void mmEM::training(param myParam)
{
vector_2str wordX;
vector_2str wordY;
bool stillTrain = true;
// reading file //
readFileXY(myParam, myParam.inputFilename, &wordX, &wordY);
// initialization //
cout << "Initialization ... ";
initialization(myParam, wordX, wordY);
// maximization //
cout << "Maximization ... " << endl;
maximization(myParam);
cout << endl;
int atIter = 0;
// still training //
while (stillTrain)
{
atIter++;
cout << "Iteration " << atIter << endl;
// for each x and y pair //
cout << "Expectation ... " ;
for (int i = 0; i < wordX.size(); i++)
{
// expectation //
expectation(myParam, wordX[i], wordY[i]);
}
cout << "Maximization ... ";
long double totalChange = maximization(myParam);
cout << "Total probability change = " << totalChange << endl;
// stop by the probability change condition //
if ((totalChange <= myParam.cutOff) && (myParam.cutOff < 1))
{
stillTrain = false;
}
// stop by the number of iteration condition //
if ((myParam.cutOff >= 1) && (atIter >= myParam.cutOff))
{
stillTrain = false;
}
}
cout << endl;
}
void mmEM::computeDataProbs(param myParam)
{
vector_2str wordX;
vector_2str wordY;
vector_Double pair_probs;
bool stillTrain = true;
// reading file //
readFileXY(myParam, myParam.inputFilename, &wordX, &wordY);
vector<double> probs_vector(global_index), counts_vector(global_index), newprobs_vector(global_index),alpha_vector(global_index);
cout << "Computing pair probs" << endl;
// for each x and y pair //
cout << "Expectation ... " ;
cout << "There are " << wordX.size() << " pairs to be for which the probability is to be computed" << endl;
string output_probs_file = myParam.outputFilename + ".probs";
cout << "Write the output probs to "<<output_probs_file<<endl;
double alignCount = 0;
double noAlignCount = 0;
for (int i = 0; i < wordX.size(); i++)
{
// computing the p(y|x) for the particular xy pair//
pair_probs.push_back(computePairProb(myParam, wordX[i], wordY[i]));
}
cout << "Write aligned data to : " << myParam.outputFilename << endl;
cout << "Write un-aligned data to : " << myParam.outputFilename << ".err" << endl;
ofstream PROBS;
PROBS.open(output_probs_file.c_str());
for (int i = 0; i < wordX.size(); i++)
{
PROBS << join(wordX[i],0,wordX[i].size(),"") <<"\t"<< join(wordY[i],0,wordY[i].size(),"") <<"\t"<< pair_probs[i]<<endl;
}
PROBS.close();
}
void mmEM::createAlignments(param myParam)
{
vector_2str wordX;
vector_2str wordY;
double lessNbest = 0;
//reading input file//
readFileXY(myParam, myParam.inputFilename, &wordX, &wordY);
cout << "There are " << wordX.size() << " pairs to be aligned" << endl;
cout << "Write aligned data to : " << myParam.outputFilename << endl;
cout << "Write un-aligned data to : " << myParam.outputFilename << ".err" << endl;
ofstream ALIGNED, NOALIGNED;
ALIGNED.open(myParam.outputFilename.c_str());
NOALIGNED.open(string(myParam.outputFilename + ".err").c_str());
double alignCount = 0;
double noAlignCount = 0;
for (unsigned int i = 0; i < wordX.size(); i++)
{
//vector<string> alignX, alignY;
vector_2str nAlignX, nAlignY;
vector<long double> nScore;
nScore = nViterbi_align(myParam, wordX[i], wordY[i], nAlignX, nAlignY);
//score = viterbi_align(myParam, wordX[i], wordY[i], &alignX, &alignY);
if (nScore.size() > 0)
{
// found n-best alignments
alignCount++;
// count number of examples that have less than n alignment candidates //
if (nScore.size() < myParam.nBest)
{
lessNbest++;
cout << join(wordX[i],0, wordX[i].size(), "") << "\t" << join(wordY[i],0,wordY[i].size(), "") << "\t has " << nScore.size() << " alignments" << endl;
}
for (int nbest = 0; nbest < nScore.size(); nbest++)
{
for (unsigned int k = 0; k < nAlignX[nbest].size(); k++)
{
ALIGNED << nAlignX[nbest][k] << myParam.sepChar;
}
