Calculate cutoff energy automatically from the value recommended in the pseudopotential files

100 views
Skip to first unread message

Sylvian Cadars

unread,
Jul 8, 2021, 10:35:06 AM7/8/21
to aiidausers

Dear aiida developpers and/or users,

Is there a way in aiida-core or in aiida-vasp to extract recommended values for the cutoff energy from the pseudopotenials ? In VASP it is recommended to calculate the cutoff energy using a ratio (say 1.3) between the cutoff energy (ENCUT) and the highest ENMAX value among those given in the pseudopotential (POTCAR) files of the elements present in the system. I would like to make this procedure automatic in a workflow, where this ratio would be used as the input parameter instead of the cutoff energy, which would then be calculated based on the system composition.

Is there a way to do this, possibly in a code-independent manner ?

Thank you.
Best regards.
Sylvian

mbe...@gmail.com

unread,
Sep 2, 2021, 9:30:05 AM9/2/21
to aiidausers

Hi Sylvian,

Sorry, it seems we missed this question! 😱Apologies for the late response, hopefully it is still of some help.

We do offer ways of adding recommended cutoffs to pseudopotentials in aiida-pseudo. There the group that contains the pseudos can store these recommended cutoffs in the extras, so you can quickly get the recommended cutoffs for a set of elements or structure, see:

https://aiida-pseudo.readthedocs.io/en/latest/design.html#recommended-cutoffs

This is then used in for example the aiida-quantumespresso plugin here:

https://github.com/aiidateam/aiida-quantumespresso/blob/c49def7a238617439feb75afd0d0cf9fbb681613/aiida_quantumespresso/workflows/pw/base.py#L189-L194

I’m not sure if I would override inputs inside a workflow. The approach which we try to follow for the Quantum ESPRESSO plugin is to provide a get_builder_from_protocol() method for each work chain that provides a builder with sensible inputs, including the recommended cutoffs for the pseudo family.

aiida-pseudo currently doesn’t support VASP pseudos, however, so these would still have to be added. If you think this is the right way forward, don’t hesitate to open an issue there. We could also implement a PotcarFamily that has its own get_recommended_cutoffs method where you can specify a ratio.

Best,
Marnik

​

Sebastiaan Huber

unread,
Sep 2, 2021, 9:52:43 AM9/2/21
to aiida...@googlegroups.com
Hi Sylvian,

Just to add to Marnik's answer: `aiida-vasp` comes with an advanced system to deal with VASP's pseudopotentials `POTCAR`s.
The original developer implemented this system because VASP's pseudopotentials are licensed.
So they wanted the provenance to be exported, without exporting the actual content of the pseudopotentials.
If you export a `VaspCalculation` it will include a representation of the pseudo potential nodes used (with their metadata such as hash etc) but not the actual content, so the data can be freely shared without infringing the license.
I am not super familiar with the rest of the functionality but a quick look through the source code does not seem to provide a method to retrieve the recommended cutoff from the pseudopotential node.
If this value is stored in the POTCAR format (I don't have an example at hand, so I cannot check) the `PotcarData` plugin could potentially implement this functionality, which might be a useful feature.
This would allow you to indeed base the cutoffs off of the recommended cutoffs in the pseudos and some factor passed in the parameters.
This discussion would therefore maybe best be continued on the `aiida-vasp` repository.

HTH,

SPH
--
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/b088696c-ead3-4e97-ac5a-e7896d08aa14n%40googlegroups.com.

Reply all
Reply to author
Forward
0 new messages