def test_all_methods(self):
# create scratch directory with files....
# having to do it all at once to minimize amount of time copying over to Scratch Directory
with ScratchDir("."):
# setup with fake Locpot object copied over
copyfile( os.path.join(file_loc, "test_path_files.tar.gz"), "./test_path_files.tar.gz")
tar = tarfile.open("test_path_files.tar.gz")
tar.extractall()
tar.close()
blocpot = Locpot.from_file( os.path.join( file_loc, "bLOCPOT.gz"))
blocpot.write_file("test_path_files/bulk/LOCPOT")
dlocpot = Locpot.from_file( os.path.join( file_loc, "dLOCPOT.gz"))
dlocpot.write_file("test_path_files/sub_1_Sb_on_Ga/charge_2/LOCPOT")
# test_from_path
sdp = SingleDefectParser.from_paths("test_path_files/sub_1_Sb_on_Ga/charge_2/",
"test_path_files/bulk/", 18.12, 2)
self.assertIsInstance(sdp, SingleDefectParser)
self.assertIsInstance(sdp.defect_entry.defect, Substitution)
self.assertEqual(sdp.defect_entry.parameters["bulk_path"],
"test_path_files/bulk/")
self.assertEqual(sdp.defect_entry.parameters["defect_path"],
"test_path_files/sub_1_Sb_on_Ga/charge_2/")
# test_freysoldt_loader
for param_key in ["axis_grid", "bulk_planar_averages", "defect_planar_averages", \
"initial_defect_structure", "defect_frac_sc_coords"]:
self.assertFalse(param_key in sdp.defect_entry.parameters.keys())
bl = sdp.freysoldt_loader()
self.assertIsInstance(bl, Locpot)
for param_key in ["axis_grid", "bulk_planar_averages", "defect_planar_averages", \
"initial_defect_structure", "defect_frac_sc_coords"]:
self.assertTrue(param_key in sdp.defect_entry.parameters.keys())
# test_kumagai_loader
for param_key in ["bulk_atomic_site_averages", "defect_atomic_site_averages", \
"site_matching_indices", "sampling_radius", "defect_index_sc_coords"]:
self.assertFalse(param_key in sdp.defect_entry.parameters.keys())
bo = sdp.kumagai_loader()
for param_key in ["bulk_atomic_site_averages", "defect_atomic_site_averages", \
"site_matching_indices", "sampling_radius", "defect_index_sc_coords"]:
self.assertTrue(param_key in sdp.defect_entry.parameters.keys())
# test_get_stdrd_metadata
sdp.get_stdrd_metadata()
for param_key in ["eigenvalues", "kpoint_weights", "bulk_energy", \
"final_defect_structure", "defect_energy", "run_metadata"]:
self.assertTrue(param_key in sdp.defect_entry.parameters.keys())
# test_get_bulk_gap_data
sdp.get_bulk_gap_data()
self.assertEqual(sdp.defect_entry.parameters["mpid"], "mp-2534")
self.assertAlmostEqual(sdp.defect_entry.parameters["gap"], 0.1887)
# test_run_compatibility
self.assertFalse('is_compatible' in sdp.defect_entry.parameters)
sdp.run_compatibility()
self.assertTrue('is_compatible' in sdp.defect_entry.parameters)
def chg_locp_average():
chg=False
print("which file would like to average: CHG or LOCPOT ?")
wait_sep()
in_str=""
while in_str=="":
in_str=input().strip()
if in_str.lower()=='chg':
filename='CHG'
check_file(filename)
grid_data = CHGCAR.from_file(filename)
head_line="#%(key1)+s %(key2)+s"%{'key1':'Distance/Ang','key2':'Average Charge/(e/Ang**3)'}
chg=True
elif in_str.lower()=='locpot':
filename='LOCPOT'
check_file(filename)
grid_data = Locpot.from_file(filename)
head_line="#%(key1)+s %(key2)+s"%{'key1':'Distance/Ang','key2':'Average Potential/(eV)'}
else:
print('unknown file file: '+in_str)
os._exit()
step_count=1
check_file(filename)
proc_str="Reading Data From "+ filename +" File ..."
