def test_cns(self):
cnv = CoordinationNumber.from_preset('VoronoiNN')
self.assertEqual(len(cnv.feature_labels()), 1)
self.assertEqual(cnv.feature_labels()[0], 'CN_VoronoiNN')
self.assertAlmostEqual(cnv.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnv.featurize(self.cscl, 0)[0], 14)
self.assertAlmostEqual(cnv.featurize(self.cscl, 1)[0], 14)
self.assertEqual(len(cnv.citations()), 2)
cnv = CoordinationNumber(VoronoiNN(), use_weights='sum')
self.assertEqual(cnv.feature_labels()[0], 'CN_VoronoiNN')
self.assertAlmostEqual(cnv.featurize(self.cscl, 0)[0], 9.2584516)
self.assertAlmostEqual(cnv.featurize(self.cscl, 1)[0], 9.2584516)
self.assertEqual(len(cnv.citations()), 2)
cnv = CoordinationNumber(VoronoiNN(), use_weights='effective')
self.assertEqual(cnv.feature_labels()[0], 'CN_VoronoiNN')
self.assertAlmostEqual(cnv.featurize(self.cscl, 0)[0], 11.648923254)
self.assertAlmostEqual(cnv.featurize(self.cscl, 1)[0], 11.648923254)
self.assertEqual(len(cnv.citations()), 2)
cnj = CoordinationNumber.from_preset('JmolNN')
self.assertEqual(cnj.feature_labels()[0], 'CN_JmolNN')
self.assertAlmostEqual(cnj.featurize(self.sc, 0)[0], 0)
self.assertAlmostEqual(cnj.featurize(self.cscl, 0)[0], 0)
self.assertAlmostEqual(cnj.featurize(self.cscl, 1)[0], 0)
self.assertEqual(len(cnj.citations()), 1)
jmnn = JmolNN(el_radius_updates={"Al": 1.55, "Cl": 1.7, "Cs": 1.7})
cnj = CoordinationNumber(jmnn)
self.assertEqual(cnj.feature_labels()[0], 'CN_JmolNN')
self.assertAlmostEqual(cnj.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnj.featurize(self.cscl, 0)[0], 8)
self.assertAlmostEqual(cnj.featurize(self.cscl, 1)[0], 8)
self.assertEqual(len(cnj.citations()), 1)
cnmd = CoordinationNumber.from_preset('MinimumDistanceNN')
self.assertEqual(cnmd.feature_labels()[0], 'CN_MinimumDistanceNN')
self.assertAlmostEqual(cnmd.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnmd.featurize(self.cscl, 0)[0], 8)
self.assertAlmostEqual(cnmd.featurize(self.cscl, 1)[0], 8)
self.assertEqual(len(cnmd.citations()), 1)
cnmok = CoordinationNumber.from_preset('MinimumOKeeffeNN')
self.assertEqual(cnmok.feature_labels()[0], 'CN_MinimumOKeeffeNN')
self.assertAlmostEqual(cnmok.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnmok.featurize(self.cscl, 0)[0], 8)
self.assertAlmostEqual(cnmok.featurize(self.cscl, 1)[0], 6)
self.assertEqual(len(cnmok.citations()), 2)
cnmvire = CoordinationNumber.from_preset('MinimumVIRENN')
self.assertEqual(cnmvire.feature_labels()[0], 'CN_MinimumVIRENN')
self.assertAlmostEqual(cnmvire.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnmvire.featurize(self.cscl, 0)[0], 8)
self.assertAlmostEqual(cnmvire.featurize(self.cscl, 1)[0], 14)
self.assertEqual(len(cnmvire.citations()), 2)
self.assertEqual(len(cnmvire.implementors()), 2)
self.assertEqual(cnmvire.implementors()[0], 'Nils E. R. Zimmermann')
def test_cns(self):
cnv = CoordinationNumber.from_preset('VoronoiNN')
self.assertEqual(len(cnv.feature_labels()), 1)
self.assertEqual(cnv.feature_labels()[0], 'CN_VoronoiNN')
self.assertAlmostEqual(cnv.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnv.featurize(self.cscl, 0)[0], 14)
self.assertAlmostEqual(cnv.featurize(self.cscl, 1)[0], 14)
self.assertEqual(len(cnv.citations()), 2)
cnv = CoordinationNumber(VoronoiNN(), use_weights='sum')
self.assertEqual(cnv.feature_labels()[0], 'CN_VoronoiNN')
self.assertAlmostEqual(cnv.featurize(self.cscl, 0)[0], 9.2584516)
self.assertAlmostEqual(cnv.featurize(self.cscl, 1)[0], 9.2584516)
self.assertEqual(len(cnv.citations()), 2)
cnv = CoordinationNumber(VoronoiNN(), use_weights='effective')
self.assertEqual(cnv.feature_labels()[0], 'CN_VoronoiNN')
self.assertAlmostEqual(cnv.featurize(self.cscl, 0)[0], 11.648923254)
self.assertAlmostEqual(cnv.featurize(self.cscl, 1)[0], 11.648923254)
self.assertEqual(len(cnv.citations()), 2)
cnj = CoordinationNumber.from_preset('JmolNN')
self.assertEqual(cnj.feature_labels()[0], 'CN_JmolNN')
self.assertAlmostEqual(cnj.featurize(self.sc, 0)[0], 0)
self.assertAlmostEqual(cnj.featurize(self.cscl, 0)[0], 0)
self.assertAlmostEqual(cnj.featurize(self.cscl, 1)[0], 0)
self.assertEqual(len(cnj.citations()), 1)
jmnn = JmolNN(el_radius_updates={"Al": 1.55, "Cl": 1.7, "Cs": 1.7})
cnj = CoordinationNumber(jmnn)
self.assertEqual(cnj.feature_labels()[0], 'CN_JmolNN')
self.assertAlmostEqual(cnj.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnj.featurize(self.cscl, 0)[0], 8)
self.assertAlmostEqual(cnj.featurize(self.cscl, 1)[0], 8)
self.assertEqual(len(cnj.citations()), 1)
cnmd = CoordinationNumber.from_preset('MinimumDistanceNN')
self.assertEqual(cnmd.feature_labels()[0], 'CN_MinimumDistanceNN')
self.assertAlmostEqual(cnmd.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnmd.featurize(self.cscl, 0)[0], 8)
self.assertAlmostEqual(cnmd.featurize(self.cscl, 1)[0], 8)
self.assertEqual(len(cnmd.citations()), 1)
cnmok = CoordinationNumber.from_preset('MinimumOKeeffeNN')
self.assertEqual(cnmok.feature_labels()[0], 'CN_MinimumOKeeffeNN')
self.assertAlmostEqual(cnmok.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnmok.featurize(self.cscl, 0)[0], 8)
self.assertAlmostEqual(cnmok.featurize(self.cscl, 1)[0], 6)
self.assertEqual(len(cnmok.citations()), 2)
cnmvire = CoordinationNumber.from_preset('MinimumVIRENN')
self.assertEqual(cnmvire.feature_labels()[0], 'CN_MinimumVIRENN')
self.assertAlmostEqual(cnmvire.featurize(self.sc, 0)[0], 6)
self.assertAlmostEqual(cnmvire.featurize(self.cscl, 0)[0], 8)
self.assertAlmostEqual(cnmvire.featurize(self.cscl, 1)[0], 14)
self.assertEqual(len(cnmvire.citations()), 2)
self.assertEqual(len(cnmvire.implementors()), 2)
self.assertEqual(cnmvire.implementors()[0], 'Nils E. R. Zimmermann')
def get_fps(structure, cutoff=10.0, processes=8):
all_descrs = []
try:
coordination_number_ = CoordinationNumber.from_preset('VoronoiNN')
voronoi_fps_ = VoronoiFingerprintModified(
cutoff=cutoff).featurize_structure(structure)
crystal_nn_fingerprint_ = CrystalNNFingerprint.from_preset('cn')
op_site_fingerprint_ = OPSiteFingerprint()
agni_fingerprints_ = AGNIFingerprints()
gaussian_symm_func_fps_ = GaussianSymmFuncModified(
).featurize_structure(structure)
pymatgen_data_ = PymatgenData()
magpie_data_ = MagpieData()
data_list = [[
structure, i, site, coordination_number_, voronoi_fps_,
crystal_nn_fingerprint_, op_site_fingerprint_, agni_fingerprints_,
gaussian_symm_func_fps_, pymatgen_data_, magpie_data_
] for i, site in enumerate(structure)]
pool = multiprocessing.Pool(processes=processes)
all_descrs = np.array(pool.map(get_all_site_descrs, data_list))
except (AttributeError, IndexError) as error:
pass
return all_descrs
def from_preset(preset, **kwargs):
"""
Create a SiteStatsFingerprint class according to a preset
Args:
preset (str) - Name of preset
kwargs - Options for SiteStatsFingerprint
"""
if preset == "CrystalNNFingerprint_cn":
return SiteStatsFingerprint(
CrystalNNFingerprint.from_preset("cn", cation_anion=False),
**kwargs)
elif preset == "CrystalNNFingerprint_cn_cation_anion":
return SiteStatsFingerprint(
CrystalNNFingerprint.from_preset("cn", cation_anion=True),
**kwargs)
elif preset == "CrystalNNFingerprint_ops":
return SiteStatsFingerprint(
CrystalNNFingerprint.from_preset("ops", cation_anion=False),
**kwargs)
elif preset == "CrystalNNFingerprint_ops_cation_anion":
return SiteStatsFingerprint(
CrystalNNFingerprint.from_preset("ops", cation_anion=True),
**kwargs)
elif preset == "OPSiteFingerprint":
return SiteStatsFingerprint(OPSiteFingerprint(), **kwargs)
elif preset == "LocalPropertyDifference_ward-prb-2017":
return SiteStatsFingerprint(
LocalPropertyDifference.from_preset("ward-prb-2017"),
