Closed christian-oreilly closed 1 year ago
Thanks for this @christian-oreilly -
I'm +1 for the use of Xarray - as I recall we already use it in the methods to find bridged channels.
Do you want to work on this or do you want me to (we could also look at it during our next meeting).
Developed in the "xarray" branch.
This issue now solved on the xarray branch.
There is an error with respect to the dimensions
https://github.com/lina-usc/pylossless/blob/50f7e52f7176485ab2e028c9d900db961c226ac1/pylossless/pipeline.py#L542-L547
flag_r_ch_inds
is actually an index along the epochs, whereas it should be along the channels. This create index-out-of-bound errors when an epochs with an index greater than the number of channels get flagged. I suspect that the same error is made across multiple use ofmarks_array2flags
. I think that the use of numpy array is error prone. I suggest replacing them with XArrays which would allow to label axes and make operation less ambiguous. I also think we need to document, at a conceptual level, the different operations of the pipeline. I'll open a separate issue for this.Unfortunately, the file that triggered that error cannot be shared openly.