[gpaw-users] H basis with p valence

[Ask Hjorth Larsen]

Hi

I don't know whether this is the easiest way but you can do a hack:

from gpaw.atom.configurations import configurations

and then modify the 'H' entry of the configurations dictionary before
running the generator.  If things were written in a more flexible way
I think one could pass the data as parameters directly, but as far as
I can see this cannot be done the way the code works right now.

If I were you I would modify the 'H' entry by grabbing the p state of
the 'B' entry and changing it in some sort of sensible way (e.g. it
must have 0 occupation).  Then the AE solver will probably find the
correct state during generation.

Regards
Ask

2012/12/3 Georg Madsen <georg.madsen at rub.de>:
> Dear all
>
> I would like to add a valence like p-state, instead of the normal
> Gaussian p-polarization, to a H-basis. This has previously worked very
> well for Mg.
>
> I tried the script below to generate it. It crashed because there are no
> "normal" p-states in the setup. A similar script worked well for Mg
> (probably because of the semicore p-state).
>
> Does anyone have a good idea ?
>
>   Best wishes
>      Georg
>
>
> from gpaw.atom.generator import Generator
> from gpaw.atom.basis import BasisMaker
>
> symbol = 'H'
> args= {'rcut': [0.9]}
>
> generator = Generator(symbol, 'LDA')
> generator.run(write_xml=False, **args)
>
> bm = BasisMaker(generator, name='dz', run=False)
>
> basis = bm.generate(zetacount=2, polarizationcount=0,
>                      energysplit=0.1, jvalues=[0,1], rcutmax=12.0)
> basis.write_xml()
>
> _______________________________________________
> gpaw-users mailing list
> gpaw-users at listserv.fysik.dtu.dk
> https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users