All Packages  Class Hierarchy  This Package  Previous  Next  Index

Interface geneticAlgorithm.Functionable

public interface Functionable
Functionable interface defines what parameters and what algorithms each GA problem should supply.

Method Index

o abcSize()
The alpha-bet size.
o better(double, double)
Compares two fitnesses and returns true if the first one is better.
o calc(Ind)
Receives one individual and returns its fitness.
o crossOver(Ind, Ind)
Receives two individuals and returns a child.
o done()
Each problem specific finilizing steps.
o finished(double)
Returns true if achieved fitness is good enough.
o generations()
Number of generations counted by the main loop.
o indSize()
returns individuals size
o maxTrials()
Maximum number of non-fitness-improving trials that a remote machine should do before stop trying.
o mutate(Ind)
Mutates a given individuals.
o mutationRatio()
1/mutationRatio of a population would be mutated.
o popSize()
Total population size.
o start()
Each problem specific initialization steps.
o subGenerations()
Number of fitness improvment steps each remote machine should perform.
o subPopulations()
Number of subpopulations.

Methods

o start 
 public abstract void start()
Each problem specific initialization steps.

o done 
 public abstract void done()
Each problem specific finilizing steps.

o indSize 
 public abstract int indSize()
returns individuals size

o abcSize 
 public abstract int abcSize()
The alpha-bet size. Each element value would be in the range 0 - (abcSize - 1).

o popSize 
 public abstract int popSize()
Total population size. Number of individuals.

o generations 
 public abstract int generations()
Number of generations counted by the main loop. Sending subpopulation to a remote machine for several subgenerations would be counted as one generation in the main loop.

o subPopulations 
 public abstract int subPopulations()
Number of subpopulations. Each subpopulation will have popSize/subPopulation individuals.

o subGenerations 
 public abstract int subGenerations()
Number of fitness improvment steps each remote machine should perform.

o maxTrials 
 public abstract int maxTrials()
Maximum number of non-fitness-improving trials that a remote machine should do before stop trying.

o mutationRatio 
 public abstract int mutationRatio()
1/mutationRatio of a population would be mutated.

o finished 
 public abstract boolean finished(double fitness)
Returns true if achieved fitness is good enough.

o better 
 public abstract boolean better(double fitness1,
                                double fitness2)
Compares two fitnesses and returns true if the first one is better.

o calc 
 public abstract double calc(Ind ind)
Receives one individual and returns its fitness.

o crossOver 
 public abstract Ind crossOver(Ind oldInd,
                               Ind ind2)
Receives two individuals and returns a child.

o mutate 
 public abstract void mutate(Ind ind)
Mutates a given individuals. 

All Packages  Class Hierarchy  This Package  Previous  Next  Index