public virtual void TrainModel()
{
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(mEvents), 100, 10);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
}
public virtual void Train()
{
if (ResolverMode.Train == mResolverMode)
{
Console.Error.WriteLine(this + " referential");
if (mDebugOn)
{
#if DNF
var writer = new System.IO.StreamWriter(mModelName + ".events", false, System.Text.Encoding.Default);
#else
var writer = new StreamWriter(new FileStream(mModelName + ".events", FileMode.Open));
#endif
foreach (SharpEntropy.TrainingEvent trainingEvent in mEvents)
{
writer.Write(trainingEvent.ToString() + "\n");
}
#if DNF
writer.Close();
#else
writer.Dispose();
#endif
}
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(mEvents), 100, 10);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
}
}
// Utilities ---------------------------------
/// <summary>
/// Trains a POS tag maximum entropy model.
/// </summary>
/// <param name="eventStream">Stream of training events</param>
/// <param name="iterations">number of training iterations to perform</param>
/// <param name="cut">cutoff value to use for the data indexer</param>
/// <returns>Trained GIS model</returns>
public static SharpEntropy.GisModel Train(SharpEntropy.ITrainingEventReader eventStream, int iterations, int cut)
{
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(iterations, new SharpEntropy.TwoPassDataIndexer(eventStream, cut));
return(new SharpEntropy.GisModel(trainer));
}
public virtual void TrainModel()
{
if (mDebugOn)
{
#if DNF
StreamWriter writer = new StreamWriter(mModelName + ".events", false, System.Text.Encoding.Default);
#else
var stream = new FileStream(mModelName + ".events", FileMode.OpenOrCreate);
StreamWriter writer = new StreamWriter(stream, System.Text.Encoding.GetEncoding(0));
#endif
foreach (SharpEntropy.TrainingEvent currentEvent in mEvents)
{
writer.Write(currentEvent.ToString() + "\n");
}
#if DNF
writer.Close();
#else
writer.Dispose();
stream.Dispose();
#endif
}
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.Smoothing = true;
trainer.TrainModel(new Util.CollectionEventReader(mEvents));
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
}
public static void Train(SharpEntropy.ITrainingEventReader eventReader, string outputFilename)
{
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer(0.1);
trainer.TrainModel(100, new SharpEntropy.TwoPassDataIndexer(eventReader, 5));
SharpEntropy.GisModel tokenizeModel = new SharpEntropy.GisModel(trainer);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(tokenizeModel, outputFilename);
}
/// <summary>Train a model based on the previously supplied evidence</summary>
public virtual void TrainModel()
{
if (DebugOn)
{
#if DNF
var writer = new System.IO.StreamWriter(ModelName + ".events", false, System.Text.Encoding.Default);
#else
var writer = new System.IO.StreamWriter(new FileStream(ModelName + ".events", FileMode.OpenOrCreate),
System.Text.Encoding.GetEncoding(0));
#endif
foreach (SharpEntropy.TrainingEvent trainingEvent in _events)
{
writer.Write(trainingEvent + "\n");
}
#if DNF
writer.Close();
#else
writer.Dispose();
#endif
}
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(_events), 100, 10);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), ModelName + ModelExtension);
}
public override void Train()
{
if (_resolverMode == ResolverMode.Train)
{
// if (DebugOn)
// {
// System.Console.Error.WriteLine(this.ToString() + " referential");
//#if DNF
// using (var writer = new System.IO.StreamWriter(_modelName + ".events", false, System.Text.Encoding.Default))
//#else
// using (var stream = new FileStream(_modelName + ".events", FileMode.OpenOrCreate))
// using (var writer = new System.IO.StreamWriter(stream, System.Text.Encoding.GetEncoding(0)))
//#endif
// {
// foreach (SharpEntropy.TrainingEvent e in _events)
// {
// writer.Write(e.ToString() + "\n");
// }
//#if DNF
// writer.Close();
//#else
// writer.Dispose();
// stream.Dispose();
//#endif
// }
// }
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(_events), 100, 10);
new BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), _modelName + ModelExtension);
NonReferentialResolver.Train();
}
}
/// <summary>
/// Train a model based on the previously supplied evidence.
