The objective of this practical work was to implement a genetic algorithm which calculates the best solution to a problem, in our case the problem was determinate the best build of a RPG character.
For more details, statement of the practical work can be found at doc/statement/Sistemas_de_Inteligencia_Artificial__TPE3.pdf.
git clone git@bitbucket.org:itba/sia-2016-08.git
cd sia-2016-08
You need to install maven and then build the jar.
sudo apt-get -y install maven
mvn clean package
Move the jar where you want and copy the tsv and txt files (configuration files, multipliers, heights and items).
cp target/ai_systems_genetic_algorithm-0.0.1-SNAPSHOT-jar-with-dependencies.jar ~/
cp doc/data/*.txt ~/
cp doc/data/*.tsv ~/
The first argument indicates the directory where the tsv and txt files are. The default value is ./ if you are running a jar or doc/data/ if you are running in eclipse.
cd ~
java -jar ./ai_systems_genetic_algorithm-0.0.1-SNAPSHOT-jar-with-dependencies.jar
In the configuration file you can indicate the parameters for the algorithm. A example configuration file is added in doc/data/configuration.txt and the default configuration will be in doc/data/defaultConfiguration.txt. In case a parameter is missing, it will be loaded from the default configuration file.
The implementation approach of this problem is generic, so you can change the character multiplaiers from the doc/data/multipliers.txt file.