evSOLve.JEvolution
Class TournamentSelection
java.lang.Object
|
+--evSOLve.JEvolution.TournamentSelection
- All Implemented Interfaces:
- Selection
- public class TournamentSelection
- extends java.lang.Object
- implements Selection
Tournament Selection draws a number of Individuals randomly from the Population
and conducts a tournament between the Individuals in the group where only the
winner survives. Each Individual takes part in 'tournamentSize' tournaments which
ensures that the size of the offspring Population is identical to the size of the
parent Population. Before each new round of tournaments Indiviudals are shuffled within
the Population so as to vary the participants in a specific tournament. In each round of
tournaments each Individual takes part in exactly one tournament.
- Version:
- 0.4
- Author:
- Helmut A. Mayer
Constructor Summary |
TournamentSelection()
Default constructor for binary tournament selection. |
TournamentSelection(int tournamentSize)
Constructor for tournament selection with a given tournament size. |
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
TournamentSelection
public TournamentSelection()
- Default constructor for binary tournament selection.
TournamentSelection
public TournamentSelection(int tournamentSize)
- Constructor for tournament selection with a given tournament size.
- Parameters:
tournamentSize
- number of Individuals in a single tournament
getTournamentSize
public int getTournamentSize()
- Gives the tournament size tournament size.
setTournamentSize
public void setTournamentSize(int tournamentSize)
throws JEvolutionException
- Sets tournament size.
- Parameters:
tournamentSize
- number of Individuals in a single tournament- Throws:
JEvolutionException
- if size < 2
doSelect
public void doSelect(Population parents,
Population offspring)
- JEvolution native and default (binary tournaments) selection method. Works for
arbitrary tournament and population sizes.
- Specified by:
doSelect
in interface Selection
- Parameters:
parents
- the Population under pressureoffspring
- the Population of survivors