[Pw_forum] lsign = .true.

[Dal Corso Andrea]

Thank you for reporting this. I have actually found a problem that might
appear in some cases in a parallel machine when lsign=.true.. This
problem is now corrected in the cvs version. From the partial input you
provided it is not possible to understand if this problem actually
occurs in your case. Please provide a 'complete' input or a simpler but
'complete' input that shows the problem (... or try with the cvs version
yourself).

Hope this helps,

Andrea


On Wed, 2008-03-26 at 14:31 -0400, EG Kim wrote:
> Dear Colleagues,
> 
> I am trying to get the wavefunction pattern, sign(psi)|psi|^2, of a band at the
> Gamma-point, using v3.2.3.  pp.x runs fine without 'lsign = .true.', but with
> 'lsign' on, the program stops with the following message in the CRASH file:
> 
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>     task #        33
>      from local_dos : error #         1
>      zero wavefunction
> 
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> 
> In the preceding pw.x (bands) run with 'wf_collect = .true.', I did calculate at
> the G-point (I checked the '[].save/K00001' directory to confirm that 'evc.dat'
> exists).  (See attached input files.)  I also tried to run pp.x in the presence
> of temporary *.wfc files (and by defining 'outdir' explicitly), but none of them
> worked.  May I kindly ask for help?
> 
> Regards,
> 
> EG Kim
> 
> 'bands' run input -------------------------------------------------------------
> 
> &control
>   calculation = 'bands'
>   pseudo_dir = '.......'
>   prefix = 'test'
>   wf_collect = .true.
> /
> &system
>   ibrav = 14
>   a = 11.15402
>   b = 16.56278
>   c = 3.90742
>   cosbc = 0.0
>   cosac = -0.0728848422
>   cosab = 0.0
>   nat = 48
>   ntyp = 2
>   ecutwfc = 25.0
>   ecutrho = 200.0
>   nbnd = 146
> /
> &electrons
>   conv_thr = 1.0D-8
>   diago_full_acc = .true.
> /
> K_POINTS
>   1
>    0.0000000   0.0000000   0.0000000   1
> 
> ATOMIC_SPECIES
> 
> pp run input ------------------------------------------------------------------
> 
> &inputpp
>   plot_num = 7
>   kpoint = 1
>   kband = 112
>   lsign = .true.
>   prefix = 'test'
>   filplot = 'test.mo'
> /
> &plot
>   iflag = 3
>   output_format = 6
>   fileout = 'test.cube'
> /
> 
> -------------------------------------------------------------------------------
> --
> + -
> | E.-G. Kim, Ph.D.
> | School of Chemistry and Biochemistry
> | 901 Atlantic Drive NW
> | Georgia Institute of Technology
> | Atlanta, GA 30332-0001
> | U.S.A.
> + -
> | +1.215.327.7315
> | egk at gatech.edu
> + -
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
-- 
Andrea Dal Corso                    Tel. 0039-040-3787428
SISSA, Via Beirut 2/4               Fax. 0039-040-3787528
34014 Trieste (Italy)               e-mail: dalcorso at sissa.it