python-neo Spike data support for OpenEphysBinaryRawIO

Hey,

I was looking for a neo IO class which supports a) lazy loading / memory efficient loading, and b) reading AND writing neo blocks. The openephysbinaryrawio class comes close, but is still missing support for spiking data (according to this line), which I definitely need. Is there any existing progress on this front? Would this issue count as a "user request" for the feature? 😄

Jan 12 '22 15:01 shashwatsridhar

Hi Shash! The good news is, there is now a dataset available with spiking data (see https://gin.g-node.org/NeuralEnsemble/ephy_testing_data/pulls/71). The bad news is that conceptually the RawIO is not designed for writing data to arbitrary formats, for this step you still need to write code outside of the OpenEphysBinaryRawIO. Luckily this format is pretty simple, so it shouldn't be too complicated. I will see when I find time to continue working on this. I already started a first draft implementation here: https://github.com/JuliaSprenger/python-neo/tree/add/openephys_spikes Feel free to test and extend this.

Jan 12 '22 16:01 JuliaSprenger

Noice, thanks Julia! I'll test and expand on your fork when I find some time.

Jan 12 '22 18:01 shashwatsridhar

Hi @shashwatsridhar! Any news on this from your side?

Mar 03 '22 13:03 JuliaSprenger

Hi Shash, you can also extend the OpenEphysIO that inherit from OpenEphyRawIoand add some writting ability.

Mar 03 '22 14:03 samuelgarcia

Hey @JuliaSprenger and @samuelgarcia, I have been neck deep in experiments that leave me with barely any time to code. This will be the case at least until May, so that's probably the earliest that I can make any significant contributions on this front :(

Mar 10 '22 14:03 shashwatsridhar