zetnormal.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 Michael Thess
  * @version 1.0
  */
package com.prudsys.pdm.Transform.OneToOne;

import com.prudsys.pdm.Adapters.PmmlVersion20.DerivedField;
import com.prudsys.pdm.Adapters.PmmlVersion20.LinearNorm;
import com.prudsys.pdm.Adapters.PmmlVersion20.NormContinuous;
import com.prudsys.pdm.Core.Category;
import com.prudsys.pdm.Core.MiningAttribute;
import com.prudsys.pdm.Core.MiningException;
import com.prudsys.pdm.Core.NumericAttribute;
import com.prudsys.pdm.Models.Statistics.Group;
import com.prudsys.pdm.Transform.OneToOneMapping;

/**
 * Realization of zeta normalization. Technically, this is also
 * a linear normalization.
 *
 * The value of mean and deviation can be set via the
 * corresponding setter methods. If a statistics model
 * object is passed, they are taken from this.
 *
 * Missing values are transformed into missing values.
 */
public class ZetNormal extends OneToOneMapping
{
  // -----------------------------------------------------------------------
  //  Constant to define deviation epsion
  // -----------------------------------------------------------------------
  /** Lower deviation bound for division. */
  public static final double DEVIATION_EPSILON = 0.0000000001;

  // -----------------------------------------------------------------------
  //  Variables declarations
  // -----------------------------------------------------------------------
  /** Mean value of attribute. */
  private double mean;

  /** Deviation of attribute. */
  private double deviation;

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

  // -----------------------------------------------------------------------
  //  Getter and setter methods
  // -----------------------------------------------------------------------
  /**
   * Sets mean value.
   *
   * @param mean new mean value
   */
  public void setMean(double mean)
  {
    this.mean = mean;
  }

  /**
   * Returns mean value.
   *
   * @return mean value
   */
  public double getMean()
  {
    return mean;
  }

  /**
   * Sets deviation value.
   *
   * @param deviation new deviation value
   */
  public void setDeviation(double deviation)
  {
    this.deviation = deviation;
  }

  /**
   * Returns deviation value.
   *
   * @return deviation value
   */
  public double getDeviation()
  {
    return deviation;
  }

  // -----------------------------------------------------------------------
  //  Transformation methods
  // -----------------------------------------------------------------------
  /**
   * Transforms the source attribute. The result is the target attribute.
   *
   * @return transformed (normalized) attribute
   * @exception MiningException could not transform attribute
   */
  public MiningAttribute transformAttribute() throws MiningException
  {
      if (getSourceAttribute() == null)
        throw new MiningException("Could not find source attribute: " + sourceName);

      if (! (getSourceAttribute() instanceof NumericAttribute))
        throw new MiningException("Source attribute '" + sourceName + "' must be numeric");

      NumericAttribute sourceAttribute      = (NumericAttribute) getSourceAttribute();
      NumericAttribute transformedAttribute = (NumericAttribute) sourceAttribute.clone();

      transformedAttribute.setName( getTargetNameDynamic() );
      transformedAttribute.setLowerBound( Double.NEGATIVE_INFINITY );
      transformedAttribute.setUpperBound( Double.POSITIVE_INFINITY );

      if (statisticsMiningModel != null) {
        Group group = statisticsMiningModel.getRootGroup();
        mean        = group.getMean();
        deviation   = group.getStandart();
      };

      return transformedAttribute;
  }

  /**
   * Transforms attribute value. The result is also a value.
   *
   * @param attributeValue value of attribute to be transformed
   * @return tranformed (normalized) value
   * @exception MiningException could not transform attribute value
   */
  public double transformAttributeValue( double attributeValue ) throws MiningException
  {
      // Missing value:
      if (Category.isMissingValue(attributeValue))
        return attributeValue;

      // Normalization:
      if (deviation < DEVIATION_EPSILON)
        return 0.0;
      double transformedValue = ( attributeValue - mean ) / deviation;

      return transformedValue;
  }

  // -----------------------------------------------------------------------
  //  Methods of PMML handling
  // -----------------------------------------------------------------------
  /**
   * Creates PMML object DerivedField of this object of NormContinuous type.
   *
   * @return DerivedField element
   * @see com.prudsys.pdm.Adapters.PmmlVersion20.NormContinuous
   * @exception MiningException could not create PMML object
   */
  public Object createPmmlObject() throws MiningException
  {
      DerivedField field = (DerivedField) super.createPmmlObject();

      NormContinuous normContinuous = new NormContinuous();
      normContinuous.setField( sourceName );
      LinearNorm linearNorm[] = new LinearNorm[2];
      linearNorm[0] = new LinearNorm();
      linearNorm[0].setOrig( "" + mean );
      linearNorm[0].setNorm( "0" );
      linearNorm[1] = new LinearNorm();
      linearNorm[1].setOrig( "" + ( deviation + mean ) );
      linearNorm[1].setNorm( "1" );
      normContinuous.setLinearNorm( linearNorm );

      field.setNormContinuous( normContinuous );
      return field;
  }

  /**
   * Creates this object from PMML object DerivedField, subobject NormContinuous.
   *
   * @param pmml DerivedField element
   * @see com.prudsys.pdm.Adapters.PmmlVersion20.NormContinuous
   * @exception MiningException could not parse PMML object
   */
  public void parsePmmlObject(Object pmml) throws MiningException
  {
      super.parsePmmlObject(pmml);

      DerivedField field = (DerivedField) pmml;

      com.prudsys.pdm.Adapters.PmmlVersion20.NormContinuous norm = field.getNormContinuous();
      sourceName = norm.getField();
      LinearNorm linearNorm[] = norm.getLinearNorm();
      mean       = Double.parseDouble( linearNorm[0].getOrig() );
      deviation  = Double.parseDouble( linearNorm[1].getOrig() ) - mean;
  }
}