search

cortex-lab / phy

phy: interactive visualization and manual spike sorting of large-scale ephys data
BSD 3-Clause "New" or "Revised" License
311 stars 157 forks source link

mismatch between channel and peak electrode #938

Closed yger closed 4 years ago

yger commented 4 years ago

Several users (and we also noticed it) reported that in latest phy version, there seems to be a mismatch between the channel where the template is peaking the most, and the number in the ClusterView. See for example the picture below, where the mismatch is visible since the min of the template is not on channel 0.

phy

rossant commented 4 years ago

Does adding show_mapped_channels = False in the params.py fixes it? The latest version takes the channel mapping into account when displaying channel labels. I don't know if that's the problem here.

The problem is that there's an ambiguity with the channel mapping: what number should be displayed in phy, the channel label before remapping, or after? This has always been confusing and it seems like different people using different software and/or hardware don't all want to see the same thing. Which is why I added this new parameter for the time being...

yger commented 4 years ago

Ok, but this does not seem to solve my issue here, and the mapping is fairly simple on this dummy example: only 30 channels, from 0 to 29, no twisted remapping involved. I'll double check

rossant commented 4 years ago

okay in this case could you send me the dataset (no need for the raw data)?

yger commented 4 years ago

Sure. Here it is. Maybe I'm missing something :-(.

silico_0.GUI.zip

rossant commented 4 years ago

Thanks, there was indeed a bug... Fix pushed on phylib and phy, could you try again?

rossant commented 4 years ago

closing for now, reopen if that didn't fix it for you!