36 #ifndef RERANKER_FEATURE_EXTRACTOR_H_ 
   37 #define RERANKER_FEATURE_EXTRACTOR_H_ 
   41 #include <unordered_map> 
   51 using std::shared_ptr;
 
   52 using std::unordered_map;
 
  141          it != candidate_set.
end();
 
  144       Extract(candidate, candidate.mutable_features());
 
  154 #define REGISTER_NAMED_FEATURE_EXTRACTOR(TYPE,NAME) \ 
  155   REGISTER_NAMED(TYPE,NAME,FeatureExtractor) 
  161 #define REGISTER_FEATURE_EXTRACTOR(TYPE) \ 
  162   REGISTER_NAMED_FEATURE_EXTRACTOR(TYPE,TYPE) 
Provides the reranker::Candidate class for representing a candidate hypothesis from an initial model...
 
vector< shared_ptr< Candidate > >::iterator iterator
 
void TearDown()
A free-floating function (within the reranker namespace) that frees statically allocated objects...
 
A class to hold a set of candidates, either for training or test. 
 
A class to represent a candidate in a set of candidates that constitutes a training instance for a re...
 
const_iterator begin() const 
 
Defines the reranker::FeatureVector class, which, as it happens, is used to store feature vectors...
 
An interface to make it easier to implement Factory-constructible types by implementing both required...
 
const_iterator end() const 
 
Provides a generic dynamic object factory. 
 
Class to hold a single training instance for a reranker, which is a set of examples, typically the n-best output of some input process, posibly including a gold-standard feature vector.