[QE-users] S Matrix not positive definite error with PBEsol

[Paolo Giannozzi]

On Tue, Jun 8, 2021 at 9:32 PM John Stoppelman <jstoppelman at comcast.net>
wrote:

> Thank you for the advice, I was using an older version of the code
> (6.2.1). I don't receive the same error when running with Quantum ESPRESSO
> 6.7, so whatever was wrong must have been fixed.
>

not necessarily so, unfortunately. That error sometimes is real and
reproducible, (S matrix is _really not_ positive definite), sometimes is
due to numerical noise and is erratic.

Paolo

From,
> John Stoppelman
>
> On 06/08/2021 9:32 AM Paolo Giannozzi <p.giannozzi at gmail.com> wrote:
>
>
> If the structure is correct and pseudopotentials reasonably good, there
> isn't much that can be done. You may want to use "ParO" or "PPCG"
> diagonalization instead of "cg" that is very slow.
>
> Paolo
>
> On Mon, Jun 7, 2021 at 6:44 PM John Stoppelman < jstoppelman at comcast.net>
> wrote:
>
> Hello,
> I am running a vc-relax calculation for a crystal structure, and I'm
> trying to compare the results between PBE and PBEsol. The SCF converges
> with PBE using the default Davidson diagonalization. When I switch to
> PBEsol pseudopotentials from the PSlibary, I receive the error:
>
> Error in routine cdiaghg (797):
> S matrix not positive definite
> When I switch the diagonalization method to 'cg', the calculation
> converges as normal, however. Are there any settings I can try in order to
> use PBEsol with the Davidson diagonalization for the SCF? I have attached
> my input file.
> Thank you for any potential help.
> _______________________________________________
> Quantum ESPRESSO is supported by MaX ( www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
>
>
> --
> Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
> Univ. Udine, via delle Scienze 206, 33100 Udine, Italy
> Phone +39-0432-558216, fax +39-0432-558222
>
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users



-- 
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 206, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20210608/980817ea/attachment.html>