/atomate

Permalink
Switch branches/tags
Nothing to show
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
Raw Blame History
620 lines (453 sloc) 22.2 KB
# coding: utf-8
from __future__ import division, print_function, unicode_literals, absolute_import
from uuid import uuid4
import numpy as np
from pymatgen.io.vasp.sets import MPRelaxSet, MPStaticSet, MPHSERelaxSet
from pymatgen.symmetry.analyzer import SpacegroupAnalyzer
from pymatgen.io.vasp.inputs import Kpoints
from atomate.vasp.config import SMALLGAP_KPOINT_MULTIPLY, STABILITY_CHECK, VASP_CMD, DB_FILE, \
ADD_WF_METADATA
from atomate.vasp.powerups import add_small_gap_multiply, add_stability_check, add_modify_incar, \
add_wf_metadata, add_common_powerups
from atomate.vasp.workflows.base.core import get_wf
from atomate.vasp.workflows.base.elastic import get_wf_elastic_constant
from atomate.vasp.workflows.base.raman import get_wf_raman_spectra
from atomate.vasp.workflows.base.gibbs import get_wf_gibbs_free_energy
from atomate.vasp.workflows.base.bulk_modulus import get_wf_bulk_modulus
from atomate.vasp.workflows.base.thermal_expansion import get_wf_thermal_expansion
from atomate.vasp.workflows.base.neb import get_wf_neb_from_endpoints, get_wf_neb_from_structure, \
get_wf_neb_from_images
__author__ = 'Anubhav Jain, Kiran Mathew'
__email__ = 'ajain@lbl.gov, kmathew@lbl.gov'
# TODO: @computron: Clarify the config dict -computron
# TODO: @computron: Allow default config dict to be loaded from file -computron
def wf_bandstructure(structure, c=None):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
wf = get_wf(structure, "bandstructure.yaml", vis=MPRelaxSet(structure, force_gamma=True),
common_params={"vasp_cmd": vasp_cmd, "db_file": db_file})
wf = add_common_powerups(wf, c)
if c.get("SMALLGAP_KPOINT_MULTIPLY", SMALLGAP_KPOINT_MULTIPLY):
wf = add_small_gap_multiply(wf, 0.5, 5, "static")
wf = add_small_gap_multiply(wf, 0.5, 5, "nscf")
if c.get("STABILITY_CHECK", STABILITY_CHECK):
wf = add_stability_check(wf, fw_name_constraint="structure optimization")
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_bandstructure_hse(structure, c=None):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
wf_src_name = "bandstructure_hse_full.yaml"
wf = get_wf(structure, wf_src_name, vis=MPHSERelaxSet(structure, force_gamma=True),
common_params={"vasp_cmd": vasp_cmd, "db_file": db_file})
wf = add_common_powerups(wf, c)
if c.get("SMALLGAP_KPOINT_MULTIPLY", SMALLGAP_KPOINT_MULTIPLY):
wf = add_small_gap_multiply(wf, 0.5, 5, "static")
if c.get("STABILITY_CHECK", STABILITY_CHECK):
wf = add_stability_check(wf, fw_name_constraint="structure optimization")
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_bandstructure_plus_hse(structure, gap_only=True, c=None):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
if gap_only:
wf_src_name = "bandstructure_hsegap.yaml"
else:
wf_src_name = "bandstructure_hse.yaml"
wf = get_wf(structure, wf_src_name, vis=MPRelaxSet(structure, force_gamma=True),
common_params={"vasp_cmd": vasp_cmd, "db_file": db_file})
wf = add_common_powerups(wf, c)
if c.get("SMALLGAP_KPOINT_MULTIPLY", SMALLGAP_KPOINT_MULTIPLY):
wf = add_small_gap_multiply(wf, 0.5, 5, "static")
wf = add_small_gap_multiply(wf, 0.5, 5, "nscf")
if c.get("STABILITY_CHECK", STABILITY_CHECK):
wf = add_stability_check(wf, fw_name_constraint="structure optimization")
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_bandstructure_plus_boltztrap(structure, c=None):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
params = []
for x in range(4): # everything but BoltzTrap task
params.append({"vasp_cmd": vasp_cmd, "db_file": db_file})
params.append({"db_file": db_file})
wf = get_wf(structure, "bandstructure_boltztrap.yaml", vis=MPRelaxSet(structure, force_gamma=True),
params=params)
wf = add_common_powerups(wf, c)
if c.get("SMALLGAP_KPOINT_MULTIPLY", SMALLGAP_KPOINT_MULTIPLY):
wf = add_small_gap_multiply(wf, 0.5, 5, "static")
wf = add_small_gap_multiply(wf, 0.5, 5, "nscf")
if c.get("STABILITY_CHECK", STABILITY_CHECK):
wf = add_stability_check(wf, fw_name_constraint="structure optimization")
