miningfilestream.java

一个数据挖掘软件ALPHAMINERR的整个过程的JAVA版源代码

/*
 *    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 2 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, write to the Free Software
 *    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */

 /**
  * Title: XELOPES Data Mining Library
  * Description: The XELOPES library is an open platform-independent and data-source-independent library for Embedded Data Mining.
  * Copyright: Copyright (c) 2002 Prudential Systems Software GmbH
  * Company: ZSoft (www.zsoft.ru), Prudsys (www.prudsys.com)
  * @author Valentine Stepanenko ([e-mail])
  * @version 1.0
  */

package com.prudsys.pdm.Input;

import java.io.BufferedReader;
import java.io.FileReader;
import java.io.IOException;
import java.io.Reader;
import java.util.Enumeration;
import java.util.Iterator;
import java.util.Vector;

import org.omg.cwm.analysis.transformation.ClassifierMap;
import org.omg.cwm.analysis.transformation.FeatureMap;
import org.omg.cwm.analysis.transformation.TransformationMap;
import org.omg.cwm.analysis.transformation.TransformationPackage;
import org.omg.cwm.objectmodel.core.CorePackage;
import org.omg.cwm.objectmodel.core.DataType;
import org.omg.cwm.resource.record.Field;
import org.omg.cwm.resource.record.RecordDef;
import org.omg.cwm.resource.record.RecordFile;
import org.omg.cwm.resource.record.RecordPackage;

import com.prudsys.pdm.Core.CategoricalAttribute;
import com.prudsys.pdm.Core.MiningAttribute;
import com.prudsys.pdm.Core.MiningDataSpecification;
import com.prudsys.pdm.Core.MiningException;
import com.prudsys.pdm.Core.NumericAttribute;

/**
 * Mining input stream class for files.
 *
 * For special file formats this class should be
 * extended.
 */
public class MiningFileStream extends MiningInputStream
{
    // -----------------------------------------------------------------------
    //  Variables declarations
    // -----------------------------------------------------------------------
    /** File name. */
    protected String fileName;

    /** File reader object. */
    protected Reader reader;

    // -----------------------------------------------------------------------
    //  Constructor
    // -----------------------------------------------------------------------
    /**
     * Empty constructor.
     */
    public MiningFileStream()
    {
    }

    /**
     * Mining file stream for a given file and given meta data.
     *
     * @param file path of file to access
     * @param metaData meta data of file data
     */
    public MiningFileStream( String file, MiningDataSpecification metaData ) throws MiningException
    {
        try
        {
            reader = new BufferedReader( new FileReader( file ) );

        }
        catch( IOException ex)
        {
            throw new MiningException( "Can't read from the file: "+file );
        }
        this.fileName = file;
        this.metaData = metaData;
    }

    /**
     * Mining file stream for a given file.
     * The meta data is automatically determined using the
     * method createMetaData.
     *
     * @param file path of file to access
     */
    public MiningFileStream( String file ) throws MiningException
    {
        this( file, null );
        metaData = recognize();
        reset();
    }

    // -----------------------------------------------------------------------
    //  Getter and setter methods
    // -----------------------------------------------------------------------
    /**
     * Returns supported stream methods.
     *
     * @return supported stream methods
     */
    public Enumeration getSupportedStreamMethods() {

      Vector<String> suppmeth = new Vector<String>();
      suppmeth.addElement("reset");
      return suppmeth.elements();
    }

    /**
     * Finds physical file model (CWM Resource Package "Record").
     *
     * @exception MiningException couldn't obtain physical model
     */
     public void findPhysicalModel() throws MiningException {

       com.prudsys.pdm.Cwm.CWMCompletePackage cwmFactory =
           com.prudsys.pdm.Cwm.CWMCompletePackage.getCWMCompletePackage();
       CorePackage cpg   = cwmFactory.getCore();
       RecordPackage rpg = cwmFactory.getRecord();
       RecordFile rfile  = rpg.getRecordFile().createRecordFile();
       rfile.setName(fileName);
       RecordDef recdef  = rpg.getRecordDef().createRecordDef();
       rfile.addRecord(recdef);
       for (int i = 0; i < metaData.getAttributesNumber(); i++) {
         Field field = rpg.getField().createField();
         MiningAttribute ma = metaData.getMiningAttribute(i);
         field.setName( ma.getName() );
         DataType dataType = cpg.getDataType().createDataType();
         String dtname = "unknownType";
         if (ma instanceof NumericAttribute) {
           if (ma.getDataType() == NumericAttribute.DOUBLE) dtname = "double";
           else if (ma.getDataType() == NumericAttribute.FLOAT) dtname = "float";
           else if (ma.getDataType() == NumericAttribute.INTEGER) dtname = "integer";
           else if (ma.getDataType() == NumericAttribute.BOOLEAN) dtname = "boolean";
           else if (ma.getDataType() == NumericAttribute.DATETIME_PRUDSYS) dtname = "datePrudsys";
           else dtname = "numeric";
         }
         else {
           if (ma.getDataType() == CategoricalAttribute.STRING) dtname = "string";
           else if (ma.getDataType() == CategoricalAttribute.BOOLEAN) dtname = "boolean";
           else dtname = "categorical";
         };
         dataType.setName(dtname);
         field.setType(dataType);
         recdef.addFeature(field);
       }

