Package weka.classifiers.pmml.producer

Class AbstractPMMLProducerHelper

java.lang.Object

weka.classifiers.pmml.producer.AbstractPMMLProducerHelper

Direct Known Subclasses:

LogisticProducerHelper

public abstract class AbstractPMMLProducerHelper
extends java.lang.Object

Abstract base class for PMMLProducer helper classes to extend.

Version:

$Revision: $

Author:

David Persons, Mark Hall (mhall{[at]}pentaho{[dot]}com)

Field Summary

Fields

Modifier and Type	Field	Description
static java.lang.String	PMML_VERSION	PMML version that the jaxbbindings were created from

Constructor Summary

Constructors

Constructor	Description
AbstractPMMLProducerHelper()

Method Summary

Modifier and Type	Method	Description
static void	addDataDictionary(Instances trainHeader, PMML pmml)	Adds a data dictionary to the supplied PMML object.
static java.lang.String[]	getNameAndValueFromUnsupervisedNominalToBinaryDerivedAttribute(Instances train, Attribute derived)	Extracts the original attribute name and value from the name of a binary indicator attribute created by unsupervised NominalToBinary.
static OPTYPE	getOPTYPE(int wekaType)	Returns an OPTYPE for a weka attribute type.
static PMML	initPMML()	Initializes a PMML object with header information.

Methods inherited from class java.lang.Object

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

Field Detail

PMML_VERSION

public static final java.lang.String PMML_VERSION

PMML version that the jaxbbindings were created from

See Also:

Constant Field Values

Constructor Detail

AbstractPMMLProducerHelper

public AbstractPMMLProducerHelper()

Method Detail

initPMML

public static PMML initPMML()

Initializes a PMML object with header information.

Returns:

an initialized PMML object

addDataDictionary

public static void addDataDictionary(Instances trainHeader,
                                     PMML pmml)

Adds a data dictionary to the supplied PMML object.

Parameters:

trainHeader - the training data header - i.e. the header of the data that enters the buildClassifier() method of the model in question

pmml - the PMML object to add the data dictionary to

getOPTYPE

public static OPTYPE getOPTYPE(int wekaType)

Returns an OPTYPE for a weka attribute type. Note that PMML only supports categorical, continuous and ordinal types.

Parameters:

wekaType - the type of the weka attribute

Returns:

the PMML type

getNameAndValueFromUnsupervisedNominalToBinaryDerivedAttribute

public static java.lang.String[] getNameAndValueFromUnsupervisedNominalToBinaryDerivedAttribute(Instances train,
                                                                                                Attribute derived)

Extracts the original attribute name and value from the name of a binary indicator attribute created by unsupervised NominalToBinary. Handles the case where one or more equals signs might be present in the original attribute name.

Parameters:

train - the original, unfiltered training header

derived - the derived attribute from which to extract the original name and value from the name created by NominalToBinary.

Returns: