def test_size(self):
n_inputs = 10
nn = NeuralNetwork()
self.assertEqual(nn.size(), 0)
x = tf.ones([1,n_inputs])
input_layer = Layer(n_units=n_inputs,activation=x)
nn.add_layer(input_layer)
self.assertEqual(nn.size(), 1)
def test_connect(self):
n_inputs = 10
n_hidden = 20
x = tf.ones([1, n_inputs])
input_layer = Layer(n_units=n_inputs,activation=x)
id_layer1 = Layer(n_hidden,activation=tf.identity)
id_layer2 = Layer(n_hidden,activation=tf.identity)
nn = NeuralNetwork()
# layer 0
nn.add_layer(input_layer)
# layer 1
nn.add_layer(id_layer1)
# layer 2
nn.add_layer(id_layer2)
self.assertEqual(nn.size(),3)
# connect input to output
self.assertEqual(nn.graph.number_of_edges(),2)
nn.connect_layers(0, 2)
self.assertEqual(nn.graph.number_of_edges(),3)
self.assertIsNot(nn.weights(0,1),nn.weights(0,2))
self.assertIsNot(nn.weights(0,1),nn.weights(1,2))
self.assertIsNot(nn.weights(1,2),nn.weights(0,2))
def test_equivalent_networks(self):
"""
Create two equivalent neural networks
nn1 = x -> 1 -> div(x/in) -> w -> sigm -> o
nn2 = x -> w -> sigm -> 0
(with shared weights)
"""
x = tf.ones([1, 4])
input_layer = Layer(n_units=4,activation=x)
weights = tf.ones([4,4])
shared_w = tf.Variable(normalised_weight_init(4,4))
nn1 = NeuralNetwork(input_layer)
l11 = Layer(4, lambda x_in: tf.div(x_in, 4))
l21 = Layer(4, tf.nn.sigmoid)
nn1.add_layer(l11,biased=False,shared_weights=weights)
nn1.add_layer(l21,biased=False,shared_weights=shared_w)
self.assertEqual(nn1.graph.number_of_edges(), 2)
self.assertEqual(nn1.size(),3)
nn2 = NeuralNetwork(Layer(4,x))
l12 = Layer(4, tf.nn.sigmoid)
nn2.add_layer(l12,biased=False,shared_weights=shared_w)
self.assertEqual(nn2.size(),2)
self.assertEqual(nn2.graph.number_of_edges(),1)
init_vars = tf.initialize_all_variables()
with tf.Session() as ss:
ss.run(init_vars)
w1 = ss.run(nn1.weights(1,2))
w2 = ss.run(nn2.weights(0,1))
# shared weights should be the same in both networks
self.assertTrue(np.array_equal(w1,w2))
out_nn1 = ss.run(nn1.output())
out_nn2 = ss.run(nn2.output())
# outputs should be the same since the networks are equivalent
self.assertTrue(np.array_equal(out_nn1,out_nn2))
def test_append(self):
n_inputs = 10
n_hidden = 20
x = tf.ones([1, n_inputs])
input_layer = Layer(n_units=n_inputs,activation=x)
nn = NeuralNetwork(input_layer)
self.assertEqual(nn.size(), 1)
# add an identity layer
id_layer = Layer(n_hidden,activation=tf.identity)
nn.add_layer(id_layer,biased=True)
self.assertEqual(nn.size(), 2)
# output node should have the last layer added
last_layer = nn._output_node[GraphKeys.LAYER]
self.assertIs(last_layer,id_layer)
def test_layer(self):
n_inputs = 10
n_hidden = 20
x = tf.ones([1, n_inputs])
input_layer = Layer(n_units=n_inputs, activation=x)
nn = NeuralNetwork(input_layer)
self.assertEqual(nn.size(), 1)
# add a layer with biases
hidden_layer = Layer(n_hidden, activation=tf.identity)
nn.add_layer(hidden_layer, biased=True)
layer_0 = nn.layer(0)
layer_1 = nn.layer(1)
self.assertEqual(layer_0, input_layer)
self.assertEqual(layer_1, hidden_layer)
