The ABINIT Users Mailing List ( CLOSED )

[Fabien Bruneval <fabien.bruneval@cea.fr>]

Dear Manuel,


1) You can find the fulfilment of the sum-rule for each momentum *q* in 
the log file of the screening run.
Changing gwencomp will change the result of the sum-rule.
2) The value of gwencomp really depends on your material, whether you 
have semicore states or not for instance...
Remember the higher the value of gwencomp, the smaller the correction.
3) For all the systems I have tried so far, the extrapolar method was 
always an improvement, sometimes a small improvement, sometimes a 
significant one. But it has been always helpful.
You should nevertheless always carry a proper band convergence study.

Enjoy the fast GW calculations!


Fabien






Manuel Pérez Jigato wrote:
> Dear all
>
> Regarding the extrapolar method, i would like to find out
>
> - Is the approximate method to determine the best value of gwencomp
> based on the completeness relation and plasma frequency, implemented 
> in abinit?
>
> - According to the paper, the suggested value is 2 Hartree.
> I guess that when the number of bands is very high, you need to 
> increase gwencomp?
>
> - Is the extrapolar method always benefitial? ie, are there no 
> requirements/conditions for a particular system to require a smaller 
> number of bands (than the standard nband using the full continuum)?
>
> thanks a lot