Merge pull request #954 from lilleswing/graph_convs (25b9527e) · Commits · 钟慕尧 / deepchem

deepchem/models/tensorgraph/models/graph_models.py

+13 −13

Original line number	Diff line number	Diff line
		@@ -12,7 +12,7 @@ from deepchem.models.tensorgraph.graph_layers import WeaveGather, \
		DTNNEmbedding, DTNNStep, DTNNGather, DAGLayer, \
		DAGGather, DTNNExtract, MessagePassing, SetGather
		from deepchem.models.tensorgraph.graph_layers import WeaveLayerFactory
		from deepchem.models.tensorgraph.layers import Dense, Concat, SoftMax, \
		from deepchem.models.tensorgraph.layers import Dense, SoftMax, \
		SoftMaxCrossEntropy, GraphConv, BatchNorm, \
		GraphPool, GraphGather, WeightedError, Dropout, BatchNormalization, Stack, Flatten, GraphCNN, GraphCNNPool
		from deepchem.models.tensorgraph.layers import L2Loss, Label, Weights, Feature
		@@ -116,7 +116,7 @@ class WeaveTensorGraph(TensorGraph):
		cost = L2Loss(in_layers=[label, regression])
		costs.append(cost)
		if self.mode == "classification":
		all_cost = Concat(in_layers=costs, axis=1)
		all_cost = Stack(in_layers=costs, axis=1)
		elif self.mode == "regression":
		all_cost = Stack(in_layers=costs, axis=1)
		self.weights = Weights(shape=(None, self.n_tasks))
		@@ -174,8 +174,8 @@ class WeaveTensorGraph(TensorGraph):
		atom_feat.append(mol.get_atom_features())
		# pair features
		pair_feat.append(
		np.reshape(mol.get_pair_features(), (n_atoms * n_atoms,
		self.n_pair_feat)))
		np.reshape(mol.get_pair_features(),
		(n_atoms * n_atoms, self.n_pair_feat)))

		feed_dict[self.atom_features] = np.concatenate(atom_feat, axis=0)
		feed_dict[self.pair_features] = np.concatenate(pair_feat, axis=0)
		@@ -310,8 +310,8 @@ class DTNNTensorGraph(TensorGraph):
		num_atoms = list(map(sum, X_b.astype(bool)[:, :, 0]))
		atom_number = [
		np.round(
		np.power(2 * np.diag(X_b[i, :num_atoms[i], :num_atoms[i]]), 1 /
		2.4)).astype(int) for i in range(len(num_atoms))
		np.power(2 * np.diag(X_b[i, :num_atoms[i], :num_atoms[i]]),
		1 / 2.4)).astype(int) for i in range(len(num_atoms))
		]
		start = 0
		for im, molecule in enumerate(atom_number):
		@@ -425,7 +425,7 @@ class DAGTensorGraph(TensorGraph):
		cost = L2Loss(in_layers=[label, regression])
		costs.append(cost)
		if self.mode == "classification":
		all_cost = Concat(in_layers=costs, axis=1)
		all_cost = Stack(in_layers=costs, axis=1)
		elif self.mode == "regression":
		all_cost = Stack(in_layers=costs, axis=1)
		self.weights = Weights(shape=(None, self.n_tasks))
		@@ -570,7 +570,7 @@ class PetroskiSuchTensorGraph(TensorGraph):
		cost = L2Loss(in_layers=[label, regression])
		costs.append(cost)
		if self.mode == "classification":
		entropy = Concat(in_layers=costs, axis=-1)
		entropy = Stack(in_layers=costs, axis=-1)
		elif self.mode == "regression":
		entropy = Stack(in_layers=costs, axis=1)
		self.my_task_weights = Weights(shape=(None, self.n_tasks))
		@@ -716,7 +716,7 @@ class GraphConvTensorGraph(TensorGraph):
		cost = L2Loss(in_layers=[label, regression])
		costs.append(cost)
		if self.mode == "classification":
		entropy = Concat(in_layers=costs, axis=-1)
		entropy = Stack(in_layers=costs, axis=1)
		elif self.mode == "regression":
		entropy = Stack(in_layers=costs, axis=1)
		self.my_task_weights = Weights(shape=(None, self.n_tasks))
		@@ -994,7 +994,7 @@ class MPNNTensorGraph(TensorGraph):
		cost = L2Loss(in_layers=[label, regression])
		costs.append(cost)
		if self.mode == "classification":
		all_cost = Concat(in_layers=costs, axis=1)
		all_cost = Stack(in_layers=costs, axis=1)
		elif self.mode == "regression":
		all_cost = Stack(in_layers=costs, axis=1)
		self.weights = Weights(shape=(None, self.n_tasks))
		@@ -1051,8 +1051,8 @@ class MPNNTensorGraph(TensorGraph):
		atom_feat.append(mol.get_atom_features())
		# pair features
		pair_feat.append(
		np.reshape(mol.get_pair_features(), (n_atoms * n_atoms,
		self.n_pair_feat)))
		np.reshape(mol.get_pair_features(),
		(n_atoms * n_atoms, self.n_pair_feat)))

		feed_dict[self.atom_features] = np.concatenate(atom_feat, axis=0)
		feed_dict[self.pair_features] = np.concatenate(pair_feat, axis=0)

deepchem/models/tensorgraph/models/test_graph_models.py

+10 −4

Original line number	Diff line number	Diff line
		@@ -11,21 +11,27 @@ from deepchem.molnet.load_function.delaney_datasets import load_delaney

		class TestGraphModels(unittest.TestCase):

		def get_dataset(self, mode='classification', featurizer='GraphConv'):
		def get_dataset(self,
		mode='classification',
		featurizer='GraphConv',
		num_tasks=2):
		data_points = 10
		tasks, all_dataset, transformers = load_delaney(featurizer)
		train, valid, test = all_dataset
		for i in range(1, num_tasks):
		tasks.append("random_task")
		w = np.ones(shape=(data_points, len(tasks)))

		if mode == 'classification':
		y = np.random.randint(0, 2, size=(data_points, len(tasks)))
		metric = deepchem.metrics.Metric(
		deepchem.metrics.roc_auc_score, np.mean, mode="classification")
		if mode == 'regression':
		else:
		y = np.random.normal(size=(data_points, len(tasks)))
		metric = deepchem.metrics.Metric(
		deepchem.metrics.mean_absolute_error, mode="regression")

		ds = NumpyDataset(train.X[:10], y, train.w[:10], train.ids[:10])
		ds = NumpyDataset(train.X[:10], y, w, train.ids[:10])

		return tasks, ds, transformers, metric

		@@ -61,7 +67,7 @@ class TestGraphModels(unittest.TestCase):

		def test_graph_conv_error_bars(self):
		tasks, dataset, transformers, metric = self.get_dataset(
		'regression', 'GraphConv')
		'regression', 'GraphConv', num_tasks=1)

		batch_size = 50
		model = GraphConvTensorGraph(

deepchem/models/tensorgraph/tests/test_layers.py

+1 −0

Original line number	Diff line number	Diff line
		@@ -280,6 +280,7 @@ class TestLayers(test_util.TensorFlowTestCase):
		with self.test_session() as sess:
		result = Log()(value).eval()
		assert np.array_equal(np.log(value), result)
		assert np.all(np.isclose(np.log(value), result, atol=0.001))

		def test_exp(self):
		"""Test that Exp can be invoked."""

Admin message