weka.core.converters

Class SVMLightLoader

java.lang.Object

weka.core.converters.AbstractLoader

weka.core.converters.AbstractFileLoader

weka.core.converters.SVMLightLoader

All Implemented Interfaces:

java.io.Serializable, BatchConverter, FileSourcedConverter, Loader, URLSourcedLoader, EnvironmentHandler, RevisionHandler

public class SVMLightLoader
extends AbstractFileLoader
implements BatchConverter, URLSourcedLoader

Reads a source that is in svm light format.

For more information about svm light see:

http://svmlight.joachims.org/

See Also:

Loader, Serialized Form

Nested Class Summary

Nested classes/interfaces inherited from interface weka.core.converters.Loader

Loader.StructureNotReadyException

Field Summary

Fields

Modifier and Type	Field and Description
static java.lang.String	FILE_EXTENSION the file extension.
protected java.util.Vector<double[]>	m_Buffer the buffer of the rows read so far.
protected java.io.Reader	m_sourceReader The reader for the source file.
protected java.lang.String	m_URL the url.

Fields inherited from class weka.core.converters.AbstractFileLoader

FILE_EXTENSION_COMPRESSED, m_env, m_File, m_sourceFile, m_structure, m_useRelativePath

Fields inherited from class weka.core.converters.AbstractLoader

m_retrieval

Fields inherited from interface weka.core.converters.Loader

BATCH, INCREMENTAL, NONE

Constructor Summary

Constructors

Constructor and Description
SVMLightLoader()

Method Summary

Modifier and Type	Method and Description
protected Attribute	determineClassAttribute() Determines the class attribute, either a binary +1/-1 or numeric attribute.
protected int	determineNumAttributes(double[] values, int num) determines the number of attributes, if the number of attributes in the given row is greater than the current amount then this number will be returned, otherwise the current number.
Instances	getDataSet() Return the full data set.
java.lang.String	getFileDescription() Returns a description of the file type.
java.lang.String	getFileExtension() Get the file extension used for svm light files.
java.lang.String[]	getFileExtensions() Gets all the file extensions used for this type of file.
Instance	getNextInstance(Instances structure) SVMLightLoader is unable to process a data set incrementally.
java.lang.String	getRevision() Returns the revision string.
Instances	getStructure() Determines and returns (if possible) the structure (internally the header) of the data set as an empty set of instances.
java.lang.String	globalInfo() Returns a string describing this Loader.
static void	main(java.lang.String[] args) Main method.
void	reset() Resets the Loader ready to read a new data set.
java.lang.String	retrieveURL() Return the current url.
void	setSource(java.io.InputStream in) Resets the Loader object and sets the source of the data set to be the supplied InputStream.
void	setSource(java.net.URL url) Resets the Loader object and sets the source of the data set to be the supplied url.
void	setURL(java.lang.String url) Set the url to load from.
protected double[]	svmlightToArray(java.lang.String row) turns a svm light row into a double array with the class as the last entry.

Methods inherited from class weka.core.converters.AbstractFileLoader

getUseRelativePath, makeOptionStr, retrieveFile, runFileLoader, setEnvironment, setFile, setSource, setUseRelativePath, useRelativePathTipText

Methods inherited from class weka.core.converters.AbstractLoader

getRetrieval, setRetrieval

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

FILE_EXTENSION

public static java.lang.String FILE_EXTENSION

the file extension.

m_URL

protected java.lang.String m_URL

the url.

m_sourceReader

protected transient java.io.Reader m_sourceReader

The reader for the source file.

m_Buffer

protected java.util.Vector<double[]> m_Buffer

the buffer of the rows read so far.

Constructor Detail

SVMLightLoader

public SVMLightLoader()

Method Detail

globalInfo

public java.lang.String globalInfo()

Returns a string describing this Loader.

Returns:

a description of the Loader suitable for displaying in the explorer/experimenter gui

getFileExtension

public java.lang.String getFileExtension()

Get the file extension used for svm light files.

