Calypso stops after the end of the 1st step

[Pablo Sucatel]

Dear All,

Sorry for my long email. There are sample input files (in blue) accompanied with questions and comments in red.

I am trying to make a test of Calypso with Quantum Espresso on SiO2. I have tried with Espresso 5.1.1 and 6.1. I do not know why but Calypso stops right after the last structure of the 1st step has been calculated. 

Here is the input for Calypso:

######## The Basic Parameters of CALYPSO ##########

SystemName = SiO2

NumberOfSpecies = 2

NameOfAtoms = Si O

NumberOfAtoms =  4 8

NumberOfFormula= 1 1 

AtomicNumber = 14 8

MaxStep = 30

#Variable stoichiometry

VSC = F

Volume = 40

@DistanceOfIon 

1. 1.

1. 1.

@End

PsoRatio = 0.3

Ialgo = 1

ICode= 4

Kgrid=0.18 0.06

kvasp = 1

Command = sh submit_long.sh

PopSize = 30

NumberOfLbest=4

NumberOfLocalOptim= 3

PickUp= F

Parallel= F

Debug=T

#### End of Calypso Input #####

Note: I have tried various values of PsoRatio from 0.7 to 0.3, but nothing changes. Same for Ialgo (1 or 2).

BTW: what is the meaning of kvasp and NumberOfLbest? How are these variables used?

Here is an example of input for Espresso (nothing particular):

&CONTROL

  title             = 'SiO2',

  calculation       = 'vc-relax',

  nstep             = 100,

  etot_conv_thr     = 1.d-6,

  forc_conv_thr     = 1.d-3,

  pseudo_dir        = './',

  outdir            = './',

  wfcdir            = './pwwfc'

  restart_mode      = 'from_scratch',

  verbosity         = 'low',

  wf_collect        = .true.,

/

&SYSTEM

  ibrav             = 0,

  celldm(1)         = 1.d0,

  nat               = 12,

  ntyp              = 2,

  nosym             = .true.,

  ecutwfc           = 40.d0,

  occupations       = 'fixed',

  nbnd              = 40,

/

&ELECTRONS

  conv_thr          = 1d-7,

  electron_maxstep  = 150,

  scf_must_converge = .false.,

  mixing_beta       = 0.1d0,

/

&IONS

  ion_dynamics      = 'bfgs',

/

&CELL

  cell_dynamics     = 'bfgs',

  press             = 0.00101325d0,

  press_conv_thr    = 0.2d0,

  cell_factor       = 3.0,

/

ATOMIC_SPECIES

Si 28.0 Si.pz-vbc.UPF

O  16.0  O.pz-mt.UPF

The pbs file I am using to submit the calculation to the queueing system (nothing particular either):

#!/bin/sh

#PBS -S /bin/sh

#PBS -N SiO2

#PBS -j oe

#PBS -l walltime=720:00:00

#PBS -l nodes=1:ppn=12

. /usr/share/Modules/init/sh

module load calypso

module load espresso-6.1/intel/openmpi/qe

cd $PBS_O_WORKDIR

tar cf arch.tar *

cd /tmp

mkdir calyp

cd calyp

tar xf $PBS_O_WORKDIR/arch.tar

calypso.x > calyp.log

rsync -atz ./* ${PBS_O_WORKDIR}/.

cd ..

rm -rf calyp

And the submit_long.sh script launched by Calypso:

#!/bin/sh

mpirun -n 12 pw.x -inp pw_input > out.pw

### to backup input and output files in directory qeout

echo "" >> step

n=$(wc -l step | awk '{print $1}')

cp pw_input qeout/pw_input_$n

cp out.pw qeout/out.pw_$n

####

At the end of the calculations I have the following files in the results directory:

CALYPSO_input.dat

CALYPSO.log

pso_ini_1

pso_opt_1

similar.dat

No *sor* file at all. How can it be? what could be the reasons?

Among the 30 structures generated 22 seem to have effectively converged, but Calypso have extracted meaningful energies for all the 30 structures, e.g.:

-285.28080208

-284.51736729

-284.83052514

-283.30807149

-284.80315154

etc.

I assume Calypso considers all the structures converged, so why is it stopping at the end of the 1st step?

See the very end of CALYPSO.log file:

   Nstru :    30

      Random Structure

      space group :  110

           4           8

      Volume :     40.00000847

  Step :   2

Here is the content of a log file that I get from the PBS queuing system. At the very end I have the following message:

forrtl: severe (59): list-directed I/O syntax error, unit 1000, file /tmp/calyp_1490560476/results/pso_opt_1

Image              PC                Routine            Line        Source             

calypso.x          000000000075A68E  Unknown               Unknown  Unknown

Any idea about where the problem comes from would be welcome.

Thank you for your help,

Best regards

Pascal (Associate professor at Aix-Marseille University, France)

[JianLv]

Dear Pascal,

Please send me all the input and output files of this calculation, then we can rerun the simulation and locate the problem.

kvass is only valid for VASP interface, which is used to kill the vasp calculations that exceed Maxtime.

NumberOfLbest controls how many local best structure are used to lead the structure evolution in PSO.

Best Regards
 
Jian Lv
 
***********************************************************************
Jian Lv PhD
College of Materials Science and Engineering,

Jilin University
2699 Qianjin Str., 130012, Changchun, China
Email: lvj...@calypso.cn or lvj...@jlu.edu.cn
***********************************************************************
CALYPSO: A  Structure Prediction Package
Webpage:  http://www.calypso.cn
***********************************************************************

--
You received this message because you are subscribed to the Google Groups "CALYPSO" group.
To unsubscribe from this group and stop receiving emails from it, send an email to CalypsoCode...@googlegroups.com.
To post to this group, send email to Calyp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/CalypsoCode/24a9550b-990d-4a5a-89a8-1cf8b7a997a5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Pablo Sucatel]

Thank you for your answer. I have just sent you an archive containing all the files.

Best regards

Pascal

[shahriar...@yahoo.de]

Hello,

I have the exact same issue, except I'm using VASP instead of QE.

Was this ever solved, and if so, what was the solution?

Thank you very much in advance for any help!

Best regards

Shahriar

[Sudha Priyanka]

I have similar issues. Any one sorted this out? Any help regarding this is highly appreciated.
 Best regards

Priyanka

[296389682]

您的邮件已收到，我会尽快处理！

This is an automatic reply, confirming that your email was received. I will reply as soon as possible. Thank you!