       physicalModel = rfile;
     }

     /**
      * Returns the CWM mapping from the physical to the logical data model.
      *
      * @return transformation of physical to logical data model
      * @throws MiningException couldn't get transformation
      */
     public org.omg.cwm.analysis.transformation.TransformationMap getPhysicalToLogicalModelTransformation()
         throws MiningException {

       com.prudsys.pdm.Cwm.CWMCompletePackage cwmFactory =
           com.prudsys.pdm.Cwm.CWMCompletePackage.getCWMCompletePackage();
       TransformationPackage tpg = cwmFactory.getTransformation();

       TransformationMap tm = tpg.getTransformationMap().createTransformationMap();
       ClassifierMap cm = tpg.getClassifierMap().createClassifierMap();
       tm.addOwnedElement(cm);

       RecordFile rfile = (RecordFile) getPhysicalModel();
       Iterator it      = rfile.getRecord().iterator();
       RecordDef recdef = (RecordDef) it.next();
       it               = recdef.getFeature().iterator();
       for (int i = 0; i < metaData.getAttributesNumber(); i++) {
         FeatureMap fm = tpg.getFeatureMap().createFeatureMap();
         fm.addSource( (Field) it.next() );
         fm.addTarget( metaData.getMiningAttribute(i) );
         cm.addFeatureMap(fm);
       };

       metaData.addOwnedElement(tm);
       physicalModel.addOwnedElement(tm);

       return tm;
     }

    /**
     * Returns file name.
     *
     * @return file name
     */
    public String getFileName()
    {
        return fileName;
    }

    // -----------------------------------------------------------------------
    //  General stream methods
    // -----------------------------------------------------------------------
    /**
     * Open mining file stream. This method can be left.
     *
     * @exception MiningException if a mining source access error occurs
     */
    public void open() throws MiningException
    {
    }

    /**
     * Close mining file stream by closing reader.
     *
     * @exception MiningException if a mining source access error occurs
     */
    public void close() throws MiningException
    {
      try
      {
          reader.close();
      }
      catch( IOException ex)
      {
          throw new MiningException( "Can't close reader from the file: "+fileName );
      };
    }

    /**
     * Recognize the input stream data specification. Not implemented.
     *
     * @return     the MiningDataSpecification
     * @exception  MiningException always thrown
     */
    public MiningDataSpecification recognize() throws MiningException
    {
        throw new MiningException( "Not implemented." );
    }

    // -----------------------------------------------------------------------
    //  Methods of cursor positioning
    // -----------------------------------------------------------------------
    /**
     * Places the cursor before first row.
     * This is done by closing and reopening the file reader.
     */
    public void reset() throws MiningException
    {
        try
        {
            if( reader!= null) reader.close();
            reader = new FileReader( fileName );
        }
        catch( IOException ex )
        {
            throw new MiningException( ex.getMessage() );
        }
    }

    /**
     * Advance cursor by one position. Not implemented.
     *
     * @return true if next vector exists, else false
     * @MiningException always thrown
     */
    public boolean next() throws MiningException
    {
        throw new MiningException( "Not implemented." );
    }

    /**
     * Move cursor to given position. Not supported.
     *
     * @param position new cursor position
     * @return true if possible, else false
     * @MiningException always thrown
     */
    public boolean move( int position ) throws MiningException
    {
        throw new MiningException( "Can't move cursor in the file source." );
    }

    // -----------------------------------------------------------------------
    //  Methods of reading from the stream
    // -----------------------------------------------------------------------
    /**
     * Reads current data vector. Not implemented.
     *
     * @return data vector at current cursor position
     * @exception MiningException always thrown
     */
    public MiningVector read() throws MiningException
    {
        throw new MiningException( "Not implemented." );
    }

    // -----------------------------------------------------------------------
    //  Other methods
    // -----------------------------------------------------------------------
    /**
     * Returns string representation of mining file stream
     * (just the file name).
     *
     * @return string representation of mining file stream
     */
    public String toString()
    {
        return fileName + "/n" + super.toString();
    }

    /**
     * Returns HTML string representation of mining file
     * stream.
     *
     * @return HTML string representation of mining file stream
     */
    public String toHtmlString()
    {
        String description = "<html>";
//        String path = "";
//        String token = "";
//        StringTokenizer st = new StringTokenizer( fileName, separator, false );
//        while( st.hasMoreTokens() )
//        {
//            token = st.nextToken();
//            path = path + token + separator;
//            description = description + "<a href=http://this?" + path + ">" +  token + separator + "</a>";
//        }
//        int d = description.lastIndexOf( separator +  "</a>" );
//        description = description.substring( 0, d ) + "</a>";
        description = description + "Data source:<br>" +
        "<a href=http:this?" + fileName + ">" +  fileName + "</a>";
        description = description + "</html>";

        return description;
    }

}