edu.cmu.minorthird.classify
Class DatasetLoader

java.lang.Object
  edu.cmu.minorthird.classify.DatasetLoader

public class DatasetLoader
extends java.lang.Object

Dataset i/o. For ordinary datasets, format is one example per line, and the format for a line is:

 type subpopid label feature1 feature2 ...
 

where

• type=b or k (for binary or ordinary examples)
• subpopid is NUL or a string, naming the subpopulation from which the example was drawn
• label is +1 or -1 for binary
• features are a feature name (for binary features) and a featureName=value for numeric features.

For example: k subpop1 2 f1=4 fOrange=1 fGreen=92 ... k subpop1 1 f1=1 fBlue=10 fGreen=2 ... k subpop2 3 f1=2 fYellow=1 fRed=42 ... or b NUL +1 f1=2 fOrange=1 fGreen=92 ... b NUL -1 f1=1 fBlue=10 fGreen=2 ... or k subpop1 2 f1 fOrange fGreen ... k subpop1 1 f1 fBlue fGreen ... k subpop2 3 f1 fYellow fRed ... For SequenceDatasets, examples from a diffrerent sequence are separated by a single line containing a "*".

Author:

William Cohen

Constructor Summary

DatasetLoader()

Method Summary

static java.lang.String	getSourceAssignedToExample(java.lang.String fileName, int lineNumber) The value that will be returned by example.getSource() for the example read in from the designated location.
java.lang.Object	load(java.io.File f) Calls loadFile.
static Dataset	loadFile(java.io.File file) Load a dataset from a file
static void	loadLinkFile(java.io.File file, RealRelationalDataset dataset) Load a link file
static Dataset	loadMulti(java.io.File file, int numDim) Load a dataset from a file
static Dataset	loadRegression(java.io.File file) Save a dataset that can be used for regression
static void	loadRelFile(java.io.File file, RealRelationalDataset dataset) Load a relational dataset from a file specifying objs
static void	loadRelTempFile(java.io.File file, RealRelationalDataset dataset) Load a relational template file
static SequenceDataset	loadSequence(java.io.File file) Load a SequenceDataset from a file
static Dataset	loadSVMStyle(java.io.File file)
static void	main(java.lang.String[] args)
static void	save(Dataset dataset, java.io.File file) Save a dataset to a file.
static void	saveRegression(Dataset dataset, java.io.File file) Save a dataset that can be used for regression
static void	saveSequence(SequenceDataset dataset, java.io.File file) Save a SequenceDataset to a file Each Example in a sequence is saved on a seperate line.

Methods inherited from class java.lang.Object

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

Constructor Detail

DatasetLoader

public DatasetLoader()

Method Detail

save

public static void save(Dataset dataset,
                        java.io.File file)
                 throws java.io.IOException

Save a dataset to a file. This should save each example in the order provided by the dataset.iterator()

Throws:

java.io.IOException

saveRegression

public static void saveRegression(Dataset dataset,
                                  java.io.File file)
                           throws java.io.IOException

Save a dataset that can be used for regression

Throws:

java.io.IOException

loadRegression

public static Dataset loadRegression(java.io.File file)
                              throws java.io.IOException,
                                     java.lang.NumberFormatException

Save a dataset that can be used for regression

Throws:

java.io.IOException

java.lang.NumberFormatException

loadFile

public static Dataset loadFile(java.io.File file)
                        throws java.io.IOException,
                               java.lang.NumberFormatException

Load a dataset from a file

Throws:

java.io.IOException

java.lang.NumberFormatException

loadRelFile

public static void loadRelFile(java.io.File file,
                               RealRelationalDataset dataset)
                        throws java.io.IOException,
                               java.lang.NumberFormatException

Load a relational dataset from a file specifying objs

Throws:

java.io.IOException

java.lang.NumberFormatException

loadLinkFile

public static void loadLinkFile(java.io.File file,
                                RealRelationalDataset dataset)
                         throws java.io.IOException,
                                java.lang.NumberFormatException

Load a link file

Throws:

java.io.IOException

java.lang.NumberFormatException

loadRelTempFile

public static void loadRelTempFile(java.io.File file,
                                   RealRelationalDataset dataset)
                            throws java.io.IOException,
                                   java.lang.NumberFormatException

Load a relational template file

Throws:

java.io.IOException

java.lang.NumberFormatException

loadMulti

public static Dataset loadMulti(java.io.File file,
                                int numDim)
                         throws java.io.IOException,
                                java.lang.NumberFormatException

Load a dataset from a file

Throws:

java.io.IOException

java.lang.NumberFormatException

saveSequence

public static void saveSequence(SequenceDataset dataset,
                                java.io.File file)
                         throws java.io.IOException

Save a SequenceDataset to a file Each Example in a sequence is saved on a seperate line. An asterix (*) alone on a line seperates the sequences. Example1,0 Example1,1 Example1,2 ... * Example2,0 Example2,1 ... *

Throws:

java.io.IOException

loadSequence

public static SequenceDataset loadSequence(java.io.File file)
                                    throws java.io.IOException,
                                           java.lang.NumberFormatException

Load a SequenceDataset from a file

Throws:

java.io.IOException

java.lang.NumberFormatException

See Also:

for format

getSourceAssignedToExample

public static java.lang.String getSourceAssignedToExample(java.lang.String fileName,
                                                          int lineNumber)

The value that will be returned by example.getSource() for the example read in from the designated location.

loadSVMStyle

public static Dataset loadSVMStyle(java.io.File file)
                            throws java.io.IOException

Throws:

java.io.IOException

load

public java.lang.Object load(java.io.File f)
                      throws java.io.IOException

Calls loadFile. The Dataset is temporarily swallowed. In other words, don't call this method.

Parameters:

f -

Throws:

java.io.IOException

main

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