jsat

Class ARFFLoader

java.lang.Object

jsat.ARFFLoader

public class ARFFLoader
extends Object

Class for loading ARFF files. ARFF is a human readable file format used by Weka. The ARFF file formal allows for attributes that have missing information, which are encoded in JSAT as Double.NaN for numeric features, and a -1 for categorical features.
About Weka

Author:

Edward Raff

Constructor Summary

Constructors

Constructor and Description
ARFFLoader()

Method Summary

Modifier and Type	Method and Description
static SimpleDataSet	loadArffFile(File file) Uses the given file path to load a data set from an ARFF file.
static SimpleDataSet	loadArffFile(Reader input) Uses the given reader to load a data set assuming it follows the ARFF file format
static void	writeArffFile(DataSet data, OutputStream os)
static void	writeArffFile(DataSet data, OutputStream os, String relation) Writes out the dataset as an ARFF file to the given stream.

Methods inherited from class java.lang.Object

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

Constructor Detail

ARFFLoader

public ARFFLoader()

Method Detail

loadArffFile

public static SimpleDataSet loadArffFile(File file)

Uses the given file path to load a data set from an ARFF file.

Parameters:

file - the path to the ARFF file to load

Returns:

the data set from the ARFF file, or null if the file could not be loaded.

loadArffFile

public static SimpleDataSet loadArffFile(Reader input)

Uses the given reader to load a data set assuming it follows the ARFF file format

Parameters:

input - the reader to load the data set from

Returns:

the data set from the stream, or null of the file could not be loaded

writeArffFile

public static void writeArffFile(DataSet data,
                                 OutputStream os)

writeArffFile

public static void writeArffFile(DataSet data,
                                 OutputStream os,
                                 String relation)

Writes out the dataset as an ARFF file to the given stream. This method will automatically handle the target variable of ClassificationDataSet and RegressionDataSet.

Parameters:

data - the dataset to write out

os - the output stream to write too

relation - the relation label to write out