The ABINIT Users Mailing List ( CLOSED )

[<valerio.olevano@polytechnique.fr>]

I insist that the mentioned warning should be eliminated
since the result continues to be correct even
if the parity is not a symmetry of the system.


On Wed, 3 Aug 2005, fabrizio cleri wrote:

> Hello everyone.
>
> I am running GW calculations on cubic SiC, a
> non-centrosymmetric (CS) zinclblende unit cell.
> Concerning the issue of CS in GW calculations, there
> was a number of postings in the past. However, it does
> not look clear to me whether the issue is still a
> problem or not.
>
> Apparently, the last response of V. Olevano
> (www.abinit.org/wws/arc/forum/2003-07/msg00017.html)
> of jul.2003 said:
>
> "If the system possesses or not the spatial inversion
> symmetry is not  important. The Warning
>
>  The inversion was not found in the symmetries list.
>  - outstat - WARNING:
>  GW prog. may give incorrect results !!
>
> given so far has been removed in the new versions as
> we checked, we are  sure the results are good."
>
> However, in the most recent versions (e.g. the 4.4.4)
> a similar warning is still there....:
> " The inversion was not found in the symmetries list.
>   - outkss - WARNING:
>   GW prog. may give incorrect results !! "
>
> And, effectively, if I run exactly the same input for
> Si and SiC, I get good results for the former, but
> "strange" results for the latter.
>
> What is the situation, actually? Is the CS "problem"
> still a problem, or maybe there is something else
> wrong in my calculation which I should look for?
>
> thanks for claryfing,
>
> fc
>
>
>
>
>
>
>
> ___________________________________
> Yahoo! Mail: gratis 1GB per i messaggi e allegati da 10MB
> http://mail.yahoo.it
>