/// </summary>
/// <seealso cref="setExtents(Context[])">
/// </seealso>
public virtual void TrainModel()
{
if (mDebugOn)
{
System.IO.StreamWriter writer = new System.IO.StreamWriter(mModelName + ".events", false, System.Text.Encoding.Default);
foreach (SharpEntropy.TrainingEvent trainingEvent in mEvents)
{
writer.Write(trainingEvent.ToString() + "\n");
}
writer.Close();
}
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(mEvents), 100, 10);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
}
/// <summary> Use this training method if you wish to supply an end of
/// sentence scanner which provides a different set of ending chars
/// other than the default ones. They are "\\.|!|\\?|\\\"|\\)".
/// </summary>
public static SharpEntropy.GisModel TrainModel(string inFile, int iterations, int cut, IEndOfSentenceScanner scanner)
{
SharpEntropy.ITrainingEventReader eventReader;
SharpEntropy.ITrainingDataReader<string> dataReader;
System.IO.StreamReader streamReader;
using (streamReader = new System.IO.StreamReader(inFile, System.Text.Encoding.UTF7))
{
dataReader = new SharpEntropy.PlainTextByLineDataReader(streamReader);
eventReader = new SentenceDetectionEventReader(dataReader, scanner);
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(eventReader, iterations, cut);
return new SharpEntropy.GisModel(trainer);
}
}
public virtual void TrainModel()
{
if (mDebugOn)
{
var writer = new StreamWriter(mModelName + ".events", false, System.Text.Encoding.Default);
foreach (var currentEvent in mEvents)
{
writer.Write(currentEvent.ToString() + "\n");
}
writer.Close();
}
var trainer = new SharpEntropy.GisTrainer();
trainer.Smoothing = true;
trainer.TrainModel(new Util.CollectionEventReader(mEvents));
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
}
public override void Train()
{
if (_resolverMode == ResolverMode.Train)
{
if (DebugOn)
{
Console.Error.WriteLine(ToString() + " referential");
using (var writer = new System.IO.StreamWriter(_modelName + ".events", false, Encoding.Default))
{
foreach (var e in _events)
{
writer.Write(e.ToString() + "\n");
}
writer.Close();
}
}
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(_events), 100, 10);
new BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), _modelName + ModelExtension);
NonReferentialResolver.Train();
}
}
public override void Train()
{
if (mResolverMode == ResolverMode.Train)
{
if (mDebugOn)
{
System.Console.Error.WriteLine(this.ToString() + " referential");
using (System.IO.StreamWriter writer = new System.IO.StreamWriter(mModelName + ".events", false, System.Text.Encoding.Default))
{
foreach (SharpEntropy.TrainingEvent e in mEvents)
{
writer.Write(e.ToString() + "\n");
}
writer.Close();
}
}
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(mEvents), 100, 10);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
mNonReferentialResolver.Train();
}
}
public static SharpEntropy.GisModel TrainModel(SharpEntropy.ITrainingEventReader eventReader, int iterations, int cut)
{
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(eventReader, iterations, cut);
return new SharpEntropy.GisModel(trainer);
}
/// <summary> Use this training method if you wish to supply an end of
/// sentence scanner which provides a different set of ending chars
/// other than the default ones. They are "\\.|!|\\?|\\\"|\\)".
/// </summary>
public static SharpEntropy.GisModel TrainModel(string inFile, int iterations, int cut, IEndOfSentenceScanner scanner)
{
SharpEntropy.ITrainingEventReader eventReader;
SharpEntropy.ITrainingDataReader<string> dataReader;
System.IO.StreamReader streamReader;
using (streamReader = new System.IO.StreamReader(inFile, System.Text.Encoding.UTF7))
{
dataReader = new SharpEntropy.PlainTextByLineDataReader(streamReader);
eventReader = new SentenceDetectionEventReader(dataReader, scanner);
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(eventReader, iterations, cut);
return new SharpEntropy.GisModel(trainer);
}
}
private static void Learn(String learnFileContent)
{
UTF8Encoding enc = new UTF8Encoding();
byte[] data = enc.GetBytes(learnFileContent);
System.IO.StreamReader trainingStreamReader = new StreamReader(new MemoryStream(data));
SharpEntropy.ITrainingEventReader eventReader = new SharpEntropy.BasicEventReader(new SharpEntropy.PlainTextByLineDataReader(trainingStreamReader));
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(eventReader);
model = new SharpEntropy.GisModel(trainer);
positiveIdx = model.GetOutcomeIndex("Positive");
}
public virtual void TrainModel()
{
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(mEvents), 100, 10);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
}
/// <summary>
/// Trains the chunker.
