def similarity(_parents, target):
featurizer = MultipleFeaturizer([
SiteStatsFingerprint.from_preset("CoordinationNumber_ward-prb-2017"),
StructuralHeterogeneity(),
ChemicalOrdering(),
MaximumPackingEfficiency(),
SiteStatsFingerprint.from_preset(
"LocalPropertyDifference_ward-prb-2017"),
StructureComposition(Stoichiometry()),
StructureComposition(ElementProperty.from_preset("magpie")),
StructureComposition(ValenceOrbital(props=["frac"])),
StructureComposition(IonProperty(fast=True)),
])
# HACK celery doesn't work with multiprocessing (used by matminer)
try:
from celery import current_task
if current_task:
featurizer.set_n_jobs(1)
except ImportError:
pass
x_target = pd.DataFrame.from_records([featurizer.featurize(target)],
columns=featurizer.feature_labels())
x_parent = pd.DataFrame.from_records(
featurizer.featurize_many(_parents, ignore_errors=True, pbar=False),
columns=featurizer.feature_labels(),
)
nulls = x_parent[x_parent.isnull().any(axis=1)].index.values
x_parent.fillna(100000, inplace=True)
x_target = x_target.reindex(sorted(x_target.columns), axis=1)
x_parent = x_parent.reindex(sorted(x_parent.columns), axis=1)
with open(os.path.join(settings.rxn_files, "scaler2.pickle"), "rb") as f:
scaler = pickle.load(f)
with open(os.path.join(settings.rxn_files, "quantiles.pickle"), "rb") as f:
quantiles = pickle.load(f)
X = scaler.transform(x_parent.append(x_target))
D = [pairwise_distances(np.array([row, X[-1]]))[0, 1] for row in X[:-1]]
_res = []
for d in D:
_res.append(np.linspace(0, 1, 101)[np.abs(quantiles - d).argmin()])
_res = np.array(_res)
_res[nulls] = -1
return _res
class DeBreuck2020Featurizer(modnet.featurizers.MODFeaturizer):
""" Featurizer presets used for the paper 'Machine learning
materials properties for small datasets' by Pierre-Paul De Breuck,
Geoffroy Hautier & Gian-Marco Rignanese, arXiv:2004.14766 (2020).
Uses most of the featurizers implemented by matminer at the time of
writing with their default hyperparameters and presets.
"""
from matminer.featurizers.composition import (
AtomicOrbitals,
AtomicPackingEfficiency,
BandCenter,
# CohesiveEnergy, - This descriptor was not used in the paper preset
# ElectronAffinity, - This descriptor was not used in the paper preset
ElectronegativityDiff,
ElementFraction,
ElementProperty,
IonProperty,
Miedema,
OxidationStates,
Stoichiometry,
TMetalFraction,
ValenceOrbital,
YangSolidSolution,
)
from matminer.featurizers.structure import (
# BagofBonds, - This descriptor was not used in the paper preset
BondFractions,
ChemicalOrdering,
CoulombMatrix,
DensityFeatures,
EwaldEnergy,
GlobalSymmetryFeatures,
MaximumPackingEfficiency,
# PartialRadialDistributionFunction,
RadialDistributionFunction,
SineCoulombMatrix,
StructuralHeterogeneity,
XRDPowderPattern,
)
from matminer.featurizers.site import (
AGNIFingerprints,
AverageBondAngle,
AverageBondLength,
BondOrientationalParameter,
ChemEnvSiteFingerprint,
CoordinationNumber,
CrystalNNFingerprint,
GaussianSymmFunc,
GeneralizedRadialDistributionFunction,
LocalPropertyDifference,
OPSiteFingerprint,
VoronoiFingerprint,
)
composition_featurizers = (
AtomicOrbitals(),
AtomicPackingEfficiency(),
BandCenter(),
ElementFraction(),
ElementProperty.from_preset("magpie"),
IonProperty(),
Miedema(),
Stoichiometry(),
TMetalFraction(),
ValenceOrbital(),
YangSolidSolution(),
)
oxide_composition_featurizers = (
ElectronegativityDiff(),
OxidationStates(),
)
structure_featurizers = (
DensityFeatures(),
GlobalSymmetryFeatures(),
RadialDistributionFunction(),
CoulombMatrix(),
# PartialRadialDistributionFunction(),
SineCoulombMatrix(),
EwaldEnergy(),
BondFractions(),
StructuralHeterogeneity(),
MaximumPackingEfficiency(),
ChemicalOrdering(),
XRDPowderPattern(),
# BagofBonds(),
)
site_featurizers = (
AGNIFingerprints(),
AverageBondAngle(VoronoiNN()),
AverageBondLength(VoronoiNN()),
BondOrientationalParameter(),
ChemEnvSiteFingerprint.from_preset("simple"),
CoordinationNumber(),
CrystalNNFingerprint.from_preset("ops"),
GaussianSymmFunc(),
GeneralizedRadialDistributionFunction.from_preset("gaussian"),
LocalPropertyDifference(),
OPSiteFingerprint(),
VoronoiFingerprint(),
)
def featurize_composition(self, df):
""" Applies the preset composition featurizers to the input dataframe,
renames some fields and cleans the output dataframe.
"""
df = super().featurize_composition(df)
_orbitals = {"s": 1, "p": 2, "d": 3, "f": 4}
df['AtomicOrbitals|HOMO_character'] = df[
'AtomicOrbitals|HOMO_character'].map(_orbitals)
df['AtomicOrbitals|LUMO_character'] = df[
'AtomicOrbitals|LUMO_character'].map(_orbitals)
df['AtomicOrbitals|HOMO_element'] = df[
'AtomicOrbitals|HOMO_element'].apply(
lambda x: -1 if not isinstance(x, str) else Element(x).Z)
df['AtomicOrbitals|LUMO_element'] = df[
'AtomicOrbitals|LUMO_element'].apply(
lambda x: -1 if not isinstance(x, str) else Element(x).Z)
df = df.replace([np.inf, -np.inf, np.nan], 0)
return modnet.featurizers.clean_df(df)
def featurize_structure(self, df):
""" Applies the preset structural featurizers to the input dataframe,
renames some fields and cleans the output dataframe.
"""
df = super().featurize_structure(df)
dist = df[
"RadialDistributionFunction|radial distribution function"].iloc[0][
'distances'][:50]
for i, d in enumerate(dist):
_rdf_key = "RadialDistributionFunction|radial distribution function|d_{:.2f}".format(
d)
df[_rdf_key] = df[
"RadialDistributionFunction|radial distribution function"].apply(
lambda x: x['distribution'][i])
df = df.drop("RadialDistributionFunction|radial distribution function",
axis=1)
_crystal_system = {
"cubic": 1,
"tetragonal": 2,
"orthorombic": 3,
"hexagonal": 4,
"trigonal": 5,
"monoclinic": 6,
"triclinic": 7
}
def _int_map(x):
if x == np.nan:
return 0
elif x:
return 1
else:
return 0
df["GlobalSymmetryFeatures|crystal_system"] = df[
"GlobalSymmetryFeatures|crystal_system"].map(_crystal_system)
df["GlobalSymmetryFeatures|is_centrosymmetric"] = df[
"GlobalSymmetryFeatures|is_centrosymmetric"].map(_int_map)
return modnet.featurizers.clean_df(df)
def featurize_site(self, df):
""" Applies the preset site featurizers to the input dataframe,
renames some fields and cleans the output dataframe.
