-
Notifications
You must be signed in to change notification settings - Fork 12
Correct order of writing in add_two_photon_series
#547
Conversation
add_two_photon_series
The changes introduce on this were already merged as of #546. We should use this PR to pin the |
Leaving this open until we get a chance to talk about it with Ben... Be prepared for the possible answer 'NWB ophys (x,y) values actually mean (cols,rows)' lol. We really should just use better language on the NWB schema, so thanks for getting that conversation started. |
|
Yep, that's what I meant. Thanks for linking. TBH my own opinion is that on the schema side, it might make the most sense to a new user to refer to their underlying image as 'width' x 'height', since those are quantities they can think about without referencing the data storage (to them it may only appear as an picture on their screen, they may never even interact with it programmatically). |
Turns out NWB expects width x height, which is the roiextractors equivalent of cols x rows, hence the transposition here. |
As discussed with @CodyCBakerPhD it seems that the
write_imaging
pipeline and in particular theadd_two_photon_series
function is writing our images inverted.This PR is connected to the proper correction in
roiextractors
and the requirements consequently pinned to them:catalystneuro/roiextractors#136