This repository contains instructional materials for PSYC 5270, a course I teach on computational systems neuroscience at the University of Virginia. Students in this course learn how to process neuroscience data and model the way the brain processes, stores, and generates information. They also gain expertise in using scientific programming (specifically Python and its ecosystem of numerical and graphics libraries) to perform analyses and build models.

The current syllabus is here.

The easiest way to run the exercises is to use mybinder. This will ensure that you have the latest version of the notebooks, all the data files, and the correct versions of required libraries. Click the badge at the top of this page to start a session. Note that this site does not save your work, so you have to be careful to download your ipynb files after you finish a session.

These instructions should work on any operating system to give you a working Python environment with all the required packages.

Install the Python 3.x version of anaconda (or miniconda if you are comfortable using the shell). Create a new Python 3.x environment called comp-neurosci (using Anaconda Navigator or by running conda create -n comp-neurosci in the shell). Activate this environment (in Anaconda Explorer, click Open Terminal under the green arrow to the right of the environment; in the shell, run conda activate comp-neurosci), then run the following commands. Run them one at a time so that you can stop and fix any errors.

conda install -n comp-neurosci git numpy scipy pandas matplotlib notebook cython ipywidgets
git clone https://github.com/melizalab/comp-neurosci
cd comp-neurosci
pip install -r requirements-conda.txt
jupyter-notebook

IMPORTANT: Before each class session, you will need to update your local copy of the repository data before starting the notebook server. From the terminal, run the following commands (again, one at a time):

conda activate comp-neurosci
cd comp-neurosci
git reset --hard HEAD
git pull --rebase
jupyter-notebook

Make sure you work on a copy of the notebook (go to File/Make a Copy... in the notebook) that you name with your UVA computing ID. The git reset --hard HEAD command will erase any work that you've done on a notebook that's under version control, but any work you do in the copy will be kept.

A Docker image is provided for running the exercises locally. These instructions should work on Windows as well as Linux or OS X. After installing Docker, run the following commands in your shell (bash for Linux and OS X; PowerShell for Windows):

docker pull dmeliza/comp-neurosci:latest
docker run --rm --name comp-neurosci -p 8888:8888 -v "$PWD":/home/jovyan/work dmeliza/comp-neurosci:latest

(In PowerShell, replace "$PWD" with ${pwd}. If this is your first time using Docker you may be asked to give permission to access the hard drive and network.)

The last command will start the docker container and the jupyter server. Copy the URL at the bottom of the output into your browser and you should have access to the notebook server. Note: files saved in the work subdirectory of the server will appear in your current working directory. Any other files or changes will be lost. Conversely, if you edit a file in the current directory it will show up in the work directory.