[ase-users] QE Bandstructure Symmetry Error
Lance Kavalsky
lance.kavalsky at mail.utoronto.ca
Thu May 9 00:59:12 CEST 2019
Hello all,
I am trying to use ase(v. 3.16.3b1) to interface with Quantum Espresso (v. 6.4) to automate generating a band structure diagram. When running I am met with an error "ValueError: Cannot find crystal symmetry". When encountering this error, espresso.pwi is printed as an empty file.
If I change the k-path to only Gamma (defined using kpts=(1,1,1)), it is able to run and give the total energy, while properly printing out the espresso.pwi file. Any advice to avoid this error would be much appreciated.
My script is pasted below.
Thanks,
Lance Kavalsky
SCRIPT BELOW:
from ase.io import read
from ase.dft.bandgap import bandgap
from ase.dft.kpoints import ibz_points,get_bandpath,special_points
from ase.dft.kpoints import get_special_points
from ase.calculators.espresso import Espresso
p = read('bphos.xyz', format='xyz')
input_data = {'control': {'pseudo_dir' : '/home/s/singhcha/mukherje/pseudo', 'disk_io':'low', 'etot_conv_thr':0.00005},
'system': {'nosym': 'true', 'ecutwfc' : 60, 'ecutrho' : 480,
'occupations' : "smearing", 'smearing' : 'marzari-vanderbilt','london_rcut':150,
'degauss' : 0.02,'vdw_corr':'DFT-D'},
'electrons' : {'mixing_mode':'plain', 'mixing_beta' : 0.4, 'electron_maxstep' : 1000}}
pseudopotentials = {'P':'P.pbe-n-kjpaw_psl.0.1.upf','Li':'Li.pbe-s-kjpaw_psl.0.2.1.UPF'}
p.set_cell([(6.561597749, 11.366025622, 0.070271656), (-6.561597749, 11.366025622, 0.070271656), (0.000000000, 0.087239675, 23.000000000)])
p.calc = Espresso(command='mpirun pw.x < espresso.pwi > espresso.pwo',pseudopotentials=pseudopotentials,input_data=input_data,
kpts = {'path' : 'GMK', 'npoints' : 15}, tprnfor=True, tstress=True, london=True)
energy=p.get_potential_energy()
bs = p.calc.band_structure()
bs.plot(emax = 10, filename = 'test.png')
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20190508/0d6a1508/attachment.html>
More information about the ase-users
mailing list