/// <summary>
/// Two argument constructor for TwoPassDataIndexer.
/// </summary>
/// <param name="eventReader">
/// An ITrainingEventReader which contains the a list of all the events
/// seen in the training data.
/// </param>
/// <param name="cutoff">
/// The minimum number of times a predicate must have been
/// observed in order to be included in the model.
/// </param>
public TwoPassDataIndexer(ITrainingEventReader eventReader, int cutoff)
{
List <ComparableEvent> eventsToCompare;
var predicateIndex = new Dictionary <string, int>();
//NotifyProgress("Indexing events using cutoff of " + cutoff + "\n");
//NotifyProgress("\tComputing event counts... ");
string tempFile = new FileInfo(Path.GetTempFileName()).FullName;
int eventCount = ComputeEventCounts(eventReader, tempFile, predicateIndex, cutoff);
//NotifyProgress("done. " + eventCount + " events");
//NotifyProgress("\tIndexing... ");
using (var fileEventReader = new FileEventReader(tempFile))
{
eventsToCompare = Index(eventCount, fileEventReader, predicateIndex);
}
if (File.Exists(tempFile))
{
File.Delete(tempFile);
}
//NotifyProgress("done.");
//NotifyProgress("Sorting and merging events... ");
SortAndMerge(eventsToCompare);
//NotifyProgress("Done indexing.");
}
/// <summary>
/// Reads events from <tt>eventStream</tt> into a dictionary. The
/// predicates associated with each event are counted and any which
/// occur at least <tt>cutoff</tt> times are added to the
/// <tt>predicatesInOut</tt> map along with a unique integer index.
/// </summary>
/// <param name="eventReader">
/// an <code>ITrainingEventReader</code> value
/// </param>
/// <param name="eventStoreFile">
/// a file name to which the events are written to for later processing.
/// </param>
/// <param name="predicatesInOut">
/// a <code>Dictionary</code> value
/// </param>
/// <param name="cutoff">
/// an <code>int</code> value
/// </param>
private int ComputeEventCounts(ITrainingEventReader eventReader, string eventStoreFile, Dictionary <string, int> predicatesInOut, int cutoff)
{
var counter = new Dictionary <string, int>();
int predicateIndex = 0;
int eventCount = 0;
using (var eventStoreWriter = new StreamWriter(eventStoreFile))
{
while (eventReader.HasNext())
{
TrainingEvent currentTrainingEvent = eventReader.ReadNextEvent();
eventCount++;
eventStoreWriter.Write(FileEventReader.ToLine(currentTrainingEvent));
string[] eventContext = currentTrainingEvent.Context;
for (int currentPredicate = 0; currentPredicate < eventContext.Length; currentPredicate++)
{
if (!predicatesInOut.ContainsKey(eventContext[currentPredicate]))
{
if (counter.ContainsKey(eventContext[currentPredicate]))
{
counter[eventContext[currentPredicate]]++;
}
else
{
counter.Add(eventContext[currentPredicate], 1);
}
if (counter[eventContext[currentPredicate]] >= cutoff)
{
predicatesInOut.Add(eventContext[currentPredicate], predicateIndex++);
counter.Remove(eventContext[currentPredicate]);
}
}
}
}
}
return(eventCount);
}