#!/usr/bin/env python
# Created by "Thieu" at 22:24, 02/03/2022 ----------%
# Email: nguyenthieu2102@gmail.com %
# Github: https://github.com/thieu1995 %
# --------------------------------------------------%
import numpy as np
from copy import deepcopy
from mealpy.optimizer import Optimizer
[docs]class OriginalCGO(Optimizer):
"""
The original version of: Chaos Game Optimization (CGO)
Links:
1. https://doi.org/10.1007/s10462-020-09867-w
Notes
~~~~~
+ 4th seed is mutation process, but it is not clear mutation on multiple variables or 1 variable
+ There is no usage of the variable alpha 4th in the paper
+ The replacement of worst solutions by generated seed are not clear (Lots of grammar errors in this section)
Examples
~~~~~~~~
>>> import numpy as np
>>> from mealpy.math_based.CGO import OriginalCGO
>>>
>>> def fitness_function(solution):
>>> return np.sum(solution**2)
>>>
>>> problem_dict1 = {
>>> "fit_func": fitness_function,
>>> "lb": [-10, -15, -4, -2, -8],
>>> "ub": [10, 15, 12, 8, 20],
>>> "minmax": "min",
>>> "verbose": True,
>>> }
>>>
>>> epoch = 1000
>>> pop_size = 50
>>> model = OriginalCGO(problem_dict1, epoch, pop_size)
>>> best_position, best_fitness = model.solve()
>>> print(f"Solution: {best_position}, Fitness: {best_fitness}")
References
~~~~~~~~~~
[1] Talatahari, S. and Azizi, M., 2021. Chaos Game Optimization: a novel metaheuristic algorithm.
Artificial Intelligence Review, 54(2), pp.917-1004.
"""
def __init__(self, problem, epoch=10000, pop_size=100, **kwargs):
"""
Args:
problem (dict): The problem dictionary
epoch (int): maximum number of iterations, default = 10000
pop_size (int): number of population size, default = 100
"""
super().__init__(problem, kwargs)
self.nfe_per_epoch = 4*pop_size
self.sort_flag = False
self.epoch = epoch
self.pop_size = pop_size
[docs] def evolve(self, epoch):
"""
The main operations (equations) of algorithm. Inherit from Optimizer class
Args:
epoch (int): The current iteration
"""
pop_new = []
for idx in range(0, self.pop_size):
s1, s2, s3 = np.random.choice(range(0, self.pop_size), 3, replace=False)
MG = (self.pop[s1][self.ID_POS] + self.pop[s2][self.ID_POS] + self.pop[s3][self.ID_POS]) / 3
## Calculating alpha based on Eq. 7
alpha1 = np.random.rand()
alpha2 = 2 * np.random.rand()
alpha3 = 1 + np.random.random() * np.random.rand()
esp = np.random.random()
# There is no usage of this variable in the paper
alpha4 = esp + esp * np.random.rand()
beta = np.random.randint(0, 2, 3)
gama = np.random.randint(0, 2, 3)
## The seed4 is mutation process, but not sure k is multiple variables or 1 variable.
## In the text said, multiple variables, but the defination of k is 1 variable. So confused
k = np.random.randint(0, self.problem.n_dims)
k_idx = np.random.choice(range(0, self.problem.n_dims), k, replace=False)
seed1 = self.pop[idx][self.ID_POS] + alpha1 * (beta[0] * self.g_best[self.ID_POS] - gama[0] * MG) # Eq. 3
seed2 = self.g_best[self.ID_POS] + alpha2 * (beta[1] * self.pop[idx][self.ID_POS] - gama[1] * MG) # Eq. 4
seed3 = MG + alpha3 * (beta[2] * self.pop[idx][self.ID_POS] - gama[2] * self.g_best[self.ID_POS]) # Eq. 5
seed4 = deepcopy(self.pop[idx][self.ID_POS]).astype(float)
seed4[k_idx] += np.random.uniform(0, 1, k)
# Check if solutions go outside the search space and bring them back
seed1 = self.amend_position(seed1)
seed2 = self.amend_position(seed2)
seed3 = self.amend_position(seed3)
seed4 = self.amend_position(seed4)
sol1 = [seed1, self.get_fitness_position(seed1)]
sol2 = [seed2, self.get_fitness_position(seed2)]
sol3 = [seed3, self.get_fitness_position(seed3)]
sol4 = [seed4, self.get_fitness_position(seed4)]
## Lots of grammar errors in this section, so confused to understand which strategy they are using
_, best_seed = self.get_global_best_solution([sol1, sol2, sol3, sol4])
pop_new.append(best_seed)
self.pop = self.greedy_selection_population(self.pop, pop_new)