/*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
/*
* PMMLClassifier.java
* Copyright (C) 2008-2012 University of Waikato, Hamilton, New Zealand
*
*/
package weka.classifiers.pmml.consumer;
import java.io.Serializable;
import org.w3c.dom.Document;
import org.w3c.dom.Element;
import org.w3c.dom.Node;
import org.w3c.dom.NodeList;
import weka.classifiers.AbstractClassifier;
import weka.core.Instances;
import weka.core.pmml.MappingInfo;
import weka.core.pmml.MiningSchema;
import weka.core.pmml.PMMLModel;
import weka.gui.Logger;
/**
* Abstract base class for all PMML classifiers.
*
* @author Mark Hall (mhall{[at]}pentaho{[dot]}com)
* @version $Revision: 8034 $
*/
public abstract class PMMLClassifier extends AbstractClassifier
implements Serializable, PMMLModel {
/** For serialization */
private static final long serialVersionUID = -5371600590320702971L;
/** PMML version */
protected String m_pmmlVersion = "?";
/** Creator application */
protected String m_creatorApplication = "?";
/** Logger */
protected Logger m_log = null;
/** The data dictionary */
protected Instances m_dataDictionary;
/** The fields and meta data used by the model */
protected MiningSchema m_miningSchema;
/** The mapping between mining schema fields and incoming instance
attributes */
protected transient MappingInfo m_fieldsMap;
/** Has the classifier been initialized (i.e. have we established
a mapping between the mining schema and the incoming instances)? */
protected transient boolean m_initialized = false;
/**
* Constructor.
*
* @param dataDictionary the data dictionary
* @param miningSchema the mining schema
*/
PMMLClassifier(Instances dataDictionary,
MiningSchema miningSchema) {
m_dataDictionary = dataDictionary;
m_miningSchema = miningSchema;
}
/**
* Set the version of PMML used for this model.
*
* @param doc the Document encapsulating the pmml
*/
public void setPMMLVersion(Document doc) {
NodeList tempL = doc.getElementsByTagName("PMML");
Node pmml = tempL.item(0);
if (pmml.getNodeType() == Node.ELEMENT_NODE) {
String version = ((Element)pmml).getAttribute("version");
if (version.length() > 0) {
m_pmmlVersion = version;
}
}
}
/**
* Set the name of the application (if specified) that created this
* model
*
* @param doc the Document encapsulating the pmml
*/
public void setCreatorApplication(Document doc) {
NodeList tempL = doc.getElementsByTagName("Header");
Node header = tempL.item(0);
if (header.getNodeType() == Node.ELEMENT_NODE) {
NodeList appL = ((Element)header).getElementsByTagName("Application");
if (appL.getLength() > 0) {
Node app = appL.item(0);
if (app.getNodeType() == Node.ELEMENT_NODE) {
String appName = ((Element)app).getAttribute("name");
if (appName != null && appName.length() > 0) {
String version = ((Element)app).getAttribute("version");
if (version != null && version.length() > 0) {
appName += " v. " + version;
}
m_creatorApplication = appName;
}
}
}
}
}
/**
* Get the data dictionary.
*
* @return the data dictionary
*/
public Instances getDataDictionary() {
return m_dataDictionary;
}
/**
* Get the mining schema for this model.
*
* @return the mining schema
*/
public MiningSchema getMiningSchema() {
return m_miningSchema;
}
/**
* Get the PMML version used for this model.
*
* @return the PMML version
*/
public String getPMMLVersion() {
return m_pmmlVersion;
}
/**
* Get the name of the application that created this model
*
* @return the name of the creating application or null
* if not specified in the pmml.
*/
public String getCreatorApplication() {
return m_creatorApplication;
}
/**
* Set a logger to use.
*
* @param log the logger to use
*/
public void setLog(Logger log) {
m_log = log;
}
/**
* Get the logger.
*
* @return the logger (or null if none is being used)
*/
public Logger getLog() {
return m_log;
}
/**
* Throw an exception - PMML models are pre-built.
*
* @param data the Instances to learn from
* @throws Exception if something goes wrong
*/
public void buildClassifier(Instances data) throws Exception {
throw new Exception("[PMMLClassifier] PMML models are pre-built "
+ "and static!");
}
/**
* Signal that a scoring run has been completed. Resets
* the initialized state to false so that a subsequent
* scoring run will trigger the mapping of the mining
* schema to incoming instances. If not called after a
* scoring run, then the classifier will assume that
* the current mapping is still valid.
*/
public void done() {
m_initialized = false;
m_fieldsMap = null;
}
/**
* Map mining schema to incoming instances.
*
* @param dataSet the structure of the incoming Instances
* @throws Exception if something goes wrong
*/
public void mapToMiningSchema(Instances dataSet) throws Exception {
if (m_fieldsMap == null) {
// PMMLUtils.mapToMiningSchema(dataSet, m_miningSchema);
m_fieldsMap = new MappingInfo(dataSet, m_miningSchema, m_log);
m_initialized = true;
}
}
/**
* Get a textual description of the mapping between mining schema
* fields and incoming data fields.
*
* @return a description of the fields mapping as a String or null if
* no mapping has been constructed yet.
*/
public String getFieldsMappingString() {
if (!m_initialized) {
return null;
}
return m_fieldsMap.getFieldsMappingString();
}
}