"""
# rename some features for backwards compatibility with pretrained models
aliases = {
"GeneralizedRadialDistributionFunction": "GeneralizedRDF",
"AGNIFingerprints": "AGNIFingerPrint",
"BondOrientationalParameter": "BondOrientationParameter",
"GaussianSymmFunc": "ChemEnvSiteFingerprint|GaussianSymmFunc",
}
df = super().featurize_site(df, aliases=aliases)
df = df.loc[:, (df != 0).any(axis=0)]
return modnet.featurizers.clean_df(df)
ignore_errors=True)
# -- end F6
# -- start F7
from matminer.featurizers.composition import ElectronAffinity
ela_feat = ElectronAffinity()
fdf = ela_feat.featurize_dataframe(fdf,
col_id='composition',
ignore_errors=True)
# -- end F7
# -- start F9
from matminer.featurizers.composition import ValenceOrbital
vlo_feat = ValenceOrbital()
fdf = vlo_feat.featurize_dataframe(fdf,
col_id='composition',
ignore_errors=True)
# -- end F9
# -- start F10
from matminer.featurizers.composition import IonProperty
iop_feat = IonProperty()
fdf = iop_feat.featurize_dataframe(fdf,
col_id='composition',
ignore_errors=True)
# -- end F10
# -- start F12
class FUTURE_PROSPECTS_2021(featurizer.extendedMODFeaturizer):
from matminer.featurizers.composition import (
AtomicOrbitals,
AtomicPackingEfficiency,
BandCenter,
CohesiveEnergy,
ElectronAffinity,
ElectronegativityDiff,
ElementFraction,
ElementProperty,
IonProperty,
Miedema,
OxidationStates,
Stoichiometry,
TMetalFraction,
ValenceOrbital,
YangSolidSolution,
)
from matminer.featurizers.structure import (
BagofBonds,
BondFractions,
ChemicalOrdering,
CoulombMatrix,
DensityFeatures,
EwaldEnergy,
GlobalSymmetryFeatures,
MaximumPackingEfficiency,
PartialRadialDistributionFunction,
RadialDistributionFunction,
SineCoulombMatrix,
StructuralHeterogeneity,
XRDPowderPattern,
)
from matminer.featurizers.site import (
AGNIFingerprints,
AverageBondAngle,
AverageBondLength,
BondOrientationalParameter,
ChemEnvSiteFingerprint,
CoordinationNumber,
CrystalNNFingerprint,
GaussianSymmFunc,
GeneralizedRadialDistributionFunction,
LocalPropertyDifference,
OPSiteFingerprint,
VoronoiFingerprint,
)
from matminer.featurizers.dos import (
DOSFeaturizer,
SiteDOS,
Hybridization,
DosAsymmetry,
)
from matminer.featurizers.bandstructure import (
BandFeaturizer,
BranchPointEnergy
)
composition_featurizers = (
AtomicOrbitals(),
AtomicPackingEfficiency(),
BandCenter(),
ElementFraction(),
ElementProperty.from_preset("magpie"),
IonProperty(),
Miedema(),
Stoichiometry(),
TMetalFraction(),
ValenceOrbital(),
YangSolidSolution(),
)
oxid_composition_featurizers = (
ElectronegativityDiff(),
OxidationStates(),
)
structure_featurizers = (
DensityFeatures(),
GlobalSymmetryFeatures(),
RadialDistributionFunction(),
CoulombMatrix(),
#PartialRadialDistributionFunction(), #Introduces a large amount of features
SineCoulombMatrix(),
EwaldEnergy(),
BondFractions(),
StructuralHeterogeneity(),
MaximumPackingEfficiency(),
ChemicalOrdering(),
XRDPowderPattern(),
)
site_featurizers = (
AGNIFingerprints(),
AverageBondAngle(VoronoiNN()),
AverageBondLength(VoronoiNN()),
BondOrientationalParameter(),
ChemEnvSiteFingerprint.from_preset("simple"),
CoordinationNumber(),
CrystalNNFingerprint.from_preset("ops"),
GaussianSymmFunc(),
GeneralizedRadialDistributionFunction.from_preset("gaussian"),
LocalPropertyDifference(),
OPSiteFingerprint(),
VoronoiFingerprint(),
)
dos_featurizers = (
DOSFeaturizer(),
SiteDOS(),
Hybridization()
)
band_featurizers = (
BandFeaturizer(),
BranchPointEnergy()
)
def __init__(self, n_jobs=None):
self._n_jobs = n_jobs
def featurize_composition(self, df):
"""Applies the preset composition featurizers to the input dataframe,
renames some fields and cleans the output dataframe.
"""
df = super().featurize_composition(df)
_orbitals = {"s": 1, "p": 2, "d": 3, "f": 4}
df["AtomicOrbitals|HOMO_character"] = df["AtomicOrbitals|HOMO_character"].map(
_orbitals
)
df["AtomicOrbitals|LUMO_character"] = df["AtomicOrbitals|LUMO_character"].map(
_orbitals
)
df["AtomicOrbitals|HOMO_element"] = df["AtomicOrbitals|HOMO_element"].apply(
lambda x: -1 if not isinstance(x, str) else Element(x).Z
)
df["AtomicOrbitals|LUMO_element"] = df["AtomicOrbitals|LUMO_element"].apply(
lambda x: -1 if not isinstance(x, str) else Element(x).Z
)
return clean_df(df)
def featurize_structure(self, df):
"""Applies the preset structural featurizers to the input dataframe,
renames some fields and cleans the output dataframe.
"""
df = super().featurize_structure(df)
dist = df["RadialDistributionFunction|radial distribution function"].iloc[0][
"distances"
][:50]
for i, d in enumerate(dist):
_rdf_key = "RadialDistributionFunction|radial distribution function|d_{:.2f}".format(
d
)
df[_rdf_key] = df[
"RadialDistributionFunction|radial distribution function"
].apply(lambda x: x["distribution"][i])
df = df.drop("RadialDistributionFunction|radial distribution function", axis=1)
_crystal_system = {
"cubic": 1,
"tetragonal": 2,
"orthorombic": 3,
"hexagonal": 4,
"trigonal": 5,
"monoclinic": 6,
"triclinic": 7,
}
def _int_map(x):
if x == np.nan:
return 0
elif x:
return 1
else:
return 0
df["GlobalSymmetryFeatures|crystal_system"] = df[
"GlobalSymmetryFeatures|crystal_system"
].map(_crystal_system)
df["GlobalSymmetryFeatures|is_centrosymmetric"] = df[
"GlobalSymmetryFeatures|is_centrosymmetric"
].map(_int_map)
return clean_df(df)
def featurize_dos(self, df):
"""Applies the presetdos featurizers to the input dataframe,
renames some fields and cleans the output dataframe.
"""
df = super().featurize_dos(df)
hotencodeColumns = ["DOSFeaturizer|vbm_specie_1","DOSFeaturizer|cbm_specie_1"]
one_hot = pd.get_dummies(df[hotencodeColumns])
df = df.drop(hotencodeColumns, axis = 1).join(one_hot)
_orbitals = {"s": 1, "p": 2, "d": 3, "f": 4}
df["DOSFeaturizer|vbm_character_1"] = df[
"DOSFeaturizer|vbm_character_1"
].map(_orbitals)
df["DOSFeaturizer|cbm_character_1"] = df[
"DOSFeaturizer|cbm_character_1"
].map(_orbitals)
# Splitting one feature into several floating features
# e.g. number;number;number into three columns
splitColumns = ["DOSFeaturizer|cbm_location_1", "DOSFeaturizer|vbm_location_1"]
for column in splitColumns:
try:
newColumns = df[column].str.split(";", n = 2, expand = True)
for i in range(0,3):
df[column + "_" + str(i)] = np.array(newColumns[i]).astype(np.float)
except:
continue
df = df.drop(splitColumns, axis=1)
df = df.drop(["dos"], axis=1)
return clean_df(df)
def featurize_bandstructure(self, df):
"""Applies the preset band structure featurizers to the input dataframe,
renames some fields and cleans the output dataframe.
"""
df = super().featurize_bandstructure(df)
def _int_map(x):
if str(x) == "False":
return 0
elif str(x) == "True":
return 1
df["BandFeaturizer|is_gap_direct"] = df[
"BandFeaturizer|is_gap_direct"
].map(_int_map)
df = df.drop(["bandstructure"], axis=1)
return clean_df(df)
def featurize_site(self, df):
"""Applies the preset site featurizers to the input dataframe,
renames some fields and cleans the output dataframe.
