-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathalgorithm.py
200 lines (162 loc) · 7.89 KB
/
algorithm.py
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
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
import random
import numpy as np
from enum import Enum
class CrossoverType(Enum):
PartialMappedCrossover = 1
OrderCrossover = 2
PositionBasedCrossover = 3
class MutationType(Enum):
Inversion = 1
Insertion = 2
Displacement = 3
ReciprocalExchange = 4
class SelectionType(Enum):
Deterministic = 1
Stochastic = 2
class GeneticAlgorithm:
def __init__(self, pop_size, number_of_genes, selection_type,
crossover_type, crossover_rate, mutation_type, mutation_rate,
compute_objective_value):
self.pop_size = pop_size
self.selection_type = selection_type
self.crossover_size = int(pop_size * crossover_rate)
if self.crossover_size % 2 == 1:
self.crossover_size -= 1
self.mutation_size = int(pop_size * mutation_rate)
self.total_size = self.pop_size + self.mutation_size + self.crossover_size
self.number_of_genes = number_of_genes
self.crossover_type = crossover_type
self.crossover_rate = crossover_rate
self.mutation_type = mutation_type
self.mutation_rate = mutation_rate
self.compute_objective_value = compute_objective_value
self.least_fitness_factor = 0.3
self.mapping = [-1 for i in range(self.number_of_genes)] # for crossover
def initialize(self):
self.selected_chromosomes = np.zeros((self.pop_size, self.number_of_genes))
self.indexs = np.arange(self.total_size)
self.chromosomes = np.zeros((self.total_size, self.number_of_genes), dtype=int)
for i in range(self.pop_size):
for j in range(self.number_of_genes):
self.chromosomes[i][j] = j
np.random.shuffle(self.chromosomes[i])
for i in range(self.pop_size, self.total_size):
for j in range(self.number_of_genes):
self.chromosomes[i][j] = -1
self.fitness = np.zeros(self.total_size)
self.objective_values = np.zeros(self.total_size)
self.best_chromosome = np.zeros(self.number_of_genes, dtype=int)
self.best_fitness = 0
def evaluate_fitness(self):
for i, chromosome in enumerate(self.chromosomes[:self.pop_size]):
self.objective_values[i] = self.compute_objective_value(chromosome)
min_obj_val = np.min(self.objective_values)
max_obj_val = np.max(self.objective_values)
range_obj_val = max_obj_val - min_obj_val
for i, obj in enumerate(self.objective_values):
self.fitness[i] = max(self.least_fitness_factor * range_obj_val, pow(10, -5)) + \
(max_obj_val - obj)
def update_best_solution(self):
best_index = np.argmax(self.fitness)
if self.best_fitness < self.fitness[best_index]:
self.best_fitness = self.fitness[best_index]
for i, gene in enumerate(self.chromosomes[best_index]):
self.best_chromosome[i] = gene
def _shuffle_index(self, length):
for i in range(length):
self.indexs[i] = i
np.random.shuffle(self.indexs[:length])
def perform_crossover_operation(self):
self._shuffle_index(self.pop_size)
child1_index = self.pop_size
child2_index = self.pop_size + 1
count_of_crossover = int(self.crossover_size / 2)
for i in range(count_of_crossover):
parent1_index = self.indexs[i]
parent2_index = self.indexs[i + 1]
if self.crossover_type == CrossoverType.PartialMappedCrossover:
self._partial_mapped_crossover(parent1_index, parent2_index, child1_index, child2_index)
self.objective_values[child1_index] = self.compute_objective_value(self.chromosomes[child1_index])
self.objective_values[child2_index] = self.compute_objective_value(self.chromosomes[child2_index])
child1_index += 2
child2_index += 2
def _partial_mapped_crossover(self, p1, p2, c1, c2):
# reset
for i in range(self.number_of_genes):
self.mapping[i] = -1
rand1 = random.randint(0, self.number_of_genes - 2)
rand2 = random.randint(rand1 + 1, self.number_of_genes - 1)
for i in range(rand1, rand2 + 1):
c1_gene = self.chromosomes[p2][i]
c2_gene = self.chromosomes[p1][i]
if c1_gene == c2_gene:
continue
elif self.mapping[c1_gene] == -1 and self.mapping[c2_gene] == -1:
self.mapping[c1_gene] = c2_gene
self.mapping[c2_gene] = c1_gene
elif self.mapping[c1_gene] == -1:
self.mapping[c1_gene] = self.mapping[c2_gene]
self.mapping[self.mapping[c2_gene]] = c1_gene
self.mapping[c2_gene] = -2
elif self.mapping[c2_gene] == -1:
self.mapping[c2_gene] = self.mapping[c1_gene]
self.mapping[self.mapping[c1_gene]] = c2_gene
self.mapping[c1_gene] = -2
else:
self.mapping[self.mapping[c1_gene]] = self.mapping[c2_gene]
self.mapping[self.mapping[c2_gene]] = self.mapping[c1_gene]
self.mapping[c1_gene] = -3
self.mapping[c2_gene] = -3
for i in range(self.number_of_genes):
if rand1 <= i <= rand2:
self.chromosomes[c1][i] = self.chromosomes[p2][i]
self.chromosomes[c2][i] = self.chromosomes[p1][i]
else:
if self.mapping[self.chromosomes[p1][i]] >= 0:
self.chromosomes[c1][i] = self.mapping[self.chromosomes[p1][i]]
else:
self.chromosomes[c1][i] = self.chromosomes[p1][i]
if self.mapping[self.chromosomes[p2][i]] >= 0:
self.chromosomes[c2][i] = self.mapping[self.chromosomes[p2][i]]
else:
self.chromosomes[c2][i] = self.chromosomes[p2][i]
def _do_roulette_wheel_selection(self, fitness_list):
sum_fitness = sum(fitness_list)
transition_probability = [fitness / sum_fitness for fitness in fitness_list]
rand = random.random()
sum_prob = 0
for i, prob in enumerate(transition_probability):
sum_prob += prob
if sum_prob >= rand:
return i
def perform_selection(self):
if self.selection_type == SelectionType.Deterministic:
index = np.argsort(self.fitness)[::-1]
elif self.selection_type == SelectionType.Stochastic:
index = [self._do_roulette_wheel_selection(self.fitness) for i in range(self.pop_size)]
else:
index = self.indexs
for i in range(self.pop_size):
for j in range(self.number_of_genes):
self.selected_chromosomes[i][j] = self.chromosomes[index[i]][j]
for i in range(self.pop_size):
for j in range(self.number_of_genes):
self.chromosomes[i][j] = self.selected_chromosomes[i][j]
def perform_mutation_operation(self):
self._shuffle_index(self.pop_size + self.crossover_size)
child1_index = self.pop_size + self.crossover_size
for i in range(self.mutation_size):
if self.mutation_type == MutationType.Inversion:
parent1_index = self.indexs[i]
self._inversion_mutation(parent1_index, child1_index)
self.objective_values[child1_index] = self.compute_objective_value(self.chromosomes[child1_index])
child1_index += 1
def _inversion_mutation(self, p1, c1):
rand1 = random.randint(0, self.number_of_genes - 2)
rand2 = random.randint(rand1 + 1, self.number_of_genes - 1)
for i in range(self.number_of_genes):
if i < rand1 or i > rand2:
self.chromosomes[c1][i] = self.chromosomes[p1][i]
else:
index = rand2 - (i - rand1)
self.chromosomes[c1][i] = self.chromosomes[p1][index]