Warning in cryssym: Symmetry library SPGLIB not available

[GENG YUAN]

Dear CP2K users,

I am currently performing some geometry optimizations for my metallic systems. There is one warning message on my output file although my simulations successfully completed with the geometry optimization completed. Below is the warning detail:

*** WARNING in cryssym. F : 163 : : Symmetry library SPGLIB not available ***

I tried to search and what I know now is the Spglib is a library for finding and handling crystal symmetries written in C. Another thing I notice is that this warning is related to the KPOINT setting. All my simulations with KPOINT setting have this warning, while this warning is gone for my simulations without KPOINT setting.

So my questions are: Why is this warning shown in my output file, is my KPOINT setting wrong? Will it affect the final energy results and how could I get rid of this warning if my simulations still need KPOINT settting. I have attached my input file here for your reference, any suggestions are welcome!

Many thanks in advance and have a nice weekend!

Sincerely,

Geng

[Pierre-André Cazade]

Dear Geng,

You compiled cp2k without spglib. This is the error message that you are getting.

Initially, Spglib was used to build the k-point grid. It is used to identify the space group of the system and build the smallest k-point grid based on it. Spglib comes with such a feature. However, Spglib is not used, in cp2k v8, to conserve the space group throughout the geometry (GEO_OPT or CELL_OPT) optimization. This has changed in the last version of the development version (v9) where the space group is conserved during optimization. It is important to note that in neither version, the space group symmetry is used to reduce the computational cost of energy and forces evaluation.

Coming back to k-points, if you have issues with spglib, you can use the keyword "SYMMETRY F". This will disable the reduction of the number of k-points based on symmetry and the call to spglib: https://manual.cp2k.org/cp2k-8_1-branch/CP2K_INPUT/FORCE_EVAL/DFT/KPOINTS.html

Finally, as a user, I would recommend you not to use k-points. Many features are not compatible with them in cp2k, and they provide inconsistent results. Use a supercell instead. In both the cell  and topology sections use the keyword MULTIPLE_UNIT_CELL to do so.

Regards,

Pierre

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/b4437744-a691-4841-bbe2-34fc7fd64855n%40googlegroups.com.

[Tiziano Müller]

Dear Pierre,

thanks for chiming it, but to clarify for the KP-related statements:

* spglib is not needed to build the k-point grid [1]
* spglib is used to reduce the number of k-points by considering the
time-reversal symmetry (when FULL_GRID is .FALSE. and SYMMETRY is .TRUE.)
* it could be used, but this is not implemented yet, to reduce the
number of k-points further by considering the space group
* correct, some features are not yet implemented for KPs [2]

Yet, we consider KP to be ready for general use (within the
aforementioned limits). If you find inconsistent results, please get in
contact with us, either by opening an issue on [3] or email.

Best regards,
Tiziano

[1]
https://github.com/cp2k/cp2k/blob/9e72863af27f901e325a68031569daed26b32cda/src/cryssym.F#L255-L303
[2] https://github.com/cp2k/cp2k/issues/225
[3] https://github.com/cp2k/cp2k/issues/

> **** WARNING in cryssym. F : 163 : : Symmetry library SPGLIB not
> available ****

> I tried to search and what I know now is the Spglib is a library for
> finding and handling crystal symmetries written in C. Another thing
> I notice is that this warning is related to the KPOINT setting. All
> my simulations with KPOINT setting have this warning, while this
> warning is gone for my simulations without KPOINT setting.
> So my questions are: Why is this warning shown in my output file, is
> my KPOINT setting wrong? Will it affect the final energy results and
> how could I get rid of this warning if my simulations still need
> KPOINT settting. I have attached my input file here for your
> reference, any suggestions are welcome!
>
> Many thanks in advance and have a nice weekend!
> Sincerely,
> Geng
>
> --
> You received this message because you are subscribed to the Google
> Groups "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to cp2k+uns...@googlegroups.com

> <mailto:cp2k+uns...@googlegroups.com>.

> To view this discussion on the web visit
> https://groups.google.com/d/msgid/cp2k/b4437744-a691-4841-bbe2-34fc7fd64855n%40googlegroups.com

> <https://groups.google.com/d/msgid/cp2k/b4437744-a691-4841-bbe2-34fc7fd64855n%40googlegroups.com?utm_medium=email&utm_source=footer>.

>
> --
> You received this message because you are subscribed to the Google
> Groups "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to cp2k+uns...@googlegroups.com

> <mailto:cp2k+uns...@googlegroups.com>.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/cp2k/CAFnf9goHrDu9y4HFyFkuNQbukv43aqy3gwWAckgzO1pTaXcf8w%40mail.gmail.com
> <https://groups.google.com/d/msgid/cp2k/CAFnf9goHrDu9y4HFyFkuNQbukv43aqy3gwWAckgzO1pTaXcf8w%40mail.gmail.com?utm_medium=email&utm_source=footer>.

--
Tiziano Müller
University of Zurich
Department of Chemistry
Winterthurerstrasse 190
CH-8057 Zürich

Tel: +41 44 63 54234
www.chem.uzh.ch
tiziano...@chem.uzh.ch

[GENG YUAN]

Dear Pierre and Tiziano,

Thanks for your clarifications, I am now more clearer about the use of k-points and spgilb. So is it fine that I could still run simulations with the k-points setting (with the warning) if my results are consistent or reasonable? As we are currently considering two simulation protocols, one is running simulations with k-points, the other is running without k-points. 

Many thanks,

Sincerely,

Geng

[Matt W]

You can check. You should get exactly the same results for an energy calculation from a supercell calculation (Nx Ny Nz) or using a kpoint grid (Nx Ny Nz), I think (this is correct with Monkhorst-Pack?). 

Geometry optimization or MD will differ as the atoms can show a reconstruction of longer wavelength.

Matt