"""
aliases = {
"GeneralizedRadialDistributionFunction": "GeneralizedRDF",
"AGNIFingerprints": "AGNIFingerPrint",
"BondOrientationalParameter": "BondOrientationParameter",
"GaussianSymmFunc": "ChemEnvSiteFingerprint|GaussianSymmFunc",
}
df = super().featurize_site(df, aliases=aliases)
df = df.loc[:, (df != 0).any(axis=0)]
return clean_df(df)
class PerovskiteProperty(BaseFeaturizer):
"""
Class to calculate perovskite features. Includes custom features from the Perovskite class and generic features from ElementProperty,
AtomicOrbitals, ValenceOrbital, and CohesiveEnergy matminer featurizers.
Options for initializing:
ordered_formula_featurizer(): for featurizing ordered formulas
cation_site_featurizer(): for featurizing unordered formulas based on user-provided cation site assignments
from_preset(): load a preset
The class can also be called manually, but be aware that different parameter sets are required for an ordered formula featurizer instance than for a cation site featurizer instance.
Parameters:
-----------
cation_site: dict of site assignments for cations, i.e. {el:site}. Elements not in cation_site are assumed to be anions on X-site
site_ox_lim: dict of oxidation state limits for each site, i.e. {site:[min,max]}. Elements on sites are limited to oxidation states within these limits
site_base_ox: dict of base oxidation state for each site, i.e. {site:ox}. Used for determining aliovalent ions and acceptor/donor dopants
ordered_formulas: if True, determine cation site assignments from order
A_site_occupancy: Number of atoms on A site. Used when ordered_formulas is True
anions: list of anions. Used when ordered_formulas is True
Parameters for ordered formula featurizer: site_ox_lim, site_base_ox, A_site_occupancy, anions
Parameters for cation site featurizer: cation_site, site_ox_lim, site_base_ox
"""
def __init__(self,
cation_site=None,
site_ox_lim={
'A': [0, 10],
'B': [0, 10],
'X': [-10, 0]
},
site_base_ox={
'A': 2,
'B': 4,
'X': -2
},
ordered_formulas=False,
A_site_occupancy=1,
anions=None):
if cation_site is None and ordered_formulas is False:
raise ValueError(
'Either cation sites must be assigned, or formulas must be ordered. Otherwise site assignments can not be determined'
)
self.cation_site = cation_site
self.site_ox_lim = site_ox_lim
self.site_base_ox = site_base_ox
self.ordered_formulas = ordered_formulas
self.A_site_occupancy = A_site_occupancy
self.anions = anions
#matminer featurizers
self.ValenceOrbital = ValenceOrbital()
self.AtomicOrbitals = AtomicOrbitalsMod()
self.CohesiveEnergy = CohesiveEnergy()
#custom ElementProperty featurizer
elemental_properties = [
'BoilingT', 'MeltingT', 'BulkModulus', 'ShearModulus', 'Row',
'Column', 'Number', 'MendeleevNumber', 'SpaceGroupNumber',
'Density', 'MolarVolume', 'FusionEnthalpy', 'HeatVaporization',
'NsUnfilled', 'NpUnfilled', 'NdUnfilled', 'NfUnfilled',
'Polarizability', 'ThermalConductivity'
]
self.ElementProperty = ElementProperty(
data_source='magpie',
features=elemental_properties,
stats=["mean", "std_dev", "range"])
self.check_matminer_featurizers()
self.featurize_options = {}
@classmethod
def from_preset(cls, preset_name):
"""
Initialize from preset
Parameters:
-----------
preset_name: name of preset to load. Currently accepts 'BCFZY'
"""
if preset_name == 'BCFZY':
#Ba(Co,Fe,Zr,Y)O_3-d system
cation_site = {
'Ba': 'A',
'Co': 'B',
'Fe': 'B',
'Zr': 'B',
'Y': 'B'
}
site_ox_lim = {'A': [2, 2], 'B': [2, 4], 'X': [-2, -2]}
site_base_ox = {'A': 2, 'B': 4, 'X': -2}
else:
raise ValueError("Invalid preset_name specified!")
return cls(cation_site, site_ox_lim, site_base_ox)
@classmethod
def ordered_formula_featurizer(cls,
A_site_occupancy=1,
anions=None,
site_ox_lim={
'A': [0, 10],
'B': [0, 10],
'X': [-10, 0]
},
site_base_ox={
'A': 2,
'B': 4,
'X': -2
}):
"""
Convenience method for instantiating a featurizer for ordered formulas
"""
return cls(cation_site=None,
site_ox_lim=site_ox_lim,
site_base_ox=site_base_ox,
ordered_formulas=True,
A_site_occupancy=A_site_occupancy,
anions=anions)
@classmethod
def cation_site_featurizer(cls,
cation_site,
site_ox_lim={
'A': [0, 10],
'B': [0, 10],
'X': [-10, 0]
},
site_base_ox={
'A': 2,
'B': 4,
'X': -2
}):
"""
Convenience method for instantiating a featurizer for unordered formulas, based on site assignments
"""
return cls(cation_site, site_ox_lim, site_base_ox)
@property
def ElementProperty_custom_labels(self):
"""
Generate custom labels for ElementProperty featurizer that follow same naming convention as Perovskite class
"""
elemental_property_label_map = {
'BoilingT': 'boil_temp',
'MeltingT': 'melt_temp',
'BulkModulus': 'bulk_mod',
'ShearModulus': 'shear_mod',
'Row': 'row',
'Column': 'column',
'Number': 'number',
'MendeleevNumber': 'mendeleev',
'SpaceGroupNumber': 'space_group',
'Density': 'density',
'MolarVolume': 'molar_vol',
'FusionEnthalpy': 'H_fus',
'HeatVaporization': 'H_vap',
'NsUnfilled': 'valence_unfilled_s',
'NpUnfilled': 'valence_unfilled_p',
'NdUnfilled': 'valence_unfilled_d',
'NfUnfilled': 'valence_unfilled_f',
'Polarizability': 'polarizability',
'ThermalConductivity': 'sigma_therm'
}
element_property_labels = list(
map(elemental_property_label_map.get,
self.ElementProperty.features))
labels = []
for attr in element_property_labels:
for stat in self.ElementProperty.stats:
if stat == 'std_dev':
stat = 'std'
labels.append(f'{attr}_{stat}')
return labels
@property
def ElementProperty_categories(self):
"""
Generate categories for ElementProperty featurizer
"""
elemental_property_category_map = {
'BoilingT': 'elemental',
'MeltingT': 'elemental',
'BulkModulus': 'elemental',
'ShearModulus': 'elemental',
'Row': 'periodic',
'Column': 'periodic',
'Number': 'periodic',
'MendeleevNumber': 'periodic',
'SpaceGroupNumber': 'periodic',
'Density': 'elemental',
'MolarVolume': 'elemental',
'FusionEnthalpy': 'elemental',
'HeatVaporization': 'elemental',
'NsUnfilled': 'electronic',
'NpUnfilled': 'electronic',
'NdUnfilled': 'electronic',
'NfUnfilled': 'electronic',
'Polarizability': 'elemental',
'ThermalConductivity': 'elemental'
}
element_property_categories = list(
map(elemental_property_category_map.get,
self.ElementProperty.features))
categories = []
for ep_cat in element_property_categories:
for stat in self.ElementProperty.stats:
categories.append(ep_cat)
return categories
@property
def ElementProperty_units(self):
"""
Generate units for ElementProperty featurizer
"""
elemental_property_unit_map = {
'BoilingT': 'temp',
'MeltingT': 'temp',
'BulkModulus': 'pressure',
'ShearModulus': 'pressure',
'Row': 'none',
'Column': 'none',
'Number': 'none',
'MendeleevNumber': 'none',
'SpaceGroupNumber': 'none',
'Density': 'density',
'MolarVolume': 'volume',
'FusionEnthalpy': 'energy',
'HeatVaporization': 'energy',
'NsUnfilled': 'none',
'NpUnfilled': 'none',
'NdUnfilled': 'none',
'NfUnfilled': 'none',
'Polarizability':
'polarizability', #complex units - doesn't matter
'ThermalConductivity': 'therm'
} #complex units - doesn't matter
element_property_units = list(
map(elemental_property_unit_map.get,
self.ElementProperty.features))
units = []
for ep_unit in element_property_units:
for stat in self.ElementProperty.stats:
units.append(ep_unit)
return units
def ElementProperty_label_check(self):
"""
Check that ElementProperty feature labels are as expected
If not, features may not align with feature labels
"""
#ElementProperty.feature_labels() code as of 2/17/19
labels = []
for attr in self.ElementProperty.features:
src = self.ElementProperty.data_source.__class__.__name__
for stat in self.ElementProperty.stats:
labels.append("{} {} {}".format(src, stat, attr))
if labels != self.ElementProperty.feature_labels():
raise Exception('ElementProperty features or labels have changed')
def set_featurize_options(
self,
sites,
ox_stats=['min', 'max', 'mean', 'median', 'std', 'range'],
ep_stats=["mean", "std_dev", "range"],
radius_type='ionic_radius',
normalize_formula=True,
silent=True,
categories=None):
"""
Set options for featurization. Since these options should be the same for all compositions in a batch, set for the featurizer instance rather than passing as args to featurize()
so that they do not have to be duplicated in every row of a DataFrame when calling featurize_dataframe().
