-
Notifications
You must be signed in to change notification settings - Fork 1
/
genetic_algorithm.py
217 lines (147 loc) · 5.31 KB
/
genetic_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
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
import numpy as np
import random
from random import randint
import copy
import matplotlib.pyplot as plt
import seaborn as sb
from map import *
import time
no_of_generations = 100
lower_limit = 0.1
N = 10
pop_size = 100
how_many_to_kill = 75
prob_mut = 0.0001
def initialise_map(lower_limit):
# N is the number of pubs
# random.random() generates a random float X ~ Unif([0,1)), we define lower_limit to avoid zeros
our_map = np.zeros((N, N))
# the adjacency matrix is a real symmetric matrix with zero entries along the diagonal
for i in range(0, N):
for j in range(0, i): # do not include i since the diagonal is zero
our_map[i][j] = random.random() * 100
our_map[j][i] = our_map[i][j]
return our_map
def create_new_route():
start = np.array([0])
intermediate_steps = np.random.permutation(np.arange(1, N - 1))
end = np.array([N - 1])
temp = np.append(start, intermediate_steps)
route = np.append(temp, end)
return route
def crossover(a, b):
c = []
for i in range(1, N - 2):
if (a[i] == b[i]):
c.append(i)
if c != []:
index = c[0]
while (index == c[0]):
index = random.randint(1, N - 2)
temp = a[c[0]]
a[c[0]] = a[index]
a[index] = temp
while (index == c[0]):
index = random.randint(1, N - 2)
temp = b[c[0]]
b[c[0]] = b[index]
b[index] = temp
return (a, b)
def mutate(a, prob_mut):
if random.random() > prob_mut:
i = random.randint(1, N - 2)
j = random.randint(1, N - 2)
temp = a[i]
a[i] = a[j]
a[j] = temp
return a
def fitness(a, our_map):
score = 0
for i in range(0, N - 1):
score += our_map[a[i]][a[i + 1]]
return score
def create_generation(pop_size, our_map):
population = []
for i in range(0, pop_size):
population.append(create_new_route())
return population
def score_population(population, our_map):
scores = []
for i in range(0, len(population)):
scores += [fitness(population[i], our_map)]
return scores
def sort_population(population, our_map):
scores = score_population(population, our_map)
np_scores = np.array(scores)
return np_scores.argsort()
def best_in_population(population, our_map):
best = sort_population(population, our_map)[0]
best_route = population[best]
return best_route
def fitness_of_best_in_population(population, our_map):
fit = fitness(best_in_population(population, our_map), our_map)
return fit
def remove_from_pop(how_many_to_kill, population, our_map):
ranked_pop = sort_population(population, our_map)
survival_of_the_fittest = ranked_pop[: (len(population) - how_many_to_kill)]
return np.array(survival_of_the_fittest)
def breeding(population, our_map):
fittest = remove_from_pop(how_many_to_kill, population, our_map)
children = []
keep = 4
for i in range(0, 4):
children.append(population[fittest[i]])
for i in range(4, len(fittest) - 1, 2):
child_1 = crossover(population[fittest[i]], population[fittest[i + 1]])[0]
child_2 = crossover(population[fittest[i]], population[fittest[i + 1]])[1]
child_1 = mutate(child_1, prob_mut)
child_2 = mutate(child_2, prob_mut)
children.append(child_1)
children.append(child_2)
while (len(children) < pop_size):
new_route = create_new_route()
children.append(new_route)
return np.array(children)
def main():
#our_map = initialise_map(lower_limit)
names_of_locations_temp = ["kelseys", "the fat pug", "the town house", "the old library", "the clarendon",
"the benjamin satchwell", "murphy's bar", "The Royal Pug", "the white house",
"The Drawing Board"]
names_of_locations = [name.lower() for name in sorted(names_of_locations_temp)]
generator = MapGenerator(names_of_locations)
locations = generator.decodeLocations()
num_to_object = {}
object_to_num = {}
for counter, location_object in enumerate(locations):
num_to_object[counter] = location_object
object_to_num[location_object] = counter
our_map = generator.adjacency_matrix_generator()
population = create_generation(pop_size, our_map)
route = []
fitnesses = []
for i in range(0, no_of_generations):
route.append(best_in_population(population, our_map))
fitnesses.append(fitness_of_best_in_population(population, our_map))
breeding(population, our_map)
population = breeding(population, our_map)
last = None
for current_route in route:
locations_to_render = [num_to_object[x] for x in current_route]
if last != locations_to_render:
generator.renderLocations(locations_to_render)
last = locations_to_render
plt.plot(np.arange(0, no_of_generations), fitnesses)
plt.ylabel('fitness')
plt.xlabel('no. of generations')
plt.show()
main()
"""
What we want it in this kinda form:
population = fitness(networks)
population = selection(networks)
population = crossover(networks)
population = mutate(networks)
Look at https://github.com/Molten-Ice/AI/blob/master/Hyperparameter%20optimisation%20using%20a%20Genetic%20algorithm
to get an idea of code style
#
"""