-
Notifications
You must be signed in to change notification settings - Fork 4
/
IMP.py
229 lines (216 loc) · 9.36 KB
/
IMP.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
218
219
220
221
222
223
224
225
226
227
228
229
from collections import defaultdict
import random
import sys
import getopt
import time
# deal with the dataset
def get_args():
graph_file = sys.argv[2]
seed_size = int(sys.argv[4])
model = sys.argv[6]
time_limit = int(sys.argv[8])
# try:
# opts, agrs = getopt.getopt(sys.argv[1:], 'i:s:m:t')
# except:
# print("Something get wrong")
# sys.exit(2)
# # print(agrs)
# graph_file = agrs[1]
# seed_size = int(agrs[3])
# model = agrs[5]
# time_limit = int(agrs[7])
return graph_file, seed_size, model, time_limit
def read_data(graph_file):
f1 = open(graph_file, 'r')
first_line = f1.readline().split()
# print(first_line)
vertex_num = int(first_line[0])
edge_num = int(first_line[1])
graph = defaultdict(dict)
out_degree = defaultdict(int)
for line in f1.readlines():
data = line.split()
out_degree[int(data[0])] += 1
graph[int(data[0])][int(data[1])] = float(data[2])
# print(graph[10])
return vertex_num, edge_num, graph, out_degree
def independent_Cascade(graph, seeds):
influnces = seeds[:] # 影响节点集合
queue = influnces[:]
while len(queue) != 0:
node = queue.pop(0)
for element in graph[node]:
# print(graph[node])
# print(element)
if element not in influnces:
probility = random.random()
if probility <= graph[node][element]:
influnces.append(element)
queue.append(element)
influnce_num = len(influnces)
return influnce_num
def linear_Threshold(graph, seeds):
influnces = seeds[:]
queue = influnces[:]
pre_node_record = defaultdict(float) # 记录前一个节点的概率总和
threshold = defaultdict(float) # 记录每个节点的阀门值
while len(queue) != 0:
node = queue.pop(0)
for element in graph[node]:
if element not in influnces:
if threshold[element] == 0: # 节点阀门值未被记录
threshold[element] = random.random()
pre_node_record[element] = pre_node_record[element] + graph[node][element] # 每次循环叠加一次不同的前一个激活节点的值
if pre_node_record[element] >= threshold[element]: # 激活
influnces.append(element)
queue.append(element)
influnce_num = len(influnces)
return influnce_num
def submodular_greedy(graph, vertex_num, seed_size, out_degree,model):
seeds = []
s = defaultdict(float)
if model == "IC":
for i in range(seed_size):
for node in range(1, vertex_num + 1):
s[node] = 0
if node not in seeds and node in out_degree:
for i in range(1000):
s[node] += independent_Cascade(graph, seeds + [node]) - independent_Cascade(graph, seeds)
s[node] /= 1000
seeds.append(max(s, key=s.get))
else:
for i in range(seed_size):
for node in range(1, vertex_num + 1):
s[node] = 0
if node not in seeds and node in out_degree:
for i in range(1000):
s[node] += linear_Threshold(graph, seeds + [node]) - linear_Threshold(graph, seeds)
s[node] /= 1000
seeds.append(max(s, key=s.get))
return seeds
def CELFII_IC(graph, vertex_num, seed_size, out_degree):
test_count = 0
seeds = []
s_n_influnece = defaultdict(float)
current_size = len(seeds) # 当前的种子群大小
while len(seeds) < seed_size:
if len(seeds) == 0: # 当种子群为空时，从所有节点中取出最优秀的种子
for node in range(1, vertex_num + 1):
s_n_influnece[node] = 0
if node in out_degree.keys(): # 拥有出度的点
for i in range(1000):
s_n_influnece[node] = s_n_influnece[node] + independent_Cascade(graph, seeds+[node])
average_influence = s_n_influnece[node] / 1000
s_n_influnece[node] = average_influence
max_seed = max(s_n_influnece, key=s_n_influnece.get) # 选出当前最大的node值
s_n_influnece.pop(max_seed)
seeds.append(max_seed)
test_count+=1
elif len(seeds)!= 0:
# print(s_n_influnece)
# print("----->", test_count)
prev_best = max(s_n_influnece, key=s_n_influnece.get)
s_n_influnece[prev_best] = 0
for i in range(1000):
new_seeds = seeds + [prev_best]
