pydruglogics.model.Evolution

Optimizer that evolves Boolean models using genetic algorithms, fitness evaluation, and parallelization.

Evolution.calculate_fitness(ga_instance, solutions, solution_idx)

Fitness function for GA, evaluates each candidate model using the provided training data.

Parameters

• ga_instance: Instance of the GA.

• solutions (list): Batch of solutions (binary vectors).

• solution_idx (int): Current solution batch index.

Returns - list: Fitness values for each solution.

Evolution.create_initial_population(population_size, num_mutations, seed=None)

Create an initial GA population by randomly mutating the reference Boolean model.

Parameters

• population_size (int): Population size.

• num_mutations (int): Number of mutations per individual.

• seed (int, optional): Seed for reproducibility.

Returns

• list: Initial population (binary vectors).

Evolution.run()

Launch the evolutionary search, returning the best Boolean models found.

Returns

• list: Best Boolean models.

Evolution.save_to_file_models(base_folder='./results/models')

Save the best models of the evolution run as .bnet files to disk.

Parameters

• base_folder (str, optional): Output directory.