
				System.out.println("Unary Rule " + binaryRule + " should have "
						+ nRightChildSubStates + " right childsubstates.");
			}
			if (binaryCounts[0][0] != null
					&& binaryCounts[0][0].length != nParentSubStates) {
				System.out.println("Unary Rule " + binaryRule + " should have "
						+ nParentSubStates + " parentsubstates.");
			}
		}
		System.out.println("Done with checks.");
	}

	/**
	 * Sum the parent symbol counter, symbolCounter. This is needed when the
	 * rule counters are altered, such as when adding randomness in optimize().
	 * <p>
	 * This assumes that the unaryRuleCounter and binaryRuleCounter contain
	 * probabilities, NOT log probabilities!
	 */
	private void tallyParentCounts() {
		symbolCounter = new CounterMap<Integer, Integer>();
		for (UnaryRule unaryRule : unaryRuleCounter.keySet()) {
			double[][] unaryCounts = unaryRuleCounter.getCount(unaryRule);
			int parentState = unaryRule.getParentState();
			isGrammarTag[parentState] = true;
			if (unaryRule.childState == parentState)
				continue;
			int nParentSubStates = numSubStates[parentState];
			double[] sum = new double[nParentSubStates];
			for (int j = 0; j < unaryCounts.length; j++) {
				if (unaryCounts[j] == null)
					continue;
				for (int i = 0; i < nParentSubStates; i++) {
					double val = unaryCounts[j][i];
					// if (val>=threshold)
					sum[i] += val;
				}
			}
			for (int i = 0; i < nParentSubStates; i++) {
				symbolCounter.incrementCount(parentState, i, sum[i]);
			}

		}
		for (BinaryRule binaryRule : binaryRuleCounter.keySet()) {
			double[][][] binaryCounts = binaryRuleCounter.getCount(binaryRule);
			int parentState = binaryRule.parentState;
			isGrammarTag[parentState] = true;
			int nParentSubStates = numSubStates[parentState];
			double[] sum = new double[nParentSubStates];
			for (int j = 0; j < binaryCounts.length; j++) {
				for (int k = 0; k < binaryCounts[j].length; k++) {
					if (binaryCounts[j][k] == null)
						continue;
					for (int i = 0; i < nParentSubStates; i++) {
						double val = binaryCounts[j][k][i];
						// if (val>=threshold)
						sum[i] += val;
					}
				}
			}
			for (int i = 0; i < nParentSubStates; i++) {
				symbolCounter.incrementCount(parentState, i, sum[i]);
			}
		}
	}

	public void tallyStateSetTree(Tree<StateSet> tree, Grammar old_grammar) {
		// Check that the top node is not split (it has only one substate)
		if (tree.isLeaf())
			return;
		if (tree.isPreTerminal())
			return;
		StateSet node = tree.getLabel();
		if (node.numSubStates() != 1) {
			System.err.println("The top symbol is split!");
			System.out.println(tree);
			System.exit(1);
		}
		// The inside score of its only substate is the (log) probability of the
		// tree
		double tree_score = node.getIScore(0);
		int tree_scale = node.getIScale();
		if (tree_score == 0) {
			System.out
					.println("Something is wrong with this tree. I will skip it.");
			return;
		}
		tallyStateSetTree(tree, tree_score, tree_scale, old_grammar);
	}

	public void tallyStateSetTree(Tree<StateSet> tree, double tree_score,
			double tree_scale, Grammar old_grammar) {
		if (tree.isLeaf())
			return;
		if (tree.isPreTerminal())
			return;
		List<Tree<StateSet>> children = tree.getChildren();
		StateSet parent = tree.getLabel();
		short parentState = parent.getState();
		int nParentSubStates = numSubStates[parentState];
		switch (children.size()) {
		case 0:
			// This is a leaf (a preterminal node, if we count the words
			// themselves),
			// nothing to do
			break;
		case 1:
			StateSet child = children.get(0).getLabel();
			short childState = child.getState();
			int nChildSubStates = numSubStates[childState];
			UnaryRule urule = new UnaryRule(parentState, childState);
			double[][] oldUScores = old_grammar.getUnaryScore(urule); // rule
			// score
			double[][] ucounts = unaryRuleCounter.getCount(urule);
			if (ucounts == null)
				ucounts = new double[nChildSubStates][];
			double scalingFactor = ScalingTools.calcScaleFactor(parent
					.getOScale()
					+ child.getIScale() - tree_scale);
			// if (scalingFactor==0){
			// System.out.println("p: "+parent.getOScale()+" c: "+child.getIScale()+" t:"+tree_scale);
			// }
			for (short i = 0; i < nChildSubStates; i++) {
				if (oldUScores[i] == null)
					continue;
				double cIS = child.getIScore(i);
				if (cIS == 0)
					continue;
				if (ucounts[i] == null)
					ucounts[i] = new double[nParentSubStates];
				for (short j = 0; j < nParentSubStates; j++) {
					double pOS = parent.getOScore(j); // Parent outside score
					if (pOS == 0)
						continue;
					double rS = oldUScores[i][j];
					if (rS == 0)
						continue;
					if (tree_score == 0)
						tree_score = 1;
					double logRuleCount = (rS * cIS / tree_score)
							* scalingFactor * pOS;
					ucounts[i][j] += logRuleCount;
				}
			}
			// urule.setScores2(ucounts);
			unaryRuleCounter.setCount(urule, ucounts);
			break;
		case 2:
			StateSet leftChild = children.get(0).getLabel();
			short lChildState = leftChild.getState();
			StateSet rightChild = children.get(1).getLabel();
			short rChildState = rightChild.getState();
			int nLeftChildSubStates = numSubStates[lChildState];
			int nRightChildSubStates = numSubStates[rChildState];
			// new double[nLeftChildSubStates][nRightChildSubStates][];
			BinaryRule brule = new BinaryRule(parentState, lChildState,
					rChildState);
			double[][][] oldBScores = old_grammar.getBinaryScore(brule); // rule
			// score
			if (oldBScores == null) {
				// rule was not in the grammar
				// parent.setIScores(iScores2);
				// break;
				oldBScores = new double[nLeftChildSubStates][nRightChildSubStates][nParentSubStates];
				ArrayUtil.fill(oldBScores, 1.0);
			}
			double[][][] bcounts = binaryRuleCounter.getCount(brule);
			if (bcounts == null)
				bcounts = new double[nLeftChildSubStates][nRightChildSubStates][];
			scalingFactor = ScalingTools.calcScaleFactor(parent.getOScale()
					+ leftChild.getIScale() + rightChild.getIScale()
					- tree_scale);
			// if (scalingFactor==0){
			// System.out.println("p: "+parent.getOScale()+" l: "+leftChild.getIScale()+" r:"+rightChild.getIScale()+" t:"+tree_scale);
			// }
			for (short i = 0; i < nLeftChildSubStates; i++) {
				double lcIS = leftChild.getIScore(i);
				if (lcIS == 0)
					continue;
				for (short j = 0; j < nRightChildSubStates; j++) {
					if (oldBScores[i][j] == null)
						continue;
					double rcIS = rightChild.getIScore(j);
					if (rcIS == 0)
						continue;
					// allocate parent array
					if (bcounts[i][j] == null)
						bcounts[i][j] = new double[nParentSubStates];
					for (short k = 0; k < nParentSubStates; k++) {
						double pOS = parent.getOScore(k); // Parent outside
						// score
						if (pOS == 0)
							continue;
						double rS = oldBScores[i][j][k];
						if (rS == 0)
							continue;
						if (tree_score == 0)
							tree_score = 1;
						double logRuleCount = (rS * lcIS / tree_score) * rcIS
								* scalingFactor * pOS;
						/*
						 * if (logRuleCount == 0) {
						 * System.out.println("rS "+rS+", lcIS "
						 * +lcIS+", rcIS "+rcIS+", tree_score "+tree_score+
						 * ", scalingFactor "+scalingFactor+", pOS "+pOS);
						 * System.out.println("Possibly underflow?"); //
						 * logRuleCount = Double.MIN_VALUE; }
						 */
						bcounts[i][j][k] += logRuleCount;
					}
				}
			}
			binaryRuleCounter.setCount(brule, bcounts);
			break;
		default:
			throw new Error("Malformed tree: more than two children");
		}

		for (Tree<StateSet> child : children) {
			tallyStateSetTree(child, tree_score, tree_scale, old_grammar);
		}
	}

	public void tallyUninitializedStateSetTree(Tree<StateSet> tree) {
		if (tree.isLeaf())
			return;
		// the lexicon handles preterminal nodes
		if (tree.isPreTerminal())
			return;
		List<Tree<StateSet>> children = tree.getChildren();
		StateSet parent = tree.getLabel();
		short parentState = parent.getState();
		int nParentSubStates = parent.numSubStates(); // numSubStates[parentState];
		switch (children.size()) {
		case 0:
			// This is a leaf (a preterminal node, if we count the words
			// themselves), nothing to do
			break;
		case 1:
			StateSet child = children.get(0).getLabel();
			short childState = child.getState();
			int nChildSubStates = child.numSubStates(); // numSubStates[childState];
			double[][] counts = new double[nChildSubStates][nParentSubStates];
			UnaryRule urule = new UnaryRule(parentState, childState, counts);
			unaryRuleCounter.incrementCount(urule, 1.0);
			break;
		case 2:
			StateSet leftChild = children.get(0).getLabel();
			short lChildState = leftChild.getState();
			StateSet rightChild = children.get(1).getLabel();
			short rChildState = rightChild.getState();
			int nLeftChildSubStates = leftChild.numSubStates(); // numSubStates[lChildState];
			int nRightChildSubStates = rightChild.numSubStates();// numSubStates[rChildState];
			double[][][] bcounts = new double[nLeftChildSubStates][nRightChildSubStates][nParentSubStates];
			BinaryRule brule = new BinaryRule(parentState, lChildState,
					rChildState, bcounts);
			binaryRuleCounter.incrementCount(brule, 1.0);
			break;
		default:
			throw new Error("Malformed tree: more than two children");
		}

		for (Tree<StateSet> child : children) {
			tallyUninitializedStateSetTree(child);
		}
	}