stats=["minimum", "maximum", "range", "mean", "avg_dev"])
elif preset == "CoordinationNumber_ward-prb-2017":
return SiteStatsFingerprint(
CoordinationNumber(nn=VoronoiNN(weight='area'),
use_weights="effective"),
stats=["minimum", "maximum", "range", "mean", "avg_dev"])
elif preset == "Composition-dejong2016_AD":
return SiteStatsFingerprint(
LocalPropertyDifference(properties=[
"Number", "AtomicWeight", "Column", "Row",
"CovalentRadius", "Electronegativity"
],
signed=False),
stats=['holder_mean::%d' % d
for d in range(0, 4 + 1)] + ['std_dev'],
)
elif preset == "Composition-dejong2016_SD":
return SiteStatsFingerprint(
LocalPropertyDifference(properties=[
"Number", "AtomicWeight", "Column", "Row",
"CovalentRadius", "Electronegativity"
],
signed=True),
stats=['holder_mean::%d' % d for d in [1, 2, 4]] + ['std_dev'],
)
elif preset == "BondLength-dejong2016":
return SiteStatsFingerprint(
AverageBondLength(VoronoiNN()),
stats=['holder_mean::%d' % d for d in range(-4, 4 + 1)] +
['std_dev', 'geom_std_dev'])
elif preset == "BondAngle-dejong2016":
return SiteStatsFingerprint(
AverageBondAngle(VoronoiNN()),
stats=['holder_mean::%d' % d for d in range(-4, 4 + 1)] +
['std_dev', 'geom_std_dev'])
else:
# TODO: Why assume coordination number? Should this just raise an error? - lw
# One of the various Coordination Number presets:
# MinimumVIRENN, MinimumDistanceNN, JmolNN, VoronoiNN, etc.
try:
return SiteStatsFingerprint(
CoordinationNumber.from_preset(preset), **kwargs)
except:
pass
raise ValueError("Unrecognized preset!")
def structure_to_convmol(structure,
properties=ELEMENTAL_PROPERTIES,
max_atoms=200,
max_features=41,
tolerance_distance=0.25):
atomic_radii = {
'At': 1.50,
'Bk': 1.70,
'Cm': 1.74,
'Fr': 2.60,
'He': 0.28,
'Kr': 1.16,
'Lr': 1.71,
'Md': 1.94,
'Ne': 0.58,
'No': 1.97,
'Rn': 1.50,
'Xe': 1.40,
}
distance_matrix = structure.distance_matrix
for index, x in np.ndenumerate(distance_matrix):
radius_1 = Element(structure._sites[
index[0]].specie).atomic_radius or atomic_radii[str(
structure._sites[index[0]].specie)]
radius_2 = Element(structure._sites[
index[1]].specie).atomic_radius or atomic_radii[str(
structure._sites[index[1]].specie)]
max_distance = radius_1 + radius_2 + tolerance_distance
if x > max_distance:
distance_matrix[index] = 0
else:
distance_matrix[index] = 1
np.fill_diagonal(distance_matrix, 1)
atom_features = []
for i, site in enumerate(structure._sites):
atom_feature_vector = []
for atom_property in properties:
min_value = np.nanmin(
np.array(list(atom_property.values()), dtype=float))
max_value = np.nanmax(
np.array(list(atom_property.values()), dtype=float))
if atom_property[str(Element(site.specie))] is not None:
atom_feature_vector.append(
(atom_property[str(Element(site.specie))] - min_value) /
(max_value - min_value))
else:
atom_feature_vector.append(None)
voronoi_min = np.array([
0.0, 10.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 0.0, 1.0,
1.0
])
voronoi_max = np.array([
120.0, 135.0, 11.0, 3.0, 11.0, 12.0, 18.0, 7.0, 17.0, 17.0, 6.0,
2.0, 6.0, 7.0
])
voronoi_fps = VoronoiFingerprint().featurize(structure, i)
i_fold_symmetry_indices = voronoi_fps[8:16]
voronoi_stats = (np.array(voronoi_fps[16:]) -
voronoi_min) / (voronoi_max - voronoi_min)
atom_feature_vector.extend(i_fold_symmetry_indices +
voronoi_stats.tolist())
coord_min = np.array([1])
coord_max = np.array([36])
coord_fps = (
(CoordinationNumber.from_preset("MinimumDistanceNN").featurize(
structure, i) - coord_min) / (coord_max - coord_min)).tolist()
atom_feature_vector.extend(coord_fps)
atom_features.append(atom_feature_vector)
atom_features = np.array(atom_features, dtype=np.float)
if np.isnan(atom_features).any():
raise ValueError('feature vector contains nan value')
return (zfill(distance_matrix, max_atoms,
max_atoms), zfill(atom_features, max_atoms,
max_features), len(structure.sites))