mne.io.read_raw_curry for Curry 9

Support & Discussions
,
1
  • MNE version: 1.8
  • operating system: macOS

Hi all,

I’m trying to load EEG files recorded with Curry 9 and got this error:
"The following required files cannot be found: ['info' 'labels']."

It seems that the function does not support Curry 9, currently interprets a .cdt.ceo file as part of a Curry 8 dataset. However, in Curry 9 , the info and labels are stored in a .cdt.dpo file instead of the .cdt.dpa that was used previously.

"Curry 8": {
        "info": ".cdt.dpa",
        "data": ".cdt",
        "labels": ".cdt.dpa",
        "events_cef": ".cdt.cef",
        "events_ceo": ".cdt.ceo",
        "hpi": ".cdt.hpi"
}

As a temporary workaround, I modified the code to use .cdt.dpo instead of .cdt.dpa , which loads the file and the output looks okay at a first glance. However, this is still under the Curry 8 dictionary and I haven’t done a more thorough check.

Has someone tried a different approach or already Preformatted textcontributed to allow compatibility with Curry9?

Thank you in advance for your help!

All the best,
Claudia

2

we’ve recently adapted Neuroscan’s reader software for distribution, and are working on incorporating that into MNE-Python. It’s not yet merged though, sadly. You can follow / comment on the linked pull request to stay abreast of when it’s available.