Download Latest Version jenetics-4.1.0.zip (25.2 MB) Get Updates
Home
Name Modified Size InfoDownloads / Week
v4.3.0 2018-10-28
0
v4.2.1 2018-09-28
0
v4.2.0 2018-05-31
0
v4.1.0 2018-02-14
3 3 weekly downloads
v4.0.0 2017-11-16
0
README.md 2017-11-16 19.4 kB
1 1 weekly downloads
Totals: 6 Items   19.4 kB 4

Jenetics

Build Status Maven Central Javadoc

Jenetics is an Genetic Algorithm, Evolutionary Algorithm and Genetic Programming library, respectively, written in Java. It is designed with a clear separation of the several concepts of the algorithm, e.g. Gene, Chromosome, Genotype, Phenotype, Population and fitness Function. Jenetics allows you to minimize and maximize the given fitness function without tweaking it. In contrast to other GA implementations, the library uses the concept of an evolution stream (EvolutionStream) for executing the evolution steps. Since the EvolutionStream implements the Java Stream interface, it works smoothly with the rest of the Java Stream API.

Other languages

  • Jenetics.Net: Experimental .NET Core port in C# of the base library.

Documentation

The library is fully documented (javadoc) and comes with an user manual (pdf).

Requirements

Runtime

  • JRE 8: Java runtime version 8 is needed for using the library, respectively for running the examples.

Build time

  • JDK 8: The Java JDK 8 must be installed.
  • Gradle 4.x: Gradle is used for building the library. (Gradle is download automatically, if you are using the Gradle Wrapper script ./gradlew, located in the base directory, for building the library.)

Test compile/execution

  • TestNG 6.11: Jenetics uses TestNG framework for unit tests.
  • Apache Commons Math 3.6: Library is used for testing statistical collectors.

Build Jenetics

For building the Jenetics library from source, download the most recent, stable package version from Github (or Sourceforge) and extract it to some build directory.

$ unzip jenetics-<version>.zip -d <builddir>

<version> denotes the actual Jenetics version and <builddir> the actual build directory. Alternatively you can check out the master branch from Github.

$ git clone https://github.com/jenetics/jenetics.git <builddir>

Jenetics uses Gradle as build system and organizes the source into sub-projects (modules). Each sub-project is located in it’s own sub-directory:

Published projects

The following projects/modules are also published to Maven.

  • jenetics: This project contains the source code and tests for the Jenetics core-module.
  • jenetics.ext: This module contains additional non-standard GA operations and data types.
  • jenetics.prog: The modules contains classes which allows to do genetic programming (GP). It seamlessly works with the existing EvolutionStream and evolution Engine.
  • jenetics.xml: XML marshalling module for the Jenetics base data structures.

Non-published projects

  • jenetics.example: This project contains example code for the core-module.
  • jenetics.doc: Contains the code of the web-site and the manual.
  • jenetics.tool: This module contains classes used for doing integration testing and algorithmic performance testing. It is also used for creating GA performance measures and creating diagrams from the performance measures.

For building the library change into the <builddir> directory (or one of the module directory) and call one of the available tasks:

  • compileJava: Compiles the Jenetics sources and copies the class files to the <builddir>/<module-dir>/build/classes/main directory.
  • jar: Compiles the sources and creates the JAR files. The artifacts are copied to the <builddir>/<module-dir>/build/libs directory.
  • javadoc: Generates the API documentation. The Javadoc is stored in the <builddir>/<module-dir>/build/docs directory
  • test: Compiles and executes the unit tests. The test results are printed onto the console and a test-report, created by TestNG, is written to <builddir>/<module-dir> directory.
  • clean: Deletes the <builddir>/build/* directories and removes all generated artifacts.

For building the library jar from the source call

$ cd <build-dir>
$ ./gradlew jar

Example

Hello World (Ones counting)

The minimum evolution Engine setup needs a genotype factory, Factory<Genotype<?>>, and a fitness Function. The Genotype implements the Factory interface and can therefore be used as prototype for creating the initial Population and for creating new random Genotypes.

import io.jenetics.BitChromosome;
import io.jenetics.BitGene;
import io.jenetics.Genotype;
import io.jenetics.engine.Engine;
import io.jenetics.engine.EvolutionResult;
import io.jenetics.util.Factory;

public class HelloWorld {
    // 2.) Definition of the fitness function.
    private static Integer eval(Genotype<BitGene> gt) {
        return gt.getChromosome()
            .as(BitChromosome.class)
            .bitCount();
    }

    public static void main(String[] args) {
        // 1.) Define the genotype (factory) suitable
        //     for the problem.
        Factory<Genotype<BitGene>> gtf =
            Genotype.of(BitChromosome.of(10, 0.5));

        // 3.) Create the execution environment.
        Engine<BitGene, Integer> engine = Engine
            .builder(HelloWorld::eval, gtf)
            .build();

        // 4.) Start the execution (evolution) and
        //     collect the result.
        Genotype<BitGene> result = engine.stream()
            .limit(100)
            .collect(EvolutionResult.toBestGenotype());

        System.out.println("Hello World:\n" + result);
    }
}

In contrast to other GA implementations, the library uses the concept of an evolution stream (EvolutionStream) for executing the evolution steps. Since the EvolutionStream implements the Java Stream interface, it works smoothly with the rest of the Java streaming API. Now let's have a closer look at listing above and discuss this simple program step by step:

  1. The probably most challenging part, when setting up a new evolution Engine, is to transform the problem domain into a appropriate Genotype (factory) representation. In our example we want to count the number of ones of a BitChromosome. Since we are counting only the ones of one chromosome, we are adding only one BitChromosome to our Genotype. In general, the Genotype can be created with 1 to n chromosomes.

