aeon_mecha

Project Aeon's main repository for manipulating acquired data. Includes modules for loading raw data, performing quality control on raw data, processing raw data, and ingesting processed data into a DataJoint MySQL database.

Set-up Instructions

The various set-up tools mentioned below do some combination of python version, environment, package, and package dependency management. For basic information on the differences between these tools, see this blog post.

Remote set-up on SWC's HPC

Prereqs

1. Ssh into the HPC and clone this repository to your home directory.

ssh <your_SWC_username>@ssh.swc.ucl.ac.uk
mkdir ~/ProjectAeon
cd ~/ProjectAeon
git clone https://github.com/SainsburyWellcomeCentre/aeon_mecha
cd aeon_mecha

Set-up

Ensure you stay in the ~/ProjectAeon/aeon_mecha directory for the rest of the set-up instructions, regