xegaX. A Family of R-Packages for Genetic and Evolutionary Algorithms with Multiple Genome Representations

Created by W.Langdon from gp-bibliography.bib Revision:1.8816

@Article{AGeyer-Schulz2025,

• author = "Andreas Geyer-Schulz",
• title = "{xegaX}. A Family of {R}-Packages for Genetic and Evolutionary Algorithms with Multiple Genome Representations",
• journal = "Archives of Data Science, Series A",
• year = "2025",
• volume = "10",
• number = "1",
• pages = "1--37",
• month = nov,
• keywords = "genetic algorithms, genetic programming, MPI, parallel computing",
• ISSN = "2363-9881",
• publisher = "KIT Scientific Publishing",
• DOI = "10.5445/IR/1000187255",
• size = "37 pages",
• abstract = "xegaX is a family of R-packages for genetic and evolutionary algorithms with multiple gene representations. At the moment, the following gene representations are supported: Binary genes, integer genes, real genes, and derivation tree genes. The package provides a common framework for genetic algorithms with binary genes (sga), genetic differential evolution algorithms (sgde), genetic algorithms with integer permutations (sgPerm), grammar-based genetic programming algorithms (sgp), and grammatical evolution algorithms (sge). The packages have a layered architecture with 4 layers: The (top-level) main program layer, the population layer which is independent of the gene representation, the gene layer which is split into gene representation dependent (initialization, crossover, mutation, and decoding) and gene representation independent (selection, evaluation) components. In addition, several innovations have been integrated into the package with the aim to improve several architectural goals simultaneously: Increased flexibility, configurability, and extensibility combined with performance improvements and scalability. For example, extensive support for parallel and distributed processing has been added: Multi-core processing on notebooks (Linux only), distributed processing on clusters of servers on a local area network (for security reasons), parallel processing on high-performance processor clusters based on rmpi. This paper will give an architectural overview of the packages as well as a description of selected innovations and their impact on the architectural goals.",

}

Genetic Programming entries for Andreas Geyer-Schulz

Citations