LSHADE

LSHADE (Improving the Search Performance of SHADE Using Linear Population Size Reduction)

class DETAlgs.lshade.LSHADE(params: LShadeData, db_conn=None, db_auto_write=False)

Bases: BaseAlg

LSHADE: Improving the Search Performance of SHADE Using Linear Population Size Reduction

Links: https://ieeexplore.ieee.org/document/6900380

References: Tanabe, R., & Fukunaga, A. S. (2014). Improving the search performance of SHADE using linear population size reduction. In 2014 IEEE Congress on Evolutionary Computation (CEC) (pp. 1658–1665). 2014 IEEE Congress on Evolutionary Computation (CEC). IEEE. https://doi.org/10.1109/cec.2014.6900380

initialize_parameters_for_epoch()

Initialize the parameters for the next epoch of the SHADE algorithm. f_table: List of scaling factors for mutation. cr_table: List of crossover rates. the_bests_to_select: List of the number of the best members to select because in crossover we need to select the best members from the population for one factor.

Returns: - f_table (List[float]): List of scaling factors for mutation. - cr_table (List[float]): List of crossover rates. - the_bests_to_select (List[int]): List of the number of the best members to possibly select.

mutate(population: Population, the_best_to_select_table: List[int], f_table: List[float]) → Population

Perform mutation step for the population in SHADE.

Parameters: - population (Population): The population to mutate. - the_best_to_select_table (List[int]): List of the number of the best members to select. - f_table (List[float]): List of scaling factors for mutation.

Returns: A new Population with mutated members.

next_epoch()

Perform the next epoch of the SHADE algorithm.

property nfe: int

Number of function evaluations performed so far.

run()

update_memory(success_f: List[float], success_cr: List[float], difference_fitness_success: List[float])

Update the memory for the crossover rates and scaling factors based on the success of the trial vectors.

Parameters: - success_f (List[float]): List of scaling factors that led to better trial vectors. - success_cr (List[float]): List of crossover rates that led to better trial vectors. - difference_fitness_success (List[float]): List of differences in objective function values (|f(u_k, G) - f(x_k, G)|).

update_population_size(nfe: int, total_nfe: int, start_pop_size: int, min_pop_size: int)

Calculate new population size using Linear Population Size Reduction (LPSR).

Parameters: - nfe (int): The current function evaluations. - total_nfe (int): The total number of function evaluations. - start_pop_size (int): The initial population size. - min_pop_size (int): The minimum population size.

write_results_to_database(results_data)

class detpy.DETAlgs.data.alg_data.LShadeData(population_size: int = 100, max_nfe: int = 100000, dimension: int = 10, additional_stop_criteria: detpy.models.stop_condition.stop_condition.StopCondition = <detpy.models.stop_condition.never_stop_condition.NeverStopCondition object at 0x7f136577ee40>, lb: list = <factory>, ub: list = <factory>, optimization_type: detpy.models.enums.optimization.OptimizationType = <OptimizationType.MINIMIZATION: 'minimization'>, boundary_constraints_fun: detpy.models.enums.boundary_constrain.BoundaryFixing = <BoundaryFixing.RANDOM: 'random'>, function: detpy.models.fitness_function.FitnessFunctionBase = None, log_population: bool = False, parallel_processing: Optional[list] = None, show_plots: bool = True, population_reduction_strategy: detpy.DETAlgs.population_reduction.population_size_reduction_strategy.PopulationSizeReductionStrategy = <detpy.DETAlgs.population_reduction.linear_population_size_reduction.LinearPopulationSizeReduction object at 0x7f136820a000>, minimum_population_size: int = 5, memory_size: int = 5, best_member_percentage: float = 0.2)

Bases: BaseData

additional_stop_criteria: StopCondition = <detpy.models.stop_condition.never_stop_condition.NeverStopCondition object>

best_member_percentage: float = 0.2

boundary_constraints_fun: BoundaryFixing = 'random'

dimension: int = 10

function: FitnessFunctionBase = None

lb: list

log_population: bool = False

max_nfe: int = 100000

memory_size: int = 5

minimum_population_size: int = 5

optimization_type: OptimizationType = 'minimization'

parallel_processing: list | None = None

population_reduction_strategy: PopulationSizeReductionStrategy = <detpy.DETAlgs.population_reduction.linear_population_size_reduction.LinearPopulationSizeReduction object>

population_size: int = 100

show_plots: bool = True

ub: list