package phrase; import gnu.trove.TIntArrayList; import io.FileUtil; import java.io.File; import java.io.IOException; import java.io.PrintStream; import java.util.List; import phrase.Corpus.Edge; public class Agree { PhraseCluster model1; C2F model2; Corpus c; private int K,n_phrases, n_words, n_contexts, n_positions1,n_positions2; /**@brief sum of loglikelihood of two * individual models */ public double llh; /**@brief Bhattacharyya distance * */ public double bdist; /** * * @param numCluster * @param corpus */ public Agree(int numCluster, Corpus corpus){ model1=new PhraseCluster(numCluster, corpus); model2=new C2F(numCluster,corpus); c=corpus; n_words=c.getNumWords(); n_phrases=c.getNumPhrases(); n_contexts=c.getNumContexts(); n_positions1=c.getNumContextPositions(); n_positions2=2; K=numCluster; } /**@brief test * */ public static void main(String args[]){ //String in="../pdata/canned.con"; String in="../pdata/btec.con"; String out="../pdata/posterior.out"; int numCluster=25; Corpus corpus = null; File infile = new File(in); try { System.out.println("Reading concordance from " + infile); corpus = Corpus.readFromFile(FileUtil.reader(infile)); corpus.printStats(System.out); } catch (IOException e) { System.err.println("Failed to open input file: " + infile); e.printStackTrace(); System.exit(1); } Agree agree=new Agree(numCluster, corpus); int iter=20; for(int i=0;i contexts = c.getEdgesForContext(context); for (int ctx=0; ctx 0; bdist += edge.getCount() * Math.log(z); arr.F.l1normalize(p); int count = edge.getCount(); //increment expected count TIntArrayList phraseToks = edge.getPhrase(); TIntArrayList contextToks = edge.getContext(); for(int tag=0;tag