-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathmprm_frog.py
212 lines (196 loc) · 8.96 KB
/
mprm_frog.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
import numpy as np
from circuit.circuit import *
import matplotlib.pyplot as plt
class JumpFrog:
def __init__(self, population, meme_size, mprm):
self.population = population
self.meme_size = meme_size
self.group_num = int(population / meme_size)
self.dim = mprm.in_num
self.global_best = None
self.mprm = mprm
self.dic = {}
def init(self):
self.frogs = np.random.randint(0, 3, size=(int(self.population / 3), self.dim))
pair_frogs_1 = (self.frogs + 1) % 3
pair_frogs_2 = (self.frogs + 2) % 3
self.frogs = np.vstack((self.frogs, pair_frogs_1, pair_frogs_2))
if self.frogs.shape[0] < self.population:
rand_frog = np.random.randint(0, 3, size=(self.population - self.frogs.shape[0], self.dim))
self.frogs = np.vstack((self.frogs, rand_frog))
self.fitness = self.get_fitness(self.frogs)
def get_fitness(self, frogs):
res = []
frogs = frogs.reshape(-1, self.dim).astype(int)
# print(frogs.shape)
for i in range(frogs.shape[0]):
res.append(self.get_one_fitness(frogs[i]))
res = np.array(res).reshape((frogs.shape[0], -1))
return -res
def get_one_fitness(self, frog):
num = get_polarity_num(frog)
if num in self.dic:
return self.dic.get(num)
self.mprm.turnTo(frog)
self.dic[num] = self.mprm.get_area()
# print("one", num, self.dic[num])
return self.dic[num]
def sort(self):
self.fitness_sub = np.argsort(-self.fitness.squeeze())
self.fitness_sub = np.array(self.fitness_sub)
def grouping(self):
self.groups = self.fitness_sub.reshape((self.meme_size, self.group_num)).T
def evolve(self, iterator_times):
global_best = self.groups[0][0]
for i in range(self.group_num):
# print("e", i)
for iter in range(iterator_times):
local_best = self.groups[i][0]
local_worst = self.groups[i][self.meme_size - 1]
for k in range(self.meme_size):
if self.fitness[self.groups[i][k]] > self.fitness[local_best]:
local_best = self.groups[i][k]
if self.fitness[self.groups[i][k]] < self.fitness[local_worst]:
local_worst = self.groups[i][k]
"""
tr = np.random.randint(0, self.meme_size, size=(1, 2))
dis = (1.5 * np.random.rand(1, self.frogs[local_best].shape[0]) *
((self.frogs[local_best] - self.frogs[tr[0][0]])
+ (self.frogs[local_worst] - self.frogs[tr[0][1]])))
"""
dis = (1.5 * np.random.rand(1, self.frogs[local_best].shape[0]) *
(self.frogs[local_best] - self.frogs[local_worst]))
"""
dis = (np.sin((np.pi/ 2) * (1 / (iterator_times - iter + 1)))
*(self.frogs[local_best] - self.frogs[local_worst])).reshape(1, -1)
"""
dis = np.round(dis)
dis[dis > 2] = 2
dis[dis < 0] = 0
"""
for it in range(dis.shape[1]):
if dis[0][it] >= 1.5:
dis[0][it] = 2
elif 0.5 <= dis[0][it] < 1.5:
dis[0][it] = 1
elif dis[0][it] < 0.5:
dis[0][it] = 0
"""
temp = self.frogs[local_worst] + dis
# temp[temp < 0] = 0
temp_f = self.get_fitness(temp)
if temp_f > self.fitness[local_worst]:
self.frogs[local_worst] = temp
self.fitness[local_worst] = temp_f
else:
dis = (1.5 * np.random.rand(1, self.frogs[global_best].shape[0]) *
(self.frogs[global_best] - self.frogs[local_worst]))
"""
dis = (np.sin((np.pi / 2) * (1 / (iterator_times - iter + 1)))
* (self.frogs[local_best] - self.frogs[local_worst])).reshape(1, -1)
"""
dis = np.round(dis)
