modification of description (a79ab39d) · Commits · 钟慕尧 / deepchem

deepchem/models/tensorflow_models/IRV.py

+8 −12

Original line number	Diff line number	Diff line
		"""TensorFlow implementation of fully connected networks.
		"""
		from __future__ import print_function
		from __future__ import division
		from __future__ import unicode_literals
		@@ -15,12 +13,12 @@ from deepchem.models.tensorflow_models.lr import TensorflowLogisticRegression

		class TensorflowMultiTaskIRVClassifier(TensorflowLogisticRegression):

		def __init__(self, n_tasks, K=10, logdir=None, penalty=0.0, n_classes=2,
		def __init__(self, n_tasks, K=10, logdir=None, n_classes=2, penalty=0.0,
		penalty_type="l2", learning_rate=0.001, momentum=.8,
		optimizer="adam", batch_size=50, verbose=True, seed=None,
		**kwargs):

		"""Initialize TensorflowMultiTaskFitTransformRegressor
		"""Initialize TensorflowMultiTaskIRVClassifier

		Parameters
		----------
		@@ -30,6 +28,8 @@ class TensorflowMultiTaskIRVClassifier(TensorflowLogisticRegression):
		Number of nearest neighbours used in classification
		logdir: str
		Location to save data
		n_classes: int
		number of different labels
		penalty: float
		Amount of penalty (l2 or l1 applied)
		penalty_type: str
		@@ -46,8 +46,6 @@ class TensorflowMultiTaskIRVClassifier(TensorflowLogisticRegression):
		Perform logging.
		seed: int
		If not none, is used as random seed for tensorflow.
		fit_transformers: list
		List of dc.trans.FitTransformer objects

		"""

		@@ -64,11 +62,9 @@ class TensorflowMultiTaskIRVClassifier(TensorflowLogisticRegression):
		**kwargs)

		def build(self, graph, name_scopes, training):
		"""Constructs the graph architecture as specified in its config.
		"""Constructs the graph architecture of IRV as described in:

		This method creates the following Placeholders:
		mol_features: Molecule descriptor (e.g. fingerprint) tensor with shape
		batch_size x n_features.
		https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2750043/
		"""
		placeholder_scope = TensorflowGraph.get_placeholder_scope(
		graph, name_scopes)

deepchem/trans/init.py

+1 −1

Original line number	Diff line number	Diff line
		@@ -14,4 +14,4 @@ from deepchem.trans.transformers import BalancingTransformer
		from deepchem.trans.transformers import CDFTransformer
		from deepchem.trans.transformers import PowerTransformer
		from deepchem.trans.transformers import CoulombFitTransformer
		from deepchem.trans.transformers import IRVFitTransformer
		No newline at end of file
		from deepchem.trans.transformers import IRVTransformer

deepchem/trans/transformers.py

+10 −7

Original line number	Diff line number	Diff line
		@@ -568,16 +568,19 @@ class CoulombFitTransformer():
		raise NotImplementedError(
		"Cannot untransform datasets with FitTransformer.")

		class IRVFitTransformer():
		"""Performs randomization and binarization operations on batches of Coulomb Matrix features during fit."""
		class IRVTransformer():
		"""Performs transform from ECFP to IRV features(K nearest neibours)."""
		def __init__(self, K, n_tasks, dataset):

		"""Initializes CoulombFitTransformer.
		"""Initializes IRVTransformer.
		Parameters:
		----------
		dataset: dc.data.Dataset object, y and w should be for one task

		K: number of nearest neighbours that count[
		dataset: dc.data.Dataset object
		train_dataset
		K: int
		number of nearest neighbours being count
		n_tasks: int
		number of tasks

		"""
		self.X = dataset.X
		@@ -612,4 +615,4 @@ class IRVFitTransformer():

		def untransform(self, z):
		raise NotImplementedError(
		"Cannot untransform datasets with FitTransformer.")
		"Cannot untransform datasets with IRVTransformer.")

examples/tox21/tox21_IRV.py

+2 −2

Original line number	Diff line number	Diff line
		@@ -22,9 +22,9 @@ train_dataset, valid_dataset, test_dataset = tox21_datasets
		K = 10
		# Fit models
		metric = dc.metrics.Metric(dc.metrics.roc_auc_score, np.mean)
		fit_transformers = [dc.trans.IRVFitTransformer(K, len(tox21_tasks), train_dataset)]
		transformers = [dc.trans.IRVTransformer(K, len(tox21_tasks), train_dataset)]

		for transformer in fit_transformers:
		for transformer in transformers:
		print("start")
		time1 = time.time()
		train_dataset = transformer.transform(train_dataset)

Admin message