nrc.fuzzy
Class ModifierFunction

java.lang.Object
  |
  +--nrc.fuzzy.ModifierFunction

All Implemented Interfaces:

java.io.Serializable

Direct Known Subclasses:

AboveModifier, BelowModifier, ExtremelyModifier, IntensifyModifier, MoreorlessModifier, NormModifier, NotModifier, PlusModifier, SlightlyModifier, SomewhatModifier, VeryModifier

public abstract class ModifierFunction

extends java.lang.Object

implements java.io.Serializable

The abstract ModifierFunction class along with the concrete implementations of this class encapsulate the set of Modifiers (or Hedges) available to operate on fuzzy values or fuzzy sets. Each concrete modifier function class provides two 'call' methods that accept as a parameter either a FuzzyValue or a FuzzySet. The method creates a new FuzzyValue or FuzzySet who's value represents the application of the linguistic modifier to the fuzzy set.

Many of the modifiers are applied by raising the membership values of the fuzzy set points to a certain exponent. For example, the modifier very is implemented in this manner using the exponent 2. However, consider the case where the fuzzy set being modified consists of {0.0/1 1.0/2 0.0/3}. In this case, the membership values, at x=1, x=2 and x=3, do not change, regardless of the value of the exponent.

However, if an additional point is inserted in this fuzzy set so that it now consists of {0.0/1 0.5/2.5 1.0/2 0.0/3}, and the modifier very is applied, the point that was originally 0.5/2.5 will now become 0.25/2.5. If we insert other additional points at arbitrary x values along the FuzzyValue, it will be found that unless the membership value of these arbitrary points is 1.0 or 0.0, the membership value will change as a result of the application of the exponent.

Because the representation of the membership function is a set of points that are joined by straight lines, it may be necessary to add additional points to the fuzzy set to enhance the precision to the resulting fuzzy set.

The points added to the fuzzy set before the application of an exponential modifier, and hence the precision of the result, can be controlled by the user in three different ways. The user can specify:

• the delta x between additional points by using the setDeltaXPrecision method. For example, if the user specifies a delta x precision of 0.1, a new point will be added at approximately every 0.1 x value increment between existing points.
• the delta y between additional points by using the setDeltaYPrecision method. Similar to delta x, a new point will be added per specified y increment.
• the number of points, in total, to be in the final FuzzyValue by using the setNumberOfPointsPrecision method.

It is interesting to note that these precisions are, in fact, imprecise themselves. To accomodate each individual FuzzyValue and try to add points with as even a spacing as possible, a specified delta x of 0.1 might be functionally applied as 0.9555, or a specified 20 additional points to add might be changed to 22 by the program.

Although the methods setDeltaXPrecision, setDeltaYPrecision, and setNumberOfPointsPrecision determine the spacing for each of the 3 possible fuzzy expanding techniques, only one of them is used at any time. Which of these techniques is used is controlled by the method setPrecisionControlType. It takes as an argument one of the following constant values: DELTA_X, DELTA_Y, or NUMBER_OF_POINTS.

The following is a visual guide to the built-in (system supplied) modifiers available to users.

A Visual Guide To Modifiers

Modifiers	Visual Representation
Unmodified: this FuzzyValue has not yet been modified. It will be the FuzzyValue to which the modifiers are applied for all of the following examples except 'norm'.
Not: the NOT modifier returns the complement of the FuzzyValue passed as its argument.
Norm: the NORM modifier returns the normalized FuzzyValue. Normalizing the FuzzyValue refers to scaling it so that at least one point on the FuzzyValue has a membership value of 1.0. FuzzyValue passed as its argument.	Before Normalization After Normalization
More_Or_Less: the MORE_OR_LESS modifier returns the expanded FuzzyValue passed as its argument, having raised all the membership values of the FuzzyValue by a factor of 1/3.
Somewhat: the SOMEWHAT modifier returns the expanded FuzzyValue passed as its argument, having raised all the membership values of the FuzzyValue by a factor of 1/2.
Plus: the PLUS modifier returns the expanded FuzzyValue passed as its argument, having raised all the membership values of the FuzzyValue by a factor of 1.25.
Very: the VERY modifier returns the expanded FuzzyValue passed as its argument, having raised all the membership values of the FuzzyValue by a factor of 2.
Extremely: the EXTREMELY modifier returns the expanded FuzzyValue passed as its argument, having raised all the membership values of the FuzzyValue by a factor of 3.
Slightly: the SLIGHTLY modifier returns the expanded FuzzyValue A passed as its argument, having performed the following modifications on it: intensify [ norm (plus A AND not very A) ]
Consider the partial breakdown, shown below, of this example. On the left, directly below, is shown this FuzzyValue after applying the plus modifier. On the right is this FuzzyValue after applying both the not and the very modifiers. Below these is depicted plus FuzzyValue AND not very FuzzyValue, showing the portion of the FuzzyValue that is selected which intuitively matches our linguistic understanding of the word "slightly".
PLUS	NOT VERY
Now, put them together. PLUS FuzzyValue AND NOT VERY FuzzyValue
Intensify: the INTENSIFY modifier returns the expanded FuzzyValue passed as its argument, having having performed the following modifications on it: if (0.0 <= y <= 0.5), y = 2*y2 if (0.5 < y <= 1.0), y = 1 - 2*(1-y)2 The has the effect of emphasizing the sections of the FuzzyValue that have a membership value greater than 0.5, and understating any sections of the FuzzyValue with a membership value less than or equal to 0.5.
Above: the ABOVE modifier identifies the first x value at which the maximum value is reached. All membership values below this point are set to zero and all membership values above this value are set to 1-y. For convex fuzzy sets this gives an intuitive result.
Below: the BELOW modifier identifies the first x value at which the maximum value is reached. All membership values above this point are set to zero and all membership values below this value are set to 1-y. For convex fuzzy sets this gives an intuitive result.

