-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathmodel.py
139 lines (111 loc) · 6.32 KB
/
model.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
from layers import GraphConvolution, InnerProductDecoder
import tensorflow as tf
from tensorflow.contrib import layers
flags = tf.app.flags
FLAGS = flags.FLAGS
class Model(object):
def __init__(self, **kwargs):
allowed_kwargs = {'name', 'logging'}
for kwarg in kwargs.keys():
assert kwarg in allowed_kwargs, 'Invalid keyword argument: ' + kwarg
for kwarg in kwargs.keys():
assert kwarg in allowed_kwargs, 'Invalid keyword argument: ' + kwarg
name = kwargs.get('name')
if not name:
name = self.__class__.__name__.lower()
self.name = name
logging = kwargs.get('logging', False)
self.logging = logging
self.vars = {}
def _build(self):
raise NotImplementedError
def build(self):
""" Wrapper for _build() """
with tf.variable_scope(self.name):
self._build()
variables = tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES, scope=self.name)
self.vars = {var.name: var for var in variables}
def fit(self):
pass
def predict(self):
pass
class MultiEncoder(Model):
def __init__(self, placeholders, num_features, features_nonzero, **kwargs):
super(MultiEncoder, self).__init__(**kwargs)
self.inputs1 = placeholders['features1']
self.inputs2 = placeholders['features2']
self.input_dim1 = num_features
self.input_dim2 = num_features
self.features_nonzero = features_nonzero
self.weight = tf.Variable(1.0e-4 * tf.ones(shape=(3025, 3025)), name="weight")
self.coef = self.weight - tf.matrix_diag(tf.diag_part(self.weight))
self.adj1 = placeholders['adjs1']
self.adj2 = placeholders['adjs2']
self.dropout = placeholders['dropout']
self.build()
def _build(self):
with tf.variable_scope('Encoder1', reuse=None):
self.hidden1 = GraphConvolution(input_dim=self.input_dim1,
output_dim=FLAGS.hidden1,
adj=self.adj1,
act=tf.nn.relu,
dropout=self.dropout,
logging=self.logging,
name='e_dense_11')(self.inputs1)
self.noise1 = gaussian_noise_layer(self.hidden1, 0.1)
self.embeddings1 = GraphConvolution(input_dim=FLAGS.hidden1,
output_dim=FLAGS.hidden2,
adj=self.adj1,
act=lambda x: x,
dropout=self.dropout,
logging=self.logging,
name='e_dense_21')(self.noise1)
self.z_mean1 = self.embeddings1
self.ZC1 = tf.matmul(self.coef, self.embeddings1)
layer_flat1, num_features1 = self.flatten_layer(self.z_mean1)
layer_full1 = tf.layers.dense(inputs=layer_flat1, units=1024, activation=None,
kernel_initializer=layers.variance_scaling_initializer(dtype=tf.float32))
self.SZ1 = tf.layers.dense(inputs=layer_full1, units=3, activation=None,
kernel_initializer=layers.variance_scaling_initializer(dtype=tf.float32))
self.reconstructions1 = InnerProductDecoder(input_dim=FLAGS.hidden2,
act=lambda x: x,
logging=self.logging)(self.ZC1)
with tf.variable_scope('Encoder2', reuse=None):
self.hidden2 = GraphConvolution(input_dim=self.input_dim2,
output_dim=FLAGS.hidden1,
adj=self.adj2,
act=tf.nn.relu,
dropout=self.dropout,
logging=self.logging,
name='e_dense_12')(self.inputs2)
self.noise2 = gaussian_noise_layer(self.hidden2, 0.1)
self.embeddings2 = GraphConvolution(input_dim=FLAGS.hidden1,
output_dim=FLAGS.hidden2,
adj=self.adj2,
act=lambda x: x,
dropout=self.dropout,
logging=self.logging,
name='e_dense_22')(self.noise2)
self.z_mean2 = self.embeddings2
self.ZC2 = tf.matmul(self.coef, self.embeddings1)
layer_flat2, num_features2 = self.flatten_layer2(self.z_mean2)
layer_full12 = tf.layers.dense(inputs=layer_flat2, units=1024, activation=None,
kernel_initializer=layers.variance_scaling_initializer(dtype=tf.float32))
self.SZ2 = tf.layers.dense(inputs=layer_full12, units=3, activation=None,
kernel_initializer=layers.variance_scaling_initializer(dtype=tf.float32))
self.reconstructions2 = InnerProductDecoder(input_dim=FLAGS.hidden2,
act=lambda x: x,
logging=self.logging)(self.ZC2)
def flatten_layer(self, layer):
layer_shape = layer.get_shape()
num_features = layer_shape[1:4].num_elements()
layer_flat = tf.reshape(layer, [-1, num_features])
return layer_flat, num_features
def flatten_layer2(self, layer):
layer_shape = layer.get_shape()
num_features = layer_shape[1:4].num_elements()
layer_flat = tf.reshape(layer, [-1, num_features])
return layer_flat, num_features
def gaussian_noise_layer(input_layer, std):
noise = tf.random_normal(shape=tf.shape(input_layer), mean=0.0, stddev=std, dtype=tf.float32)
return input_layer + noise