parcelsort

At a very early stage. Kind of like spike sorting, but intentionally overclustering. Intended for exploratory research during method development.

Prerequisites

Install and set up kachery and sortingview. Be sure to upgrade to the latest version of sortingview. Also, remember to set the FIGURL_CHANNEL environment variable.

Installation

Clone this repo.

Install parcelsort in the same conda environment as kachery and sortingview:

cd parcelsort
pip install -e .

Running the example

See devel/example1.py

Adapt this script. After running you will get a URI to the parcel sorting as well as a figURL for visualization.

Here is an example output:

https://figurl.org/fig?channel=flatiron1&figureObject=ed2456e7e5a6da94e0a7f93d07eeb415d9ac8516&label=test-parcelsort