Specified by:

getFileExtension in interface FileSourcedConverter

Returns:

the file extension

getFileExtensions

public java.lang.String[] getFileExtensions()

Gets all the file extensions used for this type of file.

Specified by:

getFileExtensions in interface FileSourcedConverter

Returns:

the file extensions

getFileDescription

public java.lang.String getFileDescription()

Returns a description of the file type.

Specified by:

getFileDescription in interface FileSourcedConverter

Returns:

a short file description

reset

public void reset()
           throws java.io.IOException

Resets the Loader ready to read a new data set.

Specified by:

reset in interface Loader

Overrides:

reset in class AbstractFileLoader

Throws:

java.io.IOException - if something goes wrong

setSource

public void setSource(java.net.URL url)
               throws java.io.IOException

Resets the Loader object and sets the source of the data set to be the supplied url.

Parameters:

url - the source url.

Throws:

java.io.IOException - if an error occurs

setURL

public void setURL(java.lang.String url)
            throws java.io.IOException

Set the url to load from.

Specified by:

setURL in interface URLSourcedLoader

Parameters:

url - the url to load from

Throws:

java.io.IOException - if the url can't be set.

retrieveURL

public java.lang.String retrieveURL()

Return the current url.

Specified by:

retrieveURL in interface URLSourcedLoader

Returns:

the current url

setSource

public void setSource(java.io.InputStream in)
               throws java.io.IOException

Resets the Loader object and sets the source of the data set to be the supplied InputStream.

Specified by:

setSource in interface Loader

Overrides:

setSource in class AbstractLoader

Parameters:

in - the source InputStream.

Throws:

java.io.IOException - if initialization of reader fails.

svmlightToArray

protected double[] svmlightToArray(java.lang.String row)
                            throws java.lang.Exception

turns a svm light row into a double array with the class as the last entry.

Parameters:

row - the row to turn into a double array

Returns:

the corresponding double array

Throws:

java.lang.Exception - if a parsing error is encountered

determineNumAttributes

protected int determineNumAttributes(double[] values,
                                     int num)
                              throws java.lang.Exception

determines the number of attributes, if the number of attributes in the given row is greater than the current amount then this number will be returned, otherwise the current number.

Parameters:

values - the parsed values

num - the current number of attributes

Returns:

the new number of attributes

Throws:

java.lang.Exception - if parsing fails

determineClassAttribute

protected Attribute determineClassAttribute()

Determines the class attribute, either a binary +1/-1 or numeric attribute.

Returns:

the generated attribute

getStructure

public Instances getStructure()
                       throws java.io.IOException

Determines and returns (if possible) the structure (internally the header) of the data set as an empty set of instances.

Specified by:

getStructure in interface Loader

Specified by:

getStructure in class AbstractLoader

Returns:

the structure of the data set as an empty set of Instances

Throws:

java.io.IOException - if an error occurs

getDataSet

public Instances getDataSet()
                     throws java.io.IOException

Return the full data set. If the structure hasn't yet been determined by a call to getStructure then method should do so before processing the rest of the data set.

Specified by:

getDataSet in interface Loader

Specified by:

getDataSet in class AbstractLoader

Returns:

the structure of the data set as an empty set of Instances

Throws:

java.io.IOException - if there is no source or parsing fails

getNextInstance

public Instance getNextInstance(Instances structure)
                         throws java.io.IOException

SVMLightLoader is unable to process a data set incrementally.

Specified by:

getNextInstance in interface Loader

Specified by:

getNextInstance in class AbstractLoader

Parameters:

structure - ignored

Returns:

never returns without throwing an exception

Throws:

java.io.IOException - always. SVMLightLoader is unable to process a data set incrementally.

getRevision

public java.lang.String getRevision()

Returns the revision string.

Specified by:

getRevision in interface RevisionHandler

Returns:

the revision

main

public static void main(java.lang.String[] args)

Main method.

Parameters:

args - should contain the name of an input file.