How to use the Pyxaid2 to read the XDATCAR from the VASP MD simulation for electronic spectrum calculation?

[misa...@gmail.com]

Hi Alexey,

I have the XDATCAR from the VASP MD simulation and want to use Pyxaid2 to calculate electronic spectrum. I have modified the parameters in the submit.pbs file, "params [\ "compute_Hprime \"] = 1", and the parameters of "minband" and "maxband" depend on the number of electronic valence from the QE pseudopotential. Meanwhile, I have modified the parameters in py-scr2.py file, "out2inp.out2inp -> xdatcar2inp.xdatcar2inp". But, the running has the following errors,

In the CRASH file, "the error opening xml data file...".

In the qe_pyxaid2_test.e44209 file, "At line 136 of file read_cards.f90 \ Fortran runtime error: End of file..."

Should I modify other files or parameters?

Best.

Misaraty

[Wei Li]

Hi Misaraty,

It's hard to say what's going wrong without knowing the details. Can you post your input files?

Wei

[misa...@gmail.com]

Hi Wei,

Thanks.

py-scr2.py,

from PYXAID2 import *

import os

user = 1 #weili or eric

#user = 2 #alexey

nsteps_per_job = 2

tot_nsteps = 4

# Step 1 - split MD trajectory on the time steps

# Provide files listed below: "GaAs-md.out" and "x0.scf.in", "x1.scf.in"

# IMPORTANT: 

# 1) use only ABSOLUTE path for PP in x.scf.in file

# 2) provided input file is just a template, so do not include coordinates

rt="/home/export/parastor/keyuser/bnulongr/apps/test/test6/"

# out2inp.out2inp(rt+"x.md.out","x0.scf.in","wd","x0.scf",0,tot_nsteps,1)  # non-relativistic setup

#out2inp.out2inp(rt+"GaAs-md.out","x1.scf.in","wd","x1.scf",0,tot_nsteps,1)  # relativistic setup

xdatcar2inp.xdatcar2inp(rt+"XDATCAR","x0.scf.in","wd","x0.scf",0,tot_nsteps,1)

# Step 2 - distribute all time steps into groups(jobs) 

# several time steps per group - this is to accelerate calculations

# creates a "customized" submit file for each job and submit it - run

# a swarm in independent calculations (trajectory pieces)

#(HTC paradigm)

# Provide the files below: 

# submit_templ.pbs - template for submit files - manually edit the variables

# x.exp.in - file for export of the wavefunction

if user==1: 

   os.system("cp submit.pbs wd")

elif user==2:

   os.system("cp submit_templ.slm wd")

os.system("cp x0.exp.in wd")

#os.system("cp x1.exp.in wd")

os.chdir("wd")

if user==1:

   #distribute.distribute(0,tot_nsteps,nsteps_per_job,"submit.pbs",["x0.exp.in","x1.exp.in"],["x0.scf","x1.scf"],1)

   distribute.distribute(0,tot_nsteps,nsteps_per_job,"submit.pbs",["x0.exp.in"],["x0.scf"],1)

   #distribute.distribute(0,tot_nsteps,nsteps_per_job,"submit.pbs",["x1.exp.in"],["x1.scf"],1)

# elif user==2:

   #distribute.distribute(0,tot_nsteps,nsteps_per_job,"submit_templ.slm",["x0.exp.in"],["x0.scf"],2)

   #distribute.distribute(0,tot_nsteps,nsteps_per_job,"submit_templ.slm",["x1.exp.in"],["x1.scf"],2)

   #distribute.distribute(0,tot_nsteps,nsteps_per_job,"submit_templ.slm",["x0.exp.in","x1.exp.in"],["x0.scf","x1.scf"],2)

submit.pbs,

#!/bin/bash

#PBS -l nodes=1:ppn=2

#PBS -N qe_pyxaid2_test

#PBS -q q_zhq_bnulongr

cd $PBS_O_WORKDIR

echo $PBS_O_WORKDIR

# the below settings works for my laptop, please customize your own environmental variables!

###### some conventional settings ###########

export PYTHONPATH=/home/export/parastor/keyuser/bnulongr/apps/libra/bin/src:$PYTHONPATH

export LD_LIBRARY_PATH=/home/export/parastor/keyuser/bnulongr/apps/libra/bin/src:$LD_LIBRARY_PATH

export PYTHONPATH=/home/export/parastor/keyuser/bnulongr/apps/pyxaid2-master:$PYTHONPATH

export LD_LIBRARY_PATH=/home/export/parastor/keyuser/bnulongr/apps/pyxaid2-master:/home/export/parastor/keyuser/bnulongr/apps/boost1.6/lib:$LD_LIBRARY_PATH

export LD_LIBRARY_PATH=/home/export/parastor/clussoft/compiler/intel/composer_xe_2015.2.164/mkl/lib/intel64:$LD_LIBRARY_PATH

