[Pw_forum] query on e-ph tutorial

GAO Zhe flux_ray12 at 163.com
Tue Jan 31 06:23:36 CET 2012


This problem happens during diagonalize the matrix. At least there are two methods to solve this problem: Firstly, changing the diagonalization method to CG, which requires to modify the PH code; secondly, increasing empty bands by "nbnd" parameter in the input file of pw.x.


--
GAO Zhe
CMC Lab, Materials Science & Engineering Department,
Seoul National University, South Korea
 



At 2012-01-30 22:48:10,"Ajit Vallabhaneni" <avallabh at purdue.edu> wrote:
>Dear Paolo,
>
>                Sorry for sending my response to your personal email. I didn't see it before.Here is my question.
>
>             I tried example 7 with the same parameters given. But the al.elph.in (with nq1=nq2=nq3=4) was terminated with the following message: 
>
>
>Band Structure Calculation
>     Davidson diagonalization with overlap
>
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>     from cdiaghg : error #        19
>     diagonalization (ZHEGV*) failed
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>
>     stopping ...
>
>
>        Is this because of the small mesh size (16x16x16)?
>
>Thanks
>Ajit
>_______________________________________________
>Pw_forum mailing list
>Pw_forum at pwscf.org
>http://www.democritos.it/mailman/listinfo/pw_forum
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20120131/ec7f58ba/attachment.html>


More information about the users mailing list