Nothing is written in the working directory

[shamik chakrabarti]

Dear users,

              I am running a simulation in quantum espresso with calculation = 'scf'. My query is that during running nothing is written in working directory. If I need to stop to change some parameter viz., mixing parameter, calculation is starting from the scratch & 'restart' can not be done as there is nothing written in the directory during running. It is only when I finished some calculation the .save file is written. What should I do?

with regards, 

--

Dr. Shamik Chakrabarti

Research Fellow 

Department of Physics

Indian Institute of Technology Patna

Bihta-801103

Patna

Bihar, India

[KK M]

share your input file tags. you may not need to copy crystal structre

--
You received this message because you are subscribed to the Google Groups "QuantumEspresso" group.
To unsubscribe from this group and stop receiving emails from it, send an email to quantumespres...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/quantumespresso/CAA%2B63%2Bo_p5PYgusFreDcf3%3DMhr3uubQsAd-d0LL%2B3w5stFjZPA%40mail.gmail.com.

[shamik chakrabarti]

---------- Forwarded message ---------
From: shamik chakrabarti <shamik1...@gmail.com>
Date: Tue, 23 Dec 2025 at 16:51
Subject: Re: Nothing is written in the working directory
To: KK M <kkmwo...@gmail.com>

The tag is as below (without crystal structure);

&CONTROL
    calculation   = "scf"
    forc_conv_thr =  1.00000e-03
    max_seconds   =  5.72800e+05
    nstep         = 100
    pseudo_dir    = "C:\cygwin_wm\opt_win\QuantumESPRESSO_7.1\pseudo"
    outdir = "C:\NDFC\NDFC_opt_Fe_Cr_FM"
    prefix = 'NDFC_opt_Fe_Cr_FM'
   



/
&SYSTEM            
    degauss     =  0.800000e-02
    ecutrho     =  3.60000e+02
    ecutwfc     =  3.00000e+01
    ibrav       = 0
    nat         = 20
    ntyp        = 7
    nspin       = 2
    occupations = "smearing"
    smearing    = "marzari-vanderbilt"
    starting_magnetization(1) = 1.14000e+00
    starting_magnetization(2) = -1.14000e+00
    starting_magnetization(3) = 5.00000e+00
    starting_magnetization(4) = -5.00000e+00
    starting_magnetization(5) = 4.00000e+00
    starting_magnetization(6) = -4.00000e+00
    starting_magnetization(7) = 0.00000e+00
   

/
&ELECTRONS
    conv_thr         =  1.00000e-06
    electron_maxstep = 600
    mixing_beta      = 0.02
    mixing_mode      = 'local-TF'
    startingpot      = "atomic"
    startingwfc      = "atomic+random"
    electron_dynamics = 'damp'
    diagonalization  = 'david'
/

&IONS
    ion_dynamics = "cg"
/
&CELL
    cell_dynamics  = "bfgs"
    press_conv_thr =  5.00000e-01
/


K_POINTS {automatic}
 2  2  2  0 0 0

regards,

[shamik chakrabarti]

---------- Forwarded message ---------
From: shamik chakrabarti <shamik1...@gmail.com>
Date: Tue, 23 Dec 2025 at 17:03
Subject: Re: Nothing is written in the working directory
To: KK M <kkmwo...@gmail.com>

I have tried using verbosity = 'high' but that does not work.  My simulation is oscillating at the moment. I will wait for sometime & then will stop it & see if any improve occurs by removing the space between the '/' as you have suggested.

On Tue, 23 Dec 2025 at 16:58, KK M <kkmwo...@gmail.com> wrote:

use verbosity = 'high' tag in &control and see if its writing the output data.

also remove space before the last '/' between each name list.

however, u can stop the run by giving system based command to stop run anytime using the run job ID. 

[shamik chakrabarti]

Removing space between '/' does not change anything............ 

[shamik chakrabarti]

---------- Forwarded message ---------
From: shamik chakrabarti <shamik1...@gmail.com>
Date: Tue, 23 Dec 2025 at 22:09
Subject: Re: Nothing is written in the working directory
To: KK M <kkmwo...@gmail.com>

There are 20 atoms in my system, when a calculation is finished the .save file is saved in the working directory. However, in between the calculation if I want to stop it for changing some parameters like mixing beta, I have found there is nothing saved in the working directory & I have to run the calculation all overagain!