Since these options change the number and meaning of features returned, it's also safest to set for the whole instance for consistency.
Parameters:
-----------
sites: list or string of sites to featurize. Any combination of 'A', 'B', 'X', and/or 'comp' accepted.
Composition-level, oxidation-state-dependent features are always calculated by the Perovskite class. Passing '' or [] will return only these features.
Specifying 'A','B', and/or 'X' sites will calculate site-level features for these sites (oxidation-state independent and dependent features, and matminer features).
Including 'comp' will calculate oxidation-state-independent features and matminer features for the full composition.
ox_stats: list of aggregate functions to apply to oxidation state combinations for feature generation using Perovskite class.
Options: 'min','max','mean','median','std','range'
ep_stats: ElementProperty stats. Options: "minimum", "maximum", "range", "mean", "avg_dev", "mode"
radius_type: Shannon radius type to use in features. Accepts 'crystal_radius' or 'ionic_radius'
normalize_formula: if True, normalize formula such that higher occupancy cation site has one formula unit (applies to Perovskite class only)
silent: if False, print informational messages from Perovksite class
categories: list of feature categories to return. If None, return all. Options: 'bonding','structure','charge','composition','electronic','elemental','periodic'
"""
feat_options = dict(sites=sites,
ox_stats=ox_stats,
radius_type=radius_type,
normalize_formula=normalize_formula,
silent=silent)
self.featurize_options.update(feat_options)
self.ElementProperty.stats = ep_stats
def featurize(self, formula):
"""
Calculate features
Parameters:
-----------
formula: chemical formula string
Returns: list of feature values
"""
if self.featurize_options == {}:
raise Exception(
'Featurize options have not been set. Use set_featurize_options before featurizing'
)
if self.ordered_formulas is True:
pvsk = Perovskite.from_ordered_formula(
formula,
self.A_site_occupancy,
self.anions,
site_ox_lim=self.site_ox_lim,
site_base_ox=self.site_base_ox,
radius_type=self.featurize_options['radius_type'],
silent=self.featurize_options['silent'])
elif self.ordered_formulas is False:
pvsk = Perovskite(formula, self.cation_site, self.site_ox_lim,
self.site_base_ox,
self.featurize_options['radius_type'],
self.featurize_options['normalize_formula'],
self.featurize_options['silent'])
pvsk_features = pvsk.featurize(self.featurize_options['sites'],
self.featurize_options['ox_stats'])
mm_features = []
for site in self.featurize_options['sites']:
vo_features = self.ValenceOrbital.featurize(
pvsk.site_composition[site]
) #avg and frac s, p , d, f electrons
vo_features += [sum(vo_features[0:3])
] #avg total valence electrons
ao_features = self.AtomicOrbitals.featurize(
pvsk.site_composition[site]
) #H**O and LUMO character and energy levels (from atomic orbitals)
ao_features = [
ao_features[i] for i in range(len(ao_features))
if i not in (0, 1, 3, 4)
] #exclude HOMO_character,HOMO_element, LUMO_character, LUMO_element - categoricals
ce_features = self.CohesiveEnergy.featurize(
pvsk.site_composition[site], formation_energy_per_atom=1e-10
) #avg elemental cohesive energy
ep_features = self.ElementProperty.featurize(
pvsk.site_composition[site]) #elemental property features
mm_features += vo_features + ao_features + ce_features + ep_features
features = list(pvsk_features) + mm_features
return features
@property
def matminer_labels(self):
"""
Feature labels for matminer-derived features
"""
labels = [
#ValenceOrbital labels
'valence_elec_s_mean',
'valence_elec_p_mean',
'valence_elec_d_mean',
'valence_elec_f_mean',
'valence_elec_s_frac',
'valence_elec_p_frac',
'valence_elec_d_frac',
'valence_elec_f_frac',
'valence_elec_tot_mean',
#AtomicOrbitals labels
#'HOMO_character',
'HOMO_energy',
#'LUMO_character',
'LUMO_energy',
'AO_gap',
#CohesiveEnergy labels
'cohesive_energy_mean'
]
#ElementProperty labels
labels += self.ElementProperty_custom_labels
return labels
@property
def matminer_categories(self):
"""
Feature categories for matminer-derived features
"""
categories = [
#ValenceOrbital categories
'electronic',
'electronic',
'electronic',
'electronic',
'electronic',
'electronic',
'electronic',
'electronic',
'electronic',
#AtomicOrbitals categories
#'HOMO_character',
'electronic',
#'LUMO_character',
'electronic',
'electronic',
#CohesiveEnergy categories
'bonding'
]
#ElementProperty categories
categories += self.ElementProperty_categories
return categories
@property
def matminer_units(self):
"""
Feature units for matminer-derived features
"""
units = [
#ValenceOrbital units
'none',
'none',
'none',
'none',
'none',
'none',
'none',
'none',
'none',
#AtomicOrbitals units
#'HOMO_character',
'energy',
#'LUMO_character',
'energy',
'energy',
#CohesiveEnergy units
'energy'
]
#ElementProperty units
units += self.ElementProperty_units
return units
def feature_labels(self):
"""
Get list of feature labels
"""
try:
pvsk_labels = Perovskite.from_preset(
'BaCoO3', 'BCFZY',
silent=True).feature_labels(self.featurize_options['sites'],
self.featurize_options['ox_stats'])
except KeyError:
raise Exception(
'Featurize options have not been set. Use set_featurize_options before accessing feature labels'
)
mm_labels = []
for site in self.featurize_options['sites']:
if site == 'comp':
site_label = 'comp'
else:
site_label = f'{site}site'
mm_labels += [
f'{site_label}_{label}' for label in self.matminer_labels
]
return pvsk_labels + mm_labels
def feature_categories(self):
"""
Get list of feature categories. For quick filtering
"""
try:
pvsk_categories = Perovskite.from_preset(
'BaCoO3', 'BCFZY', silent=True).feature_categories(
self.featurize_options['sites'],
self.featurize_options['ox_stats'])
except KeyError:
raise Exception(
'Featurize options have not been set. Use set_featurize_options before accessing feature labels'
)
mm_categories = []
for site in self.featurize_options['sites']:
mm_categories += self.matminer_categories
return pvsk_categories + mm_categories
def feature_units(self):
"""
Get list of feature labels. For dimensional analysis
"""
try:
pvsk_units = Perovskite.from_preset(
'BaCoO3', 'BCFZY',
silent=True).feature_units(self.featurize_options['sites'],
self.featurize_options['ox_stats'])
except KeyError:
raise Exception(
'Featurize options have not been set. Use set_featurize_options before accessing feature labels'
)
mm_units = []
for site in self.featurize_options['sites']:
mm_units += self.matminer_units
return pvsk_units + mm_units
def check_matminer_featurizers(self):