ALIGNED << "\t";
for (unsigned int k=0; k < nAlignY[nbest].size(); k++)
{
ALIGNED << nAlignY[nbest][k] << myParam.sepChar;
}
if (myParam.printScore)
{
ALIGNED << "\t" << nbest+1;
ALIGNED << "\t" << nScore[nbest];
}
ALIGNED << endl;
}
}
else
{
// can't be aligned
noAlignCount++;
if (myParam.errorInFile)
{
ALIGNED << "NO ALIGNMENT \t" << endl;
}
NOALIGNED << join(wordX[i],0,wordX[i].size()," ") << "\t" << join(wordY[i], 0, wordY[i].size(), " ") << endl;
}
}
ALIGNED.close();
NOALIGNED.close();
cout << "Aligned " << alignCount << " pairs" << endl;
if (noAlignCount > 0)
{
cout << "No aligned " << noAlignCount << " pairs" << endl;
}
cout << "There are " << lessNbest << " example pairs having less than " << myParam.nBest << " alignment candidates" << endl;
}
void mmEM::training(param myParam)
{
vector_2str wordX;
vector_2str wordY;
bool stillTrain = true;
// reading file //
readFileXY(myParam, myParam.inputFilename, &wordX, &wordY);
// initialization //
cout << "Initialization ... ";
initialization(myParam, wordX, wordY);
// maximization //
cout << "Maximization ... " << endl;
maximization(myParam); // lhuang: M-step after initialize (count = 1)
cout << endl;
int atIter = 0;
// still training //
global_index--;
vector<double> probs_vector(global_index), counts_vector(global_index), newprobs_vector(global_index),alpha_vector(global_index);
while (stillTrain)
{
atIter++;
cout << "Iteration " << atIter << endl;
// for each x and y pair //
cout << "Expectation ... " ;
for (int i = 0; i < wordX.size(); i++)
{
// expectation //
expectation(myParam, wordX[i], wordY[i]);
}
if (!myParam.ashish)
{
cout << "Maximization ... ";
long double totalChange = maximization(myParam); // lhuang: real M-step
cout << "Total probability change = " << totalChange << endl;
// stop by the probability change condition //
if ((totalChange <= myParam.cutOff) && (myParam.cutOff < 1))
{
stillTrain = false;
}
}
else
{
cout << "Ashish" ;
double sum_probs = 0;
for(hash_2StrDouble::iterator pos = index.begin(); pos != index.end(); pos++)
{
for (hash_StrDouble::iterator pos2 = (pos->second).begin(); pos2 != (pos->second).end(); pos2++)
{
counts_vector[index[pos->first][pos2->first]-1] = counts[pos->first][pos2->first];
// cout << pos->first << " "<< pos2->first << " "
// << counts_vector[index[pos->first][pos2->first]-1] << endl;
probs_vector[index[pos->first][pos2->first]-1] = probs[pos->first][pos2->first];
alpha_vector[index[pos->first][pos2->first]-1] = alphas[pos->first][pos2->first];
sum_probs += probs_vector[index[pos->first][pos2->first]-1];
}
}
cout << "sum " << sum_probs <<endl;
// getchar();
sum_probs = 0;
for (int i=0; i<newprobs_vector.size(); i++)
{
// cout << counts_vector[i] << " " << probs_vector[i] << " ";
sum_probs += probs_vector[i];
}
cout << "sum vector " << sum_probs <<endl;
// getchar();
projectedGradientDescentWithArmijoRule(counts_vector, probs_vector, newprobs_vector,alpha_vector,myParam.beta,myParam.pgdIter);
for(hash_2StrDouble::iterator pos = index.begin(); pos != index.end(); pos++)
{
for (hash_StrDouble::iterator pos2 = (pos->second).begin(); pos2 != (pos->second).end(); pos2++)
{
// cout << pos->first << " "<< pos2->first << " "
// << probs[pos->first][pos2->first] << " "
// << counts[pos->first][pos2->first] << " "
// << newprobs_vector[index[pos->first][pos2->first]-1] << endl;
probs[pos->first][pos2->first] = newprobs_vector[index[pos->first][pos2->first]-1];
}
}
// for (int i=0; i<newprobs_vector.size(); i++)
// {
// cout << newprobs_vector[i] << " " ;
// }
cout<< endl;
// getchar();
// clean counts //
counts.clear();
}
// stop by the number of iteration condition //
if ((myParam.cutOff >= 1) && (atIter >= myParam.cutOff))
{
stillTrain = false;
}
}
cout << endl;
}
