ec.select
Class SigmaScalingSelection

java.lang.Object
  extended by ec.BreedingSource
      extended by ec.SelectionMethod
          extended by ec.select.FitProportionateSelection
              extended by ec.select.SigmaScalingSelection
All Implemented Interfaces:
Prototype, Setup, RandomChoiceChooser, java.io.Serializable, java.lang.Cloneable

public class SigmaScalingSelection
extends FitProportionateSelection

Similar to FitProportionateSelection, but with adjustments to scale up/exaggerate differences in fitness for selection when true fitness values are very close to eachother across the population. This addreses a common problem with FitProportionateSelection wherein selection approaches random selection during late runs when fitness values do not differ by much.

Like FitProportionateSelection this is not appropriate for steady-state evolution. If you're not familiar with the relative advantages of selection methods and just want a good one, use TournamentSelection instead. Not appropriate for multiobjective fitnesses.

Note: Fitnesses must be non-negative. 0 is assumed to be the worst fitness.

Typical Number of Individuals Produced Per produce(...) call
Always 1.

Parameters

base.scaled-fitness-floor
double = some small number (defaults to 0.1)
(The sigma scaling formula sometimes returns negative values. This is unacceptable for fitness proportionate style selection so we must substitute the fitnessFloor (some value >= 0) for the sigma scaled fitness when that sigma scaled fitness <= fitnessFloor.)

Default Base
select.sigma-scaling

See Also:
Serialized Form

Field Summary
static java.lang.String P_SCALED_FITNESS_FLOOR
          Scaled fitness floor
static java.lang.String P_SIGMA_SCALING
          Default base
 
Fields inherited from class ec.select.FitProportionateSelection
fitnesses, P_FITNESSPROPORTIONATE
 
Fields inherited from class ec.SelectionMethod
INDS_PRODUCED
 
Fields inherited from class ec.BreedingSource
CHECKBOUNDARY, DEFAULT_PRODUCED, NO_PROBABILITY, P_PROB, probability, UNUSED
 
Constructor Summary
SigmaScalingSelection()
           
 
Method Summary
 Parameter defaultBase()
          Returns the default base for this prototype.
 void prepareToProduce(EvolutionState s, int subpopulation, int thread)
          A default version of prepareToProduce which does nothing.
 void setup(EvolutionState state, Parameter base)
          Sets up the BreedingPipeline.
 
Methods inherited from class ec.select.FitProportionateSelection
finishProducing, produce
 
Methods inherited from class ec.SelectionMethod
preparePipeline, produce, produces, typicalIndsProduced
 
Methods inherited from class ec.BreedingSource
clone, getProbability, pickRandom, setProbability, setupProbabilities
 
Methods inherited from class java.lang.Object
equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

P_SIGMA_SCALING

public static final java.lang.String P_SIGMA_SCALING
Default base

See Also:
Constant Field Values

P_SCALED_FITNESS_FLOOR

public static final java.lang.String P_SCALED_FITNESS_FLOOR
Scaled fitness floor

See Also:
Constant Field Values
Constructor Detail

SigmaScalingSelection

public SigmaScalingSelection()
Method Detail

defaultBase

public Parameter defaultBase()
Description copied from interface: Prototype
Returns the default base for this prototype. This should generally be implemented by building off of the static base() method on the DefaultsForm object for the prototype's package. This should be callable during setup(...).

Specified by:
defaultBase in interface Prototype
Overrides:
defaultBase in class FitProportionateSelection

setup

public void setup(EvolutionState state,
                  Parameter base)
Description copied from class: BreedingSource
Sets up the BreedingPipeline. You can use state.output.error here because the top-level caller promises to call exitIfErrors() after calling setup. Note that probability might get modified again by an external source if it doesn't normalize right.

The most common modification is to normalize it with some other set of probabilities, then set all of them up in increasing summation; this allows the use of the fast static BreedingSource-picking utility method, BreedingSource.pickRandom(...). In order to use this method, for example, if four breeding source probabilities are {0.3, 0.2, 0.1, 0.4}, then they should get normalized and summed by the outside owners as: {0.3, 0.5, 0.6, 1.0}.

Specified by:
setup in interface Prototype
Specified by:
setup in interface Setup
Overrides:
setup in class BreedingSource
See Also:
Prototype.setup(EvolutionState,Parameter)

prepareToProduce

public void prepareToProduce(EvolutionState s,
                             int subpopulation,
                             int thread)
Description copied from class: SelectionMethod
A default version of prepareToProduce which does nothing.

Overrides:
prepareToProduce in class FitProportionateSelection