Unknown Potcar Warning

There is an unresolved comment from Jan 27 of this year related to an [Unknown Potcar Warning](unknown potcar warning) that was never addressed. I am receiving the same error, but with different elements. The challenge is that this error has emerged in an atomate Gibbs workflow, but the error only appears on some runs. In other words, some runs will complete successfully while others will not. Even more curious is the fact that it appears that VASP has run successfully for some period of time before the error is raised so it is not clear how this error is actually related to the issue of VASP running. I was going to attach the launcher output in two zip files for a successful and failed run, both from the same workflow, but am not allowed as a new user. I am happy to supply these two files if someone can help me diagnose the issue. I would appreciate any help here since I cannot predict when this error will occur.

Hi @ehomer,

A recent version of pymatgen resolved cases where this error was showing up incorrectly, see Release v2022.0.15 · materialsproject/pymatgen · GitHub

Does this help? Unfortunately I’m not familiar with the Gibbs workflow otherwise.

Best,

Matt

Matt,

Thanks for your response. I was using a recent install of pymatgen (and atomate) and according to pip the error occured while running pymatgen==2022.0.16. As I mentioned, this error occurs inconsistently and it appears that VASP has run partially despite the error raised. I’m not sure how to proceed at this point.

Also, as I mentioned, I am happy to share the output of a failed and successful run, but as a new user I am not allowed to upload.

Eric

I’ll note that I just used lpad to rerun the job and it finished without issues. So it is definitely an inconsistent issue.

You should be able to upload now, sorry – Discourse can be a bit over-zealous with its spam filters.

Perhaps we can look at a zip of your failed deformation directory? Please remove POTCARs before upload.

Matt, thank you for your help with this. Attached are failed and successful runs for two structures that only differed by deformation. Upon resubmission, the failed run completed successfully.

failed.zip (115.4 KB)
successful.zip (5.9 MB)

Matt, I hope you had a nice holiday. I’m just checking in to see if I could help diagnose this issue. I’m happy to do the legwork on this but I’m not sure where to begin. If you think a short live discussion would facilitate that I am happy to do so.