def test_fit_atom_coordination(self):
a = Connectivity(depth=1, use_coordination=True)
a.fit(ALL_DATA)
self.assertEqual(
a._base_chains,
set([('C1', ), ('N3', ), ('N2', ), ('O2', ), ('N1', ), ('O1', ),
('C4', ), ('H0', ), ('H1', ), ('O0', ), ('C3', ), ('C2', )]))
def test_fit_bond(self):
a = Connectivity(depth=2)
a.fit(ALL_DATA)
self.assertEqual(
a._base_chains,
set([('H', 'O'), ('C', 'H'), ('H', 'N'), ('C', 'C'), ('H', 'H'),
('O', 'O'), ('C', 'N'), ('C', 'O')]))
def test_fit_atom_bond(self):
# This should be the exact same thing as doing it with
# use_bond_order=False
a = Connectivity(depth=1, use_bond_order=True)
a.fit(ALL_DATA)
self.assertEqual(a._base_chains,
set([('N', ), ('C', ), ('O', ), ('H', )]))
def test_unknown(self):
a = Connectivity(add_unknown=True)
expected_results = numpy.array([[1, 4, 0],
[2, 3, 4],
[25, 15, 9]])
a.fit([METHANE])
self.assertTrue((a.transform(ALL_DATA) == expected_results).all())
def test_fit_angle(self):
a = Connectivity(depth=3)
a.fit(ALL_DATA)
self.assertEqual(a._base_groups,
(('C', 'C', 'C'), ('C', 'C', 'H'), ('C', 'C', 'N'),
('C', 'C', 'O'), ('C', 'N', 'C'), ('C', 'N', 'H'),
('C', 'O', 'C'), ('C', 'O', 'H'), ('H', 'C', 'H'),
('H', 'C', 'N'), ('H', 'C', 'O'), ('H', 'N', 'H'),
('N', 'C', 'N'), ('N', 'C', 'O')))
def test_tfidf(self):
a = Connectivity(do_tfidf=True)
expected = numpy.array([[0., 0., 0., 0.], [0., 0., 0., 1.62186043],
[0., 0., 5.49306144, 1.62186043]])
a.fit(ALL_DATA)
try:
m = a.transform(ALL_DATA)
numpy.testing.assert_array_almost_equal(m, expected)
except AssertionError as e:
self.fail(e)
def test_fit_bond_bond(self):
a = Connectivity(depth=2, use_bond_order=True)
a.fit(ALL_DATA)
self.assertEqual(
a._base_groups,
((('C', 'C', '1'), ), (('C', 'C', '2'), ), (('C', 'C', '3'), ),
(('C', 'C', 'Ar'), ), (('C', 'H', '1'), ), (('C', 'N', '2'), ),
(('C', 'N', '3'), ), (('C', 'N', 'Ar'), ), (('C', 'O', '1'), ),
(('C', 'O', 'Ar'), ), (('H', 'H', '1'), ), (('H', 'N', '1'), ),
(('H', 'O', '1'), ), (('O', 'O', '1'), )))
def test_fit_dihedral(self):
# This is to test the double order flipping (CCCH vs HCCC)
a = Connectivity(depth=4)
a.fit(ALL_DATA)
self.assertEqual(
a._base_groups,
(('C', 'C', 'C', 'C'), ('C', 'C', 'C', 'H'), ('C', 'C', 'C', 'N'),
('C', 'C', 'C', 'O'), ('C', 'C', 'N', 'C'), ('C', 'C', 'N', 'H'),
('C', 'C', 'O', 'C'), ('C', 'C', 'O', 'H'), ('H', 'C', 'C', 'H'),
('H', 'C', 'C', 'N'), ('H', 'C', 'C', 'O'), ('H', 'C', 'N', 'C'),
('H', 'C', 'O', 'C'), ('N', 'C', 'C', 'O'), ('N', 'C', 'N', 'C'),
('N', 'C', 'N', 'H'), ('N', 'C', 'O', 'H'), ('O', 'C', 'N', 'C')))
def test_fit_atom_separated(self):
a = Connectivity(depth=1)
a.fit([METHANE2])
self.assertEqual(a._base_chains, set([('C', ), ('H', )]))
self.assertTrue((a.transform([METHANE2]) == numpy.array([[1,
4]])).all())
def test_fit_atom(self):
a = Connectivity(depth=1)
a.fit(ALL_DATA)
self.assertEqual(a._base_chains,
set([('N', ), ('C', ), ('O', ), ('H', )]))
def test_large_to_small_transform(self):
a = Connectivity()
a.fit([BIG])
self.assertTrue((a.transform(ALL_DATA) == ALL_ATOM).all())
def test_small_to_large_transform(self):
a = Connectivity()
a.fit([METHANE])
self.assertTrue((a.transform(ALL_DATA) == ALL_ATOM[:, :2]).all())
def test_transform(self):
a = Connectivity()
a.fit(ALL_DATA)
self.assertTrue((a.transform(ALL_DATA) == ALL_ATOM).all())
def test_fit_atom(self):
a = Connectivity(depth=1)
a.fit(ALL_DATA)
self.assertEqual(a._base_groups, (('C',), ('H',), ('N',), ('O',)))
