#!/usr/bin/env python
# Created by "Thieu" at 00:08, 27/10/2022 ----------%
# Email: nguyenthieu2102@gmail.com %
# Github: https://github.com/thieu1995 %
# --------------------------------------------------%
import numpy as np
from mealpy.optimizer import Optimizer
[docs]class OriginalAGTO(Optimizer):
"""
The original version of: Artificial Gorilla Troops Optimization (AGTO)
Links:
1. https://doi.org/10.1002/int.22535
2. https://www.mathworks.com/matlabcentral/fileexchange/95953-artificial-gorilla-troops-optimizer
Notes (parameters):
1. p1 (float): the probability of transition in exploration phase (p in the paper), default = 0.03
2. p2 (float): the probability of transition in exploitation phase (w in the paper), default = 0.8
3. beta (float): coefficient in updating equation, should be in [-5.0, 5.0], default = 3.0
Examples
~~~~~~~~
>>> import numpy as np
>>> from mealpy import FloatVar, AGTO
>>>
>>> def objective_function(solution):
>>> return np.sum(solution**2)
>>>
>>> problem_dict = {
>>> "bounds": FloatVar(n_vars=30, lb=(-10.,) * 30, ub=(10.,) * 30, name="delta"),
>>> "obj_func": objective_function,
>>> "minmax": "min",
>>> }
>>>
>>> model = AGTO.OriginalAGTO(epoch=1000, pop_size=50, p1=0.03, p2=0.8, beta=3.0)
>>> g_best = model.solve(problem_dict)
>>> print(f"Solution: {g_best.solution}, Fitness: {g_best.target.fitness}")
>>> print(f"Solution: {model.g_best.solution}, Fitness: {model.g_best.target.fitness}")
References
~~~~~~~~~~
[1] Abdollahzadeh, B., Soleimanian Gharehchopogh, F., & Mirjalili, S. (2021). Artificial gorilla troops optimizer: a new
nature‐inspired metaheuristic algorithm for global optimization problems. International Journal of Intelligent Systems, 36(10), 5887-5958.
"""
def __init__(self, epoch: int = 10000, pop_size: int = 100, p1: float = 0.03, p2: float = 0.8, beta: float = 3.0, **kwargs: object) -> None:
"""
Args:
epoch (int): maximum number of iterations, default = 10000
pop_size (int): number of population size, default = 100
"""
super().__init__(**kwargs)
self.epoch = self.validator.check_int("epoch", epoch, [1, 100000])
self.pop_size = self.validator.check_int("pop_size", pop_size, [5, 10000])
self.p1 = self.validator.check_float("p1", p1, (0, 1)) # p in the paper
self.p2 = self.validator.check_float("p2", p2, (0, 1)) # w in the paper
self.beta = self.validator.check_float("beta", beta, [-10.0, 10.0])
self.set_parameters(["epoch", "pop_size", "p1", "p2", "beta"])
self.sort_flag = False
[docs] def evolve(self, epoch):
"""
The main operations (equations) of algorithm. Inherit from Optimizer class
Args:
epoch (int): The current iteration
"""
a = (np.cos(2*self.generator.random())+1) * (1 - epoch/self.epoch)
c = a * (2 * self.generator.random() - 1)
## Exploration
pop_new = []
for idx in range(0, self.pop_size):
if self.generator.random() < self.p1:
pos_new = self.problem.generate_solution()
else:
if self.generator.random() >= 0.5:
z = self.generator.uniform(-a, a, self.problem.n_dims)
