def test_graphsage_passing_regularisers():
with pytest.raises(ValueError):
GraphSAGE(layer_sizes=[4],
n_samples=[2],
input_dim=2,
kernel_initializer="fred")
GraphSAGE(layer_sizes=[4],
n_samples=[2],
input_dim=2,
kernel_initializer="ones")
GraphSAGE(
layer_sizes=[4],
n_samples=[2],
input_dim=2,
kernel_initializer=initializers.ones(),
)
GraphSAGE(
layer_sizes=[4],
n_samples=[2],
input_dim=2,
kernel_regularizer=regularizers.l2(0.01),
)
with pytest.raises(ValueError):
GraphSAGE(layer_sizes=[4],
n_samples=[2],
input_dim=2,
kernel_regularizer="wilma")
def get_model(input_shape):
model = Sequential([
Dense(units=64,
input_shape=input_shape,
kernel_initializer=he_uniform(),
bias_initializer=ones(),
activation=relu),
Dense(units=128, activation=relu),
Dense(units=128, activation=relu),
Dense(units=128, activation=relu),
Dense(units=128, activation=relu),
Dense(units=64, activation=relu),
Dense(units=64, activation=relu),
Dense(units=64, activation=relu),
Dense(units=64, activation=relu),
Dense(units=3, activation=softmax)
])
return model
def get_regularized_model(input_shape, dropout_rate, weight_decay):
model = Sequential([
Dense(units=64,
input_shape=input_shape,
kernel_initializer=he_uniform(),
bias_initializer=ones(),
activation=relu,
kernel_regularizer=l2(weight_decay)),
Dense(units=128, activation=relu, kernel_regularizer=l2(weight_decay)),
Dense(units=128, activation=relu, kernel_regularizer=l2(weight_decay)),
Dropout(rate=dropout_rate),
Dense(units=128, activation=relu, kernel_regularizer=l2(weight_decay)),
Dense(units=128, activation=relu, kernel_regularizer=l2(weight_decay)),
BatchNormalization(),
Dense(units=64, activation=relu, kernel_regularizer=l2(weight_decay)),
Dense(units=64, activation=relu, kernel_regularizer=l2(weight_decay)),
Dropout(rate=dropout_rate),
Dense(units=64, activation=relu, kernel_regularizer=l2(weight_decay)),
Dense(units=64, activation=relu, kernel_regularizer=l2(weight_decay)),
Dense(units=3, activation=softmax)
])
return model
def build(self, input_shapes):
"""
Builds the layer
Args:
input_shapes (list of int): shapes of the layer's inputs (node features and adjacency matrix)
"""
feat_shape = input_shapes[0]
input_dim = int(feat_shape[-1])
# Variables to support integrated gradients
self.delta = self.add_weight(name="ig_delta",
shape=(),
trainable=False,
initializer=initializers.ones())
self.non_exist_edge = self.add_weight(
name="ig_non_exist_edge",
shape=(),
trainable=False,
initializer=initializers.zeros(),
)
# Initialize weights for each attention head
for head in range(self.attn_heads):
# Layer kernel
kernel = self.add_weight(
shape=(input_dim, self.units),
initializer=self.kernel_initializer,
regularizer=self.kernel_regularizer,
constraint=self.kernel_constraint,
name="kernel_{}".format(head),
)
self.kernels.append(kernel)
# # Layer bias
if self.use_bias:
bias = self.add_weight(
shape=(self.units, ),
initializer=self.bias_initializer,
regularizer=self.bias_regularizer,
constraint=self.bias_constraint,
name="bias_{}".format(head),
)
self.biases.append(bias)
# Attention kernels
attn_kernel_self = self.add_weight(
shape=(self.units, 1),
initializer=self.attn_kernel_initializer,
regularizer=self.attn_kernel_regularizer,
constraint=self.attn_kernel_constraint,
name="attn_kernel_self_{}".format(head),
)
attn_kernel_neighs = self.add_weight(
shape=(self.units, 1),
initializer=self.attn_kernel_initializer,
regularizer=self.attn_kernel_regularizer,
constraint=self.attn_kernel_constraint,
name="attn_kernel_neigh_{}".format(head),
)
self.attn_kernels.append([attn_kernel_self, attn_kernel_neighs])
self.built = True
plt.axis('off')
plt.colorbar()
plt.suptitle('Weight matrices variation')
plt.show()
model = Sequential([
Dense(
units=4,
input_shape=(4, ),
activation=relu,
trainable=False, # to freeze the layer
kernel_initializer=random_uniform(),
bias_initializer=ones()),
Dense(units=2,
activation=relu,
kernel_initializer=lecun_uniform(),
bias_initializer=zeros()),
Dense(units=4, activation=softmax)
])
model.summary()
W0_layers = get_weights(model)
b0_layers = get_biases(model)
X_train = np.random.random((100, 4))
y_train = X_train