dis[dis > 2] = 2
dis[dis < 0] = 0
"""
for it in range(dis.shape[1]):
if dis[0][it] >= 1.5:
dis[0][it] = 2
elif 0.5 <= dis[0][it] < 1.5:
dis[0][it] = 1
elif dis[0][it] < 0.5:
dis[0][it] = 0
"""
temp = self.frogs[local_worst] + dis
# temp[temp < 0] = 0
temp_f = self.get_fitness(temp)
if temp_f > self.fitness[local_worst]:
self.frogs[local_worst] = temp
self.fitness[local_worst] = temp_f
else:
new_1 = np.random.randint(0, 3, size=(1, self.dim))
new_2 = 2 - new_1
if self.get_fitness(new_2) > self.get_fitness(new_1):
new_1 = new_2
self.frogs[local_worst] = new_1
# self.frogs[local_worst] = 2 - self.frogs[local_worst]
self.fitness[local_worst] = self.get_fitness(self.frogs[local_worst])
global_best = self.elite(i, global_best)
for j in range(self.meme_size):
if self.fitness[global_best] < self.fitness[self.groups[i][j]]:
global_best = self.groups[i][j]
return global_best
def train(self, times): # 全局迭代
x = []
y = []
max_t = max(5, int(times * 0.5))
cur_t = 0
global_best = []
for i in range(times):
x.append(i)
# print("t", i)
self.fitness = self.get_fitness(self.frogs)
self.sort()
self.grouping()
global_best = self.frogs[self.groups[0][0]]
new_best = self.evolve(5)
# print(i, ":", self.fitness[new_best])
if self.get_fitness(global_best) >= self.fitness[new_best]:
y.append(self.get_fitness(global_best)[0])
cur_t += 1
if cur_t >= max_t:
"""
plt.plot(np.array(x), np.array(y))
plt.show()
"""
return global_best, self.get_fitness(global_best)
else:
y.append(self.fitness[new_best])
cur_t = 0
global_best = self.frogs[new_best]
# print("t", -self.get_fitness(global_best))
"""
plt.plot(np.array(x), np.array(y))
plt.show()
"""
# print(test(self.frogs[global_best], self.weights, self.limitation))
return global_best, self.get_fitness(global_best)
def elite(self,group_index,global_best):
local_worst = self.groups[group_index][self.meme_size - 1]
local_best = self.groups[group_index][0]
for i in range(int(0.4*self.meme_size)):
local_worst = self.groups[group_index][self.meme_size - 1]
local_best = self.groups[group_index][0]
for k in range(self.meme_size):
if self.fitness[self.groups[group_index][k]] > self.fitness[local_best]:
local_best = self.groups[group_index][k]
if self.fitness[self.groups[group_index][k]] < self.fitness[local_worst]:
local_worst = self.groups[group_index][k]
new=self.frogs[local_worst].copy()
for j in range(new.shape[0]):
new[j]=abs(np.random.normal(0, 1))*new[j]
new[new > 1] = 2
if self.fitness[local_worst]<self.get_fitness(new):
self.frogs[local_worst]=new
self.fitness[local_worst] = self.get_fitness(new)
if self.fitness[local_best]<self.fitness[local_worst]:
self.frogs[local_best]=self.frogs[local_worst]
if self.fitness[global_best]<self.fitness[local_worst]:
global_best=local_worst
for i in range(3):
new=self.frogs[local_best].copy()
for j in range(new.shape[0]):
new[j]=(1+0.1*abs(np.random.normal(0,1)))*new[j]
new[new>1]=2
if self.fitness[local_best]<self.get_fitness(new):
self.frogs[local_best]=new
if self.fitness[global_best]<self.fitness[local_best]:
global_best=local_best
return global_best
def explosion(self, local_best):
r = np.random.randint(4, self.dim)
poi = np.random.randint(0, self.dim - r)
new = self.frogs[local_best].copy()
for i in range(poi, poi + r):
new[i] = np.random.randint(0, 2)
return new