rand_idx = self.generator.integers(0, self.pop_size)
pos_new = (self.generator.random() - a) * self.pop[rand_idx].solution + c * z * self.pop[idx].solution
else:
id1, id2 = self.generator.choice(list(set(range(0, self.pop_size)) - {idx}), 2, replace=False)
pos_new = self.pop[idx].solution - c*(c*self.pop[idx].solution - self.pop[id1].solution) + \
self.generator.random() * (self.pop[idx].solution - self.pop[id2].solution)
pos_new = self.correct_solution(pos_new)
agent = self.generate_empty_agent(pos_new)
pop_new.append(agent)
if self.mode not in self.AVAILABLE_MODES:
agent.target = self.get_target(pos_new)
self.pop[idx] = self.get_better_agent(agent, self.pop[idx], self.problem.minmax)
if self.mode in self.AVAILABLE_MODES:
pop_new = self.update_target_for_population(pop_new)
self.pop = self.greedy_selection_population(self.pop, pop_new, self.problem.minmax)
_, self.g_best = self.update_global_best_agent(self.pop, save=False)
pos_list = np.array([agent.solution for agent in self.pop])
## Exploitation
pop_new = []
for idx in range(0, self.pop_size):
if a >= self.p2:
g = 2 ** c
delta = (np.abs(np.mean(pos_list, axis=0)) ** g) ** (1.0 / g)
pos_new = c*delta*(self.pop[idx].solution - self.g_best.solution) + self.pop[idx].solution
else:
if self.generator.random() >= 0.5:
h = self.generator.normal(0, 1, self.problem.n_dims)
else:
h = self.generator.normal(0, 1)
r1 = self.generator.random()
pos_new = self.g_best.solution - (2*r1-1)*(self.g_best.solution - self.pop[idx].solution) * (self.beta * h)
pos_new = self.correct_solution(pos_new)
agent = self.generate_empty_agent(pos_new)
pop_new.append(agent)
if self.mode not in self.AVAILABLE_MODES:
agent.target = self.get_target(pos_new)
self.pop[idx] = self.get_better_agent(agent, self.pop[idx], self.problem.minmax)
if self.mode in self.AVAILABLE_MODES:
pop_new = self.update_target_for_population(pop_new)
self.pop = self.greedy_selection_population(self.pop, pop_new, self.problem.minmax)
[docs]class MGTO(Optimizer):
"""
The original version of: Modified Gorilla Troops Optimization (mGTO)
Notes (parameters):
1. pp (float): the probability of transition in exploration phase (p in the paper), default = 0.03
Examples
~~~~~~~~
>>> import numpy as np
>>> from mealpy import FloatVar, AGTO
>>>
>>> def objective_function(solution):
>>> return np.sum(solution**2)
>>>
>>> problem_dict = {
>>> "bounds": FloatVar(n_vars=30, lb=(-10.,) * 30, ub=(10.,) * 30, name="delta"),
>>> "obj_func": objective_function,
>>> "minmax": "min",
>>> }
>>>
>>> model = AGTO.MGTO(epoch=1000, pop_size=50, pp=0.03)
>>> g_best = model.solve(problem_dict)
>>> print(f"Solution: {g_best.solution}, Fitness: {g_best.target.fitness}")
>>> print(f"Solution: {model.g_best.solution}, Fitness: {model.g_best.target.fitness}")
References
~~~~~~~~~~
[1] Mostafa, R. R., Gaheen, M. A., Abd ElAziz, M., Al-Betar, M. A., & Ewees, A. A. (2023). An improved gorilla
troops optimizer for global optimization problems and feature selection. Knowledge-Based Systems, 110462.
