Hi Stefan, For sure, correctly parsing CIF files is challenging, and the error messages we get from pymatgen for why a particular CIF could not be fully interpreted are not always very helpful. Some programs generate CIF but are just missing
ZjQcmQRYFpfptBannerStart
This Message Is From an External Sender
This message came from outside your organization.
ZjQcmQRYFpfptBannerEnd
Hi Stefan,
For sure, correctly parsing CIF files is challenging, and the error messages we get from pymatgen for why a particular CIF could not be fully interpreted are not always very helpful. Some
programs generate CIF but are just missing key information or use tags that are not recognized. OTOH, there are a lot of CIF files that do work.
If you’re having trouble with 40% of the CIF files, it seems likely that other people are having trouble with similar files. Sending a few of those to us might be able to help us figure out
if there is a systematic problem that we can fix.
From: Mangold, Stefan (IPS) <stefan.mangold@kit.edu>
Date: Sunday, September 29, 2024 at 6:21 AM
To: Matthew Newville <newville@cars.uchicago.edu>
Cc: XAFS Analysis using Ifeffit <ifeffit@millenia.cars.aps.anl.gov>
Subject: Re: Can't get access to WebAtoms via http://millenia.cars.aps.anl.gov/webatoms/
thanks for the feedback. The advantage of WebAtoms was, it was capable to process files, which larch can’t and it is much more „verbose“. With WebAtoms I got so much feedback, that I was able to „repair“ the files.
Typically about 40 % of cif-files, I get within cooperations from partners doing diffraction on the same samples are not compatible with larch.
Yes, we know this is down, and sorry for that. In fact, we talked about this at the most recent Larch developers meeting (which is more of a discussion with people
interested in influencing its development, so if anyone is interested in giving feedback on these tools or joining such a conversation, let me know). The current plan is to replace it with a web app that interfaces with Larch and pymatgen to generate Feff
(and hopefully FDMNES) inputs.
I hope we can get a prototype up soonish, but I think this entire area (CIF/XYS_Structure -> XAS inputs, and maybe related “databases of pre-calculated first shell
XAS scattering) are continually brought up as needing more attention.
So, if you or anyone else has thoughts or suggestions on these topics, we would love to hear them.