Add User-Defined API test (84ffef36) · Commits · 钟慕尧 / deepchem

deepchem/utils/test/test_api.py

+18 −39

Original line number	Diff line number	Diff line
		@@ -76,7 +76,7 @@ class TestAPI(unittest.TestCase):
		complex_featurizers, input_transforms,
		output_transforms, input_file, tasks,
		protein_pdb_field=None, ligand_pdb_field=None,
		shard_size=100):
		user_specified_features=None, shard_size=100):
		# Featurize input
		featurizers = compound_featurizers + complex_featurizers

		@@ -87,8 +87,10 @@ class TestAPI(unittest.TestCase):
		ligand_pdb_field=ligand_pdb_field,
		compound_featurizers=compound_featurizers,
		complex_featurizers=complex_featurizers,
		user_specified_features=user_specified_features,
		verbose=True)


		#Featurizes samples and transforms them into NumPy arrays suitable for ML.
		#returns an instance of class FeaturizedSamples()

		@@ -99,10 +101,16 @@ class TestAPI(unittest.TestCase):
		train_samples, test_samples = samples.train_test_split(
		splittype, self.train_dir, self.test_dir)

		use_user_specified_features = False
		if user_specified_features is not None:
		use_user_specified_features = True

		train_dataset = Dataset(data_dir=self.train_dir, samples=train_samples,
		featurizers=featurizers, tasks=tasks)
		featurizers=featurizers, tasks=tasks,
		use_user_specified_features=use_user_specified_features)
		test_dataset = Dataset(data_dir=self.test_dir, samples=test_samples,
		featurizers=featurizers, tasks=tasks)
		featurizers=featurizers, tasks=tasks,
		use_user_specified_features=use_user_specified_features)

		# Transforming train/test data
		train_dataset.transform(input_transforms, output_transforms)
		@@ -110,36 +118,6 @@ class TestAPI(unittest.TestCase):

		return train_dataset, test_dataset

		def user_featurize_and_split(input_file, feature_dir, samples_dir, train_dir,
		test_dir, splittype, feature_types,
		user_specified_features, input_transforms,
		output_transforms, tasks, feature_files):
		"""Featurize inputs with user-specified-features and do train-test split."""
		featurizer = DataFeaturizer(tasks=tasks,
		smiles_field="smiles",
		user_specified_features=user_specified_features,
		verbose=True)

		print("About to featurize.")
		samples = featurizer.featurize(input_file, feature_dir,
		samples_dir, shard_size=8)
		print("Completed Featurization")
		train_samples, test_samples = samples.train_test_split(
		splittype, train_dir, test_dir)
		print("Finished train test split.")
		train_dataset = Dataset(train_dir, tasks, train_samples, feature_types,
		use_user_specified_features=True)
		test_dataset = Dataset(test_dir, tasks, test_samples, feature_types,
		use_user_specified_features=True)
		print("Finished creating train test datasets")
		# Transforming train/test data
		train_dataset.transform(input_transforms, output_transforms)
		test_dataset.transform(input_transforms, output_transforms)
		print("Finished Transforming train test data.")

		return train_dataset, test_dataset


		def test_singletask_rf_ECFP_regression_API(self):
		"""Test of singletask RF ECFP regression API."""
		splittype = "scaffold"
		@@ -231,15 +209,15 @@ class TestAPI(unittest.TestCase):


		def test_singletask_mlp_USF_regression_API(self):
		"""Test of singletask MLP NNScore regression API."""
		"""Test of singletask MLP User Specified Features regression API."""
		splittype = "scaffold"
		compound_featurizers = []
		complex_featurizers = [NNScoreComplexFeaturizer()]
		complex_featurizers = []
		input_transforms = ["normalize", "truncate"]
		output_transforms = ["normalize"]
		feature_types = ["user_specified_features"]
		user_specified_features = ["evals"]
		task_types = {"label": "regression"}
		task_types = {"u0": "regression"}
		model_params = {"nb_hidden": 10, "activation": "relu",
		"dropout": .5, "learning_rate": .01,
		"momentum": .9, "nesterov": False,
		@@ -248,13 +226,14 @@ class TestAPI(unittest.TestCase):
		"nb_layers": 1, "batchnorm": False}

		input_file = "gbd3k.pkl.gz"
		protein_pdb_field = "None"
		ligand_pdb_field = "None"
		protein_pdb_field = None
		ligand_pdb_field = None
		train_dataset, test_dataset = self._featurize_train_test_split(splittype, compound_featurizers,
		complex_featurizers, input_transforms,
		output_transforms, input_file, task_types.keys(),
		protein_pdb_field=protein_pdb_field,
		ligand_pdb_field=ligand_pdb_field)
		ligand_pdb_field=ligand_pdb_field,
		user_specified_features=user_specified_features)
		model_params["data_shape"] = train_dataset.get_data_shape()

		model = SingleTaskDNN(task_types, model_params)

Admin message