forked from hyp1231/GMPT
-
Notifications
You must be signed in to change notification settings - Fork 3
/
batch.py
172 lines (137 loc) · 5.88 KB
/
batch.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
import torch
from torch_geometric.data import Data
class BatchMultiView(Data):
r"""A plain old python object modeling a batch of graphs as one big
(dicconnected) graph. With :class:`torch_geometric.data.Data` being the
base class, all its methods can also be used here.
In addition, single graphs can be reconstructed via the assignment vector
:obj:`batch`, which maps each node to its respective graph identifier.
"""
def __init__(self, batch=None, **kwargs):
super(BatchMultiView, self).__init__(**kwargs)
self.batch = batch
@staticmethod
def from_data_list(data_list):
r"""Constructs a batch object from a python list holding
:class:`torch_geometric.data.Data` objects.
The assignment vector :obj:`batch` is created on the fly."""
keys = [set(data.keys) for data in data_list]
keys = [_ for _ in set.union(*keys) if 'view' not in _]
assert 'batch' not in keys
batch = BatchMultiView()
for key in keys:
batch[key] = []
batch.batch = []
cumsum_node = 0
cumsum_edge = 0
cumsum_graph = 0
for view in ['view1_', 'view2_']:
for i, data in enumerate(data_list):
num_nodes = data[view + 'x'].shape[0]
batch.batch.append(torch.full((num_nodes, ), i + cumsum_graph, dtype=torch.long))
for key in keys:
if key in ['edge_index', 'edge_attr', 'x']:
item = data[view + key]
else:
item = data[key]
if key in ['edge_index', 'center_node_idx']:
item = item + cumsum_node
batch[key].append(item)
cumsum_node += num_nodes
cumsum_edge += data[view + 'edge_index'].shape[1]
cumsum_graph += len(data_list)
for key in keys:
batch[key] = torch.cat(
batch[key], dim=data_list[0].__cat_dim__(key, batch[key][0]))
batch.batch = torch.cat(batch.batch, dim=-1)
return batch.contiguous()
@property
def num_graphs(self):
"""Returns the number of graphs in the batch."""
return self.batch[-1].item() + 1
class BatchFinetune(Data):
r"""A plain old python object modeling a batch of graphs as one big
(dicconnected) graph. With :class:`torch_geometric.data.Data` being the
base class, all its methods can also be used here.
In addition, single graphs can be reconstructed via the assignment vector
:obj:`batch`, which maps each node to its respective graph identifier.
"""
def __init__(self, batch=None, **kwargs):
super(BatchFinetune, self).__init__(**kwargs)
self.batch = batch
@staticmethod
def from_data_list(data_list):
r"""Constructs a batch object from a python list holding
:class:`torch_geometric.data.Data` objects.
The assignment vector :obj:`batch` is created on the fly."""
keys = [set(data.keys) for data in data_list]
keys = list(set.union(*keys))
assert 'batch' not in keys
batch = BatchFinetune()
for key in keys:
batch[key] = []
batch.batch = []
cumsum_node = 0
cumsum_edge = 0
for i, data in enumerate(data_list):
num_nodes = data.num_nodes
batch.batch.append(torch.full((num_nodes, ), i, dtype=torch.long))
for key in data.keys:
item = data[key]
if key in ['edge_index', 'center_node_idx']:
item = item + cumsum_node
batch[key].append(item)
cumsum_node += num_nodes
cumsum_edge += data.edge_index.shape[1]
for key in keys:
batch[key] = torch.cat(
batch[key], dim=data_list[0].__cat_dim__(key, batch[key][0]))
batch.batch = torch.cat(batch.batch, dim=-1)
return batch.contiguous()
@property
def num_graphs(self):
"""Returns the number of graphs in the batch."""
return self.batch[-1].item() + 1
class BatchAE(Data):
r"""A plain old python object modeling a batch of graphs as one big
(dicconnected) graph. With :class:`torch_geometric.data.Data` being the
base class, all its methods can also be used here.
In addition, single graphs can be reconstructed via the assignment vector
:obj:`batch`, which maps each node to its respective graph identifier.
"""
def __init__(self, batch=None, **kwargs):
super(BatchAE, self).__init__(**kwargs)
self.batch = batch
@staticmethod
def from_data_list(data_list):
r"""Constructs a batch object from a python list holding
:class:`torch_geometric.data.Data` objects.
The assignment vector :obj:`batch` is created on the fly."""
keys = [set(data.keys) for data in data_list]
keys = list(set.union(*keys))
assert 'batch' not in keys
batch = BatchAE()
for key in keys:
batch[key] = []
batch.batch = []
cumsum_node = 0
for i, data in enumerate(data_list):
num_nodes = data.num_nodes
batch.batch.append(torch.full((num_nodes, ), i, dtype=torch.long))
for key in data.keys:
item = data[key]
if 'edge_index' in key or 'center_node_idx' in key:
item = item + cumsum_node
batch[key].append(item)
cumsum_node += num_nodes
for key in keys:
batch[key] = torch.cat(
batch[key], dim=batch.cat_dim(key))
batch.batch = torch.cat(batch.batch, dim=-1)
return batch.contiguous()
@property
def num_graphs(self):
"""Returns the number of graphs in the batch."""
return self.batch[-1].item() + 1
def cat_dim(self, key):
return -1 if 'edge_index' in key else 0