	/*
	 * public void tallyChart(Pair<double[][][][], double[][][][]> chart, double
	 * tree_score, Grammar old_grammar) { double[][][][] iScore =
	 * chart.getFirst(); double[][][][] oScore = chart.getSecond(); if
	 * (tree.isLeaf()) return; if (tree.isPreTerminal()) return;
	 * List<Tree<StateSet>> children = tree.getChildren(); StateSet parent =
	 * tree.getLabel(); short parentState = parent.getState(); int
	 * nParentSubStates = numSubStates[parentState]; switch (children.size()) {
	 * case 0: // This is a leaf (a preterminal node, if we count the words
	 * themselves), // nothing to do break; case 1: StateSet child =
	 * children.get(0).getLabel(); short childState = child.getState(); int
	 * nChildSubStates = numSubStates[childState]; UnaryRule urule = new
	 * UnaryRule(parentState, childState); double[][] oldUScores =
	 * old_grammar.getUnaryScore(urule); // rule score double[][] ucounts =
	 * unaryRuleCounter.getCount(urule); if (ucounts==null) ucounts = new
	 * double[nChildSubStates][]; double scalingFactor =
	 * Math.pow(GrammarTrainer.SCALE,
	 * parent.getOScale()+child.getIScale()-tree_scale); if (scalingFactor==0){
	 * System
	 * .out.println("p: "+parent.getOScale()+" c: "+child.getIScale()+" t:"
	 * +tree_scale); } for (short i = 0; i < nChildSubStates; i++) { if
	 * (oldUScores[i]==null) continue; double cIS = child.getIScore(i); if
	 * (cIS==0) continue; if (ucounts[i]==null) ucounts[i] = new
	 * double[nParentSubStates]; for (short j = 0; j < nParentSubStates; j++) {
	 * double pOS = parent.getOScore(j); // Parent outside score if (pOS==0)
	 * continue; double rS = oldUScores[i][j]; if (rS==0) continue; if
	 * (tree_score==0) tree_score = 1; double logRuleCount = (rS * cIS /
	 * tree_score) * scalingFactor * pOS; ucounts[i][j] += logRuleCount; } }
	 * //urule.setScores2(ucounts); unaryRuleCounter.setCount(urule, ucounts);
	 * break; case 2: StateSet leftChild = children.get(0).getLabel(); short
	 * lChildState = leftChild.getState(); StateSet rightChild =
	 * children.get(1).getLabel(); short rChildState = rightChild.getState();
	 * int nLeftChildSubStates = numSubStates[lChildState]; int
	 * nRightChildSubStates = numSubStates[rChildState]; //new
	 * double[nLeftChildSubStates][nRightChildSubStates][]; BinaryRule brule =
	 * new BinaryRule(parentState, lChildState, rChildState); double[][][]
	 * oldBScores = old_grammar.getBinaryScore(brule); // rule score if
	 * (oldBScores==null){ //rule was not in the grammar
	 * //parent.setIScores(iScores2); //break; oldBScores=new
	 * double[nLeftChildSubStates][nRightChildSubStates][nParentSubStates];
	 * ArrayUtil.fill(oldBScores,1.0); } double[][][] bcounts =
	 * binaryRuleCounter.getCount(brule); if (bcounts==null) bcounts = new
	 * double[nLeftChildSubStates][nRightChildSubStates][]; scalingFactor =
	 * Math.pow(GrammarTrainer.SCALE,
	 * parent.getOScale()+leftChild.getIScale()+rightChild
	 * .getIScale()-tree_scale); if (scalingFactor==0){
	 * System.out.println("p: "+
	 * parent.getOScale()+" l: "+leftChild.getIScale()+" r:"
	 * +rightChild.getIScale()+" t:"+tree_scale); } for (short i = 0; i <
	 * nLeftChildSubStates; i++) { double lcIS = leftChild.getIScore(i); if
	 * (lcIS==0) continue; for (short j = 0; j < nRightChildSubStates; j++) { if
	 * (oldBScores[i][j]==null) continue; double rcIS = rightChild.getIScore(j);
	 * if (rcIS==0) continue; // allocate parent array if (bcounts[i][j]==null)
	 * bcounts[i][j] = new double[nParentSubStates]; for (short k = 0; k <
	 * nParentSubStates; k++) { double pOS = parent.getOScore(k); // Parent
	 * outside score if (pOS==0) continue; double rS = oldBScores[i][j][k]; if
	 * (rS==0) continue; if (tree_score==0) tree_score = 1; double logRuleCount
	 * = (rS * lcIS / tree_score) * rcIS * scalingFactor * pOS;
	 * 
	 * bcounts[i][j][k] += logRuleCount; } } } binaryRuleCounter.setCount(brule,
	 * bcounts); break; default: throw new
	 * Error("Malformed tree: more than two children"); }
	 * 
	 * for (Tree<StateSet> child : children) { tallyStateSetTree(child,
	 * tree_score, tree_scale, old_grammar); } }
	 */
	/*
	 * private UnaryRule makeUnaryRule(Tree<String> tree) { int parent =
	 * tagNumberer.number(tree.getLabel()); int child =
	 * tagNumberer.number(tree.getChildren().get(0).getLabel()); return new
	 * UnaryRule(parent, child); }
	 * 
	 * private BinaryRule makeBinaryRule(Tree<String> tree) { int parent =
	 * tagNumberer.number(tree.getLabel()); int lChild =
	 * tagNumberer.number(tree.getChildren().get(0).getLabel()); int rChild =
	 * tagNumberer.number(tree.getChildren().get(1).getLabel()); return new
	 * BinaryRule(parent, lChild, rChild); }
	 */
	public void makeCRArrays() {
		// int numStates = closedRulesWithParent.length;
		closedSumRulesWithP = new UnaryRule[numStates][];
		closedSumRulesWithC = new UnaryRule[numStates][];
		closedViterbiRulesWithP = new UnaryRule[numStates][];
		closedViterbiRulesWithC = new UnaryRule[numStates][];

		for (int i = 0; i < numStates; i++) {
			closedSumRulesWithP[i] = (UnaryRule[]) closedSumRulesWithParent[i]
					.toArray(new UnaryRule[0]);
			closedSumRulesWithC[i] = (UnaryRule[]) closedSumRulesWithChild[i]
					.toArray(new UnaryRule[0]);
			closedViterbiRulesWithP[i] = (UnaryRule[]) closedViterbiRulesWithParent[i]
					.toArray(new UnaryRule[0]);
			closedViterbiRulesWithC[i] = (UnaryRule[]) closedViterbiRulesWithChild[i]
					.toArray(new UnaryRule[0]);
		}
	}

	public UnaryRule[] getClosedSumUnaryRulesByParent(int state) {
		if (closedSumRulesWithP == null) {
			makeCRArrays();
		}
		if (state >= closedSumRulesWithP.length) {
			return new UnaryRule[0];
		}
		return closedSumRulesWithP[state];
	}

	public UnaryRule[] getClosedSumUnaryRulesByChild(int state) {
		if (closedSumRulesWithC == null) {
			makeCRArrays();
		}
		if (state >= closedSumRulesWithC.length) {
			return new UnaryRule[0];
		}
		return closedSumRulesWithC[state];
	}

	public UnaryRule[] getClosedViterbiUnaryRulesByParent(int state) {
		if (closedViterbiRulesWithP == null) {
			makeCRArrays();
		}
		if (state >= closedViterbiRulesWithP.length) {
			return new UnaryRule[0];
		}
		return closedViterbiRulesWithP[state];
	}

	public UnaryRule[] getClosedViterbiUnaryRulesByChild(int state) {
		if (closedViterbiRulesWithC == null) {
			makeCRArrays();
		}
		if (state >= closedViterbiRulesWithC.length) {
			return new UnaryRule[0];
		}
		return closedViterbiRulesWithC[state];
	}

	@SuppressWarnings("unchecked")
	public void purgeRules() {
		Map bR = new HashMap();
		Map bR2 = new HashMap();
		for (Iterator i = bestSumRulesUnderMax.keySet().iterator(); i.hasNext();) {
			UnaryRule ur = (UnaryRule) i.next();
			if ((ur.parentState != ur.childState)) {
				bR.put(ur, ur);
				bR2.put(ur, ur);
			}
		}
		bestSumRulesUnderMax = bR;
		bestViterbiRulesUnderMax = bR2;
	}

	@SuppressWarnings("unchecked")
	public List<short[]> getBestViterbiPath(short pState, short np,
			short cState, short cp) {
		ArrayList<short[]> path = new ArrayList<short[]>();
		short[] state = new short[2];
		state[0] = pState;
		state[1] = np;
		// if we haven't built the data structure of closed paths, then
		// return the simplest possible path
		if (!findClosedPaths) {
			path.add(state);
			state = new short[2];
			state[0] = cState;
			state[1] = cp;
			path.add(state);
			return path;
		} else {
			// read the best paths off of the closedViterbiPaths list
			if (pState == cState && np == cp) {
				path.add(state);
				path.add(state);
				return path;
			}
			while (state[0] != cState || state[1] != cp) {
				path.add(state);
				state[0] = (short) closedViterbiPaths[state[0]][state[1]];
			}
			// add the destination state as well
			path.add(state);
			return path;
		}
	}

	@SuppressWarnings("unchecked")
	private void closeRulesUnderMax(UnaryRule ur) {
		short pState = ur.parentState;
		int nPSubStates = numSubStates[pState];
		short cState = ur.childState;
		double[][] uScores = ur.getScores2();
		// do all sum rules
		for (int i = 0; i < closedSumRulesWithChild[pState].size(); i++) {
			UnaryRule pr = (UnaryRule) closedSumRulesWithChild[pState].get(i);
			for (int j = 0; j < closedSumRulesWithParent[cState].size(); j++) {
				short parentState = pr.parentState;
				int nParentSubStates = numSubStates[parentState];
				UnaryRule cr = (UnaryRule) closedSumRulesWithParent[cState]
						.get(j);
				UnaryRule resultR = new UnaryRule(parentState, cr
						.getChildState());
				double[][] scores = new double[numSubStates[cr.getChildState()]][nParentSubStates];
				for (int np = 0; np < scores[0].length; np++) {
					for (int cp = 0; cp < scores.length; cp++) {
						// sum over intermediate substates
						double sum = 0;
						for (int unp = 0; unp < nPSubStates; unp++) {
							for (int ucp = 0; ucp < uScores.length; ucp++) {
								sum += pr.getScore(np, unp)
										* cr.getScore(ucp, cp)
										* ur.getScore(unp, ucp);
							}
						}
						scores[cp][np] = sum;
					}
				}
				resultR.setScores2(scores);
				// add rule to bestSumRulesUnderMax if it's better
				relaxSumRule(resultR, pState, cState);
			}
		}
		// do viterbi rules also
		for (short i = 0; i < closedViterbiRulesWithChild[pState].size(); i++) {
			UnaryRule pr = (UnaryRule) closedViterbiRulesWithChild[pState]
					.get(i);
			for (short j = 0; j < closedViterbiRulesWithParent[cState].size(); j++) {
				UnaryRule cr = (UnaryRule) closedViterbiRulesWithParent[cState]
						.get(j);
				short parentState = pr.parentState;
				int nParentSubStates = numSubStates[parentState];
				UnaryRule resultR = new UnaryRule(parentState, cr
						.getChildState());
				double[][] scores = new double[numSubStates[cr.getChildState()]][nParentSubStates];
				short[][] intermediateSubState1 = new short[nParentSubStates][numSubStates[cr
						.getChildState()]];
				short[][] intermediateSubState2 = new short[nParentSubStates][numSubStates[cr
						.getChildState()]];
				for (int np = 0; np < scores[0].length; np++) {
					for (int cp = 0; cp < scores.length; cp++) {
						// sum over intermediate substates
						double max = 0;
						for (short unp = 0; unp < nPSubStates; unp++) {
							for (short ucp = 0; ucp < uScores.length; ucp++) {
								double score = pr.getScore(np, unp)
										* cr.getScore(ucp, cp)
										* ur.getScore(unp, ucp);
								if (score > max) {
									max = score;
									intermediateSubState1[np][cp] = unp;
									intermediateSubState2[np][cp] = ucp;
								}
							}
						}
						scores[cp][np] = max;
					}
				}
				resultR.setScores2(scores);
				// add rule to bestSumRulesUnderMax if it's better
				relaxViterbiRule(resultR, pState, intermediateSubState1,
						cState, intermediateSubState2);
			}
		}
	}

	public int getUnaryIntermediate(short start, short end) {
		return closedSumPaths[start][end];
	}

