Changed one hot to be subclass of Featurizer (05c9001f) · Commits · 钟慕尧 / deepchem

deepchem/feat/molecule_featurizers/one_hot_featurizer.py

+4 −5

Original line number	Diff line number	Diff line
		@@ -5,7 +5,7 @@ import numpy as np

		from deepchem.utils.typing import RDKitMol
		from deepchem.utils.molecule_feature_utils import one_hot_encode
		from deepchem.feat.base_classes import MolecularFeaturizer
		from deepchem.feat.base_classes import Featurizer

		logger = logging.getLogger(__name__)

		@@ -16,7 +16,7 @@ ZINC_CHARSET = [
		]


		class OneHotFeaturizer(MolecularFeaturizer):
		class OneHotFeaturizer(Featurizer):
		"""Encodes SMILES as a one-hot array.

		This featurizer encodes SMILES string as a one-hot array.
		@@ -69,7 +69,7 @@ class OneHotFeaturizer(MolecularFeaturizer):
		for val in string
		])

		def _featurizeMol(self, mol: RDKitMol) -> np.ndarray:
		def featurizeMol(self, mol: RDKitMol) -> np.ndarray:
		"""Compute one-hot featurization of this molecule.

		Parameters
		@@ -88,9 +88,8 @@ class OneHotFeaturizer(MolecularFeaturizer):
		from rdkit import Chem
		except ModuleNotFoundError:
		raise ImportError("This class requires RDKit to be installed.")

		smiles = Chem.MolToSmiles(mol)
		return _featurize(smiles)
		return self._featurize(smiles)

		def pad_string(self, string: str) -> str:
		"""Pad SMILES string to `self.pad_length`

deepchem/feat/tests/test_one_hot_featurizer.py

+4 −8

Original line number	Diff line number	Diff line
		@@ -20,7 +20,6 @@ class TestOneHotFeaturizert(unittest.TestCase):
		featurizer = OneHotFeaturizer()
		feature = featurizer(string) # Implicit call to _featurize()
		assert feature.shape == (1, 100, length)

		# untransform
		undo_string = featurizer.untransform(feature[0])
		assert string == undo_string
		@@ -34,9 +33,8 @@ class TestOneHotFeaturizert(unittest.TestCase):
		smiles = 'CC(=O)Oc1ccccc1C(=O)O'
		mol = Chem.MolFromSmiles(smiles)
		featurizer = OneHotFeaturizer()
		feature = featurizer([mol]) # Implicit call to _featurizeMol()--why []?
		feature = featurizer.featurizeMol(mol) # Implicit call to _featurizeMol()--why []?
		assert feature.shape == (1, 100, length)

		# untranform
		undo_smiles = featurizer.untransform(feature[0])
		assert smiles == undo_smiles
		@@ -50,12 +48,11 @@ class TestOneHotFeaturizert(unittest.TestCase):
		featurizer = OneHotFeaturizer(max_length=120)
		feature = featurizer(string)
		assert feature.shape == (1, 120, length)

		# untranform
		undo_string = featurizer.untransform(feature[0])
		assert string == undo_string

		def test_onehot_featurizer_SMIELS_with_max_length(self):
		def test_onehot_featurizer_SMILES_with_max_length(self):
		"""
		Test one hot encoding with max_length.
		"""
		@@ -64,9 +61,8 @@ class TestOneHotFeaturizert(unittest.TestCase):
		smiles = 'CC(=O)Oc1ccccc1C(=O)O'
		mol = Chem.MolFromSmiles(smiles)
		featurizer = OneHotFeaturizer(max_length=120)
		feature = featurizer([mol])
		feature = featurizer.featurizeMol(mol)
		assert feature.shape == (1, 120, length)

		# untranform
		undo_smiles = featurizer.untransform(feature[0])
		assert smiles == undo_smiles
		@@ -80,7 +76,7 @@ class TestOneHotFeaturizert(unittest.TestCase):
		smiles = 'CN=C=O'
		mol = Chem.MolFromSmiles(smiles)
		featurizer = OneHotFeaturizer(charset=charset, max_length=100)
		feature = featurizer([mol])
		feature = featurizer(mol)
		assert np.allclose(feature[0][0], np.array([1, 0, 0, 0, 0, 0, 0]))
		assert np.allclose(feature[0][1], np.array([0, 1, 0, 0, 0, 0, 0]))
		assert np.allclose(feature[0][2], np.array([0, 0, 1, 0, 0, 0, 0]))

Admin message