JGAP (pronounced "jay-gap") is a genetic algorithms package written in Java. It is designed to require minimum effort to use "out of the box", but is highly modular and allows custom components to be easily plugged in by the more adventurous.
|Tags||Scientific/Engineering Artificial Intelligence Software Development Libraries Java Libraries|
|Operating Systems||OS Independent|
Release Notes: This release adds the Mona Lisa painting problem (GA and GP version). It adds some unit tests. CommandGene and ProgramChromosome have support for dynamization of the arity of nodes. GPConfiguration has a parameter for dynamization of nodes' arities. GPPopulation has enhanced creation of new programs. ProgramChromosome and CommandGene have new versions of the method ensureUniqueness. ProgramChromosome and GPProgram have the interface IBusinessKey (method getBusinessKey). GP-command terminal now supports primitive class types. A bug in ForLoop.clone has been fixed.
Release Notes: Greatly extended GP documentation. 12 new GP functions. Improved error handling in JGAPClientGP. Strongly improved grid computing functionality. A painted desert example (package examples.gp). SystemKit.printHelp(..) has been added. DefaultClientFeedback has been introduced. A bug in NumberKit.niceDecimalNumber, a bug in GP terminal False, and a bug in ColtRandomGenerator.nextInt() have been fixed. Test cases have been added.
Release Notes: The evolution cycle has been revamped and simplified for a better understanding. There are many smaller enhancements and architectural improvements, as well as some bugfixes. The MinimizingMakeChange example was simplified. Grid computing was improved. There are Javadoc enhancements and new JUnit tests.
Release Notes: This release introduced mutation of Genetic Programming commands, uses the XStream library for serialization issues, uses trove4J as a fast and consistent map replacement, integrated Robocode 1.5, and added cloning capabilities to many RobocodeJGAP commands. The fitness value of an original chromosome is set to clone when cloning the chromosome. Coning of application data in genes was enabled. The IComplexCommand class was made serializable. The output of generation number in GPGenotype.evolve() was fixed. All auxiliary test classes were renamed from xxxForTest to xxxForTesting.
Release Notes: Robocode 1.4.8 has been integrated. GP capabilities have been enhanced. Several bugs have been fixed. Logging of GP information has been enhanced. Javadoc has been enhanced. Some unit tests have been added.