SpikeInteface-Training-Edinburgh-May24

Material for user training day of SpikeInterface

TODO: put here some practical informations (venue, lunch, ...)

schedule

Morning : overiew and tutorials 3.5h

• 9:00 - 10:00 Spike sorting overview (Pierre)
• 10:00 - 10:45 Spikeinterface overview (Alessio)
• break
• 11:00 - 11:20 Ultra quick demo with simulated (Sam)
• 11:20 - 12:30 Long and detailed on cambridge neurotech data (Sam)

Afternoon : guided hands-on 4h

Here the goal is to give details with guided hands-on.

• 14:00 - 14:15 object interaction cookbook (Ramon) Only talk (10-15)
• 14:15 - 14:30 probe handling (Zach)
• 14:30 - 14:50 preprocessing (Alessio)
• 14:50 - 15:10 drift with generated (Pierre)
• 15:10 - 15:30 postprocessing (Chris) + plot matplolib
• break
• 15:50 - 16:10 visualization (Alessio sortingview and Sam sigui)
• 16:10 - 16:30 metrics and curation (automerge ? + reove duplicated) (Aurelien)
• 16:30 - 16:40 benchamark simulated (Sam) Only talk
• 16:50 - 17:10 Export to NWB (ramon)
• break
• 16:30 - 16:50 benchmark simulated + hybrid + real with residual (Sam)
• 16:50 - 17:10

Bonus track :

• 17:30 - 18:30 Hand on on your own dataset + docker isntall or
• 17:30 - 18:30 Container installation

=======

Installation

Procedure for Windows/Apple:

If you already have anaconda/vscode installed jump to 4.

• Step 1 : If your username/login has spaces and/or weird symbols, YOU MUST create a new user with a simpler name (no spaces, no symbols). Login with such a user name.
• Step 2: download anaconda from here https://www.anaconda.com/download For Windows users (even though it is sometimes not recommended) we advise to check “Anaconda to your path”. It will help with vscode compatibility.
• Step 3 : If you do not have a code editor we advise installing vscode. https://code.visualstudio.com/download. After installation, you can add the plugins “python” and “jupyter”
• Step 4 : Go to this page https://github.com/SpikeInterface/SpikeInteface-Training-Edinburgh-May24
• Step 5 : click on "code" (green button) and download the zip. Etxract the zip.
• Step 6 : Open the anaconda prompt (a terminal).
• Step 7 : go at the correct place where the zip is etracted. This command is a tip not the good one :cd C:/users/myusername/where_the_zip_is
• Step 8 : create the python environement conda env create -f spikeinterface_environment.yml. This can take a while and download many paquets. This need bandwith. Do not expect to do this during the tutorial

Procedure for linux ubuntu/debian style:

anaconda on linux is sometimes messing a lot for new users. Standard installation using system and pip are faster and easier to manage.

• sudo apt install python3.11 python3.11-venv python3.11-dev
• mkdir ~/.virtualenvs
• python3.11 -m venv ~/.virtualenvs/si_env
• source ~/.virtualenvs/si_env/bin/activate
• pip install --upgrade pip
• pip install spikeinterface[full]
• pip install PySide6
• pip install jupyterlab
• pip install https://github.com/NeuralEnsemble/python-neo/archive/master.zip
• pip install https://github.com/SpikeInterface/spikeinterface/archive/main.zip
• pip install https://github.com/SpikeInterface/spikeinterface-gui/archive/main.zip
• pip install https://github.com/magland/sortingview/archive/main.zip
• pip install https://github.com/NeuralEnsemble/ephyviewer/archive/master.zip

Dataset

Please download datasets here : TODO link to paste here