Open BEAGLE

Open BEAGLE is a C Evolutionary Computation (EC) framework. It provides an high-level software environment to do any kind of EC, with support for tree-based genetic programming; bit string, integer-valued vector, and real-valued vector genetic algorithms; and evolution strategy.

The Open BEAGLE architecture follows strong principles of object oriented programming, where abstractions are represented by loosely coupled objects and where it is common and easy to reuse code. Open BEAGLE is designed to provide an EC environment that is generic, user friendly, portable, efficient and robust.

Features:

Object-Oriented Foundations:
- Modern C programming approach
- Structured OO architecture
- Smart pointers for automatic memory allocation management
- XML file formats with built-in parsing facility
- Sophisticated logging mechanism with output in XML
- Parameters and algorithms dynamically configurable by files
- Generic representation of the algorithms using a plug-in mechanism
- Milestone mechanism for evolution recovery and results analysis
Open BEAGLE generic EA framework:
- Many predefined operators
- Generational, steady-state, (mu,lambda), and (mu lambda) replacement strategies
- Population composed of multiple demes
- Individuals represented on multiple genomes
- History of best-of-run individuals for the whole population and for each demes
- Complete evolution statistics
- Multiobjective optimization (currently NSGA-II and NPGA2)
- Population seeding from file
GA framework (linear representations):
- Bit string GA representation with decoder functions (binary and Gray-coded)
- Integer-valued vector GA representation
- Real-valued vector GA representation
- Evolution strategy (ES)
- Covariance Matrix Adaptation ES (CMA-ES)
- Three generic crossover operators (one-point, two-points, uniform) and two float vector specific operators (BLX-alpha and SBX)
- Four specific mutation operators
- Operators for shuffled indices genotypes for the integer-valued vector representation
- Seven illustrative examples (OneMax, ZeroMin (OneMax minimization), Function Maximization with Bit String GA, Function Maximization with Real-Valued GA, Function Maximization with ES, Multiobjective 0/1 Knapsack, Travelling Salesman Problem)
GP framework:
- Standard crossover operator
- Five mutation operators: standard (Koza GP I) mutation, swap node mutation, shrink mutation, swap subtree mutation, and random ephemerals value mutation
- Three initializations method for trees: full, grow, and half-and-half; each ramped or not
- Abstract primitive class
- Many predefined primitives
- Automatically defined functions (ADF)
- Random ephemeral constants
- Constrained GP tree operators with support for strongly-typed genetic programming
- Three illustrative examples (Symbolic Regression, Even 6-Parity, and Spambase)
Co-evolution framework:
- Co-evolution support based on multi-threading
- Multi-threading classes incapsulating OS-specific calls
- Co-evolutionary fitness evaluation operator for basic EC and GP
- Two illustrative examples (Two-Populations Iterated Prisoner's Dilemma, Co-evolutionary Symbolic Regression)