Closed chenhongbiao closed 1 year ago
Thanks for opening this issue, @chenhongbiao! Happy to hear of your team's interest in Spyglass. I'll see what we can do about the broken Dropbox link. This link has a smaller minirec
recording that might help you get started so you can see how we organize things. This repo handles generates nwbs from our raw recordings
I'm in the process of revising the docs/notebooks on #616. Feel free to get in touch about anything that's unclear as you set up your database
I don't recall the specifics, but you might have trouble down the road using MatNWB to save data and then PyNWB (with Spyglass) to access them. This may have been resolved. Same with DataJoint - while you can use both for a given project, you're less likely to hit issues (e.g., storing/retrieving longblob
data) if you commit to one, and we unfortunately don't have the bandwidth to offer Spyglass in MATLAB
@CBroz1 Thanks a lot for your fast and detailed reply. Cool, then I will check the structure of the mini recording NWB file with rec_to_nwb. Currently, we use MatNWB to create NWB files because Monkeylogic only has MATLAB toolbox. But once it's done, we will switch to Python for Spyglass.
Thanks again for your work on docs/notebooks. I read them with tutorial videos, which are really helpful.
Closed by #616
Hi, thanks for building this great data management and analysis framework. My lab is also trying to follow Spyglass steps (NWB + Datajoint + Spikeinterface).
But when I followed this tutorial notebooks, I found the download link of the big NWB data is broken. montague20200802.nwb, 8 GB: full recording, on DropBox
I notice that Spyglass requires specific structures in a NWB file in ordr to use it with predefined tables in Datajoint. So it would be great that you could fix the link or give us some guidelines about the requirements of a NWB file. We would like to integrate Blackrock ephy data and Monkeylogic Eyelink data into a NWB file by using matNWB.