Loading deepchem/feat/molecule_featurizers/one_hot_featurizer.py +41 −25 Original line number Diff line number Diff line Loading @@ -42,57 +42,73 @@ class OneHotFeaturizer(MolecularFeaturizer): self.charset = charset self.max_length = max_length def _featurize(self, mol: RDKitMol) -> np.ndarray: """Compute one-hot featurization of this molecule. def _featurize(self, string: str) -> np.ndarray: """Compute one-hot featurization of string. Parameters ---------- mol: rdkit.Chem.rdchem.Mol RDKit Mol object str: An arbitrary string to be featurized. Returns ------- np.ndarray An one hot vector encoded from SMILES. An one hot vector encoded from arbitrary input string. The shape is `(max_length, len(charset) + 1)`. The index of unknown character is `len(charset)`. """ try: from rdkit import Chem except ModuleNotFoundError: raise ImportError("This class requires RDKit to be installed.") smiles = Chem.MolToSmiles(mol) # validation if len(smiles) > self.max_length: if len(string) > self.max_length: logger.info( "The length of {} is longer than `max_length`. So we return an empty array." ) return np.array([]) smiles = self.pad_smile(smiles) string = self.pad_string(string) return np.array([ one_hot_encode(val, self.charset, include_unknown_set=True) for val in smiles for val in string ]) def pad_smile(self, smiles: str) -> str: def _featurizeMol(self, mol: RDKitMol) -> np.ndarray: """Compute one-hot featurization of this molecule. Parameters ---------- mol: rdKit.Chem.rdchem.Mol RDKit Mol object Returns ------- np.ndarray An one hot vector encoded from SMILES. The shape is '(max_length, len(charset) + 1)' The index of unknown character is 'len(charset)'. """ try: from rdkit import Chem except ModuleNotFoundError: raise ImportError("This class requires RDKit to be installed.") smiles = Chem.MolToSmiles(mol) return _featurize(smiles) def pad_string(self, string: str) -> str: """Pad SMILES string to `self.pad_length` Parameters ---------- smiles: str The smiles string to be padded. string: str The string to be padded. Returns ------- str SMILES string space padded to self.pad_length String space padded to self.pad_length """ return smiles.ljust(self.max_length) return string.ljust(self.max_length) def untransform(self, one_hot_vectors: np.ndarray) -> str: """Convert from one hot representation back to SMILES """Convert from one hot representation back to original string Parameters ---------- Loading @@ -102,13 +118,13 @@ class OneHotFeaturizer(MolecularFeaturizer): Returns ------- str SMILES string for an one hot encoded array. Original string for an one hot encoded array. """ smiles = "" string = "" for one_hot in one_hot_vectors: try: idx = np.argmax(one_hot) smiles += self.charset[idx] string += self.charset[idx] except IndexError: smiles += "" return smiles string += "" return string deepchem/feat/tests/test_one_hot_featurizer.py +39 −5 Original line number Diff line number Diff line Loading @@ -11,23 +11,51 @@ class TestOneHotFeaturizert(unittest.TestCase): Test OneHotFeaturizer. """ def test_onehot_featurizer(self): def test_onehot_featurizer_arbitrary(self): """ Test simple one hot encoding. Test simple one hot encoding for arbitrary string. """ string = "abcdefghijklmnopqrstuvwxyz" length = len(string) + 1 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 def test_onehot_featurizer_SMILES(self): """ Test simple one hot encoding for SMILES strings. """ from rdkit import Chem length = len(ZINC_CHARSET) + 1 smiles = 'CC(=O)Oc1ccccc1C(=O)O' mol = Chem.MolFromSmiles(smiles) featurizer = OneHotFeaturizer() feature = featurizer([mol]) feature = featurizer([mol]) # Implicit call to _featurizeMol()--why []? assert feature.shape == (1, 100, length) # untranform undo_smiles = featurizer.untransform(feature[0]) assert smiles == undo_smiles def test_onehot_featurizer_with_max_length(self): def test_onehot_featurizer_arbitrary_with_max_length(self): """ Test one hot encoding with max_length. """ string = "abcdefghijklmnopqrstuvwxyz" length = len(string) + 1 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): """ Test one hot encoding with max_length. """ Loading @@ -43,7 +71,7 @@ class TestOneHotFeaturizert(unittest.TestCase): undo_smiles = featurizer.untransform(feature[0]) assert smiles == undo_smiles def test_correct_transformation(self): def test_correct_transformation_SMILES(self): """ Test correct one hot encoding. """ Loading @@ -62,3 +90,9 @@ class TestOneHotFeaturizert(unittest.TestCase): # untranform undo_smiles = featurizer.untransform(feature[0]) assert smiles == undo_smiles def test_correct_transformation_arbitrary(self): """ Test correct one hot encoding. """ assert "This test case has not yet been written." Loading
