search

flatironinstitute / ironclust

Spike sorting software being developed at Flatiron Institute, based on JRCLUST (Janelia Rocket Cluster)
Apache License 2.0
28 stars 7 forks source link

using spikes clips #33

Closed ferchaure closed 4 years ago

ferchaure commented 4 years ago

Hi, Is it possible to use Ironclust in previously detected spikes? Which should be the structure of the file with the spikes to do something like that?

jamesjun commented 4 years ago

it uses PC features for sorting (trFet_spk) pope but I haven’t implemented an importing tool yet to extract features from spike clips. I think it should be easy and possible. Which format do you want to import from? Do you have an example file available to share?

-James

On Nov 11, 2019, at 10:01 AM, Fernando J. Chaure notifications@github.com wrote:

Hi, Is it possible to use Ironclust in previously detected spikes? Which should be the structure of the file with the spikes to do something like that?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

ferchaure commented 4 years ago

I'm testing a few things on single channels and for that cases I use a matrix in a .mat. You can find an example file here. I know that the structure should be more complex for polytrodes, if you show me the easiest way to import spikes to ironclus I will do the conversion.

jamesjun commented 4 years ago

I added irc2 import-clip myfile.mat command. it will output firings.mda file in the same directory where the .mat file is located. Please click on the link to learn how to parse the output.

ferchaure commented 4 years ago

Sorry I wasn't able to run the code, I though that I was doing something wrong. I tried a few things like adding a check in the line 116 of irc2 to don't enter the if when S0 already exist, etc; but I couldn't fix the the real issue.