"""
Check that features and feature order for matminer featurizers are as expected
If features or feature order have changed, featurize() may return unexpected features that do not align with feature_labels()
"""
#verify that matminer feature labels haven't changed
if self.ValenceOrbital.feature_labels() != [
'avg s valence electrons', 'avg p valence electrons',
'avg d valence electrons', 'avg f valence electrons',
'frac s valence electrons', 'frac p valence electrons',
'frac d valence electrons', 'frac f valence electrons'
]:
raise Exception('ValenceOrbital features or labels have changed')
if self.AtomicOrbitals.feature_labels() != [
'HOMO_character', 'HOMO_element', 'HOMO_energy',
'LUMO_character', 'LUMO_element', 'LUMO_energy', 'gap_AO'
]:
raise Exception('AtomicOrbitals features or labels have changed')
if self.CohesiveEnergy.feature_labels() != ['cohesive energy']:
raise Exception('CohesiveEnergy features or labels have changed')
self.ElementProperty_label_check()
def predict_log10_eps(
target: Union[Structure, Composition],
dielectric_type: str,
model_type: str,
) -> float:
"""
:param target: structure or composition to predict dielectric constants
:param dielectric_type: "el" or "ion"
:param model_type: "comp" or "comp_st"
:return: Descriptor vector
"""
if dielectric_type not in ["el", "ion"]:
raise ValueError(
f'Specify dielectric type "el" or "ion"\nInput: {dielectric_type}')
if model_type not in ["comp", "comp_st"]:
raise ValueError(
f'Specify regression_type "comp" or "comp_st"\nInput: {model_type}'
)
if model_type == "comp":
if isinstance(target, Structure):
comp = target.composition
else:
comp = target
comp_oxi = comp.add_charges_from_oxi_state_guesses()
if dielectric_type == "el":
ep = ScalarFeaturizer(ElementProperty.from_preset("matminer"),
comp)
valence = ScalarFeaturizer(ValenceOrbital(), comp)
ion_prop = ScalarFeaturizer(IonProperty(), comp)
en_diff = ScalarFeaturizer(ElectronegativityDiff(), comp_oxi)
oxi_state = ScalarFeaturizer(OxidationStates.from_preset("deml"),
comp_oxi)
atomic_orbital = ScalarFeaturizer(AtomicOrbitals(), comp)
descriptor = [
ep.get_from_label("PymatgenData minimum X"),
ep.get_from_label("PymatgenData range X"),
ep.get_from_label("PymatgenData std_dev X"),
ep.get_from_label("PymatgenData mean row"),
ep.get_from_label("PymatgenData std_dev row"),
ep.get_from_label("PymatgenData mean group"),
ep.get_from_label("PymatgenData mean block"),
ep.get_from_label("PymatgenData std_dev block"),
ep.get_from_label("PymatgenData mean atomic_mass"),
ep.get_from_label("PymatgenData std_dev atomic_mass"),
ep.get_from_label("PymatgenData std_dev atomic_radius"),
ep.get_from_label("PymatgenData minimum mendeleev_no"),
ep.get_from_label("PymatgenData range mendeleev_no"),
ep.get_from_label("PymatgenData std_dev mendeleev_no"),
ep.get_from_label("PymatgenData mean thermal_conductivity"),
ep.get_from_label("PymatgenData std_dev thermal_conductivity"),
ep.get_from_label("PymatgenData mean melting_point"),
ep.get_from_label("PymatgenData std_dev melting_point"),
valence.get_from_label("avg s valence electrons"),
valence.get_from_label("avg d valence electrons"),
valence.get_from_label("frac s valence electrons"),
valence.get_from_label("frac p valence electrons"),
valence.get_from_label("frac d valence electrons"),
ion_prop.get_from_label("avg ionic char"),
TMetalFraction().featurize(comp)[0],
en_diff.get_from_label("maximum EN difference"),
en_diff.get_from_label("range EN difference"),
en_diff.get_from_label("mean EN difference"),
en_diff.get_from_label("std_dev EN difference"),
BandCenter().featurize(comp)[0],
oxi_state.get_from_label("std_dev oxidation state"),
atomic_orbital.get_from_label("HOMO_energy"),
atomic_orbital.get_from_label("LUMO_energy"),
atomic_orbital.get_from_label("gap_AO"),
]
elif dielectric_type == "ion":
stoich = ScalarFeaturizer(Stoichiometry(), comp)
ep = ScalarFeaturizer(ElementProperty.from_preset("matminer"),
comp)
valence = ScalarFeaturizer(ValenceOrbital(), comp)
ion_prop = ScalarFeaturizer(IonProperty(), comp)
en_diff = ScalarFeaturizer(ElectronegativityDiff(), comp_oxi)
oxi_state = ScalarFeaturizer(OxidationStates.from_preset("deml"),
comp_oxi)
atomic_orbital = ScalarFeaturizer(AtomicOrbitals(), comp)
at_pack_eff = ScalarFeaturizer(AtomicPackingEfficiency(), comp)
descriptor = [
stoich.get_from_label("3-norm"),
stoich.get_from_label("5-norm"),
ep.get_from_label("PymatgenData mean X"),
ep.get_from_label("PymatgenData mean row"),
ep.get_from_label("PymatgenData std_dev row"),
ep.get_from_label("PymatgenData std_dev group"),
ep.get_from_label("PymatgenData mean block"),
ep.get_from_label("PymatgenData std_dev block"),
ep.get_from_label("PymatgenData maximum atomic_mass"),
ep.get_from_label("PymatgenData range atomic_mass"),
ep.get_from_label("PymatgenData mean atomic_mass"),
ep.get_from_label("PymatgenData std_dev atomic_mass"),
ep.get_from_label("PymatgenData maximum atomic_radius"),
ep.get_from_label("PymatgenData range atomic_radius"),
ep.get_from_label("PymatgenData mean atomic_radius"),
ep.get_from_label("PymatgenData std_dev atomic_radius"),
ep.get_from_label("PymatgenData minimum mendeleev_no"),
ep.get_from_label("PymatgenData mean mendeleev_no"),
ep.get_from_label("PymatgenData std_dev mendeleev_no"),
ep.get_from_label("PymatgenData mean thermal_conductivity"),
ep.get_from_label("PymatgenData std_dev thermal_conductivity"),
ep.get_from_label("PymatgenData mean melting_point"),
ep.get_from_label("PymatgenData std_dev melting_point"),
valence.get_from_label("avg s valence electrons"),
valence.get_from_label("frac s valence electrons"),
valence.get_from_label("frac p valence electrons"),
valence.get_from_label("frac d valence electrons"),
ion_prop.get_from_label("avg ionic char"),
TMetalFraction().featurize(comp)[0],
en_diff.get_from_label("minimum EN difference"),
en_diff.get_from_label("range EN difference"),
en_diff.get_from_label("mean EN difference"),
en_diff.get_from_label("std_dev EN difference"),
oxi_state.get_from_label("range oxidation state"),
oxi_state.get_from_label("std_dev oxidation state"),
atomic_orbital.get_from_label("LUMO_energy"),
atomic_orbital.get_from_label("gap_AO"),
at_pack_eff.get_from_label("mean simul. packing efficiency"),
at_pack_eff.get_from_label(
"mean abs simul. packing efficiency"),
at_pack_eff.get_from_label(
"dist from 1 clusters |APE| < 0.010"),
at_pack_eff.get_from_label(
"dist from 3 clusters |APE| < 0.010"),
at_pack_eff.get_from_label(
"dist from 5 clusters |APE| < 0.010"),
]
elif model_type == "comp_st":
if isinstance(target, Composition):