	@SuppressWarnings("unchecked")
	private boolean relaxSumRule(UnaryRule ur, int intState1, int intState2) {
		// TODO: keep track of path
		UnaryRule bestR = (UnaryRule) bestSumRulesUnderMax.get(ur);
		if (bestR == null) {
			bestSumRulesUnderMax.put(ur, ur);
			closedSumRulesWithParent[ur.parentState].add(ur);
			closedSumRulesWithChild[ur.childState].add(ur);
			return true;
		} else {
			boolean change = false;
			for (int i = 0; i < ur.scores[0].length; i++) {
				for (int j = 0; j < ur.scores.length; j++) {
					if (bestR.scores[j][i] < ur.scores[j][i]) {
						bestR.scores[j][i] = ur.scores[j][i];
						change = true;
					}
				}
			}
			return change;
		}
	}

	public void computePairsOfUnaries() {
		// closedSumRulesWithParent = closedViterbiRulesWithParent =
		// unaryRulesWithParent;
		for (short parentState = 0; parentState < numStates; parentState++) {
			for (short childState = 0; childState < numStates; childState++) {
				if (parentState == childState)
					continue;
				int nParentSubStates = numSubStates[parentState];
				int nChildSubStates = numSubStates[childState];
				UnaryRule resultRsum = new UnaryRule(parentState, childState);
				UnaryRule resultRmax = new UnaryRule(parentState, childState);
				double[][] scoresSum = new double[nChildSubStates][nParentSubStates];
				double[][] scoresMax = new double[nChildSubStates][nParentSubStates];
				double maxSumScore = -1;
				short bestSumIntermed = -1;
				short bestMaxIntermed = -2;
				for (int i = 0; i < unaryRulesWithParent[parentState].size(); i++) {
					UnaryRule pr = (UnaryRule) unaryRulesWithParent[parentState]
							.get(i);
					short state = pr.getChildState();
					if (state == childState) {
						double total = 0;
						double[][] scores = pr.getScores2();
						for (int cp = 0; cp < nChildSubStates; cp++) {
							if (scores[cp] == null)
								continue;
							for (int np = 0; np < nParentSubStates; np++) {
								// sum over intermediate substates
								double sum = scores[cp][np];
								scoresSum[cp][np] += sum;
								total += sum;
								if (sum > scoresMax[cp][np]) {
									scoresMax[cp][np] = sum;
									bestMaxIntermed = -1;
								}
							}
						}
						if (total > maxSumScore) {
							bestSumIntermed = -1;
							maxSumScore = total;
						}
					} else {
						for (int j = 0; j < unaryRulesWithC[childState].size(); j++) {
							UnaryRule cr = (UnaryRule) unaryRulesWithC[childState]
									.get(j);
							if (state != cr.getParentState())
								continue;
							int nMySubStates = numSubStates[state];
							double total = 0;
							for (int np = 0; np < nParentSubStates; np++) {
								for (int cp = 0; cp < nChildSubStates; cp++) {
									// sum over intermediate substates
									double sum = 0;
									double max = 0;
									for (int unp = 0; unp < nMySubStates; unp++) {
										double val = pr.getScore(np, unp)
												* cr.getScore(unp, cp);
										sum += val;
										max = Math.max(max, val);
									}
									scoresSum[cp][np] += sum;
									total += sum;
									if (max > scoresMax[cp][np]) {
										scoresMax[cp][np] = max;
										bestMaxIntermed = state;
									}
								}
							}
							if (total > maxSumScore) {
								maxSumScore = total;
								bestSumIntermed = state;
							}
						}
					}
				}
				if (maxSumScore > -1) {
					resultRsum.setScores2(scoresSum);
					addUnary(resultRsum);
					closedSumRulesWithParent[parentState].add(resultRsum);
					closedSumRulesWithChild[childState].add(resultRsum);
					closedSumPaths[parentState][childState] = bestSumIntermed;
				}
				if (bestMaxIntermed > -2) {
					resultRmax.setScores2(scoresMax);
					// addUnary(resultR);
					closedViterbiRulesWithParent[parentState].add(resultRmax);
					closedViterbiRulesWithChild[childState].add(resultRmax);
					closedViterbiPaths[parentState][childState] = bestMaxIntermed;
					/*
					 * if (bestMaxIntermed > -1){
					 * System.out.println("NEW RULE CREATED"); }
					 */
				}
			}
		}

	}

	/*
	 * @SuppressWarnings("unchecked") private boolean relaxSumRule(UnaryRule
	 * rule) { bestSumRulesUnderMax.put(rule, rule);
	 * closedSumRulesWithParent[rule.parentState].add(rule);
	 * closedSumRulesWithChild[rule.childState].add(rule); return true; }
	 */
	/**
	 * Update the best unary chain probabilities and paths with this new rule.
	 * 
	 * @param ur
	 * @param subStates1
	 * @param subStates2
	 * @return
	 */
	@SuppressWarnings("unchecked")
	private void relaxViterbiRule(UnaryRule ur, short intState1,
			short[][] intSubStates1, short intState2, short[][] intSubStates2) {
		throw new Error("Viterbi closure is broken!");
		/*
		 * UnaryRule bestR = (UnaryRule) bestViterbiRulesUnderMax.get(ur);
		 * boolean isNewRule = (bestR==null); if (isNewRule) {
		 * bestViterbiRulesUnderMax.put(ur, ur);
		 * closedViterbiRulesWithParent[ur.parentState].add(ur);
		 * closedViterbiRulesWithChild[ur.childState].add(ur); bestR = ur; } for
		 * (int i=0; i<ur.scores[0].length; i++) { for (int j=0;
		 * j<ur.scores.length; j++) { if (isNewRule || bestR.scores[j][i] <
		 * ur.scores[j][i]) { bestR.scores[j][i] = ur.scores[j][i]; // update
		 * best path information if (findClosedPaths) { short[] intermediate =
		 * null; if (ur.parentState==intState1 && intSubStates1[i][j]==i) {
		 * intermediate = new short[2]; intermediate[0] = intState2;
		 * intermediate[1] = intSubStates2[i][j]; } else { //intermediate =
		 * closedViterbiPaths
		 * [ur.parentState][intState1][i][intSubStates1[i][j]]; } if
		 * (closedViterbiPaths[ur.parentState][ur.childState]==null) {
		 * closedViterbiPaths[ur.parentState][ur.childState] = new
		 * short[numSubStates[ur.parentState]][numSubStates[ur.childState]][]; }
		 * closedViterbiPaths[ur.parentState][ur.childState][i][j] =
		 * intermediate; } } } }
		 */}

	/**
	 * Initialize the best unary chain probabilities and paths with this rule.
	 * 
	 * @param rule
	 */
	@SuppressWarnings("unchecked")
	private void relaxViterbiRule(UnaryRule rule) {
		bestViterbiRulesUnderMax.put(rule, rule);
		closedViterbiRulesWithParent[rule.parentState].add(rule);
		closedViterbiRulesWithChild[rule.childState].add(rule);
		if (findClosedPaths) {
			for (short i = 0; i < rule.scores.length; i++) {
				for (short j = 0; j < rule.scores[i].length; j++) {
					short[] pair = new short[2];
					pair[0] = rule.childState;
					pair[1] = j;
					/*
					 * if
					 * (closedViterbiPaths[rule.parentState][rule.childState]==
					 * null) {
					 * closedViterbiPaths[rule.parentState][rule.childState] =
					 * new short[rule.scores.length][rule.scores[0].length][]; }
					 * closedViterbiPaths
					 * [rule.parentState][rule.childState][i][j] = pair;
					 */
				}
			}
		}
	}

	/**
	 * 'parentRules', 'childRules' and the return value all have the same format
	 * as 'unaryRulesWithParent', but can be thought of as square matrices. All
	 * this function does is a matrix multiplication, but operating directly on
	 * this non-standard matrix representation. 'parentRules' gives the
	 * probability of going from A to B, 'childRules' from B to C, and the
	 * return value from A to C (summing out B). This function is intended
	 * primarily to compute unaryRulesWithParent^n.
	 */
	private List<UnaryRule>[] matrixMultiply(List<UnaryRule>[] parentRules,
			List<UnaryRule>[] childRules) {
		throw new Error("I'm broken by parent first");
		/*
		 * double[][][][] scores = new double[numStates][numStates][][]; for (
		 * short A=0; A<numStates; A++ ) { for ( UnaryRule rAB : parentRules[A]
		 * ) { short B = rAB.childState; double[][] scoresAB = rAB.getScores();
		 * for ( UnaryRule rBC : childRules[B] ) { short C = rBC.childState; if
		 * ( scores[A][C] == null ) { scores[A][C] = new
		 * double[numSubStates[A]][numSubStates[C]];
		 * ArrayUtil.fill(scores[A][C], Double.NEGATIVE_INFINITY); } double[][]
		 * scoresBC = rBC.getScores(); double[] scoresToAdd = new
		 * double[numSubStates[B]+1]; for ( int a = 0; a < numSubStates[A]; a++
		 * ) { for ( int c = 0; c < numSubStates[C]; c++ ) { //
		 * Arrays.fill(scoresToAdd, Double.NEGATIVE_INFINITY); // No need to
		 * here scoresToAdd[scoresToAdd.length-1] = scores[A][C][a][c]; // The
		 * current score to which to add the new contributions for ( int b = 0;
		 * b < numSubStates[B]; b++ ) { scoresToAdd[b] = scoresAB[a][b] +
		 * scoresBC[b][c]; } scores[A][C][a][c] =
		 * SloppyMath.logAdd(scoresToAdd); } } } } }
		 * 
		 * @SuppressWarnings("unchecked") List<UnaryRule>[] result = new
		 * List[numStates]; for ( short A=0; A<numStates; A++ ) { result[A] =
		 * new ArrayList<UnaryRule>(); for ( short C=0; C<numStates; C++ ) { if
		 * ( scores[A][C] != null ) { result[A].add(new
		 * UnaryRule(A,C,scores[A][C])); } } } return result;
		 */
	}

	/**
	 * rules1 += rules2 (adds rules2 into rules1, destroying rules1) No sharing
	 * of score arrays occurs because of this operation since rules2 data is
	 * either added in or copied.
	 * 
	 * @param rules1
	 * @param rules2
	 */
	private void matrixAdd(List<UnaryRule>[] rules1, List<UnaryRule>[] rules2) {
		throw new Error("I'm broken by parent first");
		/*
		 * for ( short A=0; A<numStates; A++ ) { for ( UnaryRule r2 : rules2[A]
		 * ) { short child2 = r2.getChildState(); double[][] scores2 =
		 * r2.getScores(); boolean matchFound = false; for ( UnaryRule r1 :
		 * rules1[A] ) { short child1 = r1.getChildState(); if ( child1 ==
		 * child2 ) { double[][] scores1 = r1.getScores(); for ( int a = 0; a <
		 * numSubStates[A]; a++ ) { for ( int c = 0; c < numSubStates[child1];
		 * c++ ) { scores1[a][c] = SloppyMath.logAdd(scores1[a][c],
		 * scores2[a][c]); } } matchFound = true; break; } } if (!matchFound) {
		 * // Make a (deep) copy of rule r2 UnaryRule ruleCopy = new
		 * UnaryRule(r2); double[][] scoresCopy = new
		 * double[numSubStates[A]][numSubStates[child2]]; for ( int a = 0; a <
		 * numSubStates[A]; a++ ) { for ( int c = 0; c < numSubStates[child2];
		 * c++ ) { scoresCopy[a][c] = scores2[a][c]; } }
		 * ruleCopy.setScores(scoresCopy); rules1[A].add(ruleCopy); } } }
		 */
	}

	private List<UnaryRule>[] matrixUnity() {
		throw new Error("I'm broken by parent first");
		// List<UnaryRule>[] result = new List[numStates];
		// for ( short A=0; A<numStates; A++ ) {
		// result[A] = new ArrayList<UnaryRule>();
		// double[][] scores = new double[numSubStates[A]][numSubStates[A]];
		// ArrayUtil.fill(scores, Double.NEGATIVE_INFINITY);
		// for ( int a = 0; a < numSubStates[A]; a++ ) {
		// scores[a][a] = 0;
		// }
		// UnaryRule rule = new UnaryRule(A, A, scores);
		// result[A].add(rule);
		// }
		// return result;
	}

	/**
	 * @param P
	 * @return I + P + P^2 + P^3 + ... (approximation by truncation after some
	 *         power)
	 */
	private List<UnaryRule>[] sumProductUnaryClosure(List<UnaryRule>[] P) {
		throw new Error("I'm broken by parent first");
		/*
		 * List<UnaryRule>[] R = matrixUnity(); matrixAdd(R, P); // R = I + P +
		 * P^2 + P^3 + ... List<UnaryRule>[] Q = P; // Q = P^k int maxPower = 3;
		 * for ( int i = 1; i < maxPower; i++ ) { Q = matrixMultiply(Q, P);
		 * matrixAdd(R, Q); } return R;
		 */
	}

