std::string Manager::OutputNBest()
{
arcLists.Sort();
boost::unordered_set<size_t> distinctHypos;
TrellisPaths<TrellisPath> contenders;
m_search->AddInitialTrellisPaths(contenders);
long transId = GetTranslationId();
// MAIN LOOP
stringstream out;
//Moses2::FixPrecision(out);
size_t maxIter = system.options.nbest.nbest_size * system.options.nbest.factor;
size_t bestInd = 0;
for (size_t i = 0; i < maxIter; ++i) {
if (bestInd > system.options.nbest.nbest_size || contenders.empty()) {
break;
}
//cerr << "bestInd=" << bestInd << endl;
TrellisPath *path = contenders.Get();
bool ok = false;
if (system.options.nbest.only_distinct) {
string tgtPhrase = path->OutputTargetPhrase(system);
//cerr << "tgtPhrase=" << tgtPhrase << endl;
boost::hash<std::string> string_hash;
size_t hash = string_hash(tgtPhrase);
if (distinctHypos.insert(hash).second) {
ok = true;
}
}
else {
ok = true;
}
if (ok) {
++bestInd;
out << transId << " ||| ";
path->OutputToStream(out, system);
out << "\n";
}
// create next paths
path->CreateDeviantPaths(contenders, arcLists, GetPool(), system);
delete path;
}
return out.str();
}
void OutputFeatureScores( std::ostream& out, const TrellisPath &path, const FeatureFunction *ff, std::string &lastName )
{
const StaticData &staticData = StaticData::Instance();
bool labeledOutput = staticData.IsLabeledNBestList();
// regular features (not sparse)
if (ff->GetNumScoreComponents() != ScoreProducer::unlimited) {
if( labeledOutput && lastName != ff->GetScoreProducerWeightShortName() ) {
lastName = ff->GetScoreProducerWeightShortName();
out << " " << lastName << ":";
}
vector<float> scores = path.GetScoreBreakdown().GetScoresForProducer( ff );
for (size_t j = 0; j<scores.size(); ++j) {
out << " " << scores[j];
}
}
// sparse features
else {
const FVector scores = path.GetScoreBreakdown().GetVectorForProducer( ff );
// report weighted aggregate
if (! ff->GetSparseFeatureReporting()) {
const FVector &weights = staticData.GetAllWeights().GetScoresVector();
if (labeledOutput && !boost::contains(ff->GetScoreProducerDescription(), ":"))
out << " " << ff->GetScoreProducerWeightShortName() << ":";
out << " " << scores.inner_product(weights);
}
// report each feature
else {
for(FVector::FNVmap::const_iterator i = scores.cbegin(); i != scores.cend(); i++)
out << " " << i->first << ": " << i->second;
/* if (i->second != 0) { // do not report zero-valued features
float weight = staticData.GetSparseWeight(i->first);
if (weight != 0)
out << " " << i->first << "=" << weight;
}*/
}
}
}
void GetOutputFactors(const TrellisPath &path, vector <const Factor*> &translation){
const std::vector<const Hypothesis *> &edges = path.GetEdges();
const std::vector<FactorType>& outputFactorOrder = StaticData::Instance().GetOutputFactorOrder();
assert (outputFactorOrder.size() == 1);
// print the surface factor of the translation
for (int currEdge = (int)edges.size() - 1 ; currEdge >= 0 ; currEdge--)
{
const Hypothesis &edge = *edges[currEdge];
const Phrase &phrase = edge.GetCurrTargetPhrase();
size_t size = phrase.GetSize();
for (size_t pos = 0 ; pos < size ; pos++)
{
const Factor *factor = phrase.GetFactor(pos, outputFactorOrder[0]);
translation.push_back(factor);
}
}
}
void OutputAlignment(OutputCollector* collector, size_t lineNo , const TrellisPath &path)
{
if (collector) {
OutputAlignment(collector,lineNo, path.GetEdges());
}
}