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_static(structure, c=None):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
wf = get_wf(structure, "static_only.yaml", vis=MPStaticSet(structure),
common_params={"vasp_cmd": vasp_cmd, "db_file": db_file})
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_structure_optimization(structure, c=None):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
user_incar_settings = c.get("USER_INCAR_SETTINGS")
wf = get_wf(structure, "optimize_only.yaml",
vis=MPRelaxSet(structure, force_gamma=True, user_incar_settings=user_incar_settings),
common_params={"vasp_cmd": vasp_cmd, "db_file": db_file})
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_dielectric_constant(structure, c=None):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
wf = get_wf(structure, "dielectric_constant.yaml", vis=MPRelaxSet(structure, force_gamma=True),
common_params={"vasp_cmd": vasp_cmd, "db_file": db_file})
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_dielectric_constant_no_opt(structure, c=None):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
wf = get_wf(structure, "dielectric_constant_no_opt.yaml",
common_params={"vasp_cmd": vasp_cmd, "db_file": db_file})
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_piezoelectric_constant(structure, c=None):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
wf = get_wf(structure, "piezoelectric_constant.yaml",
common_params={"vasp_cmd": vasp_cmd, "db_file": db_file})
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_elastic_constant(structure, c=None, order=2, sym_reduce=False):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
uis_optimize = {"ENCUT": 700, "EDIFF": 1e-6, "LAECHG": False}
if order > 2:
uis_optimize.update({"EDIFF": 1e-10, "EDIFFG": -0.001,
"ADDGRID": True, "LREAL": False, "ISYM": 0})
# This ensures a consistent k-point mesh across all calculations
# We also turn off symmetry to prevent VASP from changing the
# mesh internally
kpts_settings = Kpoints.automatic_density(structure, 40000, force_gamma=True)
stencils = np.linspace(-0.075, 0.075, 7)
else:
kpts_settings = {'grid_density': 7000}
stencils = None
uis_static = uis_optimize.copy()
uis_static.update({'ISIF': 2, 'IBRION': 2, 'NSW': 99, 'ISTART': 1, "PREC": "High"})
# input set for structure optimization
vis_relax = MPRelaxSet(structure, force_gamma=True, user_incar_settings=uis_optimize,
user_kpoints_settings=kpts_settings)
# optimization only workflow
wf = get_wf(structure, "optimize_only.yaml", vis=vis_relax,
params=[{"vasp_cmd": vasp_cmd, "db_file": db_file,
"name": "elastic structure optimization"}])
vis_static = MPStaticSet(structure, force_gamma=True, lepsilon=False,
user_kpoints_settings=kpts_settings,
user_incar_settings=uis_static)
# deformations wflow for elasticity calculation
wf_elastic = get_wf_elastic_constant(structure, vasp_cmd=vasp_cmd, db_file=db_file,
order=order, stencils=stencils, copy_vasp_outputs=True,
vasp_input_set=vis_static, sym_reduce=sym_reduce)
wf.append_wf(wf_elastic, wf.leaf_fw_ids)
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_elastic_constant_minimal(structure, c=None, order=2, sym_reduce=True):
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
stencil = np.arange(0.01, 0.01 * order, step=0.01)
wf = get_wf_elastic_constant(structure, vasp_cmd=vasp_cmd, db_file=db_file,
sym_reduce=sym_reduce, stencils=stencil, order=order,
copy_vasp_outputs=False)
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_raman_spectra(structure, c=None):
"""
Raman spectra workflow from the given structure and config dict
Args:
structure (Structure): input structure
c (dict): workflow config dict
Returns:
Workflow
"""
c = c or {}
modes = c.get("MODES", None)
step_size = c.get("STEP_SIZE", 0.005)
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
wf = get_wf_raman_spectra(structure, modes=modes, step_size=step_size, vasp_cmd=vasp_cmd,
db_file=db_file)
wf = add_modify_incar(wf, modify_incar_params={"incar_update": {"ENCUT": 600, "EDIFF": 1e-6}},
fw_name_constraint="static dielectric")
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_gibbs_free_energy(structure, c=None):
"""
Gibbs free energy workflow from the given structure and config dict.
