adam2392 / tng_tvb

Workspace for my code at TNG in Marseille, France.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Theoretical Neurosciences Group

Adam Li

Visiting PhD student under Whitaker and Chateaubriand Fellowship

This repo will house the code I develop in collaboration with The Virtual Brain here at Marseille, France in INS at Aix-Marseille Universite.

Prerequisites

You will need to have installed tvb library, tvb data, and peakdetection.

git clone https://github.com/the-virtual-brain/tvb-library
git clone https://github.com/the-virtual-brain/tvb-data

Download files from here for peakdetection
https://gist.github.com/sixtenbe/1178136

Installing

To install this, just simply clone from github repo along with the prerequisite files.

git clone https://github.com/adam2392/tng_tvb

Setup a virtual env and then run

source activate venv
pip install -r requirements.txt

Running Simulations

1. Locally Running on Notebook

Modify the shell file to include your local paths to tvb lib and data, and then run it.

./launch_tvb.command

To run a jupyter notebook with tvb on path.

2. Locally Running Via Script

(Untested as of 12/19/17): Run

python runmainsim.py

Running Simulations (SLURM Cluster)

In /cluster/ directory, you can modify the script to run simulations on your own cluster. Then you can just run:

./tng_job_submissions.sh

and specify some parameters, and it will run. Or if you want to vary distances with respect to a certain implantation scheme:

./tng_distance_submissions.sh

Local Modules:

  1. peakdetect Used for peakdetection on the z time series to robustly automatically get onset/offset times for seizure.
  2. tvbsim Utility module for customizing tvb simulations. For example, can move contacts, can get the closest region-contact pairs for a given electrode layout.
  3. binmarcc For running slurm jobs on MARCC cluster at JHU.
  4. bintng For running slurm jobs on the TNG cluster at INS.

About

Workspace for my code at TNG in Marseille, France.


Languages

Language:Jupyter Notebook 99.8%Language:Python 0.2%Language:Shell 0.0%Language:Makefile 0.0%