Merge branch 'master' of https://github.com/deepchem/deepchem into pddbind_load (7faea226) · Commits · 钟慕尧 / deepchem

deepchem/models/tensorgraph/fcnet.py

+2 −0

Original line number	Diff line number	Diff line
		@@ -113,6 +113,7 @@ class MultitaskClassifier(TensorGraph):
		prev_layer = layer

		# Compute the loss function for each label.
		self.neural_fingerprint = prev_layer

		logits = Reshape(
		shape=(-1, n_tasks, n_classes),
		@@ -257,6 +258,7 @@ class MultitaskRegressor(TensorGraph):
		if dropout > 0.0:
		layer = Dropout(dropout, in_layers=[layer])
		prev_layer = layer
		self.neural_fingerprint = prev_layer

		# Compute the loss function for each label.

+10 −4

Original line number	Diff line number	Diff line
		@@ -643,7 +643,7 @@ class GraphConvModel(TensorGraph):
		batch_norm3 = BatchNorm(in_layers=[dense])
		if self.dropout[-1] > 0.0:
		batch_norm3 = Dropout(self.dropout[-1], in_layers=batch_norm3)
		readout = GraphGather(
		self.neural_fingerprint = GraphGather(
		batch_size=self.batch_size,
		activation_fn=tf.nn.tanh,
		in_layers=[batch_norm3, self.degree_slice, self.membership] +
		@@ -657,7 +657,9 @@ class GraphConvModel(TensorGraph):
		logits = Reshape(
		shape=(None, n_tasks, n_classes),
		in_layers=[
		Dense(in_layers=readout, out_channels=n_tasks * n_classes)
		Dense(
		in_layers=self.neural_fingerprint,
		out_channels=n_tasks * n_classes)
		])
		logits = TrimGraphOutput([logits, weights])
		output = SoftMax(logits)
		@@ -669,13 +671,17 @@ class GraphConvModel(TensorGraph):
		labels = Label(shape=(None, n_tasks))
		output = Reshape(
		shape=(None, n_tasks),
		in_layers=[Dense(in_layers=readout, out_channels=n_tasks)])
		in_layers=[
		Dense(in_layers=self.neural_fingerprint, out_channels=n_tasks)
		])
		output = TrimGraphOutput([output, weights])
		self.add_output(output)
		if self.uncertainty:
		log_var = Reshape(
		shape=(None, n_tasks),
		in_layers=[Dense(in_layers=readout, out_channels=n_tasks)])
		in_layers=[
		Dense(in_layers=self.neural_fingerprint, out_channels=n_tasks)
		])
		log_var = TrimGraphOutput([log_var, weights])
		var = Exp(log_var)
		self.add_variance(var)

+19 −0

Original line number	Diff line number	Diff line
		@@ -62,6 +62,25 @@ class TestGraphModels(unittest.TestCase):
		assert np.allclose(scores['mean-roc_auc_score'],
		scores2['mean-roc_auc_score'])

		def test_neural_fingerprint_retrieval(self):
		tasks, dataset, transformers, metric = self.get_dataset(
		'classification', 'GraphConv')

		fp_size = 3

		batch_size = 50
		model = GraphConvModel(
		len(tasks),
		batch_size=batch_size,
		dense_layer_size=3,
		mode='classification')

		model.fit(dataset, nb_epoch=1)
		neural_fingerprints = model.predict(
		dataset, outputs=model.neural_fingerprint)
		neural_fingerprints = np.array(neural_fingerprints)[:len(dataset)]
		self.assertEqual((len(dataset), fp_size * 2), neural_fingerprints.shape)

		def test_graph_conv_regression_model(self):
		tasks, dataset, transformers, metric = self.get_dataset(
		'regression', 'GraphConv')