from detpy.DETAlgs.base import BaseAlg
from detpy.DETAlgs.data.alg_data import JADEData
from detpy.DETAlgs.methods.methods_jade import jade_adapt_mutation_factors, jade_binomial_crossing, \
jade_adapt_crossover_rates, jade_mutation, jade_selection, jade_reduce_archive, draw_norm_dist_within_bounds, \
draw_cauchy_dist_within_bounds
from detpy.models.enums.boundary_constrain import fix_boundary_constraints
[docs]
class JADE(BaseAlg):
"""
JADE
Links:
https://ieeexplore.ieee.org/document/5208221
References:
J. Zhang and A. C. Sanderson, "JADE: Adaptive Differential Evolution With Optional External Archive,"
in IEEE Transactions on Evolutionary Computation, vol. 13, no. 5, pp. 945-958, Oct. 2009,
doi: 10.1109/TEVC.2009.2014613.
"""
def __init__(self, params: JADEData, db_conn=None, db_auto_write=False):
super().__init__(JADE.__name__, params, db_conn, db_auto_write)
self.archive_size = params.archive_size
self.archive = []
self.mutation_factor_mean = params.mutation_factor_mean
self.mutation_factor_std = params.mutation_factor_std
self.mutation_factors = draw_cauchy_dist_within_bounds(self.mutation_factor_mean, self.mutation_factor_std,
self.population_size)
self.success_mutation_factors = []
self.crossover_rate_mean = params.crossover_rate_mean
self.crossover_rate_std = params.crossover_rate_std
self.crossover_rate_low = params.crossover_rate_low
self.crossover_rate_high = params.crossover_rate_high
self.crossover_rates = draw_norm_dist_within_bounds(self.crossover_rate_mean, self.crossover_rate_std,
self.crossover_rate_low, self.crossover_rate_high,
self.population_size)
self.success_crossover_rates = []
self.c = params.c
self.p = params.p
[docs]
def next_epoch(self):
# New population after mutation
v_pop = jade_mutation(self._pop, self.mutation_factors, self.p, self.archive)
# Apply boundary constrains on population in place
fix_boundary_constraints(v_pop, self.boundary_constraints_fun)
# New population after crossing
u_pop = jade_binomial_crossing(self._pop, v_pop, self.crossover_rates)
# Update values before selection
u_pop.update_fitness_values(self._function.eval, self.parallel_processing)
# Select new population
new_pop = jade_selection(self._pop, u_pop, self.mutation_factors, self.crossover_rates,
self.success_mutation_factors, self.success_crossover_rates, self.archive)
jade_reduce_archive(self.archive, self.archive_size)
self.mutation_factors, self.mutation_factor_mean = jade_adapt_mutation_factors(self.c,
self.mutation_factor_mean,
self.mutation_factor_std,
self.population_size,
self.success_mutation_factors)
self.crossover_rates, self.crossover_rate_mean = jade_adapt_crossover_rates(self.c, self.crossover_rate_mean,
self.crossover_rate_std,
self.crossover_rate_low,
self.crossover_rate_high,
self.population_size,
self.success_crossover_rates)
# Override data
self._pop = new_pop
self._epoch_number += 1