	/**
	 * Assumption: A in possibleSt ==> V[A] != null. This property is true of
	 * the result as well. The converse is not true because of a workaround for
	 * part of speech tags that we must handle here.
	 * 
	 * @param V
	 *            (considered a row vector, indexed by (state, substate))
	 * @param M
	 *            (a matrix represented in List<UnaryRule>[] (by parent) format)
	 * @param possibleSt
	 *            (a list of possible states to consider)
	 * @return U=V*M (row vector)
	 */
	public double[][] matrixVectorPreMultiply(double[][] V,
			List<UnaryRule>[] M, List<Integer> possibleSt) {
		throw new Error("I'm broken by parent first");
		/*
		 * double[][] U = new double[numStates][]; for (int pState :
		 * possibleSt){ U[pState] = new double[numSubStates[pState]];
		 * Arrays.fill(U[pState], Double.NEGATIVE_INFINITY); UnaryRule[] unaries
		 * = M[pState].toArray(new UnaryRule[0]); for ( UnaryRule ur : unaries )
		 * { int cState = ur.childState; if ( V[cState] == null ) { continue; }
		 * double[][] scores = ur.getScores(); // numSubStates[pState] *
		 * numSubStates[cState] int nParentStates = numSubStates[pState]; int
		 * nChildStates = numSubStates[cState]; double[] termsToAdd = new
		 * double[nChildStates+1]; // Could be inside the for(np) loop for (int
		 * np = 0; np < nParentStates; np++) { Arrays.fill(termsToAdd,
		 * Double.NEGATIVE_INFINITY); double currentVal = U[pState][np];
		 * termsToAdd[termsToAdd.length-1] = currentVal; for (int cp = 0; cp <
		 * nChildStates; cp++) { double iS = V[cState][cp]; if (iS ==
		 * Double.NEGATIVE_INFINITY) { continue; } double pS = scores[np][cp];
		 * termsToAdd[cp] = iS + pS; }
		 * 
		 * double newVal = SloppyMath.logAdd(termsToAdd); if (newVal >
		 * currentVal) { U[pState][np] = newVal; } } } } return U;
		 */
	}

	/**
	 * Assumption: A in possibleSt ==> V[A] != null. This property is true of
	 * the result as well. The converse is not true because of a workaround for
	 * part of speech tags that we must handle here.
	 * 
	 * @param M
	 *            (a matrix represented in List<UnaryRule>[] (by parent) format)
	 * @param V
	 *            (considered a column vector, indexed by (state, substate))
	 * @param possibleSt
	 *            (a list of possible states to consider)
	 * @return U=M*V (column vector)
	 */
	public double[][] matrixVectorPostMultiply(List<UnaryRule>[] M,
			double[][] V, List<Integer> possibleSt) {
		throw new Error("I'm broken by parent first");
		/*
		 * double[][] U = new double[numStates][]; for (int cState :
		 * possibleSt){ U[cState] = new double[numSubStates[cState]];
		 * Arrays.fill(U[cState], Double.NEGATIVE_INFINITY); } for (int pState :
		 * possibleSt){ UnaryRule[] unaries = M[pState].toArray(new
		 * UnaryRule[0]); for ( UnaryRule ur : unaries ) { int cState =
		 * ur.childState; if ( U[cState] == null ) { continue; } double[][]
		 * scores = ur.getScores(); // numSubStates[pState] *
		 * numSubStates[cState] int nParentStates = numSubStates[pState]; int
		 * nChildStates = numSubStates[cState]; double[] termsToAdd = new
		 * double[nParentStates+1]; // Could be inside the for(np) loop for (int
		 * cp = 0; cp < nChildStates; cp++) { Arrays.fill(termsToAdd,
		 * Double.NEGATIVE_INFINITY); double currentVal = U[cState][cp];
		 * termsToAdd[termsToAdd.length-1] = currentVal; for (int np = 0; np <
		 * nParentStates; np++) { double oS = V[pState][np]; if (oS ==
		 * Double.NEGATIVE_INFINITY) { continue; } double pS = scores[np][cp];
		 * termsToAdd[cp] = oS + pS; }
		 * 
		 * double newVal = SloppyMath.logAdd(termsToAdd); if (newVal >
		 * currentVal) { U[cState][cp] = newVal; } } } } return U;
		 */
	}

	/**
	 * Populates the "splitRules" accessor lists using the existing rule lists.
	 * If the state is synthetic, these lists contain all rules for the state.
	 * If the state is NOT synthetic, these lists contain only the rules in
	 * which both children are not synthetic.
	 * <p>
	 * <i>This method must be called before the grammar is used, either after
	 * training or deserializing grammar.</i>
	 */
	@SuppressWarnings("unchecked")
	public void splitRules() {
		// splitRulesWithLC = new BinaryRule[numStates][];
		// splitRulesWithRC = new BinaryRule[numStates][];
		// makeRulesAccessibleByChild();

		if (binaryRulesWithParent == null)
			return;
		splitRulesWithP = new BinaryRule[numStates][];
		splitRulesWithLC = new BinaryRule[numStates][];
		splitRulesWithRC = new BinaryRule[numStates][];

		for (int state = 0; state < numStates; state++) {
			splitRulesWithLC[state] = toBRArray(binaryRulesWithLC[state]);
			splitRulesWithRC[state] = toBRArray(binaryRulesWithRC[state]);
			splitRulesWithP[state] = toBRArray(binaryRulesWithParent[state]);
		}
		// we don't need the original lists anymore
		binaryRulesWithParent = null;
		binaryRulesWithLC = null;
		binaryRulesWithRC = null;
		makeCRArrays();
	}

	public BinaryRule[] splitRulesWithLC(int state) {
		// System.out.println("splitRulesWithLC not supported anymore.");
		// return null;
		if (state >= splitRulesWithLC.length) {
			return new BinaryRule[0];
		}
		return splitRulesWithLC[state];
	}

	public BinaryRule[] splitRulesWithRC(int state) {
		// System.out.println("splitRulesWithLC not supported anymore.");
		// return null;
		if (state >= splitRulesWithRC.length) {
			return new BinaryRule[0];
		}
		return splitRulesWithRC[state];
	}

	public BinaryRule[] splitRulesWithP(int state) {
		if (splitRulesWithP == null)
			splitRules();
		if (state >= splitRulesWithP.length) {
			return new BinaryRule[0];
		}
		return splitRulesWithP[state];
	}

	private BinaryRule[] toBRArray(List<BinaryRule> list) {
		// Collections.sort(list, Rule.scoreComparator()); // didn't seem to
		// help
		BinaryRule[] array = new BinaryRule[list.size()];
		for (int i = 0; i < array.length; i++) {
			array[i] = list.get(i);
		}
		return array;
	}

	public double[][] getUnaryScore(short pState, short cState) {
		UnaryRule r = getUnaryRule(pState, cState);
		if (r != null)
			return r.getScores2();
		if (GrammarTrainer.VERBOSE)
			System.out.println("The requested rule (" + uSearchRule
					+ ") is not in the grammar!");
		double[][] uscores = new double[numSubStates[cState]][numSubStates[pState]];
		ArrayUtil.fill(uscores, 1.0);
		return uscores;
	}

	/**
	 * @param pState
	 * @param cState
	 * @return
	 */
	public UnaryRule getUnaryRule(short pState, short cState) {
		UnaryRule uRule = new UnaryRule(pState, cState);
		UnaryRule r = unaryRuleMap.get(uRule);
		return r;
	}

	public double[][] getUnaryScore(UnaryRule rule) {
		UnaryRule r = unaryRuleMap.get(rule);
		if (r != null)
			return r.getScores2();
		if (GrammarTrainer.VERBOSE)
			System.out.println("The requested rule (" + rule
					+ ") is not in the grammar!");
		double[][] uscores = new double[numSubStates[rule.getChildState()]][numSubStates[rule
				.getParentState()]];
		ArrayUtil.fill(uscores, 1.0);
		return uscores;
	}

	public double[][][] getBinaryScore(short pState, short lState, short rState) {
		BinaryRule r = getBinaryRule(pState, lState, rState);
		if (r != null)
			return r.getScores2();
		if (GrammarTrainer.VERBOSE)
			System.out.println("The requested rule (" + bSearchRule
					+ ") is not in the grammar!");
		double[][][] bscores = new double[numSubStates[lState]][numSubStates[rState]][numSubStates[pState]];
		ArrayUtil.fill(bscores, 1.0);
		return bscores;
	}

	/**
	 * @param pState
	 * @param lState
	 * @param rState
	 * @return
	 */
	public BinaryRule getBinaryRule(short pState, short lState, short rState) {
		BinaryRule bRule = new BinaryRule(pState, lState, rState);
		BinaryRule r = binaryRuleMap.get(bRule);
		return r;
	}

	public double[][][] getBinaryScore(BinaryRule rule) {
		BinaryRule r = binaryRuleMap.get(rule);
		if (r != null)
			return r.getScores2();
		else {
			if (GrammarTrainer.VERBOSE)
				System.out.println("The requested rule (" + rule
						+ ") is not in the grammar!");
			double[][][] bscores = new double[numSubStates[rule
					.getLeftChildState()]][numSubStates[rule
					.getRightChildState()]][numSubStates[rule.getParentState()]];
			ArrayUtil.fill(bscores, 1.0);
			return bscores;
		}
	}

	public void printSymbolCounter(Numberer tagNumberer) {
		Set<Integer> set = symbolCounter.keySet();
		PriorityQueue<String> pq = new PriorityQueue<String>(set.size());
		for (Integer i : set) {
			pq
					.add((String) tagNumberer.object(i), symbolCounter
							.getCount(i, 0));
			// System.out.println(i+". "+(String)tagNumberer.object(i)+"\t
			// "+symbolCounter.getCount(i,0));
		}
		int i = 0;
		while (pq.hasNext()) {
			i++;
			int p = (int) pq.getPriority();
			System.out.println(i + ". " + pq.next() + "\t " + p);
		}
	}

	public int getSymbolCount(Integer i) {
		return (int) symbolCounter.getCount(i, 0);
	}

	private void makeRulesAccessibleByChild() {
		// first the binaries
		if (true)
			return;
		for (int state = 0; state < numStates; state++) {
			if (!isGrammarTag[state])
				continue;
			if (binaryRulesWithParent == null)
				continue;
			for (BinaryRule rule : binaryRulesWithParent[state]) {
				binaryRulesWithLC[rule.leftChildState].add(rule);
				binaryRulesWithRC[rule.rightChildState].add(rule);
			}
			// for (UnaryRule rule : unaryRulesWithParent[state]){
			// unaryRulesWithC[rule.childState].add(rule);
			// }
		}

	}

	/**
	 * Split all substates into two new ones. This produces a new Grammar with
	 * updated rules.
	 * 
	 * @param randomness
	 *            percent randomness applied in splitting rules
	 * @param mode
	 *            0: normalized (at least almost) 1: not normalized (when
	 *            splitting a log-linear grammar) 2: just noise (for log-linear
	 *            grammars with cascading regularization)
	 * @return
	 */
	public Grammar splitAllStates(double randomness, int[] counts,
			boolean moreSubstatesThanCounts, int mode) {
		if (logarithmMode) {
			throw new Error(
					"Do not split states when Grammar is in logarithm mode");
		}
		short[] newNumSubStates = new short[numSubStates.length];
		for (short i = 0; i < numSubStates.length; i++) {
			// don't split a state into more substates than times it was
			// actaully seen

			// if (!moreSubstatesThanCounts && numSubStates[i]>=counts[i]) {
			// newNumSubStates[i]=numSubStates[i];
			// }
			// else{
			newNumSubStates[i] = (short) (numSubStates[i] * 2);
			// }
		}
		boolean doNotNormalize = (mode == 1);
		newNumSubStates[0] = 1; // never split ROOT
		// create the new grammar
		Grammar grammar = new Grammar(newNumSubStates, findClosedPaths,
				smoother, this, threshold);
		Random random = GrammarTrainer.RANDOM;

		for (BinaryRule oldRule : binaryRuleMap.keySet()) {
			BinaryRule newRule = oldRule.splitRule(numSubStates,
					newNumSubStates, random, randomness, doNotNormalize, mode);
			grammar.addBinary(newRule);
		}

		for (UnaryRule oldRule : unaryRuleMap.keySet()) {
			UnaryRule newRule = oldRule.splitRule(numSubStates,
					newNumSubStates, random, randomness, doNotNormalize, mode);
			grammar.addUnary(newRule);
		}
		grammar.isGrammarTag = this.isGrammarTag;
		grammar.extendSplitTrees(splitTrees, numSubStates);
		grammar.computePairsOfUnaries();
		return grammar;
	}

	@SuppressWarnings("unchecked")
	public void extendSplitTrees(Tree<Short>[] trees, short[] oldNumSubStates) {
		this.splitTrees = new Tree[numStates];
		for (int tag = 0; tag < splitTrees.length; tag++) {
			Tree<Short> splitTree = trees[tag].shallowClone();
			for (Tree<Short> leaf : splitTree.getTerminals()) {
				List<Tree<Short>> children = leaf.getChildren();
				if (numSubStates[tag] > oldNumSubStates[tag]) {
					children
							.add(new Tree<Short>((short) (2 * leaf.getLabel())));
					children.add(new Tree<Short>(
							(short) (2 * leaf.getLabel() + 1)));
				} else {
					children.add(new Tree<Short>(leaf.getLabel()));
				}
			}
			this.splitTrees[tag] = splitTree;
		}
	}

	public int totalSubStates() {
		int count = 0;
		for (int i = 0; i < numStates; i++) {
			count += numSubStates[i];
		}
		return count;
	}

