Below, you will find a description of each featurizer, listed in tables grouped by module.

bandstructure

Features derived from a material’s electronic bandstructure.

matminer.featurizers.bandstructure

Name

Description

BranchPointEnergy

Branch point energy and absolute band edge position.

BandFeaturizer

Featurizes a pymatgen band structure object.

base

Parent classes and meta-featurizers.

matminer.featurizers.base

Name

Description

MultipleFeaturizer

Class to run multiple featurizers on the same input data.

StackedFeaturizer

Use the output of a machine learning model as features

BaseFeaturizer

Abstract class to calculate features from raw materials input data

composition

Features based on a material’s composition.

alloy

matminer.featurizers.composition.alloy

Composition featurizers specialized for use with alloys.

Name

Description

Miedema

Formation enthalpies of intermetallic compounds, from Miedema et al.

YangSolidSolution

Mixing thermochemistry and size mismatch terms of Yang and Zhang (2012)

WenAlloys

Calculate features for alloy properties.

composite

matminer.featurizers.composition.composite

Composition featurizers for composite features containing more than 1 category of general-purpose data.

Name

Description

ElementProperty

Class to calculate elemental property attributes.

Meredig

Class to calculate features as defined in Meredig et. al.

element

matminer.featurizers.composition.element

Composition featurizers for elemental data and stoichiometry.

Name

Description

ElementFraction

Class to calculate the atomic fraction of each element in a composition.

TMetalFraction

Class to calculate fraction of magnetic transition metals in a composition.

Stoichiometry

Calculate norms of stoichiometric attributes.

BandCenter

Estimation of absolute position of band center using electronegativity.

ion

matminer.featurizers.composition.ion

Composition featurizers for compositions with ionic data.

Name

Description

OxidationStates

Statistics about the oxidation states for each specie.

IonProperty

Ionic property attributes. Similar to ElementProperty.

ElectronAffinity

Calculate average electron affinity times formal charge of anion elements.

ElectronegativityDiff

Features from electronegativity differences between anions and cations.

orbital

matminer.featurizers.composition.orbital

Composition featurizers for orbital data.

Name

Description

AtomicOrbitals

Determine HOMO/LUMO features based on a composition.

ValenceOrbital

Attributes of valence orbital shells

packing

matminer.featurizers.composition.packing

Composition featurizers for determining packing characteristics.

Name

Description

AtomicPackingEfficiency

Packing efficiency based on a geometric theory of the amorphous packing

thermo

matminer.featurizers.composition.thermo

Composition featurizers for thermodynamic properties.

Name

Description

CohesiveEnergy

Cohesive energy per atom using elemental cohesive energies and

CohesiveEnergyMP

Cohesive energy per atom lookup using Materials Project

conversions

Conversion utilities.

matminer.featurizers.conversions

Name

Description

ConversionFeaturizer

Abstract class to perform data conversions.

StrToComposition

Utility featurizer to convert a string to a Composition

StructureToComposition

Utility featurizer to convert a Structure to a Composition.

StructureToIStructure

Utility featurizer to convert a Structure to an immutable IStructure.

DictToObject

Utility featurizer to decode a dict to Python object via MSON.

JsonToObject

Utility featurizer to decode json data to a Python object via MSON.

StructureToOxidStructure

Utility featurizer to add oxidation states to a pymatgen Structure.

CompositionToOxidComposition

Utility featurizer to add oxidation states to a pymatgen Composition.

CompositionToStructureFromMP

Featurizer to get a Structure object from Materials Project using the

PymatgenFunctionApplicator

Featurizer to run any function using on/from pymatgen primitives.

ASEAtomstoStructure

Convert dataframes of ase structures to pymatgen structures for further use with

dos

Features based on a material’s electronic density of states.

matminer.featurizers.dos

Name

Description

SiteDOS

report the fractional s/p/d/f dos for a particular site. a CompleteDos

DOSFeaturizer

Significant character and contribution of the density of state from a

DopingFermi

The fermi level (w.r.t. selected reference energy) associated with a

Hybridization

quantify s/p/d/f orbital character and their hybridizations at band edges

DosAsymmetry

Quantifies the asymmetry of the DOS near the Fermi level.

function

Classes for expanding sets of features calculated with other featurizers.

matminer.featurizers.function

Name

Description

FunctionFeaturizer

Features from functions applied to existing features, e.g. “1/x”

site

Features from individual sites in a material’s crystal structure.

bonding

matminer.featurizers.site.bonding

Site featurizers based on bonding.

