-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathq_net_code.py
160 lines (126 loc) · 5.2 KB
/
q_net_code.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
from __future__ import print_function
import os
import sys
import numpy as np
from copy import deepcopy
import torch
from torch.autograd import Variable
from torch.nn.parameter import Parameter
import torch.nn as nn
import torch.nn.functional as F
from torch_geometric.nn import GraphConv, TopKPooling, SAGEConv, GCNConv
from torch_geometric.nn import global_mean_pool as gap, global_max_pool as gmp
# from args import args
def node_greedy_actions(target_nodes, list_q, net):
assert len(target_nodes) == len(list_q)
actions = []
values = []
for i in range(len(target_nodes)):
region = net.list_action_space[target_nodes[i]] # 取出这个目标节点所属动作空间
if region is None:
assert list_q[i].size()[0] == net.total_nodes
else:
assert len(region) == list_q[i].size()[0]
val, act = torch.max(list_q[i], dim=0)
values.append(val)
# print(act_index)
# net.list_q = list_q
assert region is not None
if region is not None:
act = region[act.data.cpu().numpy()[0]]
act = Variable(torch.LongTensor([act]))
actions.append(act)
else:
actions.append(act)
return torch.cat(actions, dim=0).data, torch.cat(values, dim=0).data
class QNetNode(nn.Module):
def __init__(self, data, list_action_space, args,input_node_linear):
super(QNetNode, self).__init__()
self.data = data
self.list_action_space = list_action_space
self.all_selected_actions = set()
self.args = args
self.input_node_linear = input_node_linear
embed_dim = args.latent_dim
if args.bilin_q:
last_wout = embed_dim
else:
last_wout = 1
self.bias_target = Parameter(torch.Tensor(1, embed_dim))
if args.mlp_hidden:
self.linear_1 = nn.Linear(embed_dim*2, args.mlp_hidden)
self.linear_out = nn.Linear(args.mlp_hidden, last_wout)
else:
self.linear_out = nn.Linear(embed_dim, last_wout)
# self.conv1 = SAGEConv(self.data.num_features, embed_dim)
self.conv1 = GCNConv(self.data.num_features, embed_dim)
# self.pool1 = TopKPooling(embed_dim, ratio=0.8)
# self.conv2 = SAGEConv(embed_dim, embed_dim)
self.conv2 = GCNConv(embed_dim, embed_dim)
# self.pool2 = TopKPooling(embed_dim, ratio=0.8)
self.lin1 = torch.nn.Linear(128, 128)
# self.lin2 = torch.nn.Linear(128, 64)
def forward(self, time_t, states, actions, greedy_acts=False):
# input_node_linear = F.relu(self.conv1(self.data.x, self.data.edge_index))
input_node_linear = self.input_node_linear
# input_node_linear, edge_index, _, batch, _, _ = self.pool1(input_node_linear, self.data.edge_index, None, None)
target_nodes, batch_graph = zip(*states)
# print(f"batch graph length is {len(batch_graph)}")
list_pred = []
prefix_sum = []
for i in range(len(batch_graph)):
# print(i)
region = self.list_action_space[target_nodes[i]] # 看看这个region取的节点对应下面取出的embeding和一开始想要选取的节点是不是一致，否则需要更改，
node_embed = input_node_linear #.clone()
new_edges = batch_graph[i].get_new_edges()
# print(new_edges)
# print(i)
# some new edges have been added to the graph
if new_edges is not None:
# print(new_edges)
node_embed = F.relu(self.conv2(node_embed, new_edges.to(self.args.device)))
# node_embed, edge_index, _, batch, _, _ = self.pool2(node_embed, new_edges, None, None)
if not self.args.bilin_q:
node_embed[target_nodes[i]] += self.bias_target[0]
node_embed = F.relu(self.lin1(node_embed))
# graph_embed = torch.mean(node_embed, dim=0, keepdim=True)
target_embed = node_embed[target_nodes[i], :].view(-1, 1)
if region is not None:
node_embed = node_embed[region] # 取出控制的100个动作节点embeding 然后与目标节点想乘
graph_embed = torch.mean(node_embed, dim=0, keepdim=True)
# assert actions is None # 目前都是这么干的
if actions is None:
# pass
graph_embed = graph_embed.repeat(node_embed.size()[0], 1)
else:
if region is not None:
act_idx = region.index(actions[i])
else:
act_idx = actions[i]
node_embed = node_embed[act_idx, :].view(1, -1)
embed_s_a = torch.cat((node_embed, graph_embed), dim=1)
# embed_s_a = node_embed
if self.args.mlp_hidden:
embed_s_a = F.relu(self.linear_1(embed_s_a))
raw_pred = self.linear_out(embed_s_a)
if self.args.bilin_q:
raw_pred = torch.mm(raw_pred, target_embed) # (100,128) * (128,1)
list_pred.append(raw_pred)
if greedy_acts:
# sizes = [q.size() for q in list_pred]
# print(sizes)
actions, _ = node_greedy_actions(target_nodes, list_pred, self)
return actions, list_pred # 会根据本网络中预测q值最大的选择 region中的对应q值最大的那个动作list_action_space[target_node]
class NStepQNetNode(nn.Module):
def __init__(self, num_steps, data, list_action_space):
super(NStepQNetNode, self).__init__()
self.data = data
self.list_action_space = list_action_space
list_mod = []
for i in range(0, num_steps):
list_mod.append(QNetNode(data, list_action_space))
self.list_mod = nn.ModuleList(list_mod)
self.num_steps = num_steps
def forward(self, time_t, states, actions, greedy_acts=False, is_inference=False):
assert time_t >= 0 and time_t < self.num_steps
return self.list_mod[time_t](time_t, states, actions, greedy_acts, is_inference)