Support Intan "one-file-per-signal" and "one-file-per-channel" formats.

[h-mayorquin]

As in the title. We need new releases of spikeinterface and neo for this.

Adding tests here to see if anything else is broken.

Related to issues:
#789

[h-mayorquin]

With the release of neo this is ready for review.

[CodyCBakerPhD]

I thought this was a multi-file style of the format? So shouldn't there be a new interface for Intan that takes folder_path in that case?

[h-mayorquin]

I thought this was a multi-file style of the format? So shouldn't there be a new interface for Intan that takes folder_path in that case?

Good question. The file always has a header info.rhd that contains the available streams and channels with standarized names. So the format always point ot the this info file.

[CodyCBakerPhD]

Good question. The file always has a header info.rhd that contains the available streams and channels with standarized names. So the format always point ot the this info file.

Oh, I see

Sometimes the .rhd is a header + binary content

Other times it's just .rhd header and binaries live separate?

[CodyCBakerPhD]

Changes look fine and minimal to me

But what is your suggestion on release timing then? Since tests are passing, this all works with latest SI release, etc.? So we could cut a new release of NeuroConv without issue?

[h-mayorquin]

Are you planning on making a release soon? I think that @alejoe91 can help us to merge SpikeInterface/spikeinterface#2833 before the next release and then we can eliminate the ceiling.

[h-mayorquin]

Good question. The file always has a header info.rhd that contains the available streams and channels with standarized names. So the format always point ot the this info file.

Oh, I see

Sometimes the .rhd is a header + binary content

Other times it's just .rhd header and binaries live separate?

Yes.

[CodyCBakerPhD]

Are you planning on making a release soon? I think that @alejoe91 can help us to merge SpikeInterface/spikeinterface#2833 before the next release and then we can eliminate the ceiling.

Is the timeline on that in the next 2-3 weeks? It's been a while on the NeuroConv side and the GUIDE also releases in early June so would be nice to keep things synchronized

[h-mayorquin]

I think the confusion is this.

• We don't need the latest SI release for this. This will work with the current bounds that you have in main.
• But the current bounds that you have in main will make the dev tests fail.

I guess I will just eliminate the ceiling here but then you will have the dev tests failing.

[CodyCBakerPhD]

Thanks for clarifying; any ETA on the 'fix' that would otherwise happen on dev branch testing then?

[h-mayorquin]

I am not privy to spikeinterface latest plans. I will ask Alessio next time we meet.