Merge remote-tracking branch 'remotes/mine/temp3' into GPGO (bb2669be) · Commits · 钟慕尧 / deepchem

deepchem/molnet/check_availability.py

+4 −0

Original line number	Diff line number	Diff line
		@@ -162,17 +162,20 @@ CheckFeaturizer = {
		('qm7', 'tf_regression'): ['ECFP', 1024],
		('qm7', 'rf_regression'): ['ECFP', 1024],
		('qm7', 'krr'): ['ECFP', 1024],
		('qm7', 'krr_ft'): ['CoulombMatrix', 1024],
		('qm7', 'graphconvreg'): ['GraphConv', 75],
		('qm7', 'tf_regression_ft'): ['CoulombMatrix', [23, 23]],
		('qm7', 'dtnn'): ['CoulombMatrix', [23, 23]],
		('qm7', 'ani'): ['BPSymmetryFunction', [23, 4]],
		('qm7b', 'tf_regression_ft'): ['CoulombMatrix', [23, 23]],
		('qm7b', 'krr_ft'): ['CoulombMatrix', 1024],
		('qm7b', 'dtnn'): ['CoulombMatrix', [23, 23]],
		('qm8', 'tf_regression'): ['ECFP', 1024],
		('qm8', 'rf_regression'): ['ECFP', 1024],
		('qm8', 'krr'): ['ECFP', 1024],
		('qm8', 'graphconvreg'): ['GraphConv', 75],
		('qm8', 'tf_regression_ft'): ['CoulombMatrix', [26, 26]],
		('qm8', 'krr_ft'): ['CoulombMatrix', 1024],
		('qm8', 'dtnn'): ['CoulombMatrix', [26, 26]],
		('qm8', 'ani'): ['BPSymmetryFunction', [26, 4]],
		('qm9', 'tf_regression'): ['ECFP', 1024],
		@@ -180,6 +183,7 @@ CheckFeaturizer = {
		('qm9', 'krr'): ['ECFP', 1024],
		('qm9', 'graphconvreg'): ['GraphConv', 75],
		('qm9', 'tf_regression_ft'): ['CoulombMatrix', [29, 29]],
		('qm9', 'krr_ft'): ['CoulombMatrix', 1024],
		('qm9', 'dtnn'): ['CoulombMatrix', [29, 29]],
		('qm9', 'ani'): ['BPSymmetryFunction', [29, 4]]
		}

deepchem/molnet/preset_hyper_parameters.py

+2 −1

Original line number	Diff line number	Diff line
		@@ -118,7 +118,8 @@ hps['tf_regression_ft'] = {
		'fit_transformers': deepchem.trans.CoulombFitTransformer
		}
		hps['rf_regression'] = {'n_estimators': 500}
		hps['krr'] = {'alpha': 1e-3, 'gamma': 0.05}
		hps['krr'] = {'alpha': 1e-3}
		hps['krr_ft'] = {'alpha': 1e-3}
		hps['graphconvreg'] = {
		'batch_size': 128,
		'nb_epoch': 20,

deepchem/molnet/run_benchmark_models.py

+20 −3

Original line number	Diff line number	Diff line
		@@ -450,7 +450,7 @@ def benchmark_regression(train_dataset,
		assert model in [
		'tf_regression', 'tf_regression_ft', 'rf_regression', 'graphconvreg',
		'dtnn', 'dag_regression', 'xgb_regression', 'weave_regression', 'krr',
		'ani'
		'ani', 'krr_ft'
		]
		if hyper_parameters is None:
		hyper_parameters = hps[model]
		@@ -711,15 +711,32 @@ def benchmark_regression(train_dataset,

		model = deepchem.models.multitask.SingletaskToMultitask(
		tasks, model_builder)

		elif model_name == 'krr':
		# Loading hyper parameters
		alpha = hyper_parameters['alpha']
		gamma = hyper_parameters['gamma']
		nb_epoch = None

		# Building scikit learn Kernel Ridge Regression model
		def model_builder(model_dir_krr):
		sklearn_model = KernelRidge(kernel="rbf", alpha=alpha, gamma=gamma)
		sklearn_model = KernelRidge(kernel="rbf", alpha=alpha)
		return deepchem.models.SklearnModel(sklearn_model, model_dir_krr)

		model = deepchem.models.multitask.SingletaskToMultitask(
		tasks, model_builder)

		elif model_name == 'krr_ft':
		# Loading hyper parameters
		alpha = hyper_parameters['alpha']
		nb_epoch = None

		ft_transformer = deepchem.trans.CoulombFitTransformer(train_dataset)
		train_dataset = ft_transformer.transform(train_dataset)
		valid_dataset = ft_transformer.transform(valid_dataset)
		test_dataset = ft_transformer.transform(test_dataset)
		# Building scikit learn Kernel Ridge Regression model
		def model_builder(model_dir_krr):
		sklearn_model = KernelRidge(kernel="rbf", alpha=alpha)
		return deepchem.models.SklearnModel(sklearn_model, model_dir_krr)

		model = deepchem.models.multitask.SingletaskToMultitask(

deepchem/trans/transformers.py

+7 −3

Original line number	Diff line number	Diff line
		@@ -497,7 +497,7 @@ class PowerTransformer(Transformer):
		return z


		class CoulombFitTransformer():
		class CoulombFitTransformer(Transformer):
		"""Performs randomization and binarization operations on batches of Coulomb Matrix features during fit.

		Example:
		@@ -618,8 +618,12 @@ class CoulombFitTransformer():
		X = self.normalize(self.expand(self.realize(X)))
		return X

		def transform(self, dataset):
		raise NotImplementedError("Cannot transform datasets with FitTransformer")
		#def transform(self, dataset):
		# raise NotImplementedError("Cannot transform datasets with FitTransformer")

		def transform_array(self, X, y, w):
		X = self.X_transform(X)
		return (X, y, w)

		def untransform(self, z):
		raise NotImplementedError(

examples/bace_cxgb.pkl

0 → 100644

+47 −0

Original line number	Diff line number	Diff line
		(dp0
		S'reg_alpha'
		p1
		I0
		sS'colsample_bylevel'
		p2
		I1
		sS'scale_pos_weight'
		p3
		I1
		sS'learning_rate'
		p4
		F0.07681636412701465
		sS'max_delta_step'
		p5
		I1
		sS'base_score'
		p6
		F0.5
		sS'n_estimators'
		p7
		I5560
		sS'seed'
		p8
		I2016
		sS'early_stopping_rounds'
		p9
		I116
		sS'colsample_bytree'
		p10
		F0.66
		sS'min_child_weight'
		p11
		I5
		sS'subsample'
		p12
		F0.53
		sS'reg_lambda'
		p13
		I1
		sS'max_depth'
		p14
		I8
		sS'gamma'
		p15
		F0.0
		s.
		No newline at end of file

Admin message