jbnc.dataset
Class DatasetReader

java.lang.Object
  jbnc.dataset.DatasetReader

Direct Known Subclasses:

DatasetReaderInt

public class DatasetReader

extends java.lang.Object

Functions for reading data sets with test cases.

Since:

June 1, 1999

Author:

Jarek Sacha

See Also:

Dataset, NamesReader

Constructor Summary

DatasetReader()

Method Summary

protected java.util.Vector	convertCase(java.util.Vector rawData, AttributeSpecs[] names) Convert case from a raw format.
boolean	getDiscardIncompleteCases()
java.util.Vector	open(java.lang.String fileName, AttributeSpecs[] names) Reads a data file with cases - comma delimited, no header.
Dataset	open(java.lang.String fileName, java.lang.String className) Reads a data file with cases - comma delimited, with header.
void	setDiscardIncompleteCases(boolean discardIncompleteCases)

Methods inherited from class java.lang.Object

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

Constructor Detail

DatasetReader

public DatasetReader()

Method Detail

setDiscardIncompleteCases

public void setDiscardIncompleteCases(boolean discardIncompleteCases)

Parameters:

discardIncompleteCases -

getDiscardIncompleteCases

public boolean getDiscardIncompleteCases()

Returns:

discardIncompleteCases

open

public java.util.Vector open(java.lang.String fileName,
                             AttributeSpecs[] names)
                      throws java.lang.Exception

Reads a data file with cases - comma delimited, no header. This function is typically used to read files in c4.5 format, description of attributes needs to be read from the '.names' first using jbnc.dataset.NamesReader.open().

Parameters:

names - Descriptions of columns in the file.

fileName - Description of Parameter

Returns:

Vector of vectors representing cases. Each case attribute is allocated in type defined by 'names' parameters.

Throws:

java.lang.Exception - Description of Exception

open

public Dataset open(java.lang.String fileName,
                    java.lang.String className)
             throws java.lang.Exception

Reads a data file with cases - comma delimited, with header. First line in the file gives names of attributes (columns). In current implementation the attributes are assumed to be discrete.

Parameters:

fileName - Name of the file to read data from.

className - Name of the attribute representing class. If it is null, it is assumed that the last column represents class.

Returns:

Vector of vectors representing cases. Each case attribute is allocated in type defined by 'names' parameters.

Throws:

java.lang.Exception - Description of Exception

convertCase

protected java.util.Vector convertCase(java.util.Vector rawData,
                                       AttributeSpecs[] names)
                                throws java.lang.Exception

Convert case from a raw format.

Parameters:

rawData - Description of Parameter

names - Description of Parameter

Returns:

Description of the Returned Value

Throws:

java.lang.Exception - Description of Exception