[QE-users] Error with AUTOPILOT card and conjugate gradient electron_dynamics

[Nam Tran]

Dear Riccardo,


Thank you for your help.

I will definitely try that.


Regards

Nam

________________________________
From: users <users-bounces at lists.quantum-espresso.org> on behalf of Riccardo Bertossa <rbertoss at sissa.it>
Sent: Wednesday, October 24, 2018 7:33:05 AM
To: users at lists.quantum-espresso.org
Subject: Re: [QE-users] Error with AUTOPILOT card and conjugate gradient electron_dynamics


Hi,

I implemented it some months ago (see https://gitlab.com/QEF/q-e/blob/develop/CPV/src/cp_autopilot.f90 )

CG and autopilot works for sure with no parallelization options.

remember to change also orthogonalization!


best,

Riccardo Bertossa


On Tue, Oct 16, 2018 at 4:19 AM Nam Tran <vnt981 at uowmail.edu.au<https://lists.quantum-espresso.org/mailman/listinfo/users>> wrote:

> I have found that electron_dynamics = 'cg' is not implemented in the code
> (cp_autopilot.f90). Is there any special reason for this ?.
>
> ! ELECTRON_DYNAMICS
> ! electron_dynamics = 'sd' | 'verlet' | 'damp' | 'none'
> if (event_electron_dynamics(event_index)) then
>    electron_dynamics= rule_electron_dynamics(event_index)
>   frice = 0.d0
>    select case ( electron_dynamics )
>    case ('SD')
>       tsde  = .true.
>    case ('VERLET')
>       tsde  = .false.
>    case ('DAMP')
>       tsde  = .false.
>       frice = electron_damping
>    case ('NONE')
>       tsde  = .false.
>    case default
>       call auto_error(' autopilot ',' unknown electron_dynamics '//trim(electron_dynamics) )
>    end select
>    IF ( ionode ) write(*,'(4X,A,2X,A10)') 'Rule event: electron_dynamics', electron_dynamics
> endif
>
> Thanks
>
>
> ------------------------------
> *From:* users <users-bounces at lists.quantum-espresso.org<https://lists.quantum-espresso.org/mailman/listinfo/users>> on behalf of Nam
> Tran <vnt981 at uowmail.edu.au<https://lists.quantum-espresso.org/mailman/listinfo/users>>
> *Sent:* Tuesday, October 16, 2018 1:07:11 PM
> *To:* Quantum Espresso users Forum
> *Subject:* [QE-users] Error with AUTOPILOT card and conjugate gradient
> electron_dynamics
>
>
> Dear QE users and experts,
>
> I want to use  the conjugate gradient method to converge the wavefunction for
> a couple of ionic steps in order to "cool down" the electrons in my CPMD
> simulation. Therefore, I used the AUTOPILOT card to change the  electron_dynamics
> on the fly of the calculation
> e.g.
> AUTOPILOT
>     on_step = 191  : electron_dynamics='cg'
> ENDRULES
>
> However, I encountered an error which said "unknown electron_dynamics CG".
> I have tested with other options e.i "damp" but found no problem.
>
> I appreciate any help and recommendations
>
> Best Regards
>
> Nam Tran
>
>
>
> INPUT FILE
>
>
> &CONTROL
>  calculation='cp',
>  restart_mode =  'reset_counters',
>  iprint=1,
>  isave=10,
>  nstep =2000,
>  dt = 5.0,
>  ndr=91,
>  ndw=92,
>  outdir='/home/vnt/Work/Espresso/Diamond2/tmp/',
>  prefix='2DL',
>  pseudo_dir='/home/vnt/qe_pseudopotential'
> /
>
> &SYSTEM
>  ibrav = 0,
>  nat = 130,
>  ntyp = 2,
>  ecutwfc=40.0,
>  ecutrho=320.0,
>  nr1b = 24
>  nr2b = 24
>  nr3b = 24
> /
>
> &ELECTRONS
>  conv_thr = 1.d-6,
>  electron_maxstep = 500,
>  electron_dynamics='verlet',
>  emass=400, emass_cutoff=3.,
>  orthogonalization = 'ortho',
>  ortho_max = 500,
> /
>
> &IONS
>  ion_dynamics='verlet',
>  ion_temperature='nose',
>  tempw=300, fnosep=30,
> /
>
> AUTOPILOT
>     on_step = 191  : electron_dynamics='cg'
> ENDRULES
>
> ATOMIC_SPECIES
> C 12.011 C.pbe-n-rrkjus_psl.1.0.0.UPF
> O 15.999 O.pbe-n-rrkjus_psl.1.0.0.UPF
>
> CELL_PARAMETERS angstrom
> 10.057999610         0.0000000000        0.0000000000
> 0.0000000000         5.0289998055        0.0000000000
> 0.0000000000         0.0000000000        30.3360004425
>
> ATOMIC_POSITIONS angstrom
> C 1.25700 1.27200 0.17800
> C 3.77199 3.78601 0.17800
> C 1.25700 3.78601 0.17800
> .........................................................................
>
> _______________________________________________
> users mailing list
> users at lists.quantum-espresso.org<https://lists.quantum-espresso.org/mailman/listinfo/users>
> https://lists.quantum-espresso.org/mailman/listinfo/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20181024/b22ef386/attachment.html>