Open bendichter opened 1 year ago
@magland this is what we used for spikeinterface cloud, those are the default values used by spike interface: https://github.com/catalystneuro/spikeinterface_cloud/blob/main/frontend/src/pages/spikesorting/sorterSchemas.tsx
@luiztauffer I think it would be nice to define the schema in the Python class rather than in json as in https://github.com/scratchrealm/neurobass/blob/main/python/neurobass/processing_tools/spike_sorting/Kilosort3ProcessingTool.py so that the run function can use this in the context class directly.
But your format is much simpler than the one generated by pydantic (which is what I'm using now). What about defining the parameters using the class, and then creating a layer that converts that to your format, which would be used by the front-end?
BTW, here is what the neurobass GUI presents at this time, automatically based on the Pydantic-generated schema:
@magland oh this looks good I think! I just sent you the one I did for reference of the default values, but if you already have it then it's good
I currently have no ability to indicate parameters of a spike sorting pipeline. Could we expose a webform to the user so they can change the parameters from the defaults?