  2. Once this is done, the fitness function which should be maximized, can be defined. Utilizing the new language features introduced in Java 8, we simply write a private static method, which takes the genotype we defined and calculate it's fitness value. If we want to use the optimized bit-counting method, bitCount(), we have to cast the Chromosome<BitGene> class to the actual used BitChromosome class. Since we know for sure that we created the Genotype with a BitChromosome, this can be done safely. A reference to the eval method is then used as fitness function and passed to the Engine.build method.

  3. In the third step we are creating the evolution Engine, which is responsible for changing, respectively evolving, a given population. The Engine is highly configurable and takes parameters for controlling the evolutionary and the computational environment. For changing the evolutionary behavior, you can set different alterers and selectors. By changing the used Executor service, you control the number of threads, the Engine is allowed to use. An new Engine instance can only be created via its builder, which is created by calling the Engine.builder method.

  4. In the last step, we can create a new EvolutionStream from our Engine. The EvolutionStream is the model or view of the evolutionary process. It serves as a »process handle« and also allows you, among other things, to control the termination of the evolution. In our example, we simply truncate the stream after 100 generations. If you don't limit the stream, the EvolutionStream will not terminate and run forever. Since the EvolutionStream extends the java.util.stream.Stream interface, it integrates smoothly with the rest of the Java Stream API. The final result, the best Genotype in our example, is then collected with one of the predefined collectors of the EvolutionResult class.

Evolving images

This example tries to approximate a given image by semitransparent polygons. It comes with an Swing UI, where you can immediately start your own experiments. After compiling the sources with

$ ./gradlew compileTestJava

you can start the example by calling

$ ./jrun io.jenetics.example.image.EvolvingImages

Evolving images

The previous image shows the GUI after evolving the default image for about 4,000 generations. With the »Open« button it is possible to load other images for polygonization. The »Save« button allows to store polygonized images in PNG format to disk. At the button of the UI, you can change some of the GA parameters of the example.

Projects using Jenetics

  • Chronetic: Chronetic is an open-source time pattern analysis library built to describe time-series data.
  • APP4MC: Eclipse APP4MC is a platform for engineering embedded multi- and many-core software systems.

Citations

Blogs

License

The library is licensed under the Apache License, Version 2.0.

Copyright 2007-2017 Franz Wilhelmstötter

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

Release notes

4.0.0

Improvements

  • #28: Immutable population class. The original Population class has been replaced by Seq<Phenotype<G, C>>. This points to a more functional implementation of the library.
  • #119: Chromosome implementations are now fully immutable. This is an internal change only.
  • #121: Mutator class is easier now to extend. It has been extended with additional mutate methods which serves as extension points for onw Mutator implementations.
  • #123: Chromosome with variable number of genes: Most chromosomes can now be created with a variable number of genes. DoubleChromosome.of(0.0, 1.0, IntRange.of(5, 16)).
  • #172: io.jenetics.prngine library replaces the existing PRNG implementations in the io.jenetics.base module.
  • #175: Align random int range generation with io.jenetics.prngine library. This is an internal change only.
  • #180: Change library namespace from org.jenetics to io.jenetics. This is the most invasive change of this release. Users have to adopt the imports in all their code.
  • #183: Change copyright email address to ...@gmail.com
  • #200: Implementation of gene convergence termination: A termination method that stops the evolution when a user-specified percentage of the genes that make up a Genotype are deemed as converged. A gene is deemed as converged when the average value of that gene across all of the genotypes in the current population is less than a user-specified percentage away from the maximum gene value across the genotypes.
  • #253: Removal of deprecated code and classes: mainly JAXB marshalling and the LCG64ShiftRandom class.
  • #260: Clean room implementation of internal subset function. This method was a port from the C++ source written by John Burkardt. The original source has been published under the LGPL licence, which is not compatible to tha Apache 2 licence. To avoid legal issues, the affected method has been reimplemented using the Clean Room method, based on the original book, Combinatorial Algorithms for Computers and Calculators, by Albert Nijenhuis and Herbert Wilf. The io.jenetics.internal.math.comb.subset method is now fully compatible with the Apache 2 licence.
  • #262: Filter for duplicate individuals: It is now possible to intercept the stream of EvolutionResults of the evolution Engine:
final Engine<DoubleGene, Integer> engine = Engine.builder(problem)
    .mapping(EvolutionResult.toUniquePopulation())
    .build();
  • #264: Upgrade Gradle to version 4.3.
  • #266: The Seq serialization should be more robust in the case of implementation changes.
  • #268: Implementation of an EliteSelector.
  • #269: Cleanup of internal, mathematical helper functions.
  • #272: Obey Java naming convention. Two helper classes have been renamed to obey the Java naming conventions. codecs -> Codecs and limits -> Limits.
  • #279: Additional MSeq.swap method.

Bugs

  • #247: Fix the classpath of the jrun helper script.
  • #256: Buggy type signature of AnyChromosome.of method.

All Release Notes

Used software

IntelliJ

Source: README.md, updated 2017-11-16