##### path for the QE module #####

exe_qespresso=/home/export/parastor/keyuser/bnulongr/apps/qe/5.3.0/espresso-5.3.0/bin/pw.x

exe_export=/home/export/parastor/keyuser/bnulongr/apps/qe/5.3.0/espresso-5.3.0/bin/pw_export.x

exe_convert=/home/export/parastor/keyuser/bnulongr/apps/qe/5.3.0/espresso-5.3.0/bin/iotk

##### MPI #####

source /home/export/parastor/clussoft/profile.d/intelmpi.sh 

MPIRUN=/home/export/parastor/clussoft/mpi/intelmpi/5.0.2.044/intel64/bin/mpirun

NP=$(wc -l $PBS_NODEFILE | awk '{print $1}')

# These will be assigned automatically, leave them as they are

param1=

param2=

res=/home/export/parastor/keyuser/bnulongr/apps/test/test6/res

# This is invocation of the scripts which will further handle NA-MD calclculations

# on the NAC calculation step

python -c "from PYXAID2 import *

params = { }

params[\"BATCH_SYSTEM\"]=\"$MPIRUN\"

params[\"NP\"]=$NP

params[\"EXE\"]=\"$exe_qespresso\"

params[\"EXE_EXPORT\"]=\"$exe_export\"

params[\"EXE_CONVERT\"] =\"$exe_convert\"

params[\"start_indx\"]=$param1

params[\"stop_indx\"]=$param2

params[\"wd\"]=\"wd_test\"

params[\"rd\"]=\"$res\"

params[\"minband\"]=351

params[\"maxband\"]=352

params[\"minband_soc\"]=351

params[\"maxband_soc\"]=352

params[\"nac_method\"]=0

params[\"wfc_preprocess\"]=\"complete\"

params[\"do_complete\"]=1

params[\"prefix0\"]=\"x0.scf\"

params[\"prefix1\"]=\"x1.scf\"

params[\"compute_Hprime\"]=1

params[\"pptype\"]=\"US\"

print params

runMD2.runMD(params)

"

x0.scf.in,

&CONTROL

  calculation = 'scf',

  pseudo_dir = '/home/export/parastor/keyuser/bnulongr/soft/qe/upf_files',

  outdir = './',

  prefix = 'x0',

  disk_io = 'low',

  wf_collect = .true.

/

&SYSTEM

  ibrav = 0,

!  celldm(1) = 1.88973,

  nat = 78,

  ntyp = 2,

!  nspin = 2,

  nbnd = 20,

  ecutwfc = 30,

  tot_charge = 0.0,

!  starting_magnetization(1) = 0.01

  occupations = 'smearing',

  smearing = 'gaussian',

  degauss = 0.005,

 nosym = .true.,

/

&ELECTRONS

  electron_maxstep = 300,

  conv_thr = 1.D-5,

  mixing_beta = 0.45,

/

&IONS

  ion_dynamics = 'verlet',

  ion_temperature = 'andersen',

  tempw = 300.00 ,

  nraise = 1,

/

ATOMIC_SPECIES

 Zn  65.3799972534  Zn.pbe-van.UPF

 S  32.0600013733  S.pbe-n-kjpaw_psl.0.1.UPF

K_POINTS {gamma}                  

                               

CELL_PARAMETERS {angstrom}

    19.7647991179999991    0.0000000000000000    0.0000000000000000

     0.0000000000000000   19.0186996460000017    0.0000000000000000

     0.0000000000000000    0.0000000000000000   18.7049007415999995

 

 

 

x0.exp.in,

&inputpp

  prefix = 'x0',

  outdir = './',

  pseudo_dir = '/home/export/parastor/keyuser/bnulongr/soft/qe/upf_files',

  psfile(1) = 'Zn.pbe-van.UPF',

  psfile(2) = 'S.pbe-n-kjpaw_psl.0.1.UPF',

  single_file = .FALSE.,

  ascii = .TRUE.,

  uspp_spsi = .FALSE.,

/

Misaraty

[Wei Li]

Hi Misaraty,

Please switch to PYXAID1 for this purpose, 

Perhaps we will implement the functionality for transition dipole moment in PYXAID2 which is necessary for absorption spectrum calculation once we have the current issues solved.

[misa...@gmail.com]

Hi Wei,

Thanks.

The pyxaid1 for electron spectrum calculation with md.out does not contain any problems, but there was a wrong when it was switched to xdatcar.

The error was similar to Pyxaid2 with xdatcar.

Misaraty

[misa...@gmail.com]

Hi Wei,

Thanks.

Pyxaid1 with xdatcar is normal.

Best.

Misaraty