procs(proc_str,step_count,sp='-->>')
volume=grid_data.structure.volume
print("which direction would like to average: x y or z ?")
wait_sep()
in_str=""
while in_str=="":
in_str=input().strip().lower()
if in_str=="x":
selected_dir=0
elif in_str=='y':
selected_dir=1
elif in_str=='z':
selected_dir=2
else:
print("Unknow Direction!")
return
axis_grid=grid_data.get_axis_grid(selected_dir)
if chg:
aver_grid=grid_data.get_average_along_axis(selected_dir)/volume
else:
aver_grid=grid_data.get_average_along_axis(selected_dir)
data=np.vstack((axis_grid,aver_grid)).T
step_count+=1
if chg:
filename="average_CHG_"+in_str+'.dat'
else:
filename="average_LOCPOT_"+in_str+'.dat'
proc_str="Writting Average Data to "+ filename +" File ..."
procs(proc_str,step_count,sp='-->>')
write_col_data(filename,data,head_line)
def test_freysoldt_and_kumagai(self):
with ScratchDir("."):
# setup with fake Locpot object copied over
copyfile(os.path.join(TEST_DIR, "test_path_files.tar.gz"),
"test_path_files.tar.gz")
tar = tarfile.open("test_path_files.tar.gz")
tar.extractall()
tar.close()
blocpot = Locpot.from_file(os.path.join(TEST_DIR, "bLOCPOT.gz"))
blocpot.write_file(
os.path.join("test_path_files", "bulk", "LOCPOT"))
dlocpot = Locpot.from_file(os.path.join(TEST_DIR, "dLOCPOT.gz"))
dlocpot.write_file(
os.path.join("test_path_files", "sub_1_Sb_on_Ga", "charge_2",
"LOCPOT"))
fcc = freysoldt_correction_from_paths(
os.path.join("test_path_files", "sub_1_Sb_on_Ga", "charge_2"),
os.path.join("test_path_files", "bulk"),
18.12,
2,
plot=True)
self.assertAlmostEqual(fcc, -1.2435280589593547)
self.assertTrue(
os.path.exists(
os.path.join(
"test_path_files", "sub_1_Sb_on_Ga", "charge_2",
"Sub_Sb_on_Ga_mult32_chg_2_axis1_freysoldtplot.pdf")))
kcc = kumagai_correction_from_paths(
os.path.join("test_path_files", "sub_1_Sb_on_Ga", "charge_2"),
os.path.join("test_path_files", "bulk"),
18.12,
2,
plot=True)
self.assertAlmostEqual(kcc, 0.6387768530616106)
self.assertTrue(
os.path.exists(
os.path.join("test_path_files", "sub_1_Sb_on_Ga",
"charge_2",
"Sub_Sb_on_Ga_mult32_chg_2_kumagaiplot.pdf")))
def test_freysoldt_and_kumagai(self):
# create scratch directory with files....
# having to do it all at once to minimize amount of time copying over to Scratch Directory
with ScratchDir("."):
# setup with fake Locpot object copied over
copyfile(os.path.join(file_loc, "test_path_files.tar.gz"),
"./test_path_files.tar.gz")
tar = tarfile.open("test_path_files.tar.gz")
tar.extractall()
tar.close()
blocpot = Locpot.from_file(os.path.join(file_loc, "bLOCPOT.gz"))
blocpot.write_file("test_path_files/bulk/LOCPOT")
dlocpot = Locpot.from_file(os.path.join(file_loc, "dLOCPOT.gz"))
dlocpot.write_file(
"test_path_files/sub_1_Sb_on_Ga/charge_2/LOCPOT")
fcc = freysoldt_correction_from_paths(
"test_path_files/sub_1_Sb_on_Ga/charge_2/",
"test_path_files/bulk/",
18.12,
2,
plot=True)
self.assertEqual(fcc, -1.2435280589593547)
self.assertTrue(
os.path.exists(
"test_path_files/sub_1_Sb_on_Ga/charge_2/Sub_Sb_on_Ga_mult32_chg_2_axis1_freysoldtplot.pdf"
))
kcc = kumagai_correction_from_paths(
"test_path_files/sub_1_Sb_on_Ga/charge_2/",
"test_path_files/bulk/",
18.12,
2,
plot=True)
self.assertEqual(kcc, 0.6387768530616106)
self.assertTrue(
os.path.exists(
"test_path_files/sub_1_Sb_on_Ga/charge_2/Sub_Sb_on_Ga_mult32_chg_2_kumagaiplot.pdf"
))
def process_vasprun(self, dir_name, taskname, filename):
"""
Adapted from matgendb.creator
Process a vasprun.xml file.
