-
Notifications
You must be signed in to change notification settings - Fork 1
/
main_v4.py
345 lines (272 loc) · 12.7 KB
/
main_v4.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
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
from arguments import get_args
from live_video import LiveVideo
import numpy as np
import torch as th
from model.EVOLVE_GCN import EvolveGCNO, EvolveGCNH
from matplotlib import animation
import matplotlib.pyplot as plt
import gym
from time import time
import networkx as nx
from model.random_walk import Graph_RandomWalk
from collections import defaultdict
import dgl
def display_frames_as_gif(policy, frames):
patch = plt.imshow(frames[0])
plt.axis('off')
def animate(i):
patch.set_data(frames[i])
anim = animation.FuncAnimation(plt.gcf(), animate, frames=len(frames), interval=5)
anim.save('./' + policy + '_viewport_result.gif', writer='ffmpeg', fps=30)
def build_graph(edge_list, nodeFeature):
src, dst = tuple(zip(*edge_list))
src, dst = th.tensor(src).to('cuda:0'), th.tensor(dst).to('cuda:0')
u = th.cat((src, dst))
v = th.cat((dst, src))
g = dgl.graph((u, v))
g.ndata['feat'] = nodeFeature
#print(g)
return g
def compute_loss(pos_score, neg_score):
# 间隔损失
n_edges = pos_score.shape[0]
return (1 - pos_score.unsqueeze(1) + neg_score.view(n_edges, -1)).clamp(min=0).mean()
def construct_negative_graph(graph, k):
src, dst = graph.edges()
neg_src = src.repeat_interleave(k).to('cuda:0')
neg_dst = th.randint(0, graph.num_nodes(), (len(src) * k,)).to('cuda:0')
return dgl.graph((neg_src, neg_dst), num_nodes=graph.num_nodes())
def run_random_walks_n2v(graph, nodes, num_walks=10, walk_len=40):
""" In: Graph and list of nodes
Out: (target, context) pairs from random walk sampling using the sampling strategy of node2vec (deepwalk)"""
walk_len = 5
nx_G = nx.Graph()
adj = nx.adjacency_matrix(graph)
for e in graph.edges():
nx_G.add_edge(e[0], e[1])
for edge in graph.edges():
nx_G[edge[0]][edge[1]]['weight'] = adj[edge[0], edge[1]]
G = Graph_RandomWalk(nx_G, False, 1.0, 1.0)
G.preprocess_transition_probs()
walks = G.simulate_walks(num_walks, walk_len)
WINDOW_SIZE = 10
pairs = defaultdict(lambda: [])
pairs_cnt = 0
for walk in walks:
for word_index, word in enumerate(walk):
for nb_word in walk[max(word_index - WINDOW_SIZE, 0): min(word_index + WINDOW_SIZE, len(walk)) + 1]:
if nb_word != word:
pairs[word].append(nb_word)
pairs_cnt += 1
print("# nodes with random walk samples: {}".format(len(pairs)))
print("# sampled pairs: {}".format(pairs_cnt))
return pairs
def get_context_pairs(graphs, num_time_steps):
""" Load/generate context pairs for each snapshot through random walk sampling."""
print("Computing training pairs ...")
