Expand description
§evolution-ternary
Evolutionary dynamics on ternary strategy spaces.
Each agent carries a strategy vector over the alphabet {0, 1, 2}. Natural selection is sort-based fitness ranking, mutation is analogous to spreadsheet autofill, and the fitness landscape is visualised through conditional-format style heatmaps.
Re-exports§
pub use population::TernaryAgent;pub use population::TernaryPopulation;pub use fitness::FitnessFunction;pub use fitness::Environment;pub use selection::SelectionMethod;pub use selection::Selection;pub use population::SimpleRng;pub use mutation::Mutation;pub use mutation::CrossoverMethod;pub use mutation::autofill_mutate;pub use engine::EvolutionEngine;pub use engine::EvolutionConfig;pub use engine::EvolutionResult;pub use species::Species;pub use species::SpeciesTracker;pub use convergence::ConvergenceDetector;
Modules§
- convergence
- Convergence detection — fitness plateaus and species stabilization.
- engine
- Evolution engine — orchestrates selection, mutation, and generational loops.
- fitness
- Fitness functions and multi-environment support.
- mutation
- Mutation operators: point mutation, crossover (single-point, uniform).
- population
- Core population types — agents with ternary strategy vectors.
- selection
- Selection methods: tournament, roulette wheel, rank-based.
- species
- Species classification and ecology tracking.