int svc (void)
{
ACE_TRACE ("svc");
ACE_DEBUG ((LM_INFO, ACE_TEXT ("(%t) Manager started\n")));
// Create pool when you get in the first time.
create_worker_pool ();
while (!done ())
{
ACE_Time_Value tv ((long)MAX_TIMEOUT);
tv += ACE_OS::time (0);
// Get the next message
ACE_Method_Request *request = this->queue_.dequeue (&tv);
if (request == 0)
{
shut_down ();
break;
}
// Choose a worker.
Worker *worker = choose_worker ();
// Ask the worker to do the job.
worker->perform (request);
}
return 0;
}
void LDA<Scalar>::partial_fit(std::shared_ptr<corpus::Corpus> corpus) {
// Shuffle the documents for a randomized pass through
corpus->shuffle();
// Queue all the documents
for (size_t i=0; i<corpus->size(); i++) {
queue_in_.emplace_back(corpus, i);
}
// create the thread pool
create_worker_pool();
// Extract variational parameters and calculate the doc_m_step
for (size_t i=0; i<corpus->size(); i++) {
std::shared_ptr<parameters::Parameters> variational_parameters;
size_t index;
std::tie(variational_parameters, index) = extract_vp_from_queue();
// tell the thread safe event dispatcher to process the events from the
// workers
process_worker_events();
// perform the online part of m step
m_step_->doc_m_step(
corpus->at(index),
variational_parameters,
model_parameters_ // output
);
}
// destroy the thread pool
destroy_worker_pool();
// Perform any corpuswise action related to e step
e_step_->e_step();
// perform the batch part of m step
m_step_->m_step(
model_parameters_ // output
);
// inform the world that the epoch is over
get_event_dispatcher()->template dispatch<events::EpochProgressEvent<Scalar> >(model_parameters_);
}
int svc (void)
{
ACE_TRACE ("Manager::svc");
ACE_DEBUG ((LM_INFO, ACE_TEXT ("(%t) Manager started\n")));
// Create pool.
create_worker_pool ();
while (!done ())
{
ACE_Message_Block *mb = 0;
ACE_Time_Value tv ((long)MAX_TIMEOUT);
tv += ACE_OS::time (0);
// Get a message request.
if (this->getq (mb, &tv) < 0)
{
shut_down ();
break;
}
// Choose a worker.
Worker *worker = 0;
{
ACE_GUARD_RETURN (ACE_Thread_Mutex,
worker_mon, this->workers_lock_, -1);
while (this->workers_.is_empty ())
workers_cond_.wait ();
this->workers_.dequeue_head (worker);
}
// Ask the worker to do the job.
worker->putq (mb);
}
return 0;
}
typename LDA<Scalar>::MatrixX LDA<Scalar>::transform(const Eigen::MatrixXi& X) {
// cast the parameters to what is needed
auto model = std::static_pointer_cast<parameters::ModelParameters<Scalar> >(
model_parameters_
);
// make some room for the transformed data
MatrixX gammas(model->beta.rows(), X.cols());
// make a corpus to use
auto corpus = get_corpus(X);
// Queue all the documents
for (size_t i=0; i<corpus->size(); i++) {
queue_in_.emplace_back(corpus, i);
}
// create the thread pool
create_worker_pool();
// Extract variational parameters and calculate the doc_e_step
for (size_t i=0; i<corpus->size(); i++) {
std::shared_ptr<parameters::Parameters> vp;
size_t index;
std::tie(vp, index) = extract_vp_from_queue();
gammas.col(index) = std::static_pointer_cast<parameters::VariationalParameters<Scalar> >(vp)->gamma;
// tell the thread safe event dispatcher to process the events from the
// workers
process_worker_events();
}
// destroy the thread pool
destroy_worker_pool();
return gammas;
}