context_pairs_train = []
for i in range(0, num_time_steps):
context_pairs_train.append(run_random_walks_n2v(graphs[i], graphs[i].nodes()))
return context_pairs_train
if __name__ == '__main__':
args = get_args() # 从 arguments.py 获取配置参数
# graphs, adjs, labels = load_graphs_school('school')
#
if args.cuda and th.cuda.is_available():
device = th.device('cuda:0')
th.backends.cudnn.benchmark = True
else:
device = th.device('cpu')
videoUsers = []
tileNum = args.sampleRate * args.tileNum * args.tileNum # 瓦块数:一个视频帧切分成 5 * 5 = 25 个瓦块
totalUser = args.testNum + args.trainNum
# 生成所有用户的视频信息类
for index in range(totalUser):
args.userId = index + 1
videoUsers.append(LiveVideo(args))
# 为每个用户加载视频数据
for index in range(totalUser):
videoUsers[index].videoLoad()
totalTime = videoUsers[0].get_time()
# 主循环
env = gym.make('MyEnv-v1')
frames = []
his_vec = []
thredhold = args.thred * th.ones(args.testNum + args.trainNum)
for iteration in range(totalTime):
# edge list for GCN
historyRecord = [] # 用户行为的历史记录
futureRecord = [] # 用户行为未来记录
node_list = [] # 用户和瓦片的节点列表
edge_list_p = [] # 正采样图神经网络
edge_list_n = [] # 负采样图神经网络
his_and_fut_list = [] # 历史和未来图关系整合
labels = [] # 真实的观看记录标签
pre_u_embeddings = [] # 初始用户的embeddings
pre_t_embeddings = [] # 初始瓦片的embeddings
node_feature = []
# 获取所有用户历史记录
for index, client in enumerate(videoUsers):
# 获取历史的观看记录
hist_vec, _ = client.get_history()
historyRecord.append(hist_vec)
# 获取未来的观看记录
next_vec, view_point_fix = client.get_nextView()
futureRecord.append(next_vec)
# 获取综合观看记录
total_vec = hist_vec + next_vec
total_vec = np.array(total_vec).astype(np.int32)
his_and_fut_list.append(total_vec.reshape(2 * args.window, args.tileNum ** 2))
labels.append(np.array(next_vec).reshape(args.window, args.tileNum ** 2))
# 可视化选项
if index == args.visId:
view_point = view_point_fix
vec_x = (view_point_fix[-1][0] - view_point_fix[0][0]) * 100
vec_y = (view_point_fix[-1][1] - view_point_fix[0][1]) * 100
distance = (vec_x ** 2 + vec_y ** 2) ** 0.5
# 初始的用户embeddings设置
pre_u_embeddings.append(hist_vec)
# 构建用户和瓦片的节点列表
n_node = totalUser + args.tileNum ** 2
for index1 in range(args.window * 2):
node_tmp = []
for index1 in range(n_node):
node_tmp.append((index1, {}))
node_list.append(node_tmp)
for index1 in range(args.tileNum ** 2):
pre_u_embeddings.append(np.random.rand(200).tolist())
# 计算每帧用户的相似关系
for index1 in range(args.window * 2):
edge_tmp = []
for index2, value2 in enumerate(his_and_fut_list):
for index3, value3 in enumerate(his_and_fut_list[index2 + 1:]):
similarity = np.sum(np.trunc((np.sum([value3[index1], value2[index1]], axis=0))) != 1)
if similarity > args.threshold:
edge_tmp.append((index2, index2 + 1 + index3))
# 构建用户观看瓦片的关系
for index2, value2 in enumerate(his_and_fut_list):
for index3, value3 in enumerate(value2[index1]):
if value3 == 1:
edge_tmp.append((index2, totalUser + index3))
# 构建瓦片与瓦片之间的位置关系图,相邻的瓦片存在相似关系
for index2 in range(args.tileNum):
tileTmp1 = index2 * 5
for index3 in range(args.tileNum - 1):
edge_tmp.append((totalUser + tileTmp1 + index3,
totalUser + tileTmp1 + index3 + 1))
for index2 in range(args.tileNum):
tileTmp2 = index2
for index3 in range(args.tileNum - 1):
edge_tmp.append((totalUser + tileTmp2 + index3 * 5,
totalUser + tileTmp2 + (index3 + 1) * 5))
edge_list_p.append(edge_tmp)
node_feature.append(pre_u_embeddings)
futureRecord = np.array(futureRecord)
'''
for index1 in range(totalUser):
label_one = []
for index2 in range(args.window):
label_one.append(futureRecord[index1][index2::args.window])