/// Training file should be one word per line where each line consists of a
/// space-delimited triple of "word pos outcome".
/// Sentence breaks are indicated by blank lines.
/// </summary>
/// <param name="eventReader">The chunker event reader</param>
/// <param name="iterations">The number of iterations to perform</param>
/// <param name="cutoff">
/// The number of times a predicate must be seen in order
/// to be relevant for training.
/// </param>
/// <returns>Trained model</returns>
public static SharpEntropy.GisModel Train(SharpEntropy.ITrainingEventReader eventReader, int iterations, int cutoff)
{
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(iterations, new SharpEntropy.TwoPassDataIndexer(eventReader, cutoff));
return new SharpEntropy.GisModel(trainer);
}
public virtual void Train()
{
if (ResolverMode.Train == mResolverMode)
{
Console.Error.WriteLine(this + " referential");
if (mDebugOn)
{
var writer = new System.IO.StreamWriter(mModelName + ".events", false, System.Text.Encoding.Default);
foreach (SharpEntropy.TrainingEvent trainingEvent in mEvents)
{
writer.Write(trainingEvent.ToString() + "\n");
}
writer.Close();
}
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(mEvents), 100, 10);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
}
}
// Utilities -----------------------
private static SharpEntropy.GisModel Train(SharpEntropy.ITrainingEventReader eventStream, int iterations, int cut)
{
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(iterations, new SharpEntropy.TwoPassDataIndexer(eventStream, cut));
return new SharpEntropy.GisModel(trainer);
}
public override void Train()
{
if (mResolverMode == ResolverMode.Train)
{
if (mDebugOn)
{
System.Console.Error.WriteLine(this.ToString() + " referential");
using (System.IO.StreamWriter writer = new System.IO.StreamWriter(mModelName + ".events", false, System.Text.Encoding.Default))
{
foreach (SharpEntropy.TrainingEvent e in mEvents)
{
writer.Write(e.ToString() + "\n");
}
writer.Close();
}
}
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(mEvents), 100, 10);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
mNonReferentialResolver.Train();
}
}
public static void Train(SharpEntropy.ITrainingEventReader eventReader, string outputFilename)
{
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer(0.1);
trainer.TrainModel(100, new SharpEntropy.TwoPassDataIndexer(eventReader, 5));
SharpEntropy.GisModel tokenizeModel = new SharpEntropy.GisModel(trainer);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(tokenizeModel, outputFilename);
}
public virtual void TrainModel()
{
if (mDebugOn)
{
StreamWriter writer = new StreamWriter(mModelName + ".events", false, System.Text.Encoding.Default);
foreach (SharpEntropy.TrainingEvent currentEvent in mEvents)
{
writer.Write(currentEvent.ToString() + "\n");
}
writer.Close();
}
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.Smoothing = true;
trainer.TrainModel(new Util.CollectionEventReader(mEvents));
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), mModelName + mModelExtension);
}
/// <summary>Train a model based on the previously supplied evidence</summary>
public virtual void TrainModel()
{
if (DebugOn)
{
var writer = new System.IO.StreamWriter(ModelName + ".events", false, System.Text.Encoding.Default);
foreach (SharpEntropy.TrainingEvent trainingEvent in _events)
{
writer.Write(trainingEvent + "\n");
}
writer.Close();
}
var trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(new Util.CollectionEventReader(_events), 100, 10);
new SharpEntropy.IO.BinaryGisModelWriter().Persist(new SharpEntropy.GisModel(trainer), ModelName + ModelExtension);
}
private static SharpEntropy.GisModel Train(SharpEntropy.ITrainingEventReader eventReader, int iterations, int cutoff)
{
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(iterations, new SharpEntropy.TwoPassDataIndexer(eventReader, cutoff));
return(new SharpEntropy.GisModel(trainer));
}
public static SharpEntropy.GisModel TrainModel(SharpEntropy.ITrainingEventReader eventReader, int iterations, int cut)
{
SharpEntropy.GisTrainer trainer = new SharpEntropy.GisTrainer();
trainer.TrainModel(eventReader, iterations, cut);
return new SharpEntropy.GisModel(trainer);
}
