marf.Storage
Class TrainingSet

java.lang.Object
  marf.Storage.StorageManager
      marf.Storage.TrainingSet

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable, IStorageManager

public class TrainingSet
extends StorageManager

TrainingSet -- Encapsulates Subject ID and subject's clusters or a feature set.

$Id: TrainingSet.java,v 1.49 2008/02/20 07:29:06 mokhov Exp $

Since:

0.0.1, December 5, 2002

Version:

$Revision: 1.49 $

Author:

Serguei Mokhov

See Also:

Serialized Form

Field Summary

static java.lang.String	DEFAULT_TRAINING_SET_FILENAME Default TrainingSet file name of marf.training.set.
protected int	iFeatureExtractionMethod Which feature extraction method was used to determine this feature vector.
protected int	iPreprocessingMethod Which preprocessing method was applied to the sample before this feature vector was extracted.
protected int	iTrainingSetFormat
protected java.util.Vector	oTrainingSamples A collection of Clusters.
static int	TRAINING_SET_CLUSTERS
static int	TRAINING_SET_FEATURE_SETS
static int	TRAINING_SET_SAMPLES

Fields inherited from class marf.Storage.StorageManager

bDumpOnNotFound, iCurrentDumpMode, oObjectToSerialize, strFilename

Fields inherited from interface marf.Storage.IStorageManager

DUMP_BINARY, DUMP_CSV_TEXT, DUMP_GZIP_BINARY, DUMP_HTML, DUMP_SQL, DUMP_XML, MARF_INTERFACE_CODE_REVISION, STORAGE_FILE_EXTENSIONS

Constructor Summary

TrainingSet()
Construct a training set object.

Method Summary

boolean	addFeatureVector(double[] padFeatureVector, java.lang.String pstrFilename, int piSubjectID, int piPreprocessingMethod, int piFeatureExtractionMethod) Adds new feature vector to the mean and recomputes the mean.
void	backSynchronizeObject() Must to be overridden by the modules that use object serialization with the generic implementation of restore().
java.lang.Object	clone() Implements Cloneable interface for the TrainingSet object.
void	dumpCSV() If derivatives use the generic implementation of dump() with the CSV dump mode, they must override this method.
java.util.Vector	getClusters() Retrieves clusters of training samples.
int	getFeatureExtractionMethod() Returns preprocessing method used on this training set.
static java.lang.String	getMARFSourceCodeRevision() Returns source code revision information.
int	getPreprocessingMethod() Returns preprocessing method used on this training set.
int	getTrainingSetFormat()
void	restore() Retrieves the current training set from disk.
void	restoreCSV() If derivatives use the generic implementation of restore() with the CSV dump mode, they must override this method.
void	setFeatureExtractionMethod(int piFeatureExtractionMethod) Sets feature extraction method used on this training set.
void	setPreprocessingMethod(int piPreprocessingMethod) Sets feature extraction method used on this training set.
void	setTrainingSetFormat(int piTrainingSetFormat)
int	size() Gets the size of the feature vectors set.
java.lang.String	toString() Provides string representation of the training set data in addition to that of the parent StorageManager.

Methods inherited from class marf.Storage.StorageManager

dump, dumpBinary, dumpGzipBinary, dumpHTML, dumpSQL, dumpXML, enableDumpOnNotFound, equals, getDefaultExtension, getDefaultExtension, getDumpMode, getFilename, getObjectToSerialize, hashCode, restoreBinary, restoreGzipBinary, restoreHTML, restoreSQL, restoreXML, setDumpMode, setFilename

Methods inherited from class java.lang.Object

finalize, getClass, notify, notifyAll, wait, wait, wait

Field Detail

DEFAULT_TRAINING_SET_FILENAME

public static final java.lang.String DEFAULT_TRAINING_SET_FILENAME

Default TrainingSet file name of marf.training.set.

Since:

0.3.0.2

See Also:

Constant Field Values

TRAINING_SET_SAMPLES

public static final int TRAINING_SET_SAMPLES

See Also:

Constant Field Values

TRAINING_SET_CLUSTERS

public static final int TRAINING_SET_CLUSTERS

See Also:

Constant Field Values

TRAINING_SET_FEATURE_SETS

public static final int TRAINING_SET_FEATURE_SETS

See Also:

Constant Field Values

iTrainingSetFormat

protected int iTrainingSetFormat

oTrainingSamples

protected java.util.Vector oTrainingSamples

A collection of Clusters.

iPreprocessingMethod

protected int iPreprocessingMethod

Which preprocessing method was applied to the sample before this feature vector was extracted.

iFeatureExtractionMethod

protected int iFeatureExtractionMethod

Which feature extraction method was used to determine this feature vector.

