Update Coulomb Matrix featurizer tests and examples (4dc5e97c) · Commits · 钟慕尧 / deepchem

deepchem/feat/coulomb_matrices.py

+24 −2

Original line number	Diff line number	Diff line
		@@ -26,15 +26,27 @@ class CoulombMatrix(Featurizer):
		max_atoms : int
		Maximum number of atoms for any molecule in the dataset. Used to
		pad the Coulomb matrix.
		remove_hydrogens : bool, optional (default True)
		remove_hydrogens : bool, optional (default False)
		Whether to remove hydrogens before constructing Coulomb matrix.
		randomize : bool, optional (default True)
		randomize : bool, optional (default False)
		Whether to randomize Coulomb matrices to remove dependence on atom
		index order.
		upper_tri : bool, optional (default False)
		Whether to return the upper triangular portion of the Coulomb matrix.
		n_samples : int, optional (default 1)
		Number of random Coulomb matrices to generate if randomize is True.
		seed : int, optional
		Random seed.

		Example:

		>>> featurizers = dc.feat.CoulombMatrix(max_atoms=23)
		>>> input_file = "input.sdf"
		>>> tasks = ["task0"]
		>>> featurizer = dc.data.SDFLoader(tasks, smiles_field="smiles", mol_field="mol",
		featurizer=featurizers)
		>>> dataset = featurizer.featurize(input_file)

		"""
		conformers = True
		name = 'coulomb_matrix'
		@@ -176,6 +188,16 @@ class CoulombMatrixEig(CoulombMatrix):
		Number of random Coulomb matrices to generate if randomize is True.
		seed : int, optional
		Random seed.

		Example:

		>>> featurizers = dc.feat.CoulombMatrixEig(max_atoms=23)
		>>> input_file = "input.sdf"
		>>> tasks = ["task0"]
		>>> featurizer = dc.data.SDFLoader(tasks, smiles_field="smiles", mol_field="mol",
		featurizer=featurizers)
		>>> dataset = featurizer.featurize(input_file)

		"""

		conformers = True

deepchem/feat/tests/test_coulomb_matrices.py

+19 −2

Original line number	Diff line number	Diff line
		@@ -28,14 +28,31 @@ class TestCoulombMatrix(unittest.TestCase):
		"""
		Test CoulombMatrix.
		"""
		f = cm.CoulombMatrix(self.mol.GetNumAtoms())
		rval = f([self.mol])
		assert rval.shape == (1, self.mol.GetNumConformers(), self.mol.GetNumAtoms(), self.mol.GetNumAtoms())

		def test_coulomb_matrix_padding(self):
		"""
		Test CoulombMatrix with padding.
		"""
		max_atoms = self.mol.GetNumAtoms() * 2
		f = cm.CoulombMatrix(max_atoms=max_atoms)
		rval = f([self.mol])
		assert rval.shape == (1, self.mol.GetNumConformers(), max_atoms, max_atoms)

		def test_upper_tri_coulomb_matrix(self):
		"""
		Test upper triangular CoulombMatrix.
		"""
		f = cm.CoulombMatrix(self.mol.GetNumAtoms(), upper_tri=True)
		rval = f([self.mol])
		size = np.triu_indices(self.mol.GetNumAtoms())[0].size
		assert rval.shape == (1, self.mol.GetNumConformers(), size)

		def test_coulomb_matrix_padding(self):
		def test_upper_tri_coulomb_matrix_padding(self):
		"""
		Test CoulombMatrix with padding.
		Test upper triangular CoulombMatrix with padding.
		"""
		f = cm.CoulombMatrix(max_atoms=self.mol.GetNumAtoms() * 2, upper_tri=True)
		rval = f([self.mol])

Admin message