Genetic Algorithm Solver Module

This module provides a framework for solving optimization problems using genetic algorithms (GA). It was developed for the PW3/TP3 as part of the profesionnal programming course run by Ms T. Drumond.

1. Classes

Individual

Represents an individual in the GA population.

Methods:

• init(chromosome: list, fitness: float): Initializes an individual.
• lt(other): Comparator based on fitness.
• repr(): String representation.

GAProblem

Defines a GA problem. Users need to implement the functions inside this class for their specific problems.

Methods:

• single_chromosome_initialization(): Initialize a chromosome.
• computeFiness(chromosome): Compute fitness.
• reproduction(population): Reproduction method.
• mutation(new_chrom, mutation_rate): Mutation method.

GASolver

Implements the GA solving process with generic functions.

Methods:

• init(problem: GAProblem, selection_rate=0.5, mutation_rate=0.1): Initializes the solver.
• reset_population(pop_size=5000): Initializes the population.
• selection(): Selection process.
• evolve_for_one_generation(): Evolution for one generation.
• get_best_individual(): Returns the best individual.
• show_generation_summary(): Prints population summary.
• evolve_until(max_nb_of_generations=500, threshold_fitness=None): Evolution until termination

2. Usage

• Define problem by subclassing GAProblem.
• Implement functions in GAProblem for a given problem.
• Instantiate problem.
• Create a solver instance with the problem.
• Customize solver parameters.
• Reset population.
• Evolve population.

3. Example Implementations

Two example implementations are provided for guidance in the examples folder:

mastermind_problem.py: Solves the Mastermind game using the module. tsp_problem.py: Solves the Traveling Salesman Problem using the module.