deepchem/feat/molecule_featurizers/one_hot_featurizer.py +41 −25 Original line number Diff line number Diff line Loading @@ -42,57 +42,73 @@ class OneHotFeaturizer(MolecularFeaturizer): self.charset = charset self.max_length = max_length def _featurize(self, mol: RDKitMol) -> np.ndarray: """Compute one-hot featurization of this molecule. def _featurize(self, string: str) -> np.ndarray: """Compute one-hot featurization of string. Parameters ---------- mol: rdkit.Chem.rdchem.Mol RDKit Mol object str: An arbitrary string to be featurized. Returns ------- np.ndarray An one hot vector encoded from SMILES. An one hot vector encoded from arbitrary input string. The shape is `(max_length, len(charset) + 1)`. The index of unknown character is `len(charset)`. """ try: from rdkit import Chem except ModuleNotFoundError: raise ImportError("This class requires RDKit to be installed.") smiles = Chem.MolToSmiles(mol) # validation if len(smiles) > self.max_length: if len(string) > self.max_length: logger.info( "The length of {} is longer than `max_length`. So we return an empty array." ) return np.array([]) smiles = self.pad_smile(smiles) string = self.pad_string(string) return np.array([ one_hot_encode(val, self.charset, include_unknown_set=True) for val in smiles for val in string ]) def pad_smile(self, smiles: str) -> str: def _featurizeMol(self, mol: RDKitMol) -> np.ndarray: """Compute one-hot featurization of this molecule. Parameters ---------- mol: rdKit.Chem.rdchem.Mol RDKit Mol object Returns ------- np.ndarray An one hot vector encoded from SMILES. The shape is '(max_length, len(charset) + 1)' The index of unknown character is 'len(charset)'. """ try: from rdkit import Chem except ModuleNotFoundError: raise ImportError("This class requires RDKit to be installed.") smiles = Chem.MolToSmiles(mol) return _featurize(smiles) def pad_string(self, string: str) -> str: """Pad SMILES string to `self.pad_length` Parameters ---------- smiles: str The smiles string to be padded. string: str The string to be padded. Returns ------- str SMILES string space padded to self.pad_length String space padded to self.pad_length """ return smiles.ljust(self.max_length) return string.ljust(self.max_length) def untransform(self, one_hot_vectors: np.ndarray) -> str: """Convert from one hot representation back to SMILES """Convert from one hot representation back to original string Parameters ---------- Loading @@ -102,13 +118,13 @@ class OneHotFeaturizer(MolecularFeaturizer): Returns ------- str SMILES string for an one hot encoded array. Original string for an one hot encoded array. """ smiles = "" string = "" for one_hot in one_hot_vectors: try: idx = np.argmax(one_hot) smiles += self.charset[idx] string += self.charset[idx] except IndexError: smiles += "" return smiles string += "" return string
deepchem/feat/tests/test_one_hot_featurizer.py +39 −5 Original line number Diff line number Diff line Loading @@ -11,23 +11,51 @@ class TestOneHotFeaturizert(unittest.TestCase): Test OneHotFeaturizer. """ def test_onehot_featurizer(self): def test_onehot_featurizer_arbitrary(self): """ Test simple one hot encoding. Test simple one hot encoding for arbitrary string. """ string = "abcdefghijklmnopqrstuvwxyz" length = len(string) + 1 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 def test_onehot_featurizer_SMILES(self): """ Test simple one hot encoding for SMILES strings. """ from rdkit import Chem length = len(ZINC_CHARSET) + 1 smiles = 'CC(=O)Oc1ccccc1C(=O)O' mol = Chem.MolFromSmiles(smiles) featurizer = OneHotFeaturizer() feature = featurizer([mol]) feature = featurizer([mol]) # Implicit call to _featurizeMol()--why []? assert feature.shape == (1, 100, length) # untranform undo_smiles = featurizer.untransform(feature[0]) assert smiles == undo_smiles def test_onehot_featurizer_with_max_length(self): def test_onehot_featurizer_arbitrary_with_max_length(self): """ Test one hot encoding with max_length. """ string = "abcdefghijklmnopqrstuvwxyz" length = len(string) + 1 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): """ Test one hot encoding with max_length. """ Loading @@ -43,7 +71,7 @@ class TestOneHotFeaturizert(unittest.TestCase): undo_smiles = featurizer.untransform(feature[0]) assert smiles == undo_smiles def test_correct_transformation(self): def test_correct_transformation_SMILES(self): """ Test correct one hot encoding. """ Loading @@ -62,3 +90,9 @@ class TestOneHotFeaturizert(unittest.TestCase): # untranform undo_smiles = featurizer.untransform(feature[0]) assert smiles == undo_smiles def test_correct_transformation_arbitrary(self): """ Test correct one hot encoding. """ assert "This test case has not yet been written."