"""
vasprun_file = os.path.join(dir_name, filename)
vrun = Vasprun(vasprun_file)
d = vrun.as_dict()
# rename formula keys
for k, v in {"formula_pretty": "pretty_formula",
"composition_reduced": "reduced_cell_formula",
"composition_unit_cell": "unit_cell_formula"}.items():
d[k] = d.pop(v)
for k in ["eigenvalues", "projected_eigenvalues"]: # large storage space breaks some docs
if k in d["output"]:
del d["output"][k]
comp = Composition(d["composition_unit_cell"])
d["formula_anonymous"] = comp.anonymized_formula
d["formula_reduced_abc"] = comp.reduced_composition.alphabetical_formula
d["dir_name"] = os.path.abspath(dir_name)
d["completed_at"] = str(datetime.datetime.fromtimestamp(os.path.getmtime(vasprun_file)))
d["density"] = vrun.final_structure.density
# replace 'crystal' with 'structure'
d["input"]["structure"] = d["input"].pop("crystal")
d["output"]["structure"] = d["output"].pop("crystal")
for k, v in {"energy": "final_energy", "energy_per_atom": "final_energy_per_atom"}.items():
d["output"][k] = d["output"].pop(v)
if self.parse_dos == True or (str(self.parse_dos).lower() == "auto" and vrun.incar.get("NSW", 1) == 0):
try:
d["dos"] = vrun.complete_dos.as_dict()
except:
raise ValueError("No valid dos data exist in {}.".format(dir_name))
# Band structure parsing logic
if str(self.bandstructure_mode).lower() == "auto":
# if line mode nscf
if vrun.incar.get("ICHARG", 0) > 10 and vrun.kpoints.num_kpts > 0:
bs_vrun = BSVasprun(vasprun_file, parse_projected_eigen=True)
bs = bs_vrun.get_band_structure(line_mode=True)
# else if nscf
elif vrun.incar.get("ICHARG", 0) > 10:
bs_vrun = BSVasprun(vasprun_file, parse_projected_eigen=True)
bs = bs_vrun.get_band_structure()
# else just regular calculation
else:
bs = vrun.get_band_structure()
# only save the bandstructure if not moving ions
if vrun.incar["NSW"] == 0:
d["bandstructure"] = bs.as_dict()
# legacy line/True behavior for bandstructure_mode
elif self.bandstructure_mode:
bs_vrun = BSVasprun(vasprun_file, parse_projected_eigen=True)
bs = bs_vrun.get_band_structure(line_mode=(str(self.bandstructure_mode).lower() == "line"))
d["bandstructure"] = bs.as_dict()
# parse bandstructure for vbm/cbm/bandgap but don't save
else:
bs = vrun.get_band_structure()
# Parse electronic information if possible.
# For certain optimizers this is broken and we don't get an efermi resulting in the bandstructure
try:
bs_gap = bs.get_band_gap()
d["output"]["vbm"] = bs.get_vbm()["energy"]
d["output"]["cbm"] = bs.get_cbm()["energy"]
d["output"]["bandgap"] = bs_gap["energy"]
d["output"]["is_gap_direct"] = bs_gap["direct"]
d["output"]["is_metal"] = bs.is_metal()
if not bs_gap["direct"]:
d["output"]["direct_gap"] = bs.get_direct_band_gap()
if isinstance(bs, BandStructureSymmLine):
d["output"]["transition"] = bs_gap["transition"]
except Exception:
if self.bandstructure_mode is True:
import traceback
logger.error(traceback.format_exc())
logger.error("Error in " + os.path.abspath(dir_name) + ".\n" + traceback.format_exc())
raise
logger.warning("Error in parsing bandstructure")
if vrun.incar["IBRION"] == 1:
logger.warning("Vasp doesn't properly output efermi for IBRION == 1")
d["task"] = {"type": taskname, "name": taskname}
d["output_file_paths"] = self.process_raw_data(dir_name, taskname=taskname)
if "locpot" in d["output_file_paths"] and self.parse_locpot:
locpot = Locpot.from_file(os.path.join(dir_name, d["output_file_paths"]["locpot"]))
d["output"]["locpot"] = {i: locpot.get_average_along_axis(i) for i in range(3)}
if hasattr(vrun, "force_constants"):
# phonon-dfpt
d["output"]["force_constants"] = vrun.force_constants.tolist()
d["output"]["normalmode_eigenvals"] = vrun.normalmode_eigenvals.tolist()
d["output"]["normalmode_eigenvecs"] = vrun.normalmode_eigenvecs.tolist()
return d
def process_vasprun(self, dir_name, taskname, filename):
"""
Adapted from matgendb.creator
Process a vasprun.xml file.
