ec.gp.ge
Class GEProblem

java.lang.Object
  extended by ec.Problem
      extended by ec.gp.ge.GEProblem
All Implemented Interfaces:
GroupedProblemForm, Prototype, Setup, SimpleProblemForm, java.io.Serializable, java.lang.Cloneable

public class GEProblem
extends Problem
implements SimpleProblemForm, GroupedProblemForm

GEProblem is a special replacement for Problem which performs GE mapping. You do not subclass from GEProblem. Rather, create a GPProblem subclass and set it to be the 'problem' parameter of the GEProblem. The GEProblem will convert the GEIndividual into a GPIndividual, then pass this GPIndividual to the GPProblem to be evaluated.

The procedure is as follows. Let's say your GPProblem is the Artificial Ant problem. Instead of saying...

eval.problem = ec.app.ant.Ant
eval.problem = ec.app.ant.Ant
eval.problem.data = ec.app.ant.AntData
eval.problem.moves = 400
eval.problem.file = santafe.trl

... you instead make your problem a GEProblem like this:

eval.problem = ec.gp.ge.GEProblem

... and then you hang the Ant problem, and all its subsidiary data, as the 'problem' parameter from the GEProblem like so:

eval.problem.problem = ec.app.ant.Ant
eval.problem.problem.data = ec.app.ant.AntData
eval.problem.problem.moves = 400
eval.problem.problem.file = santafe.trl

Everything else should be handled for you. GEProblem is also compatible with the MasterProblem procedure for distributed evaluation, and is also both a SimpleProblemForm and a GroupedProblemForm. We've got you covered.

Parameters

base.problem
classname, inherits from GPProblem
(The GPProblem which actually performs the evaluation of the mapped GPIndividual)

See Also:
Serialized Form

Field Summary
static java.lang.String P_PROBLEM
           
 GPProblem problem
           
 
Constructor Summary
GEProblem()
           
 
Method Summary
 boolean canEvaluate()
          Asynchronous Steady-State EC only: Returns true if the problem is ready to evaluate.
 java.lang.Object clone()
          Creates a new individual cloned from a prototype, and suitable to begin use in its own evolutionary context.
 void closeContacts(EvolutionState state, int result)
          Called to shut down remote evaluation network contacts when the run is completed.
 void describe(EvolutionState state, Individual ind, int subpopulation, int threadnum, int log)
          Part of SimpleProblemForm.
 void evaluate(EvolutionState state, Individual[] ind, boolean[] updateFitness, boolean countVictoriesOnly, int[] subpops, int threadnum)
          Default version assumes that every individual is a GEIndividual.
 void evaluate(EvolutionState state, Individual ind, int subpopulation, int threadnum)
          Evaluates the individual in ind, if necessary (perhaps not evaluating them if their evaluated flags are true), and sets their fitness appropriately.
 void finishEvaluating(EvolutionState state, int threadnum)
          Will be called by the Evaluator after prepareToEvaluate(...) is called and then a series of individuals are evaluated.
 void initializeContacts(EvolutionState state)
          Called to set up remote evaluation network contacts when the run is started.
 void postprocessPopulation(EvolutionState state, Population pop, boolean countVictoriesOnly)
          Finish processing the population (such as fitness information) after evaluation.
 void prepareToEvaluate(EvolutionState state, int threadnum)
          May be called by the Evaluator prior to a series of individuals to evaluate, and then ended with a finishEvaluating(...).
 void preprocessPopulation(EvolutionState state, Population pop, boolean countVictoriesOnly)
          Set up the population pop (such as fitness information) prior to evaluation.
 void reinitializeContacts(EvolutionState state)
          Called to reinitialize remote evaluation network contacts when the run is restarted from checkpoint.
 void setup(EvolutionState state, Parameter base)
          Sets up the object by reading it from the parameters stored in state, built off of the parameter base base.
 
Methods inherited from class ec.Problem
defaultBase, describe
 
Methods inherited from class java.lang.Object
equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

P_PROBLEM

public static final java.lang.String P_PROBLEM
See Also:
Constant Field Values

problem

public GPProblem problem
Constructor Detail

GEProblem

public GEProblem()
Method Detail

setup

public void setup(EvolutionState state,
                  Parameter base)
Description copied from interface: Prototype
Sets up the object by reading it from the parameters stored in state, built off of the parameter base base. If an ancestor implements this method, be sure to call super.setup(state,base); before you do anything else.

For prototypes, setup(...) is typically called once for the prototype instance; cloned instances do not receive the setup(...) call. setup(...) may be called more than once; the only guarantee is that it will get called at least once on an instance or some "parent" object from which it was ultimately cloned.

Specified by:
setup in interface Prototype
Specified by:
setup in interface Setup
Overrides:
setup in class Problem

clone

public java.lang.Object clone()
Description copied from interface: Prototype
Creates a new individual cloned from a prototype, and suitable to begin use in its own evolutionary context.

Typically this should be a full "deep" clone. However, you may share certain elements with other objects rather than clone hem, depending on the situation:

Implementations.

