private List <ComparableEvent> Index(List <TrainingEvent> events, Dictionary <string, int> predicateIndex)
{
var map = new Dictionary <string, int>();
int eventCount = events.Count;
int outcomeCount = 0;
var eventsToCompare = new List <ComparableEvent>(eventCount);
var indexedContext = new List <int>();
for (int eventIndex = 0; eventIndex < eventCount; eventIndex++)
{
TrainingEvent currentTrainingEvent = events[eventIndex];
string[] eventContext = currentTrainingEvent.Context;
ComparableEvent comparableEvent;
int outcomeIndex;
string outcome = currentTrainingEvent.Outcome;
if (map.ContainsKey(outcome))
{
outcomeIndex = map[outcome];
}
else
{
outcomeIndex = outcomeCount++;
map.Add(outcome, outcomeIndex);
}
for (int currentEventContext = 0; currentEventContext < eventContext.Length; currentEventContext++)
{
string predicate = eventContext[currentEventContext];
if (predicateIndex.ContainsKey(predicate))
{
indexedContext.Add(predicateIndex[predicate]);
}
}
// drop events with no active features
if (indexedContext.Count > 0)
{
comparableEvent = new ComparableEvent(outcomeIndex, indexedContext.ToArray());
eventsToCompare.Add(comparableEvent);
}
else
{
//"Dropped event " + oEvent.Outcome + ":" + oEvent.Context);
}
// recycle the list
indexedContext.Clear();
}
SetOutcomeLabels(ToIndexedStringArray(map));
SetPredicateLabels(ToIndexedStringArray(predicateIndex));
return(eventsToCompare);
}
private List <ComparableEvent> Index(int eventCount, ITrainingEventReader eventReader, Dictionary <string, int> predicateIndex)
{
var outcomeMap = new Dictionary <string, int>();
int outcomeCount = 0;
var eventsToCompare = new List <ComparableEvent>(eventCount);
var indexedContext = new List <int>();
while (eventReader.HasNext())
{
TrainingEvent currentTrainingEvent = eventReader.ReadNextEvent();
string[] eventContext = currentTrainingEvent.Context;
ComparableEvent comparableEvent;
int outcomeId;
string outcome = currentTrainingEvent.Outcome;
if (outcomeMap.ContainsKey(outcome))
{
outcomeId = outcomeMap[outcome];
}
else
{
outcomeId = outcomeCount++;
outcomeMap.Add(outcome, outcomeId);
}
for (int currentPredicate = 0; currentPredicate < eventContext.Length; currentPredicate++)
{
string predicate = eventContext[currentPredicate];
if (predicateIndex.ContainsKey(predicate))
{
indexedContext.Add(predicateIndex[predicate]);
}
}
// drop events with no active features
if (indexedContext.Count > 0)
{
comparableEvent = new ComparableEvent(outcomeId, indexedContext.ToArray());
eventsToCompare.Add(comparableEvent);
}
else
{
//"Dropped event " + currentTrainingEvent.Outcome + ":" + currentTrainingEvent.Context);
}
// recycle the list
indexedContext.Clear();
}
SetOutcomeLabels(ToIndexedStringArray(outcomeMap));
SetPredicateLabels(ToIndexedStringArray(predicateIndex));
return(eventsToCompare);
}
/// <summary>
/// Sorts and uniques the array of comparable events. This method
/// will alter the eventsToCompare array -- it does an in place
/// sort, followed by an in place edit to remove duplicates.
/// </summary>
/// <param name="eventsToCompare">
/// a List of <code>ComparableEvent</code> values
/// </param>
protected internal virtual void SortAndMerge(List <ComparableEvent> eventsToCompare)
{
eventsToCompare.Sort();
int eventCount = eventsToCompare.Count;
int uniqueEventCount = 1; // assertion: eventsToCompare.length >= 1
if (eventCount <= 1)
{
return; // nothing to do; edge case (see assertion)
}
ComparableEvent comparableEvent = eventsToCompare[0];
for (int currentEvent = 1; currentEvent < eventCount; currentEvent++)
{
ComparableEvent eventToCompare = eventsToCompare[currentEvent];
if (comparableEvent.Equals(eventToCompare))
{
comparableEvent.SeenCount++; // increment the seen count
eventsToCompare[currentEvent] = null; // kill the duplicate
}
else
{
comparableEvent = eventToCompare; // a new champion emerges...
uniqueEventCount++; // increment the # of unique events
}
}
//NotifyProgress("done. Reduced " + eventCount + " events to " + uniqueEventCount + ".");
mContexts = new int[uniqueEventCount][];
mOutcomeList = new int[uniqueEventCount];
mNumTimesEventsSeen = new int[uniqueEventCount];
for (int currentEvent = 0, currentStoredEvent = 0; currentEvent < eventCount; currentEvent++)
{
ComparableEvent eventToStore = eventsToCompare[currentEvent];
if (null == eventToStore)
{
continue; // this was a dupe, skip over it.
}
mNumTimesEventsSeen[currentStoredEvent] = eventToStore.SeenCount;
mOutcomeList[currentStoredEvent] = eventToStore.Outcome;
mContexts[currentStoredEvent] = eventToStore.GetPredicateIndexes();
++currentStoredEvent;
}
}