"""
def __init__(self, epoch: int = 10000, pop_size: int = 100, pp: float = 0.03, **kwargs: object) -> None:
"""
Args:
epoch (int): maximum number of iterations, default = 10000
pop_size (int): number of population size, default = 100
pp (float): the probability of transition in exploration phase (p in the paper), default = 0.03
"""
super().__init__(**kwargs)
self.epoch = self.validator.check_int("epoch", epoch, [1, 100000])
self.pop_size = self.validator.check_int("pop_size", pop_size, [5, 10000])
self.pp = self.validator.check_float("p1", pp, (0, 1)) # p in the paper
self.set_parameters(["epoch", "pop_size", "pp"])
self.sort_flag = False
[docs] def amend_solution(self, solution: np.ndarray) -> np.ndarray:
condition = np.logical_and(self.problem.lb <= solution, solution <= self.problem.ub)
random_pos = self.generator.uniform(self.problem.lb, self.problem.ub)
return np.where(condition, solution, random_pos)
[docs] def evolve(self, epoch):
"""
The main operations (equations) of algorithm. Inherit from Optimizer class
Args:
epoch (int): The current iteration
"""
F = 1 + np.cos(2 * self.generator.random())
C = F * (1 - epoch / self.epoch)
L = C * self.generator.choice([-1, 1])
## Elite opposition-based learning
pos_list = np.array([agent.solution for agent in self.pop])
d_lb, d_ub = np.min(pos_list, axis=0), np.max(pos_list, axis=0)
pos_list = d_lb + d_ub - pos_list
pop_new = []
for idx in range(0, self.pop_size):
pos_new = self.correct_solution(pos_list[idx])
agent = self.generate_empty_agent(pos_new)
pop_new.append(agent)
if self.mode not in self.AVAILABLE_MODES:
pop_new[-1].target = self.get_target(pos_new)
if self.mode in self.AVAILABLE_MODES:
pop_new = self.update_target_for_population(pop_new)
self.pop = pop_new
_, self.g_best = self.update_global_best_agent(self.pop, save=False)
## Exploration
pop_new = []
for idx in range(0, self.pop_size):
if self.generator.random() < self.pp:
pos_new = self.problem.generate_solution()
else:
if self.generator.random() >= 0.5:
rand_idx = self.generator.integers(0, self.pop_size)
pos_new = (self.generator.random() - C) * self.pop[rand_idx].solution + L * self.generator.uniform(-C, C) * self.pop[idx].solution
else:
id1, id2 = self.generator.choice(list(set(range(0, self.pop_size)) - {idx}), 2, replace=False)
pos_new = self.pop[idx].solution - L*(L*self.pop[idx].solution - self.pop[id1].solution) + \
self.generator.random() * (self.pop[idx].solution - self.pop[id2].solution)
pos_new = self.correct_solution(pos_new)
agent = self.generate_empty_agent(pos_new)
pop_new.append(agent)
if self.mode not in self.AVAILABLE_MODES:
agent.target = self.get_target(pos_new)
self.pop[idx] = self.get_better_agent(agent, self.pop[idx], self.problem.minmax)
if self.mode in self.AVAILABLE_MODES:
pop_new = self.update_target_for_population(pop_new)
self.pop = self.greedy_selection_population(self.pop, pop_new, self.problem.minmax)
_, self.g_best = self.update_global_best_agent(self.pop, save=False)
pos_list = np.array([agent.solution for agent in self.pop])
## Exploitation
pop_new = []
for idx in range(0, self.pop_size):
if np.abs(C) >= 1:
g = self.generator.choice([-0.5, 2.])
M = (np.abs(np.mean(pos_list, axis=0)) ** g) ** (1.0 / g)
# print(M)
p = self.generator.uniform(0, 1, self.problem.n_dims)
pos_new = L * M * (self.pop[idx].solution - self.g_best.solution) * (0.01 * np.tan(np.pi*( p - 0.5)))
else:
Q = 2 * self.generator.random() - 1
v = self.generator.uniform(0, 1)
pos_new = self.g_best.solution - Q * (self.g_best.solution - self.pop[idx].solution) * np.tan(v * np.pi/2)
pos_new = self.correct_solution(pos_new)
agent = self.generate_empty_agent(pos_new)
pop_new.append(agent)
if self.mode not in self.AVAILABLE_MODES:
agent.target = self.get_target(pos_new)
self.pop[idx] = self.get_better_agent(agent, self.pop[idx], self.problem.minmax)
if self.mode in self.AVAILABLE_MODES:
pop_new = self.update_target_for_population(pop_new)
self.pop = self.greedy_selection_population(self.pop, pop_new, self.problem.minmax)