"""
vasprun_file = os.path.join(dir_name, filename)
vrun = Vasprun(vasprun_file, parse_potcar_file=self.parse_potcar_file)
d = vrun.as_dict()
# rename formula keys
for k, v in {
"formula_pretty": "pretty_formula",
"composition_reduced": "reduced_cell_formula",
"composition_unit_cell": "unit_cell_formula"
}.items():
d[k] = d.pop(v)
for k in ["eigenvalues", "projected_eigenvalues"
]: # large storage space breaks some docs
if k in d["output"]:
del d["output"][k]
comp = Composition(d["composition_unit_cell"])
d["formula_anonymous"] = comp.anonymized_formula
d["formula_reduced_abc"] = comp.reduced_composition.alphabetical_formula
d["dir_name"] = os.path.abspath(dir_name)
d["completed_at"] = str(
datetime.datetime.fromtimestamp(os.path.getmtime(vasprun_file)))
d["density"] = vrun.final_structure.density
# replace 'crystal' with 'structure'
d["input"]["structure"] = d["input"].pop("crystal")
d["output"]["structure"] = d["output"].pop("crystal")
for k, v in {
"energy": "final_energy",
"energy_per_atom": "final_energy_per_atom"
}.items():
d["output"][k] = d["output"].pop(v)
# Process bandstructure and DOS
if self.bandstructure_mode != False:
bs = self.process_bandstructure(vrun)
if bs:
d["bandstructure"] = bs
if self.parse_dos != False:
dos = self.process_dos(vrun)
if dos:
d["dos"] = dos
# Parse electronic information if possible.
# For certain optimizers this is broken and we don't get an efermi resulting in the bandstructure
try:
bs = vrun.get_band_structure()
bs_gap = bs.get_band_gap()
d["output"]["vbm"] = bs.get_vbm()["energy"]
d["output"]["cbm"] = bs.get_cbm()["energy"]
d["output"]["bandgap"] = bs_gap["energy"]
d["output"]["is_gap_direct"] = bs_gap["direct"]
d["output"]["is_metal"] = bs.is_metal()
if not bs_gap["direct"]:
d["output"]["direct_gap"] = bs.get_direct_band_gap()
if isinstance(bs, BandStructureSymmLine):
d["output"]["transition"] = bs_gap["transition"]
except Exception:
logger.warning("Error in parsing bandstructure")
if vrun.incar["IBRION"] == 1:
logger.warning(
"Vasp doesn't properly output efermi for IBRION == 1")
if self.bandstructure_mode is True:
logger.error(traceback.format_exc())
logger.error("Error in " + os.path.abspath(dir_name) + ".\n" +
traceback.format_exc())
raise
# Should roughly agree with information from .get_band_structure() above, subject to tolerances
# If there is disagreement, it may be related to VASP incorrectly assigning the Fermi level
try:
band_props = vrun.eigenvalue_band_properties
d["output"]["eigenvalue_band_properties"] = {
"bandgap": band_props[0],
"cbm": band_props[1],
"vbm": band_props[2],
"is_gap_direct": band_props[3]
}
except Exception:
logger.warning("Error in parsing eigenvalue band properties")