On Tue, 23 Dec 2025 at 19:36, KK M <kkmwo...@gmail.com> wrote:

more over see the below code for SCF as a sample,

if u use these many tags, should be enough i think, but i am not sure about your system, but check. in this example also some tags not needed for SCF, but i used in my case.

how many atoms your system has?

-------------

&CONTROL
  calculation = 'scf'
  etot_conv_thr =   3.0000000000d-04
  forc_conv_thr =   1.0000000000d-04
  outdir = './out/'
  prefix = 'wat-in-co'
  pseudo_dir = '../pseudo'
  tprnfor = .true.
  tstress = .true.
  verbosity = 'high'
  max_seconds = 10800
  nstep = 120
  restart_mode = 'from_scratch'
/
&SYSTEM
  degauss =   2.0000000000d-02
  ecutrho =   3.6000000000d+02
  ecutwfc =   4.5000000000d+01
  ibrav = 0
  A = 7.5435
  nat = 25
  nosym = .true.
  ntyp = 6
  occupations = 'smearing'
  smearing = 'cold'
  input_dft='PBE'
! ##  nbnd = 220
/
&ELECTRONS
 conv_thr =         1.6000000000d-08
 electron_maxstep = 1800
 mixing_beta =      2.0000000000d-01
 mixing_mode='plain'
/

----------------------------------------

On Wed, Dec 24, 2025 at 12:00 AM KK M <kkmwo...@gmail.com> wrote:

it doesnt change anything much but will help you focus on the tags and the code. there sometimes might be hidden codes (ASCII text) may interfere. but your code is running means there is no such issue.

another thing do not use so many tags. all tags may not be required. use basic essential tags only and you are using only SCF. why are you taking so much max_seconds?

[shamik chakrabarti]

Dear Sir,

             The replies to your queries are as follows,

1. No the outdirectory & working directory is different.

2. I am using only my laptop.

3. max_seconds are kept as it is as I have copied vc-relax file & change it to scf file.

4.  I have learned that ecutrho should be 10-12 times of ecutwfc for USPP, that's why I kept it as 12 times higher than ecutwfc

5. I will comment out unnecessary tags.

6. I will use single quotes as you have advised

Looking forward to your further suggestions if any.

with regards,

On Wed, 24 Dec 2025 at 09:08, KK M <kkmwo...@gmail.com> wrote:

Firstly you reply to the google groups also so that it is useful for everyone.

second point is, as per my understanding, the .save file will be written at successful end of the scf step.

though your run is working, but few my queries.

1. is the outdir path same as working directory where your input file exists?

2. are u using MPI super computer or only your desktop system?

3. use of max_seconds   =  5.72800e+05  is very high time 159 hours. why so much for just 20 atoms system?

4. ecutrho = 360 is high for ecutwfc of 30.  just 10 times of ecutwfc = ecutrho is much more than enough.

5. also remove or comment out some special tags and first see if ur system is completing SCF with minimum tags. later you keep add other constraints.

6. why you putting double quotes like calculation   = "scf"? use single quotes only unless your QE version demands.

sometimes QE doesnt show error for formatting, but at same time doesnt bother to giver error nor warning, it simply ignores. so we do not know if there is any issue with the tags. we think all are ok, if not showing any error.

  occupations = 'smearing'

  smearing = 'cold'
  input_dft='PBE'
! ##  nbnd = 220
/
&ELECTRONS
 conv_thr =         1.6000000000d-08
 electron_maxstep = 1800
 mixing_beta =      2.0000000000d-01
 mixing_mode='plain'
/

----------------------------------------

On Wed, Dec 24, 2025 at 12:00 AM KK M <kkmwo...@gmail.com> wrote:

it doesnt change anything much but will help you focus on the tags and the code. there sometimes might be hidden codes (ASCII text) may interfere. but your code is running means there is no such issue.

another thing do not use so many tags. all tags may not be required. use basic essential tags only and you are using only SCF. why are you taking so much max_seconds?

[KK M]

Thanks for your explanation. 

you change input file.

for SCF you dont need &ions, and &cell name lists. remove them.

if u using only laptop there is no need of max_seconds tag as per i know.

you see my sample scf input file, and follow similar tags and remove rest of them. 

[shamik chakrabarti]

Ok Sir, thank you so much for your support! :)

with regards,