Name

Description

BondOrientationalParameter

Averages of spherical harmonics of local neighbors

AverageBondLength

Determines the average bond length between one specific site

AverageBondAngle

Determines the average bond angles of a specific site with

chemical

matminer.featurizers.site.chemical

Site featurizers based on local chemical information, rather than geometry alone.

Name

Description

ChemicalSRO

Chemical short range ordering, deviation of local site and nominal structure compositions

EwaldSiteEnergy

Compute site energy from Coulombic interactions

LocalPropertyDifference

Differences in elemental properties between site and its neighboring sites.

SiteElementalProperty

Elemental properties of atom on a certain site

external

matminer.featurizers.site.external

Site featurizers requiring external libraries for core functionality.

Name

Description

SOAP

Smooth overlap of atomic positions (interface via DScribe).

fingerprint

matminer.featurizers.site.fingerprint

Site featurizers that fingerprint a site using local geometry.

Name

Description

AGNIFingerprints

Product integral of RDF and Gaussian window function, from Botu et al.

OPSiteFingerprint

Local structure order parameters computed from a site’s neighbor env.

CrystalNNFingerprint

A local order parameter fingerprint for periodic crystals.

VoronoiFingerprint

Voronoi tessellation-based features around target site.

ChemEnvSiteFingerprint

Resemblance of given sites to ideal environments

misc

matminer.featurizers.site.misc

Miscellaneous site featurizers.

Name

Description

IntersticeDistribution

Interstice distribution in the neighboring cluster around an atom site.

CoordinationNumber

Number of first nearest neighbors of a site.

rdf

matminer.featurizers.site.rdf

Site featurizers based on distribution functions.

Name

Description

GaussianSymmFunc

Gaussian symmetry function features suggested by Behler et al.

GeneralizedRadialDistributionFunction

Compute the general radial distribution function (GRDF) for a site.

AngularFourierSeries

Compute the angular Fourier series (AFS), including both angular and radial info

structure

Generating features based on a material’s crystal structure.

bonding

matminer.featurizers.structure.bonding

Structure featurizers based on bonding.

Name

Description

BondFractions

Compute the fraction of each bond in a structure, based on NearestNeighbors.

BagofBonds

Compute a Bag of Bonds vector, as first described by Hansen et al. (2015).

GlobalInstabilityIndex

The global instability index of a structure.

StructuralHeterogeneity

Variance in the bond lengths and atomic volumes in a structure

MinimumRelativeDistances

Determines the relative distance of each site to its closest neighbor.

composite

matminer.featurizers.structure.composite

Structure featurizers producing more than one kind of structure feature data.

Name

Description

JarvisCFID

Classical Force-Field Inspired Descriptors (CFID) from Jarvis-ML.

matrix

matminer.featurizers.structure.matrix

Structure featurizers generating a matrix for each structure. Most matrix structure featurizers contain the ability to flatten matrices to be dataframe-friendly.

Name

Description

CoulombMatrix

The Coulomb matrix, a representation of nuclear coulombic interaction.

SineCoulombMatrix

A variant of the Coulomb matrix developed for periodic crystals.

OrbitalFieldMatrix

Representation based on the valence shell electrons of neighboring atoms.

misc

matminer.featurizers.structure.misc

Miscellaneous structure featurizers.

Name

Description

EwaldEnergy

Compute the energy from Coulombic interactions.

StructureComposition

Features related to the composition of a structure

XRDPowderPattern

1D array representing powder diffraction of a structure as calculated by

order

matminer.featurizers.structure.order

Structure featurizers based on packing or ordering.

Name

Description

DensityFeatures

Calculates density and density-like features

ChemicalOrdering

How much the ordering of species in the structure differs from random

MaximumPackingEfficiency

Maximum possible packing efficiency of this structure

StructuralComplexity

Shannon information entropy of a structure.

rdf

matminer.featurizers.structure.rdf

Structure featurizers implementing radial distribution functions.

Name

Description

RadialDistributionFunction

Calculate the radial distribution function (RDF) of a crystal structure.

PartialRadialDistributionFunction

Compute the partial radial distribution function (PRDF) of an xtal structure

ElectronicRadialDistributionFunction

Calculate the inherent electronic radial distribution function (ReDF)

sites

matminer.featurizers.structure.sites

Structure featurizers based on aggregating site features.

Name

Description

SiteStatsFingerprint

Computes statistics of properties across all sites in a structure.

symmetry

matminer.featurizers.structure.symmetry

Structure featurizers based on symmetry.

Name

Description

GlobalSymmetryFeatures

Determines symmetry features, e.g. spacegroup number and crystal system

Dimensionality

Returns dimensionality of structure: 1 means linear chains of atoms OR