# store run name and location ,e.g. relax1, relax2, etc.
d["task"] = {"type": taskname, "name": taskname}
# include output file names
d["output_file_paths"] = self.process_raw_data(dir_name,
taskname=taskname)
# parse axially averaged locpot
if "locpot" in d["output_file_paths"] and self.parse_locpot:
locpot = Locpot.from_file(
os.path.join(dir_name, d["output_file_paths"]["locpot"]))
d["output"]["locpot"] = {
i: locpot.get_average_along_axis(i)
for i in range(3)
}
if self.store_volumetric_data:
for file in self.store_volumetric_data:
if file in d["output_file_paths"]:
try:
# assume volumetric data is all in CHGCAR format
data = Chgcar.from_file(
os.path.join(dir_name,
d["output_file_paths"][file]))
d[file] = data.as_dict()
except:
raise ValueError("Failed to parse {} at {}.".format(
file, d["output_file_paths"][file]))
# parse force constants
if hasattr(vrun, "force_constants"):
d["output"]["force_constants"] = vrun.force_constants.tolist()
d["output"][
"normalmode_eigenvals"] = vrun.normalmode_eigenvals.tolist()
d["output"][
"normalmode_eigenvecs"] = vrun.normalmode_eigenvecs.tolist()
# perform Bader analysis using Henkelman bader
if self.parse_bader and "chgcar" in d["output_file_paths"]:
suffix = '' if taskname == 'standard' else ".{}".format(taskname)
bader = bader_analysis_from_path(dir_name, suffix=suffix)
d["bader"] = bader
return d
#%%
from matplotlib import cm
import numpy as np
from pymatgen.core.sites import Site
from pymatgen.io.vasp import Locpot
import matplotlib.pyplot as plt
import os
os.chdir('/home/jinho93/new/oxides/perobskite/lanthanum-aluminate/periodic_step/vasp/my015/4/lvhar')
loc = Locpot.from_file('LOCPOT')
os.chdir('/home/jinho93/new/oxides/perobskite/lanthanum-aluminate/periodic_step/vasp/my015/from-gulp/421/locpot')
# os.chdir('/home/jinho93/new/oxides/perobskite/lanthanum-aluminate/periodic_step/vasp/my015/from-gulp/421/opti/loc')
loc2 = Locpot.from_file('LOCPOT')
os.chdir('/home/jinho93/new/oxides/perobskite/lanthanum-aluminate/periodic_step/vasp/stengel/015/4/lvhar')
loc3 = Locpot.from_file('LOCPOT')
tloc = [loc, loc2, loc3]
# tloc = [loc, loc2]
#%%
def charge(pot):
dat = np.sum(pot.data['total'], axis=0) / pot.dim[0]
dat = np.roll(dat, int(pot.dim[1] * 0.8) // 2, axis=0)
dat = np.swapaxes(dat, 0, 1)
grad = np.gradient(dat)
u = grad[0]
v = grad[1]
uu = np.gradient(u)
vv = np.gradient(v)
return(uu, vv)
u, v = charge(loc)
#%%
import matplotlib.pyplot as plt
from pymatgen.core.sites import Site
from pymatgen.io.vasp import Locpot
import os
os.chdir(
'/home/jinho93/new/oxides/perobskite/lanthanum-aluminate/periodic_step/vasp/stengel/015/4/lvhar'
)
loc = Locpot.from_file('LOCPOT')
s = loc.structure
from scipy.interpolate import griddata
import numpy as np
#%%
dat = loc.data['total']
x = []
y = []
z = []
site: Site
on = False
for site in sorted(s.sites, key=lambda x: x.c):
if site.species_string == "O":
# or site.species_string == "Al":
x.append(site.b)
y.append(site.c)
z.append(dat[int(dat.shape[0] * site.a),
int(dat.shape[1] * site.b),
def process_vasprun(self, dir_name, taskname, filename):
"""
Adapted from matgendb.creator
Process a vasprun.xml file.