raise ValueError(
'comp_st (Using compositional and structural descriptor) is specified, '
'but target is composition')
comp: Composition = target.composition
comp_oxi = comp.add_charges_from_oxi_state_guesses()
target_orig = deepcopy(target)
target.add_oxidation_state_by_guess()
if dielectric_type == "el":
ep = ScalarFeaturizer(ElementProperty.from_preset("matminer"),
comp)
valence = ScalarFeaturizer(ValenceOrbital(), comp)
en_diff = ScalarFeaturizer(ElectronegativityDiff(), comp_oxi)
atomic_orbital = ScalarFeaturizer(AtomicOrbitals(), comp)
density = ScalarFeaturizer(DensityFeatures(), target)
dist_btw_nn = MinimumRelativeDistances().featurize(target_orig)
opsf = SiteFeaturizer(OPSiteFingerprint(), target)
voro_fp = SiteFeaturizer(VoronoiFingerprint(use_symm_weights=True),
target)
gsf = SiteFeaturizer(GaussianSymmFunc(), target)
lpd = SiteFeaturizer(
LocalPropertyDifference.from_preset("ward-prb-2017"), target)
descriptor = [
ep.get_from_label("PymatgenData std_dev X"),
ep.get_from_label("PymatgenData mean block"),
ep.get_from_label("PymatgenData std_dev atomic_mass"),
valence.get_from_label("frac d valence electrons"),
TMetalFraction().featurize(comp)[0],
en_diff.get_from_label("maximum EN difference"),
en_diff.get_from_label("mean EN difference"),
atomic_orbital.get_from_label("HOMO_energy"),
atomic_orbital.get_from_label("LUMO_energy"),
density.get_from_label("density"),
np.mean(dist_btw_nn),
np.std(dist_btw_nn),
opsf.get_from_label_func("tetrahedral CN_4", np.max),
opsf.get_from_label_func("rectangular see-saw-like CN_4",
np.max),
np.max([
EwaldSiteEnergy(accuracy=4).featurize(target, i)
for i in range(target.num_sites)
]),
voro_fp.get_from_label_func("Voro_area_std_dev", np.max),
voro_fp.get_from_label_func("Voro_area_std_dev", np.mean),
voro_fp.get_from_label_func("Voro_dist_minimum", np.min),
voro_fp.get_from_label_func("Voro_dist_minimum", np.std),
gsf.get_from_label_func("G2_20.0", np.std),
gsf.get_from_label_func("G2_80.0", np.max),
gsf.get_from_label_func("G4_0.005_4.0_-1.0", np.mean),
lpd.get_from_label_func("local difference in NdValence",
np.mean),
lpd.get_from_label_func("local difference in NValence",
np.min),
lpd.get_from_label_func("local difference in NValence",
np.std),
lpd.get_from_label_func("local difference in NdUnfilled",
np.mean),
lpd.get_from_label_func("local difference in NUnfilled",
np.min),
lpd.get_from_label_func("local difference in NUnfilled",
np.mean),
lpd.get_from_label_func("local difference in GSmagmom",
np.mean)
]
elif dielectric_type == "ion":
ep = ScalarFeaturizer(ElementProperty.from_preset("matminer"),
comp)
atomic_orbitals = ScalarFeaturizer(AtomicOrbitals(), comp)
density = ScalarFeaturizer(DensityFeatures(), target)
str_het = ScalarFeaturizer(StructuralHeterogeneity(), target)
opsf = SiteFeaturizer(OPSiteFingerprint(), target)
voro_fp = SiteFeaturizer(VoronoiFingerprint(use_symm_weights=True),
target)
gsf = SiteFeaturizer(GaussianSymmFunc(), target)
lpd = SiteFeaturizer(
LocalPropertyDifference.from_preset("ward-prb-2017"), target)
descriptor = [
ep.get_from_label("PymatgenData std_dev row"),
ep.get_from_label("PymatgenData mean thermal_conductivity"),
ep.get_from_label("PymatgenData std_dev melting_point"),
TMetalFraction().featurize(comp)[0],
atomic_orbitals.get_from_label("gap_AO"),
density.get_from_label("density"),
density.get_from_label("packing fraction"),
str_het.get_from_label("mean neighbor distance variation"),
str_het.get_from_label("avg_dev neighbor distance variation"),
opsf.get_from_label_func("sgl_bd CN_1", np.mean),
opsf.get_from_label_func("bent 150 degrees CN_2", np.mean),
opsf.get_from_label_func("linear CN_2", np.mean),
opsf.get_from_label_func("trigonal planar CN_3", np.mean),
opsf.get_from_label_func("pentagonal planar CN_5", np.std),
opsf.get_from_label_func("octahedral CN_6", np.max),
opsf.get_from_label_func("octahedral CN_6", np.std),
opsf.get_from_label_func("q6 CN_12", np.mean),
np.max([
EwaldSiteEnergy(accuracy=4).featurize(target, i)
for i in range(target.num_sites)
]),
voro_fp.get_from_label_func("Symmetry_weighted_index_4",
np.std),
voro_fp.get_from_label_func("Voro_vol_maximum", np.mean),
voro_fp.get_from_label_func("Voro_area_std_dev", np.mean),
voro_fp.get_from_label_func("Voro_area_minimum", np.std),
voro_fp.get_from_label_func("Voro_area_maximum", np.min),
voro_fp.get_from_label_func("Voro_dist_std_dev", np.mean),
gsf.get_from_label_func("G2_80.0", np.min),
gsf.get_from_label_func("G4_0.005_4.0_1.0", np.std),
lpd.get_from_label_func("local difference in Number", np.max),
lpd.get_from_label_func("local difference in MendeleevNumber",
np.max),
lpd.get_from_label_func("local difference in MendeleevNumber",
np.min),
lpd.get_from_label_func("local difference in AtomicWeight",
np.max),
lpd.get_from_label_func("local difference in AtomicWeight",
np.mean),
lpd.get_from_label_func("local difference in MeltingT",
np.mean),
lpd.get_from_label_func("local difference in Row", np.max),
lpd.get_from_label_func(
"local difference in Electronegativity", np.min),
lpd.get_from_label_func("local difference in NValence",
np.std),
lpd.get_from_label_func("local difference in NsUnfilled",
np.mean),
lpd.get_from_label_func("local difference in NdUnfilled",
np.max),
lpd.get_from_label_func("local difference in NdUnfilled",
np.std),
lpd.get_from_label_func("local difference in NUnfilled",
np.max),
lpd.get_from_label_func("local difference in NUnfilled",
np.min),
lpd.get_from_label_func("local difference in NUnfilled",
np.mean),
lpd.get_from_label_func("local difference in NUnfilled",
np.std),
lpd.get_from_label_func("local difference in GSvolume_pa",
np.max),
lpd.get_from_label_func("local difference in GSvolume_pa",
np.min),
lpd.get_from_label_func("local difference in SpaceGroupNumber",
np.max),
]
with open(
f"{os.path.dirname(__file__)}/{dielectric_type}_{model_type}.joblib",
"rb") as fr:
model: RandomForestRegressor = joblib.load(fr)
with open(
f"{os.path.dirname(__file__)}/{dielectric_type}_{model_type}_scaler.joblib",
"rb") as fr:
scaler: StandardScaler = joblib.load(fr)
descriptor = scaler.transform([descriptor])
return model.predict(descriptor)[0]
def featurize_structures(self, featurizer=None, **kwargs):
"""
Featurizes the hypothetical structures available from
hypo_structures method. Hypothetical structures for which
featurization fails is removed and valid structures are
made available as valid_structures
Args:
featurizer (Featurizer): A MatMiner Featurizer.
Defaults to MultipleFeaturizer with PRB Ward
Voronoi descriptors.
**kwargs (dict): kwargs passed to featurize_many
method of featurizer.