Args:
structure (Structure): input structure
c (dict): workflow config dict
Returns:
Workflow
"""
c = c or {}
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
eos = c.get("EOS", "vinet")
qha_type = c.get("QHA_TYPE", "debye_model")
# min and max temp in K, default setting is to compute the properties at 300K only
t_min = c.get("T_MIN", 300.0)
t_max = c.get("T_MAX", 300.0)
t_step = c.get("T_STEP", 100.0)
pressure = c.get("PRESSURE", 0.0)
poisson = c.get("POISSON", 0.25)
anharmonic_contribution = c.get("ANHARMONIC_CONTRIBUTION", False)
metadata = c.get("METADATA", None)
# 21 deformed structures: from -10% to +10%
defos = [(np.identity(3) * (1 + x)).tolist() for x in np.linspace(-0.1, 0.1, 21)]
deformations = c.get("DEFORMATIONS", defos)
user_kpoints_settings = {"grid_density": 7000}
tag = "gibbs group: >>{}<<".format(str(uuid4()))
# input set for structure optimization
vis_relax = MPRelaxSet(structure, force_gamma=True)
v = vis_relax.as_dict()
v.update({"user_kpoints_settings": user_kpoints_settings})
vis_relax = vis_relax.__class__.from_dict(v)
# optimization only workflow
wf = get_wf(structure, "optimize_only.yaml",
params=[{"vasp_cmd": vasp_cmd, "db_file": db_file,
"name": "{} structure optimization".format(tag)}],
vis=vis_relax)
# static input set for the transmute firework
uis_static = {
"ISIF": 2,
"ISTART": 1,
}
lepsilon = False
if qha_type not in ["debye_model"]:
lepsilon = True
try:
from phonopy import Phonopy
except ImportError:
raise RuntimeError("'phonopy' package is NOT installed but is required for the final "
"analysis step; you can alternatively switch to the qha_type to "
"'debye_model' which does not require 'phonopy'.")
vis_static = MPStaticSet(structure, force_gamma=True, lepsilon=lepsilon,
user_kpoints_settings=user_kpoints_settings,
user_incar_settings=uis_static)
# get gibbs workflow and chain it to the optimization workflow
wf_gibbs = get_wf_gibbs_free_energy(structure, user_kpoints_settings=user_kpoints_settings,
deformations=deformations, vasp_cmd=vasp_cmd, db_file=db_file,
eos=eos, qha_type=qha_type, pressure=pressure, poisson=poisson,
t_min=t_min, t_max=t_max, t_step=t_step, metadata=metadata,
anharmonic_contribution=anharmonic_contribution,
tag=tag, vasp_input_set=vis_static)
# chaining
wf.append_wf(wf_gibbs, wf.leaf_fw_ids)
wf = add_modify_incar(wf, modify_incar_params={"incar_update": {"ENCUT": 600,
"EDIFF": 1e-6,
"LAECHG": False}})
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_bulk_modulus(structure, c=None):
"""
Bulk modulus workflow from the given structure and config dict.
