/// <summary>
/// Repositions the stream at the beginning and the previously seen object
/// sequence will be repeated exactly. This method can be used to re-read the
/// stream if multiple passes over the objects are required.
/// </summary>
public void Reset()
{
sequenceStream.Reset();
}
public AbstractModel TrainModel(int trainIterations, ISequenceStream trainStream, int cutoff,
bool trainUseAverage) {
iterations = trainIterations;
useAverage = trainUseAverage;
sequenceStream = trainStream;
info.Append("Trained using Perceptron Sequence algorithm.\n\n");
var di = new OnePassDataIndexer(new SequenceStreamEventStream(trainStream), cutoff, false);
trainStream.Reset();
numSequences = 0;
while (trainStream.Read() != null) {
numSequences++;
}
outcomeList = di.GetOutcomeList();
predLabels = di.GetPredLabels();
pMap = new IndexHashTable<string>(predLabels, 0.7d);
// Incorporation indexed data for training...
numEvents = di.GetNumEvents();
outcomeLabels = di.GetOutcomeLabels();
oMap = new Dictionary<string, int>();
for (var i = 0; i < outcomeLabels.Length; i++) {
oMap.Add(outcomeLabels[i], i);
}
outcomeList = di.GetOutcomeList();
numPreds = predLabels.Length;
numOutcomes = outcomeLabels.Length;
if (trainUseAverage) {
updates = new int[numPreds][][];
for (var i = 0; i < numPreds; i++) {
updates[i] = new int[numOutcomes][];
for (var j = 0; j < numOutcomes; j++) {
updates[i][j] = new int[3];
}
}
}
// done.
Display("done.\n");
info.Append("Number of Event Tokens: {0}\n" +
" Number of Outcomes: {1}\n" +
" Number of Predicates: {2}\n", numEvents, numOutcomes, numPreds);
Display("\tNumber of Event Tokens: " + numEvents);
Display("\t Number of Outcomes: " + numOutcomes);
Display("\t Number of Predicates: " + numPreds);
param = new MutableContext[numPreds];
if (trainUseAverage)
averageParams = new MutableContext[numPreds];
allOutcomesPattern = new int[numOutcomes];
for (var i = 0; i < numOutcomes; i++) {
allOutcomesPattern[i] = i;
}
for (var pi = 0; pi < numPreds; pi++) {
param[pi] = new MutableContext(allOutcomesPattern, new double[numOutcomes]);
if (trainUseAverage)
averageParams[pi] = new MutableContext(allOutcomesPattern, new double[numOutcomes]);
for (var aoi = 0; aoi < numOutcomes; aoi++) {
param[pi].SetParameter(aoi, 0.0d);
if (trainUseAverage)
averageParams[pi].SetParameter(aoi, 0.0d);
}
}
Display("Computing model parameters...");
FindParameters();
Display("...done.");
/*************** Create and return the model ******************/
// ReSharper disable CoVariantArrayConversion
if (trainUseAverage)
return new PerceptronModel(averageParams, predLabels, outcomeLabels) {
info = info
};
return new PerceptronModel(param, predLabels, outcomeLabels) {
info = info
};
// ReSharper restore CoVariantArrayConversion
}
private void NextIteration(int iteration)
{
iteration--; //move to 0-based index
var numCorrect = 0;
var oei = 0;
var si = 0;
var featureCounts = new Dictionary <int, Dictionary <string, float> >();
for (var oi = 0; oi < numOutcomes; oi++)
{
featureCounts[oi] = new Dictionary <string, float>();
}
// ReSharper disable once CoVariantArrayConversion
var model = new PerceptronModel(param, pMap, outcomeLabels);
sequenceStream.Reset();
Sequence sequence;
while ((sequence = sequenceStream.Read()) != null)
{
var taggerEvents = sequenceStream.UpdateContext(sequence, model);
var update = false;
for (var ei = 0; ei < sequence.Events.Length; ei++, oei++)
{
if (!taggerEvents[ei].Outcome.Equals(sequence.Events[ei].Outcome))
{
update = true;
//break;
}
else
{
numCorrect++;
}
}
if (update)
{
for (var oi = 0; oi < numOutcomes; oi++)
{
featureCounts[oi].Clear();
}
//training feature count computation
for (var ei = 0; ei < sequence.Events.Length; ei++, oei++)
{
var contextStrings = sequence.Events[ei].Context;
var values = sequence.Events[ei].Values;
var oi = oMap[sequence.Events[ei].Outcome];
for (var ci = 0; ci < contextStrings.Length; ci++)
{
float value = 1;
if (values != null)
{
value = values[ci];
}
if (featureCounts[oi].ContainsKey(contextStrings[ci]))
{
featureCounts[oi][contextStrings[ci]] += value;
}
else
{
featureCounts[oi][contextStrings[ci]] = value;
}
}
}
//evaluation feature count computation
foreach (var taggerEvent in taggerEvents)
{
var contextStrings = taggerEvent.Context;
var values = taggerEvent.Values;
var oi = oMap[taggerEvent.Outcome];
for (var ci = 0; ci < contextStrings.Length; ci++)
{
float value = 1;
