Skip to content

Analyzing & Modifying Structures

Tip

Simmate toolkit still uses pymatgen. Therefore, this tutorial also serves as a guide to using their package. See also:

Quick Start

  1. Ensure you have the POSCAR file of NaCl from the previous tutorial.

  2. You can load the structure into python: 

    from simmate.toolkit import Structure

structure = Structure.from_file("POSCAR")

  3. Access various properties of the structure, lattice, and composition: 

    # explore structure-based properties
structure.density
structure.distance_matrix
structure.cart_coords
structure.num_sites

# access the structure's composition and its properties
composition = structure.composition
composition.reduced_formula
composition.elements

# access the structure's lattice and its properties
lattice = structure.lattice
lattice.volume
lattice.matrix
lattice.beta

  4. Create new structures using some transformation or analysis: 

    structure.add_oxidation_state_by_guess()
structure.make_supercell([2,2,2])

  5. Export your final structure to a new file format: 

    structure.to(filename="NaCl.cif", fmt="cif")

Extra Examples

Looking for advanced features? Simmate is gradually incorporating these into our toolkit module, but many more are available through PyMatGen and MatMiner (which are preinstalled for you).

Random Structure Creation

Creating a random structure from a spacegroup and composition:

from simmate.toolkit import Composition
from simmate.toolkit.creators import RandomSymStructure

composition = Composition("Ca2N")
creator = RandomSymStructure(composition)

structure = creator.create_structure(spacegroup=166)

Fingerprints (MatMiner)

Matminer is useful for analyzing structures and creating machine-learning inputs. One common analysis is the generating a RDF fingerprint to help analyze bonding and compare structures:

from matminer.featurizers.structure.rdf import RadialDistributionFunction

rdf_analyzer = RadialDistributionFunction(bin_size=0.1)
rdf = rdf_analyzer.featurize(structure)

Structure Matching (PyMatGen)

Pymatgen currently offers the most functionality. One common function is checking if two structures are symmetrically equivalent (under some tolerance):

from pymatgen.analysis.structure_matcher import StructureMatcher

matcher = StructureMatcher()
is_matching = matcher.fit(structure1, structure2)