Each individual fuzzy modifier is created as a subclass of the ModifierFunction class. If created according to the instructions provided these fuzzy modifer classes when Constructed will automatically call the Modifiers.add method to add themselves to the list of available modifiers. So the preferred way to add new modifiers is:

• Create the modifier class

  
        public class VeryModifier extends ModifierFunction
        {
          public VeryModifier() // use a default name for the modifier
          {
            // required to set the name of the modifier 
            // and add to list in Modifiers class
            super( "very" ); 
          }
  
          public VeryModifier(String s) // use a supplied name for the modifier
          {
            // required to set the name of the modifier 
            // and add to list in Modifiers class
            super( s );
          }
  
          public FuzzyValue call(FuzzyValue fv)
          {
            FuzzyValue fvNew = null;
            // executes the 'call' method below for FuzzySet to do the modifier function
            FuzzySet fs = call(fv.getFuzzySet()); 
            try {
              fvNew = new FuzzyValue(fv.getFuzzyVariable(), fs);
            }
            catch (XValueOutsideUODException e)
            { // we know that the modifers like 'very' do not expand the x value range
              // of the fuzzy set so if get this exception there must be an internal error
              System.err.println("Internal error in Modifier function '"+getName()+
                                 "': " + e);
              System.exit(100);
            }
            // must do this to set the linguistic expression for the new fuzzy value
            fvNew.unaryModifyLinguisticExpression(getName(), fv.getLinguisticExpression());
            return(fvNew);
          }
  
          public FuzzySet call(FuzzySet fs)
          { // the real work is done here -- on the fuzzy set
            return(concentrateDilute(fs, 2.0));
          }
        }
   

• Create an instance of the new modifier class. This automatically adds the class instance to the list of modifier functions maintained by the Modifiers class.
• Once this has been done the modifier can be used directly by the user

  
        myVeryMod = new VeryModifier("very");
            ...
        aModifiedFuzzyValue = myVeryMod.call(aFuzzyValue);
   OR
        String mod = "very";
            ...
        aModifiedFuzzyValue = Modifers.call(mod, aFuzzyValue);
   

  The second form uses the Modifiers class to perform a 'variable' modifier method and is used internally by the fuzzy package when creating fuzzy values or fuzzy terms using 'linguistic expressions'.

For convenience an instance of each of the built-in (system supplied) modifiers is created in the Modifiers class and a method is created for each one allowing them to be accessed in a simpler fashion:

      aModifiedFuzzyValue = Modifers.very(aFuzzyValue);
 

NOTE: all modifier names are case insensitive ('very' is the same as 'VERY')

See Also:

Modifiers, Serialized Form

Field Summary

static int	DELTA_X A constant used in setting the precisionControlType so that delta x spacing will be used when expanding fuzzy sets
static int	DELTA_Y A constant used in setting the precisionControlType so that delta y spacing will be used when expanding fuzzy sets
protected static double	deltaX This class variable represents a delta x spacing to be used in the expansion of the FuzzyValue/FuzzySet as required by certain fuzzy modifier functions.
protected static double	deltaY This class variable represents a delta y spacing to be used in the expansion of the FuzzyValue/FuzzySet as required by certain fuzzy modifier functions.
static int	NUMBER_OF_POINTS A constant used in setting the precisionControlType so that number of points setting will be used when expanding fuzzy sets
protected static int	numberOfPoints This class variable represents the number of points to be used in the expansion of the FuzzyValue/FuzzySet as required by certain fuzzy modifier functions.
protected static int	precisionControlType This class variable designates the type of precision control to be used in the expansion of fuzzy sets.