	/**
	 * Tally the probability of seeing each substate. This data is needed for
	 * tallyMergeScores. mergeWeights is indexed as [state][substate]. This data
	 * should be normalized before being used by another function.
	 * 
	 * @param tree
	 * @param mergeWeights
	 *            The probability of seeing substate given state.
	 */
	public void tallyMergeWeights(Tree<StateSet> tree, double mergeWeights[][]) {
		if (tree.isLeaf())
			return;
		StateSet label = tree.getLabel();
		short state = label.getState();
		double probs[] = new double[label.numSubStates()];
		double total = 0, tmp;
		for (short i = 0; i < label.numSubStates(); i++) {
			tmp = label.getIScore(i) * label.getOScore(i);
			// TODO: put in the scale parameters???
			probs[i] = tmp;
			total += tmp;
		}
		if (total == 0)
			total = 1;
		for (short i = 0; i < label.numSubStates(); i++) {
			mergeWeights[state][i] += probs[i] / total;
		}
		for (Tree<StateSet> child : tree.getChildren()) {
			tallyMergeWeights(child, mergeWeights);
		}
	}

	/*
	 * normalize merge weights. assumes that the mergeWeights are given as logs.
	 * the normalized weights are returned as probabilities.
	 */
	public void normalizeMergeWeights(double[][] mergeWeights) {
		for (int state = 0; state < mergeWeights.length; state++) {
			double sum = 0;
			for (int subState = 0; subState < numSubStates[state]; subState++) {
				sum += mergeWeights[state][subState];
			}
			if (sum == 0)
				sum = 1;
			for (int subState = 0; subState < numSubStates[state]; subState++) {
				mergeWeights[state][subState] /= sum;
			}
		}
	}

	/**
	 * Calculate the log likelihood gain of merging pairs of split states
	 * together. This information is returned in deltas[state][merged substate].
	 * It requires mergeWeights to be calculated by tallyMergeWeights.
	 * 
	 * @param tree
	 * @param deltas
	 *            The log likelihood gained by merging pairs of substates.
	 * @param mergeWeights
	 *            The probability of seeing substate given state.
	 */
	public void tallyMergeScores(Tree<StateSet> tree, double[][][] deltas,
			double[][] mergeWeights) {
		if (tree.isLeaf())
			return;
		StateSet label = tree.getLabel();
		short state = label.getState();
		double[] separatedScores = new double[label.numSubStates()];
		double[] combinedScores = new double[label.numSubStates()];
		double combinedScore;
		// calculate separated scores

		double separatedScoreSum = 0, tmp;
		// don't need to deal with scale factor because we divide below
		for (int i = 0; i < label.numSubStates(); i++) {
			tmp = label.getIScore(i) * label.getOScore(i);
			combinedScores[i] = separatedScores[i] = tmp;
			separatedScoreSum += tmp;
		}
		// calculate merged scores
		for (short i = 0; i < numSubStates[state]; i++) {
			for (short j = (short) (i + 1); j < numSubStates[state]; j++) {
				short[] map = new short[2];
				map[0] = i;
				map[1] = j;
				double[] tmp1 = new double[2], tmp2 = new double[2];
				double mergeWeightSum = 0;
				for (int k = 0; k < 2; k++) {
					mergeWeightSum += mergeWeights[state][map[k]];
				}
				if (mergeWeightSum == 0)
					mergeWeightSum = 1;
				for (int k = 0; k < 2; k++) {
					tmp1[k] = label.getIScore(map[k])
							* mergeWeights[state][map[k]] / mergeWeightSum;
					tmp2[k] = label.getOScore(map[k]);
				}
				combinedScore = (tmp1[0] + tmp1[1]) * (tmp2[0] + tmp2[1]);
				combinedScores[i] = combinedScore;
				combinedScores[j] = 0;
				if (combinedScore != 0 && separatedScoreSum != 0)
					deltas[state][i][j] += Math.log(separatedScoreSum
							/ ArrayUtil.sum(combinedScores));
				for (int k = 0; k < 2; k++)
					combinedScores[map[k]] = separatedScores[map[k]];
				if (Double.isNaN(deltas[state][i][j])) {
					System.out.println(" deltas[" + tagNumberer.object(state)
							+ "][" + i + "][" + j + "] = NaN");
					System.out.println(Arrays.toString(separatedScores) + " "
							+ Arrays.toString(tmp1) + " "
							+ Arrays.toString(tmp2) + " " + combinedScore + " "
							+ Arrays.toString(mergeWeights[state]));
				}
			}
		}

		for (Tree<StateSet> child : tree.getChildren()) {
			tallyMergeScores(child, deltas, mergeWeights);
		}
	}

	/**
	 * This merges the substate pairs indicated by
	 * mergeThesePairs[state][substate pair]. It requires merge weights
	 * calculated by tallyMergeWeights.
	 * 
	 * @param mergeThesePairs
	 *            Which substate pairs to merge.
	 * @param mergeWeights
	 *            The probability of seeing each substate.
	 */
	public Grammar mergeStates(boolean[][][] mergeThesePairs,
			double[][] mergeWeights) {
		if (logarithmMode) {
			throw new Error("Do not merge grammars in logarithm mode!");
		}
		short[] newNumSubStates = new short[numSubStates.length];
		short[][] mapping = new short[numSubStates.length][];
		// invariant: if partners[state][substate][0] == substate, it's the 1st
		// one
		short[][][] partners = new short[numSubStates.length][][];
		calculateMergeArrays(mergeThesePairs, newNumSubStates, mapping,
				partners, numSubStates);
		// create the new grammar
		Grammar grammar = new Grammar(newNumSubStates, findClosedPaths,
				smoother, this, threshold);
		// for (Rule r : allRules) {
		// if (r instanceof BinaryRule) {
		for (BinaryRule oldRule : binaryRuleMap.keySet()) {
			// BinaryRule oldRule = r;
			short pS = oldRule.getParentState(), lcS = oldRule
					.getLeftChildState(), rcS = oldRule.getRightChildState();
			double[][][] oldScores = oldRule.getScores2();
			// merge binary rule
			double[][][] newScores = new double[newNumSubStates[lcS]][newNumSubStates[rcS]][newNumSubStates[pS]];
			for (int i = 0; i < numSubStates[pS]; i++) {
				if (partners[pS][i][0] == i) {
					int parentSplit = partners[pS][i].length;
					for (int j = 0; j < numSubStates[lcS]; j++) {
						if (partners[lcS][j][0] == j) {
							int leftSplit = partners[lcS][j].length;
							for (int k = 0; k < (numSubStates[rcS]); k++) {
								if (partners[rcS][k][0] == k) {
									int rightSplit = partners[rcS][k].length;
									double[][][] scores = new double[leftSplit][rightSplit][parentSplit];
									for (int js = 0; js < leftSplit; js++) {
										for (int ks = 0; ks < rightSplit; ks++) {
											if (oldScores[partners[lcS][j][js]][partners[rcS][k][ks]] == null)
												continue;
											for (int is = 0; is < parentSplit; is++) {
												scores[js][ks][is] = oldScores[partners[lcS][j][js]][partners[rcS][k][ks]][partners[pS][i][is]];
											}
										}
									}
									if (rightSplit == 2) {
										for (int is = 0; is < parentSplit; is++) {
											for (int js = 0; js < leftSplit; js++) {
												scores[js][0][is] = scores[js][1][is] = scores[js][0][is]
														+ scores[js][1][is];
											}
										}
									}
									if (leftSplit == 2) {
										for (int is = 0; is < parentSplit; is++) {
											for (int ks = 0; ks < rightSplit; ks++) {
												scores[0][ks][is] = scores[1][ks][is] = scores[0][ks][is]
														+ scores[1][ks][is];
											}
										}
									}
									if (parentSplit == 2) {
										for (int js = 0; js < leftSplit; js++) {
											for (int ks = 0; ks < rightSplit; ks++) {
												double mergeWeightSum = mergeWeights[pS][partners[pS][i][0]]
														+ mergeWeights[pS][partners[pS][i][1]];
												if (SloppyMath
														.isDangerous(mergeWeightSum))
													mergeWeightSum = 1;
												scores[js][ks][0] = scores[js][ks][1] = ((scores[js][ks][0] * mergeWeights[pS][partners[pS][i][0]]) + (scores[js][ks][1] * mergeWeights[pS][partners[pS][i][1]]))
														/ mergeWeightSum;
											}
										}
									}
									for (int is = 0; is < parentSplit; is++) {
										for (int js = 0; js < leftSplit; js++) {
											for (int ks = 0; ks < rightSplit; ks++) {
												newScores[mapping[lcS][partners[lcS][j][js]]][mapping[rcS][partners[rcS][k][ks]]][mapping[pS][partners[pS][i][is]]] = scores[js][ks][is];
											}
										}
									}
								}
							}
						}
					}
				}
			}
			BinaryRule newRule = new BinaryRule(oldRule);
			newRule.setScores2(newScores);
			grammar.addBinary(newRule);
		}
		// } else if (r instanceof UnaryRule) {
		for (UnaryRule oldRule : unaryRuleMap.keySet()) {
			// UnaryRule oldRule = (UnaryRule) r;
			short pS = oldRule.getParentState(), cS = oldRule.getChildState();
			// merge unary rule
			double[][] newScores = new double[newNumSubStates[cS]][newNumSubStates[pS]];
			double[][] oldScores = oldRule.getScores2();
			boolean allZero = true;
			for (int i = 0; i < numSubStates[pS]; i++) {
				if (partners[pS][i][0] == i) {
					int parentSplit = partners[pS][i].length;
					for (int j = 0; j < numSubStates[cS]; j++) {
						if (partners[cS][j][0] == j) {
							int childSplit = partners[cS][j].length;
							double[][] scores = new double[childSplit][parentSplit];
							for (int js = 0; js < childSplit; js++) {
								if (oldScores[partners[cS][j][js]] == null)
									continue;
								for (int is = 0; is < parentSplit; is++) {
									scores[js][is] = oldScores[partners[cS][j][js]][partners[pS][i][is]];
								}
							}
							if (childSplit == 2) {
								for (int is = 0; is < parentSplit; is++) {
									scores[0][is] = scores[1][is] = scores[0][is]
											+ scores[1][is];
								}
							}
							if (parentSplit == 2) {
								for (int js = 0; js < childSplit; js++) {
									double mergeWeightSum = mergeWeights[pS][partners[pS][i][0]]
											+ mergeWeights[pS][partners[pS][i][1]];
									if (SloppyMath.isDangerous(mergeWeightSum))
										mergeWeightSum = 1;
									scores[js][0] = scores[js][1] = ((scores[js][0] * mergeWeights[pS][partners[pS][i][0]]) + (scores[js][1] * mergeWeights[pS][partners[pS][i][1]]))
											/ mergeWeightSum;
								}
							}
							for (int is = 0; is < parentSplit; is++) {
								for (int js = 0; js < childSplit; js++) {
									newScores[mapping[cS][partners[cS][j][js]]][mapping[pS][partners[pS][i][is]]] = scores[js][is];
									allZero = allZero && (scores[js][is] == 0);
								}
							}
						}
					}
				}
			}
			// if (allZero){
			// System.out.println("Maybe an underflow? Rule: "+oldRule);
			// System.out.println(ArrayUtil.toString(newScores));
			// System.out.println(ArrayUtil.toString(oldScores));
			// System.out.println(Arrays.toString(mergeWeights[pS]));
			// }
			UnaryRule newRule = new UnaryRule(oldRule);
			newRule.setScores2(newScores);
			grammar.addUnary(newRule);
		}
		grammar.pruneSplitTree(partners, mapping);
		grammar.isGrammarTag = this.isGrammarTag;
		grammar.closedSumRulesWithParent = grammar.closedViterbiRulesWithParent = grammar.unaryRulesWithParent;
		grammar.closedSumRulesWithChild = grammar.closedViterbiRulesWithChild = grammar.unaryRulesWithC;

