dgbowl / yadg

yadg: yet another datagram
https://dgbowl.github.io/yadg
GNU General Public License v3.0
36 stars 14 forks source link

incompatible with eclab 11.60 #175

Open lg391 opened 4 months ago

lg391 commented 4 months ago

I had to upgrade eclab to 11.60 and now am getting errors which seem similar to when 11.50 was first released. anything I can do to fix this? thanks

ml-evs commented 4 months ago

Hi @lg391, are you able to share any specific errors, or even better, a small data file that breaks it? (you should be able to simply drag it into the comment box here)

lg391 commented 4 months ago

hi @ml-evs, yes the error is: raise NotImplementedError("Unknown parameter offset or technique dtype.") NotImplementedError: Unknown parameter offset or technique dtype. I can try to add a small data file later. seems like an issue that has happened previously when eclab has updated to 11.50 for example

ml-evs commented 4 months ago

Thanks, happy to help debug if you can share a file, and thanks for the report

PeterKraus commented 3 months ago

Hi @lg391, any chance you could share an mpr file (and ideally also a mpt file generated from it) which causes such crashes?

axnj2 commented 3 months ago

Hi @lg391, any chance you could share an mpr file (and ideally also a mpt file generated from it) which causes such crashes?

Hello, I happen to have a very similar issue and a file that fails. (I'm just not certain of the EC-Lab version since this file was send to me by a colleague).

This is supposed to be a file from an EIS I wasn't sure from the documentation if this project explicitly supports this yet.

LFP-237-form_C01.mpr.zip

From what I've been able to figure out it fails both in process_settings() when trying to extract "technique parameters" and completely in process_data()(I gave up trying to skip failed steps when the data didn't seam to be extracted).

(all of this in the file mpr.pyof course).

minimal code :

import yadg
data = yadg.extractors.extract(filetype="eclab.mpr", path="LFP-237-form_C01.mpr")
print(data)
PeterKraus commented 3 months ago

Yes, this file looks strange. It seems to be a GCPL file, not EIS. When I open it with EC-Lab v11.52, I cannot export it to mpt - only the first 800 or so rows are dumped.

If you could confirm with your colleague this file was generated with EC-Lab 11.60 or newer, and ideally also supply a mpt file (text export), that would be wonderful.

I have a very hacked-together version of yadg that can parse the above file, but I'm not quite sure it's right... If you're feeling very brave, try building yadg using the branch in the PR below.

axnj2 commented 3 months ago

I've found the settings file which indicates these files where generated with EC-Lab v11.46 so that file might not be entirely relevant to this issue how do you want to proceed @PeterKraus ?

On my side I've managed (on a windows machine) to export the files to mpt with the EC-Lab demo version 11.60. So I'm going to continue with those it should work better.

LFP-237-IMP.mps.zip

axnj2 commented 3 months ago

So the first file was in fact not and EIS experiment but a "Galvanostatic Cycling with Potential Limitation" as written in the mpt file bellow.

LFP_237_form_C01.mpt.zip

PeterKraus commented 3 months ago

OK, this is an EC-Lab 11.46 file which has a format we previously didn't know about. I'll rename the PR and merge in due course. It's unrelated to EC-Lab 11.60.

As an aside, this mpr file can only be exported to mpt properly using EC-Lab older than 11.52 (and likely also 11.50) - if you attempt to export this mpr file with a newer EC-Lab, you only get around 800 lines, instead of the 11400 or so that you should be getting.

PeterKraus commented 2 months ago

@lg391 Could you please share a mpr file that doesn't work with yadg-5.1.2? I got my hands on some mpr files generated using EC-Lab 11.60 and yadg-5.1.2 can parse them just fine.