Constructor Summary

ModifierFunction(java.lang.String name)
All Modifier Functions must have a String name that they are known by so they can be used in linguistic expressions such as "very cold".

Method Summary

abstract nrc.fuzzy.FuzzySet	call(nrc.fuzzy.FuzzySet fs) A concrete implementation of a Modifier Function will need to supply this method, since it will be called to perform the modifier operation on the Fuzzy Set.
abstract nrc.fuzzy.FuzzyValue	call(nrc.fuzzy.FuzzyValue fv) A concrete implementation of a Modifier Function will need to supply this method, since it will be called to perform the modifier operation on the Fuzzy Value.
static nrc.fuzzy.FuzzySet	concentrateDilute(nrc.fuzzy.FuzzySet a, double power) Returns a new FuzzySet object which represents the expansion of the FuzzySet argument, with all the membership values raised to the specified power.
static nrc.fuzzy.FuzzySet	expandSet(nrc.fuzzy.FuzzySet a) This method expands a fuzzy set to (possibly) include more points as required for the implementation of certain fuzzy modifier functions.
java.lang.String	getName() Retrieves the string name of the ModifierFunction.
static void	setDeltaXPrecision(double precision) Sets the delta x precision value used for fuzzy set expansion.
static void	setDeltaYPrecision(double precision) Sets the delta y precision value for fuzzy set expansion.
static void	setNumberOfPointsPrecision(int numPoints) Sets the number of points used for FuzzyValue/FuzzySet expansion.
static void	setPrecisionControlType(int type) Sets the type of precision control that will be used by the expandSet method that expands the number of points in a fuzzy set.

Methods inherited from class java.lang.Object

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

Field Detail

deltaX

protected static double deltaX

This class variable represents a delta x spacing to be used in the expansion of the FuzzyValue/FuzzySet as required by certain fuzzy modifier functions. The deltaX value is used only when precisionControlType is set to DELTA_X. The smaller the delta value, the greater the precision when a fuzzy set is modified (i.e. more points will be generated in the resulting fuzzy set).

See Also:

expandSet(nrc.fuzzy.FuzzySet), precisionControlType, Modifiers

deltaY

protected static double deltaY

This class variable represents a delta y spacing to be used in the expansion of the FuzzyValue/FuzzySet as required by certain fuzzy modifier functions. The deltaY value is used only when precisionControlType is set to DELTA_Y. The smaller the delta value, the greater the precision when a fuzzy set is modified (i.e. more points will be generated in the resulting fuzzy set).

See Also:

expandSet(nrc.fuzzy.FuzzySet), precisionControlType, Modifiers

numberOfPoints

protected static int numberOfPoints

This class variable represents the number of points to be used in the expansion of the FuzzyValue/FuzzySet as required by certain fuzzy modifier functions. The numberOfPoints value is used only when precisionControlType is set to NUMBER_OF_POINTS. The lager the number of points, the greater the precision when a fuzzy set is modified (i.e. more points will be generated in the resulting fuzzy set).

See Also:

expandSet(nrc.fuzzy.FuzzySet), precisionControlType, Modifiers

precisionControlType

protected static int precisionControlType

This class variable designates the type of precision control to be used in the expansion of fuzzy sets. The values allowed are DELTA_X, DELTA_Y, and NUMBER_OF_POINTS.

See Also:

expandSet(nrc.fuzzy.FuzzySet), Modifiers

DELTA_X

public static final int DELTA_X

A constant used in setting the precisionControlType so that delta x spacing will be used when expanding fuzzy sets

See Also:

precisionControlType, Constant Field Values

DELTA_Y

public static final int DELTA_Y

A constant used in setting the precisionControlType so that delta y spacing will be used when expanding fuzzy sets

See Also:

precisionControlType, Constant Field Values

NUMBER_OF_POINTS

public static final int NUMBER_OF_POINTS

A constant used in setting the precisionControlType so that number of points setting will be used when expanding fuzzy sets

See Also:

precisionControlType, Constant Field Values

Constructor Detail

ModifierFunction

public ModifierFunction(java.lang.String name)

All Modifier Functions must have a String name that they are known by so they can be used in linguistic expressions such as "very cold". The concrete implementations of ModifierFucntion must supply a constructor with a string name argument. Normally the constructor will look like:

    VeryModifier( String s )
    {
       super(s);
    }
 (/pre>
 The super constructor will take care of recording the string name of
 the modifier in the class and will also add the modifier to the list
 (hash table) of modifiers in the Modifiers class.