labels.append(label_one)
'''
k = args.input_dim
node_feature = th.tensor(node_feature).to(th.float32).to('cuda:0')
TP, TN, FP, FN = 0, 0, 0, 0
PredictedTile = 0
startT1 = time()
for index1 in range(totalUser - 1):
n_graphs = []
graphs = []
for index2 in range(args.window * 2):
Graph = build_graph(edge_list_p[index2],
node_feature[index2])
nGraph = construct_negative_graph(Graph, 1)
nGraph.ndata['feat'] = node_feature[index2]
graphs.append(Graph)
n_graphs.append(nGraph)
if args.model == 'EvolveGCN-O':
model = EvolveGCNO(in_feats=k,
n_hidden=args.n_hidden,
num_layers=args.n_layers,
n_classes=args.n_output)
elif args.model == 'EvolveGCN-H':
model = EvolveGCNH(in_feats=k,
num_layers=args.n_layers)
model = model.to(device)
optimizer = th.optim.Adam(model.parameters(), lr=args.lr)
for i in range(args.window, args.window * 2 - 1):
g_list = graphs[i - args.window:i]
ng_list = n_graphs[i - args.window:i]
for epoch in range(args.epochGCN):
model.train()
# get predictions which has label
pos_score, neg_score = model(g_list, ng_list, node_feature[i - args.window:i])
loss = compute_loss(pos_score, neg_score)
optimizer.zero_grad()
loss.backward()
print(loss)
optimizer.step()
# 设置端点处
node_embeddings = model.saga(graphs[i - args.window:i + 1], node_feature[i - args.window:i + 1])
user_embeddings = node_embeddings[0:totalUser]
tile_embeddings = node_embeddings[totalUser - 1:-1]
thredhold[index1] = sum(his_and_fut_list[index1][i]) * 2
if thredhold[index1] == 0:
thredhold[index1] = 1
result = model.predict(user_embeddings[index1].reshape(1, args.n_output), tile_embeddings, thredhold[index1])
for index2, value2 in enumerate(his_and_fut_list[index1][i]):
if value2 == 1 and result[0, index2] == 1: # result[index1, index2]
TP += 1
PredictedTile += 1
elif value2 == 1 and result[0, index2] == 0: # result[index1, index2]
FP += 1
elif value2 == 0 and result[0, index2] == 1: # result[index1, index2]
FN += 1
PredictedTile += 1
elif value2 == 0 and result[0, index2] == 0: # result[index1, index2]
TN += 1
"""
for index, node_one in enumerate(node_embeddings):
user_embeddings = node_one[0:totalUser]
tile_embeddings = node_one[totalUser - 1:-1]
result = model.predict(user_embeddings[index1].reshape(1, k), tile_embeddings, thredhold[index1])
for index2, value2 in enumerate(labels[index1][index]):
if value2 == 1 and result[0, index2] == 1: # result[index1, index2]
TP += 1
PredictedTile += 1
elif value2 == 1 and result[0, index2] == 0: # result[index1, index2]
FP += 1
elif value2 == 0 and result[0, index2] == 1: # result[index1, index2]
FN += 1
PredictedTile += 1
elif value2 == 0 and result[0, index2] == 0: # result[index1, index2]
TN += 1
"""
endT1 = time()
totalT1 = endT1 - startT1
if TP + TN + FP + FN == 0:
accuracy = 0
else:
accuracy = (TP + TN) / (TP + TN + FP + FN)
if (TP + FP) == 0:
precision = 0
else:
precision = TP / (TP + FP)
if (TP + FN) == 0:
recall = 0
else:
recall = TP / (TP + FN)
avePreTile = PredictedTile / 8 # / (8 * args.testNum)
if precision >= 0.8 and recall < 0.6:
thredhold[index1] += -1
elif precision < 0.8:
thredhold[index1] += +1
aveTime = totalT1 # / (args.testNum + args.trainNum)
print("accuracy:", str(accuracy),
"precision:", str(precision),
"recall:", str(recall),
"predicted tile:", str(avePreTile),
"Train Time:", str(aveTime))
sortedValues = [str(accuracy), str(precision), str(recall), str(avePreTile), str(aveTime)]
videoUsers[index1].allWriter.writerCSVA(sortedValues)
display_frames_as_gif(args.policy, frames)