marginal_profit = independent_Cascade(graph, seeds + [prev_best]) - independent_Cascade(graph, seeds)
s_n_influnece[prev_best] += marginal_profit
s_n_influnece[prev_best] = s_n_influnece[prev_best] / 1000
current_seed = max(s_n_influnece, key=s_n_influnece.get)
if current_seed == prev_best:
seeds.append(current_seed)
s_n_influnece.pop(current_seed)
else:
continue
return seeds
def CELFII_LT(graph, vertex_num, seed_size, out_degree):
seeds = []
s_n_influnece = defaultdict(float)
current_size = len(seeds) # 当前的种子群大小
while len(seeds) < seed_size:
if len(seeds) == 0: # 当种子群为空时，从所有节点中取出最优秀的种子
for node in range(1, vertex_num + 1):
s_n_influnece[node] = 0
if node in out_degree: # 拥有出度的点
for i in range(1000):
single_node = []
single_node.append(node)
s_n_influnece[node] = s_n_influnece[node] + linear_Threshold(graph, single_node)
average_influence = s_n_influnece[node] / 1000
s_n_influnece[node] = average_influence
max_seed = max(s_n_influnece, key=s_n_influnece.get)
s_n_influnece.pop(max_seed)
seeds.append(max_seed)
else:
prev_best = max(s_n_influnece, key=s_n_influnece.get)
s_n_influnece[prev_best] = 0
for i in range(1000):
new_seeds = seeds + [prev_best]
marginal_profit = linear_Threshold(graph, new_seeds) - linear_Threshold(graph, seeds)
s_n_influnece[prev_best] = s_n_influnece[prev_best] + marginal_profit
s_n_influnece[prev_best] = s_n_influnece[prev_best] / 1000
current_seed = max(s_n_influnece, key=s_n_influnece.get)
if current_seed == prev_best:
seeds.append(current_seed)
s_n_influnece.pop(current_seed)
else:
continue
return seeds
def CELFII(graph, vertex_num, seed_size, out_degree, model):
if model == "IC":
seeds = CELFII_IC(graph, vertex_num, seed_size, out_degree)
else:
seeds = CELFII_LT(graph, vertex_num, seed_size, out_degree)
return seeds
def calculate_average(graph, seeds, model):
if model == "IC":
count = 0
total_influence = 0
while count < 1000:
total_influence += independent_Cascade(graph, seeds)
count += 1
IC_average = total_influence/count
average_result = IC_average
else:
count = 0
total_influence = 0
while count < 1000:
total_influence += linear_Threshold(graph, seeds)
count += 1
LT_average = total_influence / count
average_result = LT_average
return average_result
def get_final_seeds(graph, vertex_num, seed_size, out_degree, model, time_limit):
time_budget = time_limit
start_time = time.time()
final_seeds = []
total_influence = 0
while time.time() - start_time < time_budget - 2:
final_seeds = CELFII(graph, vertex_num, seed_size, out_degree, model)
# total_influence = calculate_average(graph, final_seeds, model)
if len(final_seeds) > 0:
break
# print("total_influence: ", total_influence)
# print("final_seeds: ", final_seeds)
for i in final_seeds:
print(i)
# ----------------------------测试区域----------------------------------
start_time = time.time()
graph_file, seed_size, model, time_limit = get_args()
vertex_num, edge_num, graph, out_degree = read_data(graph_file)
# print(vertex_num, edge_num, graph, out_degree)
seeds = [56, 58, 53, 62, 28, 48, 50, 61, 60, 45]
seeds2 = [56, 58, 53, 50, 28, 48, 62, 61, 60, 45]
# [56, 58, 53, 62, 28]
# [56, 58, 53, 50, 28]
# [56, 58, 48, 53, 62]print(calculate_average(graph, seeds, model))
# # print(calculate_average(graph, seeds2, model))
# print(linear_Threshold(graph, seeds))
#
# print(CELF2(graph, range(1, vertex_num+1), seed_size, out_degree, model))
# print(CELFII_IC(graph, vertex_num, seed_size, out_degree))
# print(CELFII_LT(graph, vertex_num, seed_size, out_degree))
# print(submodular_greedy(graph, vertex_num, seed_size, out_degree, model))
# print(CELFII(graph, vertex_num, seed_size, out_degree, model))
get_final_seeds(graph, vertex_num, seed_size, out_degree, model, time_limit)
# print(time.time() - start_time)