Constructor Detail

TrainingSet

public TrainingSet()

Construct a training set object.

Method Detail

getClusters

public final java.util.Vector getClusters()

Retrieves clusters of training samples.

Returns:

vector of training samples.

getPreprocessingMethod

public final int getPreprocessingMethod()

Returns preprocessing method used on this training set.

Returns:

the method

getFeatureExtractionMethod

public final int getFeatureExtractionMethod()

Returns preprocessing method used on this training set.

Returns:

piPreprocessingMethod the method

setPreprocessingMethod

public final void setPreprocessingMethod(int piPreprocessingMethod)

Sets feature extraction method used on this training set.

Parameters:

piPreprocessingMethod - the method

setFeatureExtractionMethod

public final void setFeatureExtractionMethod(int piFeatureExtractionMethod)

Sets feature extraction method used on this training set.

Parameters:

piFeatureExtractionMethod - the method

addFeatureVector

public final boolean addFeatureVector(double[] padFeatureVector,
                                      java.lang.String pstrFilename,
                                      int piSubjectID,
                                      int piPreprocessingMethod,
                                      int piFeatureExtractionMethod)

Adds new feature vector to the mean and recomputes the mean.

Parameters:

padFeatureVector - vector to add

pstrFilename - to check to avoid training on the same file

piSubjectID - for which subject that vector is

piPreprocessingMethod - preprocessing method used

piFeatureExtractionMethod - feature extraction method used

Returns:

true if the vector was added; false otherwise

size

public final int size()

Gets the size of the feature vectors set.

Returns:

number of training samples in the set

restoreCSV

public void restoreCSV()
                throws StorageException

Description copied from class: StorageManager

If derivatives use the generic implementation of restore() with the CSV dump mode, they must override this method.

Specified by:

restoreCSV in interface IStorageManager

Overrides:

restoreCSV in class StorageManager

Throws:

StorageException - in case of I/O or otherwise error

Since:

0.3.0.6

See Also:

IStorageManager.restoreCSV()

restore

public void restore()
             throws StorageException

Retrieves the current training set from disk.

Specified by:

restore in interface IStorageManager

Overrides:

restore in class StorageManager

Throws:

StorageException - in a case of I/O or otherwise error

See Also:

IStorageManager.restore()

dumpCSV

public void dumpCSV()
             throws StorageException

Description copied from class: StorageManager

If derivatives use the generic implementation of dump() with the CSV dump mode, they must override this method.

Specified by:

dumpCSV in interface IStorageManager

Overrides:

dumpCSV in class StorageManager

Throws:

StorageException - in case of I/O or otherwise error

Since:

0.3.0.6

See Also:

IStorageManager.dumpCSV()

getTrainingSetFormat

public int getTrainingSetFormat()

Returns:

Returns the iTrainingSetFormat.

Since:

0.3.0.6

setTrainingSetFormat

public void setTrainingSetFormat(int piTrainingSetFormat)

Parameters:

piTrainingSetFormat - The iTrainingSetFormat to set.

Since:

0.3.0.6

backSynchronizeObject

public void backSynchronizeObject()

Description copied from class: StorageManager

Must to be overridden by the modules that use object serialization with the generic implementation of restore(). By default this method is unimplemented.

Overrides:

backSynchronizeObject in class StorageManager

Since:

0.3.0.6

See Also:

StorageManager.backSynchronizeObject()

clone

public java.lang.Object clone()

Implements Cloneable interface for the TrainingSet object. Performs a "deep" copy of this object including all clusters and the feature set.

Overrides:

clone in class StorageManager

Since:

0.3.0.5

See Also:

Object.clone()

toString

public java.lang.String toString()

Provides string representation of the training set data in addition to that of the parent StorageManager.

Overrides:

toString in class StorageManager

Since:

0.3.0.6

See Also:

StorageManager.toString()

getMARFSourceCodeRevision

public static java.lang.String getMARFSourceCodeRevision()

Returns source code revision information.

Returns:

revision string

Since:

0.3.0.2