		return grammar;
	}

	/**
	 * @param mergeThesePairs
	 * @param partners
	 */
	private void pruneSplitTree(short[][][] partners, short[][] mapping) {
		for (int tag = 0; tag < splitTrees.length; tag++) {
			Tree<Short> splitTree = splitTrees[tag];
			int maxDepth = splitTree.getDepth();
			for (Tree<Short> preTerminal : splitTree.getAtDepth(maxDepth - 2)) {
				List<Tree<Short>> children = preTerminal.getChildren();
				ArrayList<Tree<Short>> newChildren = new ArrayList<Tree<Short>>(
						2);
				for (int i = 0; i < children.size(); i++) {
					Tree<Short> child = children.get(i);
					int curLoc = child.getLabel();
					if (partners[tag][curLoc][0] == curLoc) {
						newChildren.add(new Tree<Short>(mapping[tag][curLoc]));
					}
				}
				preTerminal.setChildren(newChildren);
			}
		}
	}

	public static void checkNormalization(Grammar grammar) {
		double[][] psum = new double[grammar.numSubStates.length][];
		for (int pS = 0; pS < grammar.numSubStates.length; pS++) {
			psum[pS] = new double[grammar.numSubStates[pS]];
		}
		boolean[] sawPS = new boolean[grammar.numSubStates.length];
		for (UnaryRule ur : grammar.unaryRuleMap.values()) {
			int pS = ur.getParentState();
			sawPS[pS] = true;
			int cS = ur.getChildState();
			double[][] scores = ur.getScores2();
			for (int ci = 0; ci < grammar.numSubStates[cS]; ci++) {
				if (scores[ci] == null)
					continue;
				for (int pi = 0; pi < grammar.numSubStates[pS]; pi++) {
					psum[pS][pi] += scores[ci][pi];
				}
			}
		}
		for (BinaryRule br : grammar.binaryRuleMap.values()) {
			int pS = br.getParentState();
			sawPS[pS] = true;
			int lcS = br.getLeftChildState();
			int rcS = br.getRightChildState();
			double[][][] scores = br.getScores2();
			for (int lci = 0; lci < grammar.numSubStates[lcS]; lci++) {
				for (int rci = 0; rci < grammar.numSubStates[rcS]; rci++) {
					if (scores[lci][rci] == null)
						continue;
					for (int pi = 0; pi < grammar.numSubStates[pS]; pi++) {
						psum[pS][pi] += scores[lci][rci][pi];
					}
				}
			}
		}
		System.out.println();
		System.out.println("Checking for substates whose probs don't sum to 1");
		for (int pS = 0; pS < grammar.numSubStates.length; pS++) {
			if (!sawPS[pS])
				continue;
			for (int pi = 0; pi < grammar.numSubStates[pS]; pi++) {
				if (Math.abs(1 - psum[pS][pi]) > 0.001)
					System.out.println(" state " + pS + " substate " + pi
							+ " gives bad psum: " + psum[pS][pi]);
			}
		}
	}

	/**
	 * @param mergeThesePairs
	 * @param newNumSubStates
	 * @param mapping
	 * @param partners
	 */
	public static void calculateMergeArrays(boolean[][][] mergeThesePairs,
			short[] newNumSubStates, short[][] mapping, short[][][] partners,
			short[] numSubStates) {
		for (short state = 0; state < numSubStates.length; state++) {
			short mergeTarget[] = new short[mergeThesePairs[state].length];
			Arrays.fill(mergeTarget, (short) -1);
			short count = 0;
			mapping[state] = new short[numSubStates[state]];
			partners[state] = new short[numSubStates[state]][];
			for (short j = 0; j < numSubStates[state]; j++) {
				if (mergeTarget[j] != -1) {
					mapping[state][j] = mergeTarget[j];
				} else {
					partners[state][j] = new short[1];
					partners[state][j][0] = j;
					mapping[state][j] = count;
					count++;
					// assume we're only merging pairs, so we only see things to
					// merge
					// with this substate when this substate isn't being merged
					// with anything
					// earlier
					for (short k = (short) (j + 1); k < numSubStates[state]; k++) {
						if (mergeThesePairs[state][j][k]) {
							mergeTarget[k] = mapping[state][j];
							partners[state][j] = new short[2];
							partners[state][j][0] = j;
							partners[state][j][1] = k;
							partners[state][k] = partners[state][j];
						}
					}
				}
			}
			newNumSubStates[state] = count;
		}
		newNumSubStates[0] = 1; // never split or merge ROOT
	}

	public void fixMergeWeightsEtc(boolean[][][] mergeThesePairs,
			double[][] mergeWeights, boolean[][][] complexMergePairs) {
		short[] newNumSubStates = new short[numSubStates.length];
		short[][] mapping = new short[numSubStates.length][];
		// invariant: if partners[state][substate][0] == substate, it's the 1st
		// one
		short[][][] partners = new short[numSubStates.length][][];
		calculateMergeArrays(mergeThesePairs, newNumSubStates, mapping,
				partners, numSubStates);
		for (int tag = 0; tag < numSubStates.length; tag++) {
			double[] newMergeWeights = new double[newNumSubStates[tag]];
			for (int i = 0; i < numSubStates[tag]; i++) {
				newMergeWeights[mapping[tag][i]] += mergeWeights[tag][i];
			}
			mergeWeights[tag] = newMergeWeights;

			boolean[][] newComplexMergePairs = new boolean[newNumSubStates[tag]][newNumSubStates[tag]];
			boolean[][] newMergeThesePairs = new boolean[newNumSubStates[tag]][newNumSubStates[tag]];
			for (int i = 0; i < complexMergePairs[tag].length; i++) {
				for (int j = 0; j < complexMergePairs[tag].length; j++) {
					newComplexMergePairs[mapping[tag][i]][mapping[tag][j]] = newComplexMergePairs[mapping[tag][i]][mapping[tag][j]]
							|| complexMergePairs[tag][i][j];
					newMergeThesePairs[mapping[tag][i]][mapping[tag][j]] = newMergeThesePairs[mapping[tag][i]][mapping[tag][j]]
							|| mergeThesePairs[tag][i][j];
				}
			}
			complexMergePairs[tag] = newComplexMergePairs;
			mergeThesePairs[tag] = newMergeThesePairs;
		}
	}

	public void logarithmMode() {
		// System.out.println("The gramar is in logarithmMode!");
		if (logarithmMode)
			return;
		logarithmMode = true;
		for (UnaryRule r : unaryRuleMap.keySet()) {
			logarithmModeRule(unaryRuleMap.get(r));
		}
		for (BinaryRule r : binaryRuleMap.keySet()) {
			logarithmModeRule(binaryRuleMap.get(r));
		}
		// Leon thinks the following sets of rules are already covered above,
		// but he wants to take no chances
		logarithmModeBRuleListArray(binaryRulesWithParent);
		logarithmModeBRuleListArray(binaryRulesWithLC);
		logarithmModeBRuleListArray(binaryRulesWithRC);
		logarithmModeBRuleArrayArray(splitRulesWithLC);
		logarithmModeBRuleArrayArray(splitRulesWithRC);
		logarithmModeBRuleArrayArray(splitRulesWithP);
		logarithmModeURuleListArray(unaryRulesWithParent);
		logarithmModeURuleListArray(unaryRulesWithC);
		logarithmModeURuleListArray(sumProductClosedUnaryRulesWithParent);
		logarithmModeURuleListArray(closedSumRulesWithParent);
		logarithmModeURuleListArray(closedSumRulesWithChild);
		logarithmModeURuleListArray(closedViterbiRulesWithParent);
		logarithmModeURuleListArray(closedViterbiRulesWithChild);
		logarithmModeURuleArrayArray(closedSumRulesWithP);
		logarithmModeURuleArrayArray(closedSumRulesWithC);
		logarithmModeURuleArrayArray(closedViterbiRulesWithP);
		logarithmModeURuleArrayArray(closedViterbiRulesWithC);
	}

	/**
	 * 
	 */
	private void logarithmModeBRuleListArray(List<BinaryRule>[] a) {
		if (a != null) {
			for (List<BinaryRule> l : a) {
				if (l == null)
					continue;
				for (BinaryRule r : l) {
					logarithmModeRule(r);
				}
			}
		}
	}

	/**
	 * 
	 */
	private void logarithmModeURuleListArray(List<UnaryRule>[] a) {
		if (a != null) {
			for (List<UnaryRule> l : a) {
				if (l == null)
					continue;
				for (UnaryRule r : l) {
					logarithmModeRule(r);
				}
			}
		}
	}

	/**
	 * 
	 */
	private void logarithmModeBRuleArrayArray(BinaryRule[][] a) {
		if (a != null) {
			for (BinaryRule[] l : a) {
				if (l == null)
					continue;
				for (BinaryRule r : l) {
					logarithmModeRule(r);
				}
			}
		}
	}

	/**
	 * 
	 */
	private void logarithmModeURuleArrayArray(UnaryRule[][] a) {
		if (a != null) {
			for (UnaryRule[] l : a) {
				if (l == null)
					continue;
				for (UnaryRule r : l) {
					logarithmModeRule(r);
				}
			}
		}
	}

	/**
	 * @param r
	 */
	private static void logarithmModeRule(BinaryRule r) {
		if (r == null || r.logarithmMode)
			return;
		r.logarithmMode = true;
		double[][][] scores = r.getScores2();
		for (int i = 0; i < scores.length; i++) {
			for (int j = 0; j < scores[i].length; j++) {
				if (scores[i][j] == null)
					continue;
				for (int k = 0; k < scores[i][j].length; k++) {
					scores[i][j][k] = Math.log(scores[i][j][k]);
				}
			}
		}
		r.setScores2(scores);
	}

	/**
	 * @param r
	 */
	private static void logarithmModeRule(UnaryRule r) {
		if (r == null || r.logarithmMode)
			return;
		r.logarithmMode = true;
		double[][] scores = r.getScores2();
		for (int j = 0; j < scores.length; j++) {
			if (scores[j] == null)
				continue;
			for (int k = 0; k < scores[j].length; k++) {
				scores[j][k] = Math.log(scores[j][k]);
			}
		}
		r.setScores2(scores);
	}

	public boolean isLogarithmMode() {
		return logarithmMode;
	}

	public final boolean isGrammarTag(int n) {
		return isGrammarTag[n];
	}

	public Grammar projectGrammar(double[] condProbs, int[][] fromMapping,
			int[][] toSubstateMapping) {
		short[] newNumSubStates = new short[numSubStates.length];
		for (int state = 0; state < numSubStates.length; state++) {
			newNumSubStates[state] = (short) toSubstateMapping[state][0];
		}

		Grammar grammar = new Grammar(newNumSubStates, findClosedPaths,
				smoother, this, threshold);
		for (BinaryRule oldRule : binaryRuleMap.keySet()) {
			short pcS = oldRule.getParentState(), lcS = oldRule
					.getLeftChildState(), rcS = oldRule.getRightChildState();
			double[][][] oldScores = oldRule.getScores2();
			// merge binary rule
			double[][][] newScores = new double[newNumSubStates[lcS]][newNumSubStates[rcS]][newNumSubStates[pcS]];
			for (int lS = 0; lS < numSubStates[lcS]; lS++) {
				for (int rS = 0; rS < numSubStates[rcS]; rS++) {
					if (oldScores[lS][rS] == null)
						continue;
					for (int pS = 0; pS < numSubStates[pcS]; pS++) {
						newScores[toSubstateMapping[lcS][lS + 1]][toSubstateMapping[rcS][rS + 1]][toSubstateMapping[pcS][pS + 1]] += condProbs[fromMapping[pcS][pS]]
								* oldScores[lS][rS][pS];
					}
				}
			}
			BinaryRule newRule = new BinaryRule(oldRule, newScores);
			grammar.addBinary(newRule);
		}
		for (UnaryRule oldRule : unaryRuleMap.keySet()) {
			short pcS = oldRule.getParentState(), ccS = oldRule.getChildState();
			double[][] oldScores = oldRule.getScores2();
			double[][] newScores = new double[newNumSubStates[ccS]][newNumSubStates[pcS]];
			for (int cS = 0; cS < numSubStates[ccS]; cS++) {
				if (oldScores[cS] == null)
					continue;
				for (int pS = 0; pS < numSubStates[pcS]; pS++) {
					newScores[toSubstateMapping[ccS][cS + 1]][toSubstateMapping[pcS][pS + 1]] += condProbs[fromMapping[pcS][pS]]
							* oldScores[cS][pS];
				}
			}
			UnaryRule newRule = new UnaryRule(oldRule, newScores);
			grammar.addUnary(newRule);
			// grammar.closedSumRulesWithParent[newRule.parentState].add(newRule);
			// grammar.closedSumRulesWithChild[newRule.childState].add(newRule);
		}

		grammar.computePairsOfUnaries();
		// grammar.splitRules();
		grammar.makeCRArrays();
		grammar.isGrammarTag = this.isGrammarTag;
		// System.out.println(grammar.toString());
		return grammar;
	}

	public Grammar copyGrammar(boolean noUnaryChains) {
		short[] newNumSubStates = numSubStates.clone();