Args:
structure (Structure): input structure
c (dict): workflow config dict
Returns:
Workflow
"""
c = c or {}
eos = c.get("EOS", "vinet")
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
user_kpoints_settings = {"grid_density": 7000}
# 6 deformations
deformations = [(np.identity(3) * (1 + x)).tolist() for x in np.linspace(-0.05, 0.05, 6)]
tag = "bulk_modulus group: >>{}<<".format(str(uuid4()))
# input set for structure optimization
vis_relax = MPRelaxSet(structure, force_gamma=True)
v = vis_relax.as_dict()
v.update({"user_kpoints_settings": user_kpoints_settings})
vis_relax = vis_relax.__class__.from_dict(v)
# static input set for the transmute firework
uis_static = {
"ISIF": 2,
"ISTART": 1,
"IBRION": 2,
"NSW": 99
}
# optimization only workflow
wf = get_wf(structure, "optimize_only.yaml",
params=[{"vasp_cmd": vasp_cmd, "db_file": db_file,
"name": "{} structure optimization".format(tag)}],
vis=vis_relax)
vis_static = MPStaticSet(structure, force_gamma=True, lepsilon=False,
user_kpoints_settings=user_kpoints_settings,
user_incar_settings=uis_static)
# get the deformations wflow for bulk modulus calculation
wf_bm = get_wf_bulk_modulus(structure, eos=eos, user_kpoints_settings=user_kpoints_settings,
deformations=deformations, vasp_cmd=vasp_cmd, db_file=db_file, tag=tag,
vasp_input_set=vis_static)
# chain it
wf.append_wf(wf_bm, wf.leaf_fw_ids)
wf = add_modify_incar(wf, modify_incar_params={"incar_update": {"ENCUT": 600, "EDIFF": 1e-6}})
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_thermal_expansion(structure, c=None):
"""
Thermal expansion coefficient workflow from the given structure and config dict.
Args:
structure (Structure): input structure
c (dict): workflow config dict
Returns:
Workflow
"""
c = c or {}
eos = c.get("EOS", "vinet")
vasp_cmd = c.get("VASP_CMD", VASP_CMD)
db_file = c.get("DB_FILE", DB_FILE)
pressure = c.get("PRESSURE", 0.0)
user_kpoints_settings = {"grid_density": 7000}
# 10 deformations
deformations = [(np.identity(3) * (1 + x)).tolist() for x in np.linspace(-0.1, 0.1, 10)]
tag = "thermal_expansion group: >>{}<<".format(str(uuid4()))
# input set for structure optimization
vis_relax = MPRelaxSet(structure, force_gamma=True)
v = vis_relax.as_dict()
v.update({"user_kpoints_settings": user_kpoints_settings})
vis_relax = vis_relax.__class__.from_dict(v)
# optimization only workflow
wf = get_wf(structure, "optimize_only.yaml",
params=[{"vasp_cmd": vasp_cmd, "db_file": db_file,
"name": "{} structure optimization".format(tag)}],
vis=vis_relax)
wf_thermal = get_wf_thermal_expansion(structure, user_kpoints_settings=user_kpoints_settings,
deformations=deformations, vasp_cmd=vasp_cmd, db_file=db_file,
eos=eos, pressure=pressure, tag=tag)
# chain it
wf.append_wf(wf_thermal, wf.leaf_fw_ids)
wf = add_modify_incar(wf, modify_incar_params={"incar_update": {"ENCUT": 600, "EDIFF": 1e-6}})
wf = add_common_powerups(wf, c)
if c.get("ADD_WF_METADATA", ADD_WF_METADATA):
wf = add_wf_metadata(wf, structure)
return wf
def wf_nudged_elastic_band(structures, parent, c=None):
"""
Nudged elastic band (NEB) workflow from the given structures and config dict.