Specified by:
clone in interface Prototype
Overrides:
clone in class Problem

prepareToEvaluate

public void prepareToEvaluate(EvolutionState state,
                              int threadnum)
Description copied from class: Problem
May be called by the Evaluator prior to a series of individuals to evaluate, and then ended with a finishEvaluating(...). If this is the case then the Problem is free to delay modifying the individuals or their fitnesses until at finishEvaluating(...). If no prepareToEvaluate(...) is called prior to evaluation, the Problem must complete its modification of the individuals and their fitnesses as they are evaluated as stipulated in the relevant evaluate(...) documentation for SimpleProblemForm or GroupedProblemForm. The default method does nothing. Note that prepareToEvaluate() can be called *multiple times* prior to finishEvaluating() being called -- in this case, the subsequent calls may be ignored.

Overrides:
prepareToEvaluate in class Problem

finishEvaluating

public void finishEvaluating(EvolutionState state,
                             int threadnum)
Description copied from class: Problem
Will be called by the Evaluator after prepareToEvaluate(...) is called and then a series of individuals are evaluated. However individuals may be evaluated without prepareToEvaluate or finishEvaluating being called at all. See the documentation for prepareToEvaluate for more information. The default method does nothing.

Overrides:
finishEvaluating in class Problem

initializeContacts

public void initializeContacts(EvolutionState state)
Description copied from class: Problem
Called to set up remote evaluation network contacts when the run is started. By default does nothing.

Overrides:
initializeContacts in class Problem

reinitializeContacts

public void reinitializeContacts(EvolutionState state)
Description copied from class: Problem
Called to reinitialize remote evaluation network contacts when the run is restarted from checkpoint. By default does nothing.

Overrides:
reinitializeContacts in class Problem

closeContacts

public void closeContacts(EvolutionState state,
                          int result)
Description copied from class: Problem
Called to shut down remote evaluation network contacts when the run is completed. By default does nothing.

Overrides:
closeContacts in class Problem

canEvaluate

public boolean canEvaluate()
Description copied from class: Problem
Asynchronous Steady-State EC only: Returns true if the problem is ready to evaluate. In most cases, the default is true.

Overrides:
canEvaluate in class Problem

preprocessPopulation

public void preprocessPopulation(EvolutionState state,
                                 Population pop,
                                 boolean countVictoriesOnly)
Description copied from interface: GroupedProblemForm
Set up the population pop (such as fitness information) prior to evaluation. Although this method is not static, you should not use it to write to any instance variables in the GroupedProblem instance; this is because it's possible that the instance used is in fact the prototype, and you will have no guarantees that your instance variables will remain valid during the evaluate(...) process. Do not assume that pop will be the same as state.pop -- it may not. state is only provided to give you access to EvolutionState features. Typically you'd use this method to set the Fitness values of all Individuals to 0.

countVictoriesOnly will be set if Individuals' fitness is to be based on whether they're the winner of a test, instead of based on the specifics of the scores in the tests. This really only happens for Single-Elimination Tournament one-population competitive coevolution.

Specified by:
preprocessPopulation in interface GroupedProblemForm

postprocessPopulation

public void postprocessPopulation(EvolutionState state,
                                  Population pop,
                                  boolean countVictoriesOnly)
Description copied from interface: GroupedProblemForm
Finish processing the population (such as fitness information) after evaluation. Although this method is not static, you should not use it to write to any instance variables in the GroupedProblem instance; this is because it's possible that the instance used is in fact the prototype, and you will have no guarantees that your instance variables will remain valid during the evaluate(...) process. Do not assume that pop will be the same as state.pop -- it may not. state is only provided to give you access to EvolutionState features.

countVictoriesOnly will be set if Individuals' fitness is to be based on whether they're the winner of a test, instead of based on the specifics of the scores in the tests. This really only happens for Single-Elimination Tournament one-population competitive coevolution. If this is set, probably would leave the Fitnesses as they are here (they've been set and incremented in evaluate(...)), but if it's not set, you may want to set the Fitnesses to the maximum or average or the various trials performed.

Specified by:
postprocessPopulation in interface GroupedProblemForm

evaluate

public void evaluate(EvolutionState state,
                     Individual[] ind,
                     boolean[] updateFitness,
                     boolean countVictoriesOnly,
                     int[] subpops,
                     int threadnum)
Default version assumes that every individual is a GEIndividual. The underlying problem.evaluate() must be prepared for the possibility that some GPIndividuals handed it are in fact null, meaning that they couldn't be extracted from the GEIndividual string. You should assign them bad fitness in some appropriate way.

Specified by:
evaluate in interface GroupedProblemForm

evaluate

public void evaluate(EvolutionState state,
                     Individual ind,
                     int subpopulation,
                     int threadnum)
Description copied from interface: SimpleProblemForm
Evaluates the individual in ind, if necessary (perhaps not evaluating them if their evaluated flags are true), and sets their fitness appropriately.

Specified by:
evaluate in interface SimpleProblemForm

describe

public void describe(EvolutionState state,
                     Individual ind,
                     int subpopulation,
                     int threadnum,
                     int log)
Description copied from class: Problem
Part of SimpleProblemForm. Included here so you don't have to write the default version, which usually does nothing.

Specified by:
describe in interface SimpleProblemForm
Overrides:
describe in class Problem