See Also:
Modifiers

Method Detail

setDeltaXPrecision

public static void setDeltaXPrecision(double precision)

Sets the delta x precision value used for fuzzy set expansion. It is important to note that setting the delta x precision value does not set the precision control type. The default precision control type is DELTA_Y, therefore, to set the delta x precision and have it applied by certain modifier functions (that use expandSet for example), the setPrecisionControlType method must also be used.

Parameters:

precision - the value of the desired precision for expanding FuzzyValues/FuzzySets. For example, if this argument has the value 0.1, additional points will be added to the FuzzyValue at x increments of approximately 0.1.

See Also:

expandSet(nrc.fuzzy.FuzzySet), Modifiers, precisionControlType

setDeltaYPrecision

public static void setDeltaYPrecision(double precision)

Sets the delta y precision value for fuzzy set expansion. It is important to note that setting the delta y precision does not set the precision control type. The default precision control type is DELTA_Y, but, to be sure to set the delta y precision and have it applied by certain modifier functions (that use expandSet for example), the setPrecisionControlType method should be used.

Parameters:

precision - the value of the desired precision for expanding FuzzyValues/FuzzySets. For example, if this argument has the value 0.1, additional points will be added to the FuzzyValue/FuzzySet at y increments of approximately 0.1.

See Also:

expandSet(nrc.fuzzy.FuzzySet), Modifiers, precisionControlType

setNumberOfPointsPrecision

public static void setNumberOfPointsPrecision(int numPoints)

Sets the number of points used for FuzzyValue/FuzzySet expansion. It is important to note that setting the number of points does not set the precision control type. The default precision control type is DELTA_Y, therefore, to set the number of points precision and have it applied by certain modifier functions (that use expandSet for example), the setPrecisionControlType method must also be used.

Parameters:

numPoints - the desired integer number of points to be added to each FuzzyValue. Must be > 4.

See Also:

expandSet(nrc.fuzzy.FuzzySet), Modifiers, precisionControlType

setPrecisionControlType

public static void setPrecisionControlType(int type)

Sets the type of precision control that will be used by the expandSet method that expands the number of points in a fuzzy set. See the class description for a complete justification as to why expansion of the set and precision control are required, and for an explanation of the three different types of precision control. The argument for this method must be one of the following three constants: DELTA_X, DELTA_Y, NUMBER_OF_POINTS.

Note that setting the individual precisions does not affect which precision is used in expansion calculations. Only this method designates which type of precision control will be used.

Parameters:

type - the precision control type desired for the expansion of the FuzzyValue, if expansion is required. This argument must be one of the following constants: DELTA_X, DELTA_Y, NUMBER_OF_POINTS.

See Also:

expandSet(nrc.fuzzy.FuzzySet), Modifiers, precisionControlType

getName

public final java.lang.String getName()

Retrieves the string name of the ModifierFunction.

call

public abstract nrc.fuzzy.FuzzyValue call(nrc.fuzzy.FuzzyValue fv)

A concrete implementation of a Modifier Function will need to supply this method, since it will be called to perform the modifier operation on the Fuzzy Value.

Parameters:

fv - the fuzzy value to be modified

call

public abstract nrc.fuzzy.FuzzySet call(nrc.fuzzy.FuzzySet fs)

A concrete implementation of a Modifier Function will need to supply this method, since it will be called to perform the modifier operation on the Fuzzy Set.

Parameters:

fs - the fuzzy set to be modified

expandSet

public static nrc.fuzzy.FuzzySet expandSet(nrc.fuzzy.FuzzySet a)

This method expands a fuzzy set to (possibly) include more points as required for the implementation of certain fuzzy modifier functions. The expansion is controlled by the setting of precisionControlType.

Parameters:

a - the fuzzy set to be expanded.

Returns:

the expanded fuzzy set

See Also:

Modifiers, precisionControlType, setPrecisionControlType(int)

concentrateDilute

public static nrc.fuzzy.FuzzySet concentrateDilute(nrc.fuzzy.FuzzySet a,
                                                   double power)

Returns a new FuzzySet object which represents the expansion of the FuzzySet argument, with all the membership values raised to the specified power.

Parameters:

a - the FuzzySet to expand and either concentrate or dilute via exponential methods

power - the power (exponent) to which the membership values are raised

Returns:

a new FuzzySet object representing the concentration or dilution of the FuzzySet argument by the exponent argument.