		Grammar grammar = new Grammar(newNumSubStates, findClosedPaths,
				smoother, this, threshold);
		for (BinaryRule oldRule : binaryRuleMap.keySet()) {
			BinaryRule newRule = new BinaryRule(oldRule);
			grammar.addBinary(newRule);
		}
		for (UnaryRule oldRule : unaryRuleMap.keySet()) {
			UnaryRule newRule = new UnaryRule(oldRule);
			grammar.addUnary(newRule);
		}
		if (noUnaryChains) {
			closedSumRulesWithParent = closedViterbiRulesWithParent = unaryRulesWithParent;
			closedSumRulesWithChild = closedViterbiRulesWithChild = unaryRulesWithC;

		} else
			grammar.computePairsOfUnaries();
		grammar.makeCRArrays();
		grammar.isGrammarTag = this.isGrammarTag;
		/*
		 * grammar.ruleIndexer = ruleIndexer; grammar.startIndex = startIndex;
		 * grammar.nEntries = nEntries; grammar.toBeIgnored = toBeIgnored;
		 */
		return grammar;
	}

	public Grammar projectTo0LevelGrammar(double[] condProbs,
			int[][] fromMapping, int[][] toMapping) {
		int newNumStates = fromMapping[fromMapping.length - 1][0];
		// all rules have the same parent in this grammar
		double[][] newBinaryProbs = new double[newNumStates][newNumStates];
		double[] newUnaryProbs = new double[newNumStates];

		short[] newNumSubStates = new short[numSubStates.length];
		Arrays.fill(newNumSubStates, (short) 1);
		Grammar grammar = new Grammar(newNumSubStates, findClosedPaths,
				smoother, this, threshold);

		// short[] newNumSubStates = new short[newNumStates];
		// grammar.numSubStates = newNumSubStates;
		// grammar.numStates = (short)newNumStates;

		for (BinaryRule oldRule : binaryRuleMap.keySet()) {
			short pcS = oldRule.getParentState(), lcS = oldRule
					.getLeftChildState(), rcS = oldRule.getRightChildState();
			double[][][] oldScores = oldRule.getScores2();
			// merge binary rule
			// double[][][] newScores = new double[1][1][1];
			for (int lS = 0; lS < numSubStates[lcS]; lS++) {
				for (int rS = 0; rS < numSubStates[rcS]; rS++) {
					if (oldScores[lS][rS] == null)
						continue;
					for (int pS = 0; pS < numSubStates[pcS]; pS++) {
						newBinaryProbs[toMapping[lcS][lS]][toMapping[rcS][rS]] +=
						// newBinaryProbs[lcS][rcS] +=
						condProbs[fromMapping[pcS][pS]] * oldScores[lS][rS][pS];
					}
				}
			}
			// BinaryRule newRule = new BinaryRule(oldRule);
			// newRule.setScores2(newScores);
			// grammar.addBinary(newRule);
		}
		for (UnaryRule oldRule : unaryRuleMap.keySet()) {
			short pcS = oldRule.getParentState(), ccS = oldRule.getChildState();
			double[][] oldScores = oldRule.getScores2();
			for (int cS = 0; cS < numSubStates[ccS]; cS++) {
				if (oldScores[cS] == null)
					continue;
				for (int pS = 0; pS < numSubStates[pcS]; pS++) {
					// newScores[0][0] +=
					// condProbs[fromMapping[pcS][pS]]*oldScores[cS][pS];
					newUnaryProbs[toMapping[ccS][cS]] +=
					// newUnaryProbs[ccS] +=
					condProbs[fromMapping[pcS][pS]] * oldScores[cS][pS];

				}
			}
			// UnaryRule newRule = new UnaryRule(oldRule);
			// newRule.setScores2(newScores);
			// grammar.addUnary(newRule);
			// grammar.closedSumRulesWithParent[newRule.parentState].add(newRule);
			// grammar.closedSumRulesWithChild[newRule.childState].add(newRule);
		}

		for (short lS = 0; lS < newBinaryProbs.length; lS++) {
			for (short rS = 0; rS < newBinaryProbs.length; rS++) {
				if (newBinaryProbs[lS][rS] > 0) {
					double[][][] newScores = new double[1][1][1];
					newScores[0][0][0] = newBinaryProbs[lS][rS];
					BinaryRule newRule = new BinaryRule((short) 0, lS, rS,
							newScores);
					// newRule.setScores2(newScores);
					grammar.addBinary(newRule);
				}
			}
		}

		for (short cS = 0; cS < newUnaryProbs.length; cS++) {
			if (newUnaryProbs[cS] > 0) {
				double[][] newScores = new double[1][1];
				newScores[0][0] = newUnaryProbs[cS];
				UnaryRule newRule = new UnaryRule((short) 0, cS, newScores);
				// newRule.setScores2(newScores);
				grammar.addUnary(newRule);
			}
		}

		grammar.computePairsOfUnaries();
		grammar.makeCRArrays();
		grammar.isGrammarTag = this.isGrammarTag;
		// System.out.println(grammar.toString());
		return grammar;
	}

	public double[] computeConditionalProbabilities(int[][] fromMapping,
			int[][] toMapping) {
		double[][] transitionProbs = computeProductionProbabilities(fromMapping);
		// System.out.println(ArrayUtil.toString(transitionProbs));
		double[] expectedCounts = computeExpectedCounts(transitionProbs);
		// System.out.println(Arrays.toString(expectedCounts));
		/*
		 * for (int state=0; state<mapping.length-1; state++){ for (int
		 * substate=0; substate<mapping[state].length; substate++){
		 * System.out.println
		 * ((String)tagNumberer.object(state)+"_"+substate+" "+
		 * expectedCounts[mapping[state][substate]]); } }
		 */

		double[] condProbs = new double[expectedCounts.length];
		for (int projectedState = 0; projectedState < toMapping[toMapping.length - 1][0]; projectedState++) {
			double sum = 0;
			for (int state = 0; state < fromMapping.length - 1; state++) {
				for (int substate = 0; substate < fromMapping[state].length; substate++) {
					if (toMapping[state][substate] == projectedState)
						sum += expectedCounts[fromMapping[state][substate]];
				}
			}
			for (int state = 0; state < fromMapping.length - 1; state++) {
				for (int substate = 0; substate < fromMapping[state].length; substate++) {
					if (toMapping[state][substate] == projectedState)
						condProbs[fromMapping[state][substate]] = expectedCounts[fromMapping[state][substate]]
								/ sum;
				}
			}
		}
		return condProbs;
	}

	public int[][] computeToMapping(int level, int[][] toSubstateMapping) {
		if (level == -1)
			return computeMapping(-1);
		short[] numSubStates = this.numSubStates;
		int[][] mapping = new int[numSubStates.length + 1][];
		int k = 0;
		for (int state = 0; state < numSubStates.length; state++) {
			mapping[state] = new int[numSubStates[state]];
			int oldVal = -1;
			for (int substate = 0; substate < numSubStates[state]; substate++) {
				if (substate != 0
						&& oldVal != toSubstateMapping[state][substate + 1])
					k++;
				mapping[state][substate] = k;
				oldVal = toSubstateMapping[state][substate + 1];
			}
			k++;
		}
		mapping[numSubStates.length] = new int[1];
		mapping[numSubStates.length][0] = k;
		// System.out.println("The merged grammar will have "+k+" substates.");
		return mapping;
	}

	public int[][] computeMapping(int level) {
		// level -1 -> 0-bar states
		// level 0 -> x-bar states
		// level 1 -> each (state,substate) gets its own index
		short[] numSubStates = this.numSubStates;
		int[][] mapping = new int[numSubStates.length + 1][];
		int k = 0;
		for (int state = 0; state < numSubStates.length; state++) {
			mapping[state] = new int[numSubStates[state]];
			Arrays.fill(mapping[state], -1);
			// if (!grammar.isGrammarTag(state)) continue;
			for (int substate = 0; substate < numSubStates[state]; substate++) {
				if (level >= 1)
					mapping[state][substate] = k++;
				else if (level == -1) {
					if (this.isGrammarTag(state))
						mapping[state][substate] = 0;
					else
						mapping[state][substate] = state;
				} else
					/* level==0 */
					mapping[state][substate] = state;
			}
		}
		mapping[numSubStates.length] = new int[1];
		mapping[numSubStates.length][0] = (level < 1) ? numSubStates.length : k;
		// System.out.println("The grammar has "+mapping[numSubStates.length][0]+" substates.");
		return mapping;
	}

	public int[][] computeSubstateMapping(int level) {
		// level 0 -> merge all substates
		// level 1 -> merge upto depth 1 -> keep upto 2 substates
		// level 2 -> merge upto depth 2 -> keep upto 4 substates
		short[] numSubStates = this.numSubStates;
		// for (int i=0; i<numSubStates.length; i++)
		// System.out.println(i+" "+numSubStates[i]+" "+splitTrees[i].toString());
		int[][] mapping = new int[numSubStates.length][];
		for (int state = 0; state < numSubStates.length; state++) {
			mapping[state] = new int[numSubStates[state] + 1];
			int k = 0;
			if (level >= 0) {
				Arrays.fill(mapping[state], -1);
				Tree<Short> hierarchy = splitTrees[state];
				List<Tree<Short>> subTrees = hierarchy.getAtDepth(level);
				for (Tree<Short> subTree : subTrees) {
					List<Short> leaves = subTree.getYield();
					for (Short substate : leaves) {
						// System.out.println(substate+" "+numSubStates[state]+" "+state);
						if (substate == numSubStates[state])
							System.out.print("Will crash.");
						mapping[state][substate + 1] = k;
					}
					k++;
				}
			} else {
				k = 1;
			}
			mapping[state][0] = k;
		}
		return mapping;
	}

	public void computeReverseSubstateMapping(int level, int[][] lChildMap,
			int[][] rChildMap) {
		// level 1 -> how do the states from depth 1 expand to depth 2
		for (int state = 0; state < numSubStates.length; state++) {
			Tree<Short> hierarchy = splitTrees[state];
			List<Tree<Short>> subTrees = hierarchy.getAtDepth(level);
			lChildMap[state] = new int[subTrees.size()];
			rChildMap[state] = new int[subTrees.size()];
			for (Tree<Short> subTree : subTrees) {
				int substate = subTree.getLabel();
				if (subTree.isLeaf()) {
					lChildMap[state][substate] = substate;
					rChildMap[state][substate] = substate;
					continue;
				}
				boolean first = true;
				int nChildren = subTree.getChildren().size();
				for (Tree<Short> child : subTree.getChildren()) {
					if (first) {
						lChildMap[state][substate] = child.getLabel();
						first = false;
					} else
						rChildMap[state][substate] = child.getLabel();
					if (nChildren == 1)
						rChildMap[state][substate] = child.getLabel();
				}
			}
		}
	}

	private double[] computeExpectedCounts(double[][] transitionProbs) {
		// System.out.println(ArrayUtil.toString(transitionProbs));
		double[] expectedCounts = new double[transitionProbs.length];
		double[] tmpCounts = new double[transitionProbs.length];
		expectedCounts[0] = 1;
		tmpCounts[0] = 1;
		// System.out.print("Computing expected counts");
		int iter = 0;
		double diff = 1;
		double sum = 1; // 1 for the root
		while (diff > 1.0e-10 && iter < 50) {
			iter++;
			for (int state = 1; state < expectedCounts.length; state++) {
				for (int pState = 0; pState < expectedCounts.length; pState++) {
					tmpCounts[state] += expectedCounts[pState]
							* transitionProbs[pState][state];
				}

