def test_cart_dists(self):
sm = StructureMatcher()
l = Lattice.orthorhombic(1, 2, 3)
s1 = np.array([[0.13, 0.25, 0.37], [0.1, 0.2, 0.3]])
s2 = np.array([[0.11, 0.22, 0.33]])
s3 = np.array([[0.1, 0.2, 0.3], [0.11, 0.2, 0.3]])
s4 = np.array([[0.1, 0.2, 0.3], [0.1, 0.6, 0.7]])
mask = np.array([[False, False]])
mask2 = np.array([[False, True]])
mask3 = np.array([[False, False], [False, False]])
mask4 = np.array([[False, True], [False, True]])
n1 = (len(s1) / l.volume) ** (1/3)
n2 = (len(s2) / l.volume) ** (1/3)
self.assertRaises(ValueError, sm._cart_dists, s2, s1, l, mask.T, n2)
self.assertRaises(ValueError, sm._cart_dists, s1, s2, l, mask.T, n1)
d, ft, s = sm._cart_dists(s1, s2, l, mask, n1)
self.assertTrue(np.allclose(d, [0]))
self.assertTrue(np.allclose(ft, [-0.01, -0.02, -0.03]))
self.assertTrue(np.allclose(s, [1]))
#check that masking best value works
d, ft, s = sm._cart_dists(s1, s2, l, mask2, n1)
self.assertTrue(np.allclose(d, [0]))
self.assertTrue(np.allclose(ft, [0.02, 0.03, 0.04]))
self.assertTrue(np.allclose(s, [0]))
#check that averaging of translation is done properly
d, ft, s = sm._cart_dists(s1, s3, l, mask3, n1)
self.assertTrue(np.allclose(d, [0.08093341]*2))
self.assertTrue(np.allclose(ft, [0.01, 0.025, 0.035]))
self.assertTrue(np.allclose(s, [1, 0]))
#check distances are large when mask allows no 'real' mapping
d, ft, s = sm._cart_dists(s1, s4, l, mask4, n1)
self.assertTrue(np.min(d) > 1e8)
self.assertTrue(np.min(ft) > 1e8)
def test_cart_dists(self):
sm = StructureMatcher()
l = Lattice.orthorhombic(1, 2, 3)
s1 = np.array([[0.13, 0.25, 0.37], [0.1, 0.2, 0.3]])
s2 = np.array([[0.11, 0.22, 0.33]])
s3 = np.array([[0.1, 0.2, 0.3], [0.11, 0.2, 0.3]])
s4 = np.array([[0.1, 0.2, 0.3], [0.1, 0.6, 0.7]])
mask = np.array([[False, False]])
mask2 = np.array([[False, True]])
mask3 = np.array([[False, False], [False, False]])
mask4 = np.array([[False, True], [False, True]])
n1 = (len(s1) / l.volume)**(1 / 3)
n2 = (len(s2) / l.volume)**(1 / 3)
self.assertRaises(ValueError, sm._cart_dists, s2, s1, l, mask.T, n2)
self.assertRaises(ValueError, sm._cart_dists, s1, s2, l, mask.T, n1)
d, ft, s = sm._cart_dists(s1, s2, l, mask, n1)
self.assertTrue(np.allclose(d, [0]))
self.assertTrue(np.allclose(ft, [-0.01, -0.02, -0.03]))
self.assertTrue(np.allclose(s, [1]))
#check that masking best value works
d, ft, s = sm._cart_dists(s1, s2, l, mask2, n1)
self.assertTrue(np.allclose(d, [0]))
self.assertTrue(np.allclose(ft, [0.02, 0.03, 0.04]))
self.assertTrue(np.allclose(s, [0]))
#check that averaging of translation is done properly
d, ft, s = sm._cart_dists(s1, s3, l, mask3, n1)
self.assertTrue(np.allclose(d, [0.08093341] * 2))
self.assertTrue(np.allclose(ft, [0.01, 0.025, 0.035]))
self.assertTrue(np.allclose(s, [1, 0]))
#check distances are large when mask allows no 'real' mapping
d, ft, s = sm._cart_dists(s1, s4, l, mask4, n1)
self.assertTrue(np.min(d) > 1e8)
self.assertTrue(np.min(ft) > 1e8)
