[gpaw-users] : RuntimeError: Ghost!

[Jens Jørgen Mortensen]

On Tue, 2010-11-16 at 13:15 -0600, Yuelin Li wrote:
> Yes, I tried the empty_state flag, works for Ca, but not for Mn and O.

But you shouldn't use empty_states for Mn and O where there is no ghost
- then you can also stick with the standard radii.

Jens Jørgen

>  I 
> was able to avoid the ghost by changing the cut off radius but the 
> radius are different for the 0.5, 1.0, and 0.0 states.
> 
> As far as the trend is concerned, xas is getting popular and core states 
> will be on high demand.
> 
> Best,
> 
> Yuelin
> 
> Ph. D
> Advanced Photon Source, Argonne National Laboratory
> 630 252 7863
> http://sector7.xor.aps.anl.gov/~ylli/
> 
> On 11/16/2010 2:17 AM, Jens Jørgen Mortensen wrote:
> > On Mon, 2010-11-15 at 22:57 +0100, Mathias Ljungberg wrote:
> >> Hi Yuelin,
> >>
> >> I suspect that you don't really have a ghost state. What happens is
> >> that with the core hole you get a bound d-state, which is not so
> >> strange since the potential becomes more attractive than for Ca.
> >> Looking at generator.py I believe that the code mistakes this for a
> >> ghost state.
> >>
> >> I think the solution is to add a bound d-projector. I could only do
> >> this however, by modifying lines  77-85 in all_electron.py since the
> >> bound states seem fixed for a given element. Since the code has
> >> changed a lot recently I don't want to ruin anything, maybe someone
> >> else can take a look at how to add new bound states in a safe way?
> >
> > You can do this:
> >
> >    gpaw-setup --core-hole=1s,0.5 Ca --empty-states=3d
> >
> > or like this in your script:
> >
> >    g.run(name=name, empty_states='3d', **parameters[el])
> >
> >> A thought: since a number of people are starting to use core holes,
> >> maybe we should start to distribute core hole setups?
> >
> > That's a good idea!
> >
> > Jens Jørgen
> >
> >> Mathias
> >>
> >>
> >>
> >> On Nov 15, 2010, at 5:16 PM, Yuelin Li wrote:
> >>
> >>> Ask,
> >>>
> >>> thanks. You are suggesting that one has to re-generate the
> >>> pseud-potential for Ca? I think each element is unique in this sense
> >>> how
> >>> can parameters for other elements helpful? And what elements do you
> >>> have
> >>> in mind?
> >>>
> >>> Best,
> >>>
> >>> Yuelin
> >>>
> >>> On 11/13/2010 3:16 PM, Ask Hjorth Larsen wrote:
> >>>> Hi Yuelin
> >>>>
> >>>> On Sat, 13 Nov 2010, Yuelin Li wrote:
> >>>>
> >>>>> Hi Ask,
> >>>>>
> >>>>> Thanks. I will read the paper to understand the issue. On the other
> >>>>> hand, I
> >>>>> wonder if you have a quick fix for the current script, i.e., the
> >>>>> "some
> >>>>> other
> >>>>> parameters" that can be used. Thanks,
> >>>>>
> >>>>> Best,
> >>>>>
> >>>>> Yuelin
> >>>>
> >>>> Unfortunately the best advice I can give is to look at the setups of
> >>>> elements with similar electron configurations. It's non-trivial to
> >>>> generate pseudopotentials and PAW setups. I don't know whether the
> >>>> article really provides any solutions, but it does explain what the
> >>>> issue is.
> >>>>
> >>>> Regards
> >>>> Ask
> >>> _______________________________________________
> >>> gpaw-users mailing list
> >>> gpaw-users at listserv.fysik.dtu.dk
> >>> https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users
> >>
> >> _______________________________________________
> >> gpaw-users mailing list
> >> gpaw-users at listserv.fysik.dtu.dk
> >> https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users
> >