ibrav

[Witty Yonng]

Hello everyone,

I found that AiiDA will automatically set "ibrav" to 0 in quantumespresso, and it will cause problems in terms of symmetry. I'm wondering whether we can set "ibrav" and "celldm" as we did in running quantumespresso by hand. It will save resources and avoid many problems.

Best,

Witty

[Timo Reents]

Hi Witty,

yes, you can specify it in the parameters input namespace of the PW CalcJob.

Just include "ibrav" in the "SYSTEM" dictionary of the parameters.

I also used it some time ago but for me it did not really make a difference. QE was able to detect the same symmetries when using ibrav=0. Moreover, it was not always easy to derive the correct ibrav in a high-throughput study.

Best,

Timo

--
AiiDA is supported by the NCCR MARVEL (http://nccr-marvel.ch/), funded by the Swiss National Science Foundation, and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/).
 
Before posting your first question, please see the posting guidelines at http://www.aiida.net/?page_id=356 .
---
You received this message because you are subscribed to the Google Groups "aiidausers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to aiidausers+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/aiidausers/2e2c2393-ea57-44e4-bca3-92cd84274bbcn%40googlegroups.com.

[Witty Yonng]

Yes, I have tried set ibrav to a specific int, and it will raise error. It shows that the mapping from conventional cell to primitive cell in aiida is not the same as in quantumespresso. 

[mbe...@gmail.com]

Hi Witty,

Sorry for the slow response!

I haven’t used the feature much to be honest, but just gave it a go and setting ibrav works as Timo described. Could you send me the .xyz file of the structure you are using and which ibrav?

Best,
Marnik

​

[Timo Reents]

Hi Witty,  Hi Marnik,

I forgot to reply to your last message, Witty. I remember that I had similar problems concerning the mapping between AiiDA and QE. In some cases this was related to a simple rotation. The corresponding validation method in the "qe_tools" library compares only the components of the lattice vectors.

    https://github.com/aiidateam/qe-tools/blob/main/src/qe_tools/converters/_structure.py#L138-L156

Back then, I wrote a function to check whether the difference in the cells is related to a rotation. If this was the case, I rotated the cell accordingly so that the validation did not raise an error. Even with this additional functionalities, I was not able to set the correct ibrav for all structures. If I remember correctly, I had problems especially in the triclinic cases although these are the structures with the lowest symmetry constraints.

Best,

Timo

To view this discussion on the web visit https://groups.google.com/d/msgid/aiidausers/24551c5d-be09-4cc5-9e53-98f7598200fbn%40googlegroups.com.

[Witty Yonng]

Hi all,

I find it's a general problem. When we use pymatgen or ase to read cif files, it will get a primitive cell which is not coincide with the primitive cell defined by QuantumEspresso. If we convert the primitive cell form to the expected QE format (https://www.quantum-espresso.org/Doc/INPUT_PW.html#idm224), it will not raise error.

Best,

Witty

[mbe...@gmail.com]

Hi Witty & Timo,

Thanks for your comments! I’d have to investigate to see how to fix these issues, if possible. (your function may be of use here, Timo 🙃). Can either of you send me an example structure (in .cif/.xyz format) where setting ibrav fails?

That said, AiiDA will always print the lattice vectors with sufficient #digits that QE should pick up the symmetry. @Witty: if you have an example where this is not the case, please send it to me.

Best,
Marnik

​