Merge pull request #6 from deepchem/master (db47d907) · Commits · 钟慕尧 / deepchem

contrib/DiabeticRetinopathy/data.py

+1 −1

Original line number	Diff line number	Diff line
		@@ -46,7 +46,7 @@ def load_images_DR(split='random', seed=None):
		(-1, 1))
		image_full_paths = [os.path.join(images_path, n) for n in image_names]

		classes, cts = np.unique(all_labels.values(), return_counts=True)
		classes, cts = np.unique(list(all_labels.values()), return_counts=True)
		weight_ratio = dict(zip(classes, np.max(cts) / cts.astype(float)))
		weights = np.array([weight_ratio[l[0]] for l in labels]).reshape((-1, 1))

deepchem/models/tensorgraph/layers.py

+1 −0

Original line number	Diff line number	Diff line
		@@ -2,6 +2,7 @@
		from __future__ import division
		import random
		import string
		import warnings
		from collections import Sequence
		from copy import deepcopy

deepchem/models/tensorgraph/tensor_graph.py

+13 −2

Original line number	Diff line number	Diff line
		@@ -974,16 +974,27 @@ class TensorGraph(Model):

		def get_layer_variables(self, layer):
		"""Get the list of trainable variables in a layer of the graph."""
		if tfe.in_eager_mode():
		return layer.variables
		if not self.built:
		self.build()
		with self._get_tf("Graph").as_default():
		if tfe.in_eager_mode():
		return layer.variables
		if layer.variable_scope == '':
		return []
		return tf.get_collection(
		tf.GraphKeys.TRAINABLE_VARIABLES, scope=layer.variable_scope)

		def get_layer_variable_values(self, layer):
		"""Get the variable values associated with a given layer """

		layer_variables = self.get_layer_variables(layer)
		with self._get_tf("Graph").as_default():
		if tfe.in_eager_mode():
		return [v.numpy() for v in layer_variables]
		if len(layer_variables) == 0:
		return []
		return self.session.run(layer_variables)

		def get_variables(self):
		"""Get the list of all trainable variables in the graph."""
		if not self.built:

deepchem/models/tensorgraph/tests/test_tensor_graph.py

+41 −1

Original line number	Diff line number	Diff line
		@@ -12,7 +12,7 @@ import deepchem as dc
		from deepchem.data import NumpyDataset
		from deepchem.data.datasets import Databag
		from deepchem.models.tensorgraph.layers import Dense, SoftMaxCrossEntropy, ReduceMean, ReduceSum, SoftMax, Constant, Variable
		from deepchem.models.tensorgraph.layers import Feature, Label
		from deepchem.models.tensorgraph.layers import Feature, Label, Input
		from deepchem.models.tensorgraph.layers import ReduceSquareDifference, Add, GRU
		from deepchem.models.tensorgraph.tensor_graph import TensorGraph
		from deepchem.models.tensorgraph.optimizers import GradientDescent, ExponentialDecay, Adam
		@@ -576,3 +576,43 @@ class TestTensorGraph(unittest.TestCase):
		(1, n_features))).flatten()
		self.assertAlmostEqual(
		pred1[task], (pred2 + norm * delta)[task], places=4)

		def test_get_layer_variable_values(self):
		"""Test to get the variable values associated with a layer"""
		# Test for correct value return (normal mode)
		tg = dc.models.TensorGraph()
		var = Variable([10.0, 12.0])
		tg.add_output(var)
		expected = [10.0, 12.0]
		obtained = tg.get_layer_variable_values(var)[0]
		np.testing.assert_array_equal(expected, obtained)

		# Test for shapes (normal mode)
		tg = dc.models.TensorGraph()
		input_tensor = Input(shape=(10, 100))
		output = Dense(out_channels=20, in_layers=[input_tensor])
		tg.add_output(output)
		expected_shape = (100, 20)
		obtained_shape = tg.get_layer_variable_values(output)[0].shape
		assert expected_shape == obtained_shape

		def test_get_layer_variable_values_eager(self):
		"""Tests to get variable values associated with a layer in eager mode"""

		with context.eager_mode():
		# Test for correct value return (eager mode)
		tg = dc.models.TensorGraph()
		var = Variable([10.0, 12.0])
		tg.add_output(var)
		expected = [10.0, 12.0]
		obtained = tg.get_layer_variable_values(var)[0]
		np.testing.assert_array_equal(expected, obtained)

		# Test for shape (eager mode)
		tg = dc.models.TensorGraph()
		input_tensor = Input(shape=(10, 100))
		output = Dense(out_channels=20, in_layers=[input_tensor])
		tg.add_output(output)
		expected_shape = (100, 20)
		obtained_shape = tg.get_layer_variable_values(output)[0].shape
		assert expected_shape == obtained_shape

deepchem/molnet/load_function/pdbbind_datasets.py

+1 −1

Original line number	Diff line number	Diff line
		@@ -244,7 +244,7 @@ def load_pdbbind(featurizer="grid", split="random", subset="core", reload=True):
		# No transformations of data
		transformers = []
		if split == None:
		return tasks, (dataset, None, None), transformers
		return pdbbind_tasks, (dataset, None, None), transformers

		# TODO(rbharath): This should be modified to contain a cluster split so
		# structures of the same protein aren't in both train/test

Admin message