if (values != null)
{
value = values[ci];
}
float c;
if (featureCounts[oi].ContainsKey(contextStrings[ci]))
{
c = featureCounts[oi][contextStrings[ci]] - value;
}
else
{
c = -1 * value;
}
if (c.Equals(0f))
{
featureCounts[oi].Remove(contextStrings[ci]);
}
}
}
for (var oi = 0; oi < numOutcomes; oi++)
{
foreach (var feature in featureCounts[oi].Keys)
{
var pi = pMap[feature];
if (pi != -1)
{
param[pi].UpdateParameter(oi, featureCounts[oi][feature]);
if (useAverage)
{
if (updates[pi][oi][VALUE] != 0)
{
averageParams[pi].UpdateParameter(oi,
updates[pi][oi][VALUE] *
(numSequences * (iteration - updates[pi][oi][ITER]) +
(si - updates[pi][oi][EVENT])));
}
updates[pi][oi][VALUE] = (int)param[pi].Parameters[oi];
updates[pi][oi][ITER] = iteration;
updates[pi][oi][EVENT] = si;
}
}
}
}
// ReSharper disable once CoVariantArrayConversion
model = new PerceptronModel(param, pMap, outcomeLabels);
}
si++;
}
//finish average computation
var totIterations = (double)iterations * si;
if (useAverage && iteration == iterations - 1)
{
for (var pi = 0; pi < numPreds; pi++)
{
var predParams = averageParams[pi].Parameters;
for (var oi = 0; oi < numOutcomes; oi++)
{
if (updates[pi][oi][VALUE] != 0)
{
predParams[oi] += updates[pi][oi][VALUE] *
(numSequences * (iterations - updates[pi][oi][ITER]) -
updates[pi][oi][EVENT]);
}
if (predParams[oi].Equals(0d))
{
continue;
}
predParams[oi] /= totIterations;
averageParams[pi].SetParameter(oi, predParams[oi]);
}
}
}
trainingAccuracy = (double)numCorrect / numEvents;
Display(string.Format("{0,-4} {1} of {2} - {3}",
iteration,
numCorrect,
numEvents,
trainingAccuracy));
}
public AbstractModel TrainModel(int trainIterations, ISequenceStream trainStream, int cutoff,
bool trainUseAverage)
{
iterations = trainIterations;
useAverage = trainUseAverage;
sequenceStream = trainStream;
info.Append("Trained using Perceptron Sequence algorithm.\n\n");
var di = new OnePassDataIndexer(new SequenceStreamEventStream(trainStream), cutoff, false);
trainStream.Reset();
numSequences = 0;
while (trainStream.Read() != null)
{
numSequences++;
}
outcomeList = di.GetOutcomeList();
predLabels = di.GetPredLabels();
pMap = new IndexHashTable <string>(predLabels, 0.7d);
// Incorporation indexed data for training...
numEvents = di.GetNumEvents();
outcomeLabels = di.GetOutcomeLabels();
oMap = new Dictionary <string, int>();
for (var i = 0; i < outcomeLabels.Length; i++)
{
oMap.Add(outcomeLabels[i], i);
}
outcomeList = di.GetOutcomeList();
numPreds = predLabels.Length;
numOutcomes = outcomeLabels.Length;
if (trainUseAverage)
{
updates = new int[numPreds][][];
for (var i = 0; i < numPreds; i++)
{
updates[i] = new int[numOutcomes][];
for (var j = 0; j < numOutcomes; j++)
{
updates[i][j] = new int[3];
}
}
}
// done.
Display("done.\n");
info.Append("Number of Event Tokens: {0}\n" +
" Number of Outcomes: {1}\n" +
" Number of Predicates: {2}\n", numEvents, numOutcomes, numPreds);
Display("\tNumber of Event Tokens: " + numEvents);
Display("\t Number of Outcomes: " + numOutcomes);
Display("\t Number of Predicates: " + numPreds);
param = new MutableContext[numPreds];
if (trainUseAverage)
{
averageParams = new MutableContext[numPreds];
}
allOutcomesPattern = new int[numOutcomes];
for (var i = 0; i < numOutcomes; i++)
{
allOutcomesPattern[i] = i;
}
for (var pi = 0; pi < numPreds; pi++)
{
param[pi] = new MutableContext(allOutcomesPattern, new double[numOutcomes]);
if (trainUseAverage)
{
averageParams[pi] = new MutableContext(allOutcomesPattern, new double[numOutcomes]);
}
for (var aoi = 0; aoi < numOutcomes; aoi++)
{
param[pi].SetParameter(aoi, 0.0d);
if (trainUseAverage)
{
averageParams[pi].SetParameter(aoi, 0.0d);
}
}
}
Display("Computing model parameters...");
FindParameters();
Display("...done.");
/*************** Create and return the model ******************/
// ReSharper disable CoVariantArrayConversion
if (trainUseAverage)
{
return new PerceptronModel(averageParams, predLabels, outcomeLabels)
{
info = info
}
}
;
return(new PerceptronModel(param, predLabels, outcomeLabels)
{
info = info
});
// ReSharper restore CoVariantArrayConversion
}
/// <summary>
/// Repositions the stream at the beginning and the previously seen object
/// sequence will be repeated exactly. This method can be used to re-read the
/// stream if multiple passes over the objects are required.
/// </summary>
public void Reset()
{
enumerator = null;
sequenceStream.Reset();
}