com.rubecula.darwin.domain.fitness
Class FitnessFunction_ScaledNormal

java.lang.Object
  com.rubecula.darwin.foundation.Function_
      com.rubecula.darwin.domain.fitness.FitnessFunction_
          com.rubecula.darwin.domain.fitness.FitnessFunction_ScaledNormal

All Implemented Interfaces:

FitnessFunction, Function, com.rubecula.jexpression.EvalExpression, com.rubecula.jexpression.EvalExpressionMutable, com.rubecula.jexpression.Valuable, net.sf.tostring0.Identifiable

public class FitnessFunction_ScaledNormal
extends FitnessFunction_

Lifespan: permanent. This class implements an approximate normal distribution but it is scaled so that the value at the mean is unity rather than the area under the entire curve being unit. To be honest, I'm not sure whether this has a sound mathematical basis for use in evolutionary computations but I'm using it for now.

Author:

Robin Hillyard

Field Summary

Fields inherited from class com.rubecula.darwin.foundation.Function_

evaluator, LOG

Fields inherited from interface com.rubecula.darwin.domain.helper.FitnessFunction

PREFIX_VAR_RPN, VAR_SHAPE_FACTOR, VAR_TARGET, VAR_VALUE

Constructor Summary

FitnessFunction_ScaledNormal()

FitnessFunction_ScaledNormal(com.rubecula.jexpression.Evaluator evaluator)

Method Summary

java.lang.String	getIdentifier()
protected double	standardFunction(java.lang.Object... arguments)
protected java.lang.CharSequence[]	standardFunctionTokens(com.rubecula.jexpression.Notation notation)

Methods inherited from class com.rubecula.darwin.domain.fitness.FitnessFunction_

getFitness

Methods inherited from class com.rubecula.darwin.foundation.Function_

addListener, addSymbol, equals, evaluate, getErrorInfo, getEvaluator, getExpression, getExpressionMap, getExpressionTerms, getValue, hashCode, isDifferent, isMutable, normalDistributionDensityFunction, normalDistributionDensityFunctionScaled, onFunctionChange, removeVariable, reset, setEvaluator, setExpression, setExpression

Methods inherited from class java.lang.Object

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

Constructor Detail

FitnessFunction_ScaledNormal

public FitnessFunction_ScaledNormal()

FitnessFunction_ScaledNormal

public FitnessFunction_ScaledNormal(com.rubecula.jexpression.Evaluator evaluator)

Parameters:

evaluator -

Method Detail

getIdentifier

public java.lang.String getIdentifier()

Returns:

"Fitness function: scaled normal distribution"

See Also:

Identifiable.getIdentifier()

standardFunction

protected double standardFunction(java.lang.Object... arguments)
                           throws FunctionException

Specified by:

standardFunction in class Function_

Parameters:

arguments - these arguments must be, in order,

1. Number: value
2. Number: mean
3. Number: standard deviation

Note that we do not protect against cast exceptions.

Returns:

result of invoking Function_.normalDistributionDensityFunctionScaled(double, double, double) .

Throws:

FunctionException - XXX

See Also:

Function_.standardFunction(Object[])

standardFunctionTokens

protected java.lang.CharSequence[] standardFunctionTokens(com.rubecula.jexpression.Notation notation)

Specified by:

standardFunctionTokens in class Function_

Parameters:

notation - the notation type used by the evaluator.

Returns:

the sequence of tokens that corresponds to the Function_.standardFunction(Object...). This is called by Function_.reset().

See Also:

Function_.standardFunctionTokens(Notation)