"""
vasprun_file = os.path.join(dir_name, filename)
vrun = Vasprun(vasprun_file)
d = vrun.as_dict()
# rename formula keys
for k, v in {"formula_pretty": "pretty_formula",
"composition_reduced": "reduced_cell_formula",
"composition_unit_cell": "unit_cell_formula"}.items():
d[k] = d.pop(v)
for k in ["eigenvalues", "projected_eigenvalues"]: # large storage space breaks some docs
if k in d["output"]:
del d["output"][k]
comp = Composition(d["composition_unit_cell"])
d["formula_anonymous"] = comp.anonymized_formula
d["formula_reduced_abc"] = comp.reduced_composition.alphabetical_formula
d["dir_name"] = os.path.abspath(dir_name)
d["completed_at"] = str(datetime.datetime.fromtimestamp(os.path.getmtime(vasprun_file)))
d["density"] = vrun.final_structure.density
# replace 'crystal' with 'structure'
d["input"]["structure"] = d["input"].pop("crystal")
d["output"]["structure"] = d["output"].pop("crystal")
for k, v in {"energy": "final_energy", "energy_per_atom": "final_energy_per_atom"}.items():
d["output"][k] = d["output"].pop(v)
# Process bandstructure and DOS
if self.bandstructure_mode != False:
bs = self.process_bandstructure(vrun)
if bs:
d["bandstructure"] = bs
if self.parse_dos != False:
dos = self.process_dos(vrun)
if dos:
d["dos"] = dos
# Parse electronic information if possible.
# For certain optimizers this is broken and we don't get an efermi resulting in the bandstructure
try:
bs = vrun.get_band_structure()
bs_gap = bs.get_band_gap()
d["output"]["vbm"] = bs.get_vbm()["energy"]
d["output"]["cbm"] = bs.get_cbm()["energy"]
d["output"]["bandgap"] = bs_gap["energy"]
d["output"]["is_gap_direct"] = bs_gap["direct"]
d["output"]["is_metal"] = bs.is_metal()
if not bs_gap["direct"]:
d["output"]["direct_gap"] = bs.get_direct_band_gap()
if isinstance(bs, BandStructureSymmLine):
d["output"]["transition"] = bs_gap["transition"]
except Exception:
logger.warning("Error in parsing bandstructure")
if vrun.incar["IBRION"] == 1:
logger.warning("Vasp doesn't properly output efermi for IBRION == 1")
if self.bandstructure_mode is True:
logger.error(traceback.format_exc())
logger.error("Error in " + os.path.abspath(dir_name) + ".\n" + traceback.format_exc())
raise
# store run name and location ,e.g. relax1, relax2, etc.
d["task"] = {"type": taskname, "name": taskname}
# include output file names
d["output_file_paths"] = self.process_raw_data(dir_name, taskname=taskname)
# parse axially averaged locpot
if "locpot" in d["output_file_paths"] and self.parse_locpot:
locpot = Locpot.from_file(os.path.join(dir_name, d["output_file_paths"]["locpot"]))
d["output"]["locpot"] = {i: locpot.get_average_along_axis(i) for i in range(3)}
if self.parse_chgcar != False:
# parse CHGCAR file only for static calculations
# TODO require static run later
# if self.parse_chgcar == True and vrun.incar.get("NSW", 0) < 1:
try:
chgcar = self.process_chgcar(os.path.join(dir_name, d["output_file_paths"]["chgcar"]))
except:
raise ValueError("No valid charge data exist")
d["chgcar"] = chgcar
if self.parse_aeccar != False:
try:
chgcar = self.process_chgcar(os.path.join(dir_name, d["output_file_paths"]["aeccar0"]))
except:
raise ValueError("No valid charge data exist")
d["aeccar0"] = chgcar
try:
chgcar = self.process_chgcar(os.path.join(dir_name, d["output_file_paths"]["aeccar2"]))
except:
raise ValueError("No valid charge data exist")
d["aeccar2"] = chgcar
# parse force constants
if hasattr(vrun, "force_constants"):
d["output"]["force_constants"] = vrun.force_constants.tolist()
d["output"]["normalmode_eigenvals"] = vrun.normalmode_eigenvals.tolist()
d["output"]["normalmode_eigenvecs"] = vrun.normalmode_eigenvecs.tolist()
# Try and perform bader
if self.parse_bader:
try:
bader = bader_analysis_from_path(dir_name, suffix=".{}".format(taskname))
except Exception as e:
bader = "Bader analysis failed: {}".format(e)
d["bader"] = bader
return d