Package ec.steadystate

Class SteadyStateEvaluator

java.lang.Object

ec.Evaluator

ec.simple.SimpleEvaluator

ec.steadystate.SteadyStateEvaluator

All Implemented Interfaces:

Setup, Singleton, Serializable

public class SteadyStateEvaluator
extends SimpleEvaluator

This subclass of Evaluator performs the evaluation portion of Steady-State Evolution and (in distributed form) Asynchronous Evolution. The procedure is as follows. We begin with an empty Population and one by one create new Indivdiuals and send them off to be evaluated. In basic Steady-State Evolution the individuals are immediately evaluated and we wait for them; but in Asynchronous Evolution the individuals are evaluated for however long it takes and we don't wait for them to finish. When individuals return they are added to the Population until it is full. No duplicate individuals are allowed.

At this point the system switches to its "steady state": individuals are bred from the population one by one, and sent off to be evaluated. Once again, in basic Steady-State Evolution the individuals are immediately evaluated and we wait for them; but in Asynchronous Evolution the individuals are evaluated for however long it takes and we don't wait for them to finish. When an individual returns, we mark an individual in the Population for death, then replace it with the new returning individual. Note that during the steady-state, Asynchronous Evolution could be still sending back some "new" individuals created during the initialization phase, not "bred" individuals.

The determination of how an individual is marked for death is done by the SteadyStateBreeder.

When SteadyStateEvaluator sends indivduals off to be evaluated, it stores them in an internal queue, along with the subpopulation in which they were destined. This tuple is defined by QueueIndividual.java

See Also:

• Serialized Form

Field Summary

Fields inherited from class ec.simple.SimpleEvaluator

C_AUTO, cloneProblem, MERGE_BEST, MERGE_MEAN, MERGE_MEDIAN, mergeForm, numTests, P_CHUNK_SIZE, P_CLONE_PROBLEM, P_MERGE, P_NUM_TESTS, pool, V_AUTO, V_BEST, V_MEAN, V_MEDIAN

Fields inherited from class ec.Evaluator

masterproblem, P_IAMSLAVE, P_MASTERPROBLEM, p_problem, P_PROBLEM, runComplete

Constructor Summary

Constructors

Constructor	Description
SteadyStateEvaluator()

Method Summary

Modifier and Type	Method	Description
boolean	canEvaluate()	Returns true if we're ready to evaluate an individual.
void	evaluateIndividual(EvolutionState state, Individual ind, int subpop)	Submits an individual to be evaluated by the Problem, and adds it and its subpopulation to the queue.
Individual	getNextEvaluatedIndividual(EvolutionState state)	Returns an evaluated individual is in the queue and ready to come back to us.
int	getSubpopulationOfEvaluatedIndividual()	Returns the subpopulation of the last evaluated individual returned by getNextEvaluatedIndividual, or potentially -1 if getNextEvaluatedIndividual was never called or hasn't returned an individual yet.
boolean	isIdealFitness(EvolutionState state, Individual ind)	The SimpleEvaluator determines that a run is complete by asking each individual in each population if he's optimal; if he finds an individual somewhere that's optimal, he signals that the run is complete.
void	prepareToEvaluate(EvolutionState state, int thread)
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.simple.SimpleEvaluator

evalPopChunk, evaluatePopulation, runComplete

Methods inherited from class ec.Evaluator

closeContacts, initializeContacts, postEvaluationGlobalUpdate, postEvaluationLocalUpdate, reinitializeContacts, setRunComplete

Methods inherited from class java.lang.Object

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

Constructor Details

SteadyStateEvaluator

public SteadyStateEvaluator()

Method Details

setup

public void setup(EvolutionState state,
 Parameter base)

Description copied from interface: Setup

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.

Specified by:

setup in interface Setup

Overrides:

setup in class SimpleEvaluator

prepareToEvaluate

public void prepareToEvaluate(EvolutionState state,
 int thread)

evaluateIndividual

public void evaluateIndividual(EvolutionState state,
 Individual ind,
 int subpop)

Submits an individual to be evaluated by the Problem, and adds it and its subpopulation to the queue.

canEvaluate

public boolean canEvaluate()

Returns true if we're ready to evaluate an individual. Ordinarily this is ALWAYS true, except in the asynchronous evolution situation, where we may not have a processor ready yet.

getNextEvaluatedIndividual

public Individual getNextEvaluatedIndividual(EvolutionState state)

Returns an evaluated individual is in the queue and ready to come back to us. Ordinarily this is ALWAYS true at the point that we call it, except in the asynchronous evolution situation, where we may not have a job completed yet, in which case NULL is returned. Once an individual is returned by this function, no other individual will be returned until the system is ready to provide us with another one. NULL will be returned otherwise.

getSubpopulationOfEvaluatedIndividual

public int getSubpopulationOfEvaluatedIndividual()

Returns the subpopulation of the last evaluated individual returned by getNextEvaluatedIndividual, or potentially -1 if getNextEvaluatedIndividual was never called or hasn't returned an individual yet.

isIdealFitness

public boolean isIdealFitness(EvolutionState state,
 Individual ind)

The SimpleEvaluator determines that a run is complete by asking each individual in each population if he's optimal; if he finds an individual somewhere that's optimal, he signals that the run is complete.