Skip to content

quick_guide.rst

Latest commit

 

History

History
99 lines (72 loc) · 3.19 KB

quick_guide.rst

File metadata and controls

99 lines (72 loc) · 3.19 KB

Quick guide

The quick guide is designed to give a brief overview of the steps required to run a RAFFLE calculation. It is assumed that the user has already installed RAFFLE and has a basic understanding of the command line. For a more detailed explanation of the parameters and options available, please see the :doc:`Main tutorial page </tutorials/index>` for individual tutorials detailing each step.

RAFFLE is a random structure search package designed primarily for interfaces. The tutorials will use bulk systems to demonstrate its functionality, but it is recommended to use packages such as AIRSS for bulk random structure search.

Single iteration

Here is a script to run a single iteration of the RAFFLE generator for structure search.

# Single iteration of RAFFLE structure search
from ase.io import read, write
from raffle.generator import raffle_generator

generator = raffle_generator()

host = read("host.xyz")
generator.set_host(host)
generator.set_grid(grid_spacing=0.1)
generator.set_bounds([[0, 0, 0.5], [1, 1, 0.75]])
generator.distributions.set_element_energies(
    { 'C': -9.063733 }
)

database = read("database.xyz", index=":")
generator.distributions.create(database)

structures, status = generator.generate(
    num_structures = 1,
    stoichiometry = { 'C': 2 },
)

write("output.xyz", structures)

This script will generate a single structure with a stoichiometry of two carbon atoms. The host structure is read from a file, and the grid spacing and bounds are set. The element energies are set to the energy of carbon, and the database is read from a file. The generator is then run, and the structures are written to an output file.

Iterative structure search

Here is an example of how to run an iterative structure search with RAFFLE and check for convergence.

# Iterative RAFFLE structure search
from ase.io import read, write
from ase.optimize import FIRE
from raffle.generator import raffle_generator
from mace.calculators import mace_mp

generator = raffle_generator()
generator.set_history_len(10)
mace = mace_mp(model="medium", dispersion=False, default_dtype="float32", device='cpu')

host = read("host.xyz")
generator.set_host(host)
generator.set_grid(grid_spacing=0.1)
generator.set_bounds([[0, 0, 0.5], [1, 1, 0.75]])
generator.distributions.set_element_energies(
    { 'C': -9.063733 }
)

database = read("database.xyz", index=":")
generator.distributions.create(database)

num_structures_old = 0
for i in range(10):
    generator.generate(
        num_structures = 2,
        stoichiometry = { 'C': 2 }
        calc = calc
    )
    structures, status = generator.get_structures()
    for structure in structures:
        optimiser = FIRE(structure)
        optimiser.run(fmax=0.05)

    generator.update(structures)
    num_structures_old = len(structures)
    if generator.is_converged():
        break

write("output.xyz", structures)