			}
			diff = 0;
			sum = 1;
			for (int state = 1; state < expectedCounts.length; state++) {
				// tmpCounts[state] /= sum;
				diff += (Math.abs(expectedCounts[state] - tmpCounts[state]));
				expectedCounts[state] = tmpCounts[state];
				sum += tmpCounts[state];
				tmpCounts[state] = 0;
			}
			expectedCounts[0] = 1;
			tmpCounts[0] = 1;
			// System.out.println(Arrays.toString(tmpCounts));
			// System.out.println(diff);
			// System.out.print(".");
			// System.out.print(diff);
		}
		// System.out.println("done.\nExpected total count: "+sum);
		// System.out.println(Arrays.toString(expectedCounts));
		return expectedCounts;
		// System.out.println(grammar.toString());
	}

	private double[][] computeProductionProbabilities(int[][] mapping) {
		short[] numSubStates = this.numSubStates;
		int totalStates = mapping[numSubStates.length][0];
		// W_ij is the probability of state i producing state j
		double[][] W = new double[totalStates][totalStates];

		for (int state = 0; state < numSubStates.length; state++) {
			// if (!grammar.isGrammarTag(state)) continue;
			BinaryRule[] parentRules = this.splitRulesWithP(state);
			for (int i = 0; i < parentRules.length; i++) {
				BinaryRule r = parentRules[i];
				int lState = r.leftChildState;
				int rState = r.rightChildState;
				/*
				 * if (lState==15||rState==15){ System.out.println("Found one");
				 * }
				 */
				double[][][] scores = r.getScores2();
				for (int lS = 0; lS < numSubStates[lState]; lS++) {
					for (int rS = 0; rS < numSubStates[rState]; rS++) {
						if (scores[lS][rS] == null)
							continue;
						for (int pS = 0; pS < numSubStates[state]; pS++) {
							W[mapping[state][pS]][mapping[lState][lS]] += scores[lS][rS][pS];
							W[mapping[state][pS]][mapping[rState][rS]] += scores[lS][rS][pS];
						}
					}
				}
			}
			List<UnaryRule> uRules = this.getUnaryRulesByParent(state);
			for (UnaryRule r : uRules) {
				int cState = r.childState;
				if (cState == state)
					continue;
				/*
				 * if (cState==15){ System.out.println("Found one"); }
				 */
				double[][] scores = r.getScores2();
				for (int cS = 0; cS < numSubStates[cState]; cS++) {
					if (scores[cS] == null)
						continue;
					for (int pS = 0; pS < numSubStates[state]; pS++) {
						W[mapping[state][pS]][mapping[cState][cS]] += scores[cS][pS];
					}
				}
			}
		}
		return W;
	}

	public void computeProperClosures() {
		int[][] map = new int[numStates][];
		int index = 0;
		for (int state = 0; state < numStates; state++) {
			map[state] = new int[numSubStates[state]];
			for (int substate = 0; substate < numSubStates[state]; substate++) {
				map[state][substate] = index++;
			}
		}

		double[][][] sumClosureMatrix = new double[10][index][index];
		// initialize
		for (int parentState = 0; parentState < numStates; parentState++) {
			for (int i = 0; i < unaryRulesWithParent[parentState].size(); i++) {
				UnaryRule rule = unaryRulesWithParent[parentState].get(i);
				short childState = rule.getChildState();
				double[][] scores = rule.getScores2();
				for (int childSubState = 0; childSubState < numSubStates[childState]; childSubState++) {
					if (scores[childSubState] == null)
						continue;
					for (int parentSubState = 0; parentSubState < numSubStates[parentState]; parentSubState++) {
						sumClosureMatrix[0][map[parentState][parentSubState]][map[childState][childSubState]] = scores[childSubState][parentSubState];
					}
				}
			}
		}
		// now loop until convergence = length 10 for now
		for (int length = 1; length < 10; length++) {
			for (short interState = 0; interState < numStates; interState++) {
				for (int i = 0; i < unaryRulesWithParent[interState].size(); i++) {
					UnaryRule rule = unaryRulesWithParent[interState].get(i);
					short endState = rule.getChildState();
					double[][] scores = rule.getScores2();

					// loop over substates
					for (int startState = 0; startState < numStates; startState++) {
						// we have a start and an end and need to loop over the
						// intermediate state,substates
						for (int startSubState = 0; startSubState < numSubStates[startState]; startSubState++) {
							for (int endSubState = 0; endSubState < numSubStates[endState]; endSubState++) {
								double ruleScore = 0;
								if (scores[endSubState] == null)
									continue;
								for (int interSubState = 0; interSubState < numSubStates[interState]; interSubState++) {
									ruleScore += sumClosureMatrix[length - 1][map[startState][startSubState]][map[interState][interSubState]]
											* scores[endSubState][interSubState];
								}
								sumClosureMatrix[length][map[startState][startSubState]][map[endState][endSubState]] += ruleScore;
							}
						}
					}
				}
			}
		}

		// now sum up the paths of different lengths
		double[][] sumClosureScores = new double[index][index];
		for (int length = 0; length < 10; length++) {
			for (int startState = 0; startState < index; startState++) {
				for (int endState = 0; endState < index; endState++) {
					sumClosureScores[startState][endState] += sumClosureMatrix[length][startState][endState];
				}
			}
		}

		// reset the lists of unaries
		closedSumRulesWithParent = new List[numStates];
		closedSumRulesWithChild = new List[numStates];
		for (short startState = 0; startState < numStates; startState++) {
			closedSumRulesWithParent[startState] = new ArrayList<UnaryRule>();
			closedSumRulesWithChild[startState] = new ArrayList<UnaryRule>();
		}

		// finally create rules and add them to the arrays
		for (short startState = 0; startState < numStates; startState++) {
			for (short endState = 0; endState < numStates; endState++) {
				if (startState == endState)
					continue;
				boolean atLeastOneNonZero = false;
				double[][] scores = new double[numSubStates[endState]][numSubStates[startState]];
				for (int startSubState = 0; startSubState < numSubStates[startState]; startSubState++) {
					for (int endSubState = 0; endSubState < numSubStates[endState]; endSubState++) {
						double score = sumClosureScores[map[startState][startSubState]][map[endState][endSubState]];
						if (score > 0) {
							scores[endSubState][startSubState] = score;
							atLeastOneNonZero = true;
						}
					}
				}
				if (atLeastOneNonZero) {
					UnaryRule newUnary = new UnaryRule(startState, endState,
							scores);
					addUnary(newUnary);
					closedSumRulesWithParent[startState].add(newUnary);
					closedSumRulesWithChild[endState].add(newUnary);
				}
			}
		}
		if (closedSumRulesWithP == null) {
			closedSumRulesWithP = new UnaryRule[numStates][];
			closedSumRulesWithC = new UnaryRule[numStates][];
		}
		for (int i = 0; i < numStates; i++) {
			closedSumRulesWithP[i] = (UnaryRule[]) closedSumRulesWithParent[i]
					.toArray(new UnaryRule[0]);
			closedSumRulesWithC[i] = (UnaryRule[]) closedSumRulesWithChild[i]
					.toArray(new UnaryRule[0]);
		}

	}

	/**
	 * @param output
	 */
	public void writeSplitTrees(Writer w) {
		PrintWriter out = new PrintWriter(w);
		for (int state = 1; state < numStates; state++) {
			String tag = (String) tagNumberer.object(state);
			if (isGrammarTag[state] && tag.endsWith("^g"))
				tag = tag.substring(0, tag.length() - 2);
			out.write(tag + "\t" + splitTrees[state].toString() + "\n");
		}
		out.flush();
		out.close();
	}

	public int[][] getClosedSumPaths() {
		return closedSumPaths;
	}
	
	public void overwriteWithMaxent() {
		//this is the map which maps each subrule to its location in binary rule
		HashMap<String, Pair<BinaryRule, Triple>> str2brule = new HashMap<String, Pair<BinaryRule,Triple>>();
		
		//create a eventstream
		BinaryRuleEventStream binaryRuleEventStream = new BinaryRuleEventStream();
		
		for (BinaryRule binaryRule : binaryRuleCounter.keySet()) {
			double[][][] score2 = binaryRuleCounter.getCount(binaryRule);
			for (Triple triple : binaryRule.getAllSubRules(score2)) {
				String bruleStr = binaryRule.getStrSubRule(triple);
				
				str2brule.put(bruleStr, new Pair<BinaryRule, Triple>(binaryRule, triple));
				
				for (int i=0; i<Math.round(1+binaryRule.getCountForSubRule(score2, triple)); i++) {
					binaryRuleEventStream.addEvent(bruleStr, BinaryRuleEventStream.getContext(bruleStr));
				}
				
				//reset this count to zero
				binaryRule.setProbForSubRule(score2, triple, 0.0);
			}
		}
		
		//now train the maxent model
		GISModel bruleMaxentModel = GIS.trainModel(binaryRuleEventStream, 10, 0);
		
		for (BinaryRule binaryRule : binaryRuleCounter.keySet()) {
			double[][][] score2 = binaryRuleCounter.getCount(binaryRule);
			for (Triple triple : binaryRule.getAllSubRules(score2)) {
				
				String bruleStr = binaryRule.getStrSubRule(triple);
				
				double[] outcomes = bruleMaxentModel.eval(BinaryRuleEventStream.getContext(bruleStr));
				
				for (int i = 0; i < outcomes.length; i++) {
					Pair<BinaryRule, Triple> pair = str2brule.get(bruleMaxentModel.getOutcome(i));
				
					//select the outcomes which match this binaryrule (ignoring all others)
					pair.getFirst().incProbForSubRule(score2, pair.getSecond(), outcomes[i]);
				}
			}
		}		
		
		//this is the map which maps each subrule to its location in binary rule
		HashMap<String, Pair<UnaryRule, Pair<Integer, Integer>>> str2urule = new HashMap<String, Pair<UnaryRule,Pair<Integer, Integer>>>();
		
		//create a eventstream
		UnaryRuleEventStream unaryRuleEventStream = new UnaryRuleEventStream();
		
		for (UnaryRule unaryRule : unaryRuleCounter.keySet()) {
			double[][] scores2 = unaryRuleCounter.getCount(unaryRule);
			for (Pair<Integer, Integer> pair : unaryRule.getAllSubRules(scores2)) {
				String uruleStr = unaryRule.getStrSubRule(pair);
				
				str2urule.put(uruleStr, new Pair<UnaryRule, Pair<Integer, Integer>>(unaryRule, pair));
				
				for (int i=0; i<Math.round(1+unaryRule.getCountForSubRule(scores2, pair)); i++) {
					unaryRuleEventStream.addEvent(uruleStr, UnaryRuleEventStream.getContext(uruleStr));
				}
				
				//reset this count to zero
				unaryRule.setProbForSubRule(scores2, pair, 0.0);
			}
		}
		
		//now train the maxent model
		GISModel uruleMaxentModel = GIS.trainModel(unaryRuleEventStream);
		
		for (UnaryRule unaryRule : unaryRuleCounter.keySet()) {
			double[][] scores2 = unaryRuleCounter.getCount(unaryRule);
			for (Pair<Integer, Integer> pair : unaryRule.getAllSubRules(scores2)) {
				
				String uruleStr = unaryRule.getStrSubRule(pair);
				
				double[] outcomes = uruleMaxentModel.eval(UnaryRuleEventStream.getContext(uruleStr));
				
				for (int i = 0; i < outcomes.length; i++) {
					Pair<UnaryRule, Pair<Integer, Integer>> pair2 = str2urule.get(uruleMaxentModel.getOutcome(i));
					
					//select the outcomes which match this binaryrule (ignoring all others)
					pair2.getFirst().incProbForSubRule(scores2, pair2.getSecond(), outcomes[i]);
				}
			}
		}		
	}
}