Returns:
pandas.DataFrame: features
"""
# Note the redundancy here is for pandas to work
if self.hypo_structures is None:
warnings.warn("No structures available. Generating structures.")
self.get_structures()
print("Generating features")
featurizer = featurizer if featurizer else MultipleFeaturizer([
SiteStatsFingerprint.from_preset(
"CoordinationNumber_ward-prb-2017"),
StructuralHeterogeneity(),
ChemicalOrdering(),
MaximumPackingEfficiency(),
SiteStatsFingerprint.from_preset(
"LocalPropertyDifference_ward-prb-2017"),
StructureComposition(Stoichiometry()),
StructureComposition(ElementProperty.from_preset("magpie")),
StructureComposition(ValenceOrbital(props=['frac'])),
StructureComposition(IonProperty(fast=True))
])
features = featurizer.featurize_many(
self.hypo_structures['pmg_structures'],
ignore_errors=True,
**kwargs)
n_species, formula = [], []
for s in self.hypo_structures['pmg_structures']:
n_species.append(len(s.composition.elements))
formula.append(s.composition.formula)
self._features_df = pd.DataFrame.from_records(
features, columns=featurizer.feature_labels())
self._features_df.index = self.hypo_structures.index
self._features_df['N_species'] = n_species
self._features_df['Composition'] = formula
self.features = self._features_df.dropna(axis=0, how='any')
self.features = self.features.reindex(sorted(self.features.columns),
axis=1)
self._valid_structure_labels = list(self.features.index)
self.valid_structures = self.hypo_structures.loc[
self._valid_structure_labels]
print("{} out of {} structures were successfully featurized.".format(
self.features.shape[0], self._features_df.shape[0]))
return self.features
def AddFeatures(df): # Add features by Matminer
from matminer.featurizers.conversions import StrToComposition
df = StrToComposition().featurize_dataframe(df, "formula")
from matminer.featurizers.composition import ElementProperty
ep_feat = ElementProperty.from_preset(preset_name="magpie")
df = ep_feat.featurize_dataframe(
df, col_id="composition"
) # input the "composition" column to the featurizer
from matminer.featurizers.conversions import CompositionToOxidComposition
from matminer.featurizers.composition import OxidationStates
df = CompositionToOxidComposition().featurize_dataframe(df, "composition")
os_feat = OxidationStates()
df = os_feat.featurize_dataframe(df, "composition_oxid")
from matminer.featurizers.composition import ElectronAffinity
ea_feat = ElectronAffinity()
df = ea_feat.featurize_dataframe(df,
"composition_oxid",
ignore_errors=True)
from matminer.featurizers.composition import BandCenter
bc_feat = BandCenter()
df = bc_feat.featurize_dataframe(df,
"composition_oxid",
ignore_errors=True)
from matminer.featurizers.composition import CohesiveEnergy
ce_feat = CohesiveEnergy()
df = ce_feat.featurize_dataframe(df,
"composition_oxid",
ignore_errors=True)
from matminer.featurizers.composition import Miedema
m_feat = Miedema()
df = m_feat.featurize_dataframe(df, "composition_oxid", ignore_errors=True)
from matminer.featurizers.composition import TMetalFraction
tmf_feat = TMetalFraction()
df = tmf_feat.featurize_dataframe(df,
"composition_oxid",
ignore_errors=True)
from matminer.featurizers.composition import ValenceOrbital
vo_feat = ValenceOrbital()
df = vo_feat.featurize_dataframe(df,
"composition_oxid",
ignore_errors=True)
from matminer.featurizers.composition import YangSolidSolution
yss_feat = YangSolidSolution()
df = yss_feat.featurize_dataframe(df,
"composition_oxid",
ignore_errors=True)
from matminer.featurizers.structure import GlobalSymmetryFeatures
# This is the border between compositional features and structural features. Comment out the following featurizers to use only compostional features.
gsf_feat = GlobalSymmetryFeatures()
df = gsf_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.structure import StructuralComplexity
sc_feat = StructuralComplexity()
df = sc_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.structure import ChemicalOrdering
co_feat = ChemicalOrdering()
df = co_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.structure import MaximumPackingEfficiency
mpe_feat = MaximumPackingEfficiency()
df = mpe_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.structure import MinimumRelativeDistances
mrd_feat = MinimumRelativeDistances()
df = mrd_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.structure import StructuralHeterogeneity
sh_feat = StructuralHeterogeneity()
df = sh_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.structure import SiteStatsFingerprint
from matminer.featurizers.site import AverageBondLength
from pymatgen.analysis.local_env import CrystalNN
bl_feat = SiteStatsFingerprint(
AverageBondLength(CrystalNN(search_cutoff=20)))
df = bl_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.site import AverageBondAngle
ba_feat = SiteStatsFingerprint(
AverageBondAngle(CrystalNN(search_cutoff=20)))
df = ba_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.site import BondOrientationalParameter
bop_feat = SiteStatsFingerprint(BondOrientationalParameter())
df = bop_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.site import CoordinationNumber
cn_feat = SiteStatsFingerprint(CoordinationNumber())
df = cn_feat.featurize_dataframe(df, "structure", ignore_errors=True)
from matminer.featurizers.structure import DensityFeatures
df_feat = DensityFeatures()
df = df_feat.featurize_dataframe(df, "structure", ignore_errors=True)
return (df)
OxidCompositions = CompositionToOxidComposition()
print(OxidCompositions.feature_labels())
df = OxidCompositions.featurize_dataframe(df, 'composition')
#CohesiveEnergy
from matminer.featurizers.composition import CohesiveEnergy
cohesive_energy = CohesiveEnergy()
cohesive_energy.set_n_jobs(28)
labels.append(cohesive_energy.feature_labels())
df = cohesive_energy.featurize_dataframe(df,
'composition',
ignore_errors=True)
#ValenceOrbital
from matminer.featurizers.composition import ValenceOrbital
valence_orbital = ValenceOrbital()
valence_orbital.set_n_jobs(28)
labels.append(valence_orbital.feature_labels())
df = valence_orbital.featurize_dataframe(df,
'composition',
ignore_errors=True)
#AtomicOrbital
from matminer.featurizers.composition import AtomicOrbitals
atomic_orbitals = AtomicOrbitals()
atomic_orbitals.set_n_jobs(28)
labels.append(atomic_orbitals.feature_labels())
df = atomic_orbitals.featurize_dataframe(df,
'composition',
ignore_errors=True)
class BCA_Featurizer(BaseFeaturizer):
def __init__(self,radius_type='ionic_radius',normalize_formula=False):
self.radius_type = radius_type
self.normalize_formula = normalize_formula
self.ValenceOrbital = ValenceOrbital()
self.AtomicOrbitals = AtomicOrbitalsMod()
self.CohesiveEnergy = CohesiveEnergy()
self.BandCenter = BandCenter()
self.ValenceOrbitalEnergy = ValenceOrbitalEnergy()
#custom ElementProperty featurizer
elemental_properties = ['BoilingT', 'MeltingT',
'BulkModulus', 'ShearModulus',
'Row', 'Column', 'Number', 'MendeleevNumber', 'SpaceGroupNumber',
'Density','MolarVolume',
'FusionEnthalpy','HeatVaporization',
'Polarizability',
'ThermalConductivity']
self.ElementProperty = ElementProperty(data_source='magpie',features=elemental_properties,
stats=["mean", "std_dev"])
#check matminer featurizers
self.check_matminer_featurizers()
def featurize(self,composition):
bca = BCA(composition,self.radius_type,self.normalize_formula)
bca_features = bca.featurize()
vo_features = self.ValenceOrbital.featurize(bca.metal_composition) #avg and frac s, p , d, f electrons for metals
vo_features += [sum(vo_features[0:3])] #avg total valence electrons for metals
ao_features = self.AtomicOrbitals.featurize(bca.metal_composition) #H**O and LUMO character and energy levels for metals from atomic orbitals)
ao_features = [ao_features[i] for i in range(len(ao_features)) if i not in (0,1,3,4)]#exclude HOMO_character,HOMO_element, LUMO_character, LUMO_element - categoricals
ce_features = self.CohesiveEnergy.featurize(bca.metal_composition,formation_energy_per_atom=1e-10) #avg metal elemental cohesive energy
bc_features = self.BandCenter.featurize(bca.metal_composition) + self.BandCenter.featurize(bca.composition)
ve_features = self.ValenceOrbitalEnergy.featurize(bca.metal_composition) + self.ValenceOrbitalEnergy.featurize(bca.composition)
ep_features = self.ElementProperty.featurize(bca.metal_composition) + self.ElementProperty.featurize(bca.composition)
mm_features = vo_features + ao_features + ce_features + bc_features + ve_features + ep_features
return list(bca_features.values()) + mm_features
@property
def ElementProperty_custom_labels(self):
"""
Generate custom labels for ElementProperty featurizer that follow same naming convention as Perovskite class
"""
elemental_property_label_map = {'BoilingT':'boil_temp','MeltingT':'melt_temp',
'BulkModulus':'bulk_mod','ShearModulus':'shear_mod',
'Row':'row','Column':'column','Number':'number','MendeleevNumber':'mendeleev','SpaceGroupNumber':'space_group',
'Density':'density','MolarVolume':'molar_vol',
'FusionEnthalpy':'H_fus','HeatVaporization':'H_vap',
'Polarizability':'polarizability',
'ThermalConductivity':'sigma_therm'}
element_property_labels = list(map(elemental_property_label_map.get,self.ElementProperty.features))
labels = []
for attr in element_property_labels:
for stat in self.ElementProperty.stats:
if stat=='std_dev':
stat = 'std'
labels.append(f'M_{attr}_{stat}')
for attr in element_property_labels:
for stat in self.ElementProperty.stats:
if stat=='std_dev':
stat = 'std'
labels.append(f'BCA_{attr}_{stat}')
return labels
@property
def ElementProperty_units(self):
"""
Generate units for ElementProperty featurizer that follow same naming convention as Perovskite class
"""
elemental_property_unit_map = {'BoilingT':'temperature','MeltingT':'temperature',
'BulkModulus':'pressure','ShearModulus':'pressure',
'Row':'none','Column':'none','Number':'none','MendeleevNumber':'none','SpaceGroupNumber':'none',
'Density':'density','MolarVolume':'volume',
'FusionEnthalpy':'energy','HeatVaporization':'energy',
'Polarizability':'polarizability',
'ThermalConductivity':'therm'}
element_property_units = list(map(elemental_property_unit_map.get,self.ElementProperty.features))
units = []
for ep_unit in element_property_units:
for stat in self.ElementProperty.stats:
units.append(ep_unit)
return units*2
def ElementProperty_label_check(self):
"""
Check that ElementProperty feature labels are as expected
If not, features may not align with feature labels
"""
#ElementProperty.feature_labels() code as of 1/24/20
labels = []
for attr in self.ElementProperty.features:
src = self.ElementProperty.data_source.__class__.__name__
for stat in self.ElementProperty.stats:
labels.append("{} {} {}".format(src, stat, attr))
if labels!=self.ElementProperty.feature_labels():
raise Exception('ElementProperty features or labels have changed')
@property
def matminer_labels(self):
"""
Feature labels for matminer-derived features
"""
labels = [
#ValenceOrbital labels
'M_ValenceElec_s_mean',
'M_ValenceElec_p_mean',
'M_ValenceElec_d_mean',
'M_ValenceElec_f_mean',
'M_ValenceElec_s_frac',
'M_ValenceElec_p_frac',
'M_ValenceElec_d_frac',
'M_ValenceElec_f_frac',
'M_ValenceElec_tot_mean',
#AtomicOrbitals labels
#'M_HOMO_character',
'M_HOMO_energy',
#'M_LUMO_character',
'M_LUMO_energy',
'M_AO_gap',
#CohesiveEnergy labels
'M_cohesive_energy_mean',
#BandCenter labels
'M_BandCenter',
'BCA_BandCenter',
#ValenceOrbitalEnergy labels
'M_ValenceEnergy_mean',
'BCA_ValenceEnergy_mean'
]
labels += self.ElementProperty_custom_labels
return labels
@property
def matminer_units(self):
"""
Feature units for matminer-derived features
"""
units = [
#ValenceOrbital units
'none',
'none',
'none',
'none',
'none',
'none',
'none',
'none',
'none',
#AtomicOrbitals units
#'M_HOMO_character',
'energy',
#'M_LUMO_character',
'energy',
'energy',
#CohesiveEnergy units
'energy',
#BandCenter units
'energy',
'energy',
#ValenceOrbitalEnergy units
'energy',
'energy'
]
units += self.ElementProperty_units
return units
def feature_labels(self):
bca_feature_labels = list(BCA(mg.Composition('BaO'),self.radius_type,self.normalize_formula).featurize().keys())
return bca_feature_labels + self.matminer_labels
def feature_units(self):
bca_units = BCA(mg.Composition('BaO')).feature_units()
return bca_units + self.matminer_units
def check_matminer_featurizers(self):
"""
Check that features and feature order for matminer featurizers are as expected
If features or feature order have changed, featurize() may return unexpected features that do not align with feature_labels()
"""
#verify that matminer feature labels haven't changed
if self.ValenceOrbital.feature_labels() != ['avg s valence electrons',
'avg p valence electrons',
'avg d valence electrons',
'avg f valence electrons',
'frac s valence electrons',
'frac p valence electrons',
'frac d valence electrons',
'frac f valence electrons']:
raise Exception('ValenceOrbital features or labels have changed')
if self.AtomicOrbitals.feature_labels() != ['HOMO_character',
'HOMO_element',
'HOMO_energy',
'LUMO_character',
'LUMO_element',
'LUMO_energy',
'gap_AO']:
raise Exception('AtomicOrbitals features or labels have changed')
if self.CohesiveEnergy.feature_labels() != ['cohesive energy']:
raise Exception('CohesiveEnergy features or labels have changed')
if self.BandCenter.feature_labels() != ['band center']:
raise Exception('BandCenter features or labels have changed')
self.ElementProperty_label_check()
def citations(self):
featurizers = [self.ValenceOrbital, self.AtomicOrbitals, self.CohesiveEnergy, self.BandCenter, self.ValenceOrbitalEnergy, BCA(mg.Composition('BaO'))]
return list(np.unique(sum([f.citations() for f in featurizers],[])))
'NdUnfilled', 'NfUnfilled', 'NUnfilled', 'GSvolume_pa',
'SpaceGroupNumber', 'GSbandgap', 'GSmagmom')
#The following features will be created by using matminer package.
featurizer = MultipleFeaturizer([
SiteStatsFingerprint(CoordinationNumber().from_preset('VoronoiNN'),
stats=('mean', 'std_dev', 'minimum', 'maximum')),
StructuralHeterogeneity(),
ChemicalOrdering(),
MaximumPackingEfficiency(),
SiteStatsFingerprint(
LocalPropertyDifference(properties=element_properties),
stats=('mean', 'std_dev', 'minimum', 'maximum', 'range')),
StructureComposition(Stoichiometry()),
StructureComposition(ElementProperty.from_preset("magpie")),
StructureComposition(ValenceOrbital(props=['frac'])),
StructureComposition(IonProperty(fast=True))
])
#Generate VT based features from the material's crystal lat_params.
feature_data = featurizer.featurize_dataframe(df,
col_id=['structure'],
ignore_errors=True)
#"lat_params","compound possible" and "material_id" are not resonable physical features, so we drop these three columns
feature_data = feature_data.drop(
["structure", "compound possible", "material_id"], axis=1)
#write the data into a csv file for later use
feature_data.to_csv("data_delta_e_data.csv", index=False)
from sklearn.model_selection import KFold, cross_val_score
from sklearn.model_selection import train_test_split, KFold, ShuffleSplit
from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor
def feature_labels(self):
# Since we have more features than just element fractions, append 'fraction' to element symbols for clarity
element_fraction_features = [e + " fraction" for e in ElementFraction().feature_labels()]
valence_orbital_features = ValenceOrbital().feature_labels()
return element_fraction_features+self._element_property_feature_labels+valence_orbital_features