'is_optimized' default False
'neb_round' default 1
Notes:
Length of Structure list and "is_optimized" are used to determine the workflow:
1 structure # The parent structure & two endpoint indexes provided; need relaxation.
# The parent structure & two endpoint indexes provided; no need to relax.
2 structures # Two endpoints provided; need to relax two endpoints.
# Two relaxed endpoints provided; no need to relax two endpoints.
>=3 structures # All images including two endpoints are provided.
Args:
structures ([Structure]):
1) The parent structure
2) Two endpoint structures
3) An initial NEB path that comprises both images and endpoints
parent (Structure): parent structure used to get two endpoints.
c (dict): workflow config dict, basic format:
{"fireworks": [], "common_params": {}, "additional_ep_params": {},
"additional_neb_params": {}}. When the length of structures is 1, "site_indices" key
must be included in c. Note that "fireworks" is a list corresponding to the order of
execution.
Returns:
Workflow
"""
if not(isinstance(structures, list) and len(structures) > 0):
raise ValueError("structures must be a list of Structure objects!")
# config initialization
c = c or {}
# TODO: @shyuep - config dict params are typically capitalized and intended to be global
# params unless namespaced. e.g. NEB.COMMON_PARAMS if it's only settings for NEB
# workflows. -computron
spec = c.get("common_params", {})
is_optimized = spec.get("is_optimized", False)
# TODO: @shyuep: the whole point of preset workflows is that they are supposed to be simple.
# e.g., give a structure or a list of structures, and forget about the rest. Here it looks like
# one needs to construct some kind of complicated configuration dictionary. Pretty sure no one
# apart from the people in your group have any idea how to use these functions or set up this
# config dict (which should not be necessary in the first place). -computron
if c.get("fireworks"): # Check config dict file
fw_list = [f['fw'] for f in c.get("fireworks")]
neb_round = len([f for f in fw_list if "NEBFW" in f])
if neb_round < 1:
raise ValueError("At least one NEB Fireworks (NEBFW) is needed in the config dict!")
if len(structures) == 1:
assert "site_indices" in spec, "Site indices not provided in config dict!"
assert len(spec["site_indices"]) == 2, "Only two site indices should be provided!"
if is_optimized:
assert len(fw_list) == neb_round + 1
else:
assert len(fw_list) == neb_round + 2
elif len(structures) == 2:
if is_optimized:
assert len(fw_list) == neb_round
else:
assert len(fw_list) == neb_round + 1
else: # Default settings if config dict is not provided.
neb_round = 1
# Get user_incar_settings, user_kpoints_settings & additional_cust_args
user_incar_settings = [{}] * (neb_round + 2)
user_kpoints_settings = [{"grid_density": 1000}] * (neb_round + 2)
additional_cust_args = [{}] * (neb_round + 2)
# TODO: @shyuep - I have no idea what's going on here -computron
if "fireworks" in c:
for i in range(1, len(c["fireworks"]) + 1):
user_incar_settings[-i] = c["fireworks"][-i].get("user_incar_settings", {})
user_kpoints_settings[-i] = c["fireworks"][-i].get("user_kpoints_settings",
{"grid_density": 1000})
additional_cust_args[-i] = c["fireworks"][-i].get("additional_cust_args", {})
kwargs = {"user_incar_settings": user_incar_settings,
"user_kpoints_settings": user_kpoints_settings,
"additional_cust_args": additional_cust_args}
# Assign workflow using the number of given structures
if len(structures) == 1:
wf = get_wf_neb_from_structure(structure=structures[0],
additional_spec=spec, **kwargs)
elif len(structures) == 2:
wf = get_wf_neb_from_endpoints(parent=parent, endpoints=structures,
additional_spec=spec, **kwargs)
else: # len(structures) >= 3
wf = get_wf_neb_from_images(parent=parent, images=structures,
additional_spec=spec, **kwargs)
return wf