Genetic-Algorithm/genetic_algorithm.py at main · arindam2590/Genetic-Algorithm · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
import random
import json
import numpy as np


class GeneticAlgorithm:
    def __init__(self, maze, source, pop_size, mut_rate):
        with open('Utils/config.json', 'r') as file:
            self.params = json.load(file)

        self.maze = maze
        self.src = source
        self.population_size = pop_size
        self.lower_bound = self.params['CHROMOSOME_LOWER']
        self.upper_bound = self.params['CHROMOSOME_UPPER']
        self.mutation_rate = mut_rate
        self.tour_size_percent = self.params['TOURNAMENT_PERCENT']
        self.parent_percent = self.params['PARENT_PERCENT']
        self.tournament_size = None
        self.n_parents = None

    def generate_initial_population(self):
        population = []
        for i in range(self.population_size):
            path_length = np.random.randint(self.lower_bound, self.upper_bound)
            chromosome = []

            previous_move = random.choice(['U', 'D', 'L', 'R'])
            chromosome.append(previous_move)
            for _ in range(1, path_length):
                next_move = random.choice(['U', 'D', 'L', 'R'])
                while (previous_move == 'U' and next_move == 'D') or (previous_move == 'D' and next_move == 'U') or \
                        (previous_move == 'L' and next_move == 'R') or (previous_move == 'R' and next_move == 'L'):
                    next_move = random.choice(['U', 'D', 'L', 'R'])
                chromosome.append(next_move)
                previous_move = next_move
            population.append(''.join(chromosome))
        print(f'Info: Initial Population has been created. \nInfo: Initial Population Size: {self.population_size}')
        print(f'-' * 101)
        return population

    # def select_parents(self, population, fitness_scores):
    #     self.population_size = len(population)
    #     self.tournament_size = int(self.tour_size_percent * self.population_size)
    #     self.n_parents = int(self.parent_percent * self.tournament_size)
    #
    #     print(f'Info: Size of the Tournament - {self.tournament_size}')
    #     parent1, parent2 = [], []
    #     # Select two parents from the population using tournament selection
    #     for i in range(self.n_parents):
    #         parent1.append(self.tournament_selection(population, fitness_scores))
    #         parent2.append(self.tournament_selection(population, fitness_scores))
    #     return parent1, parent2

    def select_parents(self, population, fitness_scores):
        filtered_population, filtered_fitness = zip(*[(individual, fit) for individual, fit in
                                                      zip(population, fitness_scores) if fit >= 0]) \
            if any(fit > 0 for fit in fitness_scores) else ([], [])

        if len(filtered_population) == 0:
            print(f"Warning: All individuals had negative fitness! Regenerating population...")
            exit(0)

        sorted_population = [x for _, x in sorted(zip(filtered_fitness, filtered_population), reverse=True)]
        sorted_fitness = sorted(filtered_fitness, reverse=True)
        print(sorted_fitness)

        self.population_size = len(sorted_population)
        self.tournament_size = int(self.tour_size_percent * self.population_size)
        self.n_parents = int(self.parent_percent * self.tournament_size)

        print(f'Info: Size of the Tournament - {self.tournament_size}, Sorted Population Size: {self.population_size}')
        parent1, parent2 = [], []
        for i in range(self.n_parents):
            parent1.append(self.tournament_selection(sorted_population))
            parent2.append(self.tournament_selection(sorted_population))
        return parent1, parent2

    def tournament_selection(self, population):
        # Randomly select 'tournament_size' individuals (by index) from the population.
        contestant_indices = random.sample(range(self.population_size), self.tournament_size)
        best_index = contestant_indices[0]
        best_parent = population[best_index]
        return best_parent

    # def crossover(self, parent1, parent2):
    #     # Perform crossover between two parents to create a child
    #     split_point = random.randint(1, min(len(parent1), len(parent2)) - 1)
    #     child1 = parent1[:split_point] + parent2[split_point:]
    #     child2 = parent2[:split_point] + parent1[split_point:]
    #     return child1, child2

    def crossover(self, parent1, parent2):
        k1 = self.find_k_point(parent1)
        k2 = self.find_k_point(parent2)

        child = None
        if k1 == 0 and k2 == 0:
            k = random.randint(1, min(len(parent1), len(parent2)) - 1)
            child = parent1[:k] + parent2[k:]
        elif k1 == k2 > 0:
            child = parent1[:k1] + parent2[k2:]
        elif k1 > k2:
            if k2 == 0:
                k = random.randint(1, len(parent2))
                child = parent1[:k1] + parent2[k:]
            else:
                child = parent1[:k1] + parent2[k2:]
        elif k1 < k2:
            if k1 == 0:
                k = random.randint(1, len(parent1))
                child = parent2[:k2] + parent1[k:]
            else:
                child = parent2[:k2] + parent1[k1:]

        return child

    def find_k_point(self, chromosome):
        x, y = self.src
        k = 0
        for gene in chromosome:
            x, y = x + (gene == 'D') - (gene == 'U'), y + (gene == 'R') - (gene == 'L')
            if x < 0 or x >= self.maze.shape[0] or y < 0 or y >= self.maze.shape[1] or self.maze[x][y] == 1:
                return k
            k += 1
        return k

    def mutate(self, chromosome):
        new_chromosome = ''
        for i in range(len(chromosome)-1):
            gene = chromosome[i]
            next_gene = chromosome[i+1]
            if (gene == 'L' and next_gene == 'R') or (gene == 'R' and next_gene == 'L') or \
                (gene == 'U' and next_gene == 'D') or (gene == 'D' and next_gene == 'U'):
                if random.random() < self.mutation_rate:
                    possible_gene = random.choice(['U', 'D', 'L', 'R'])
                # new_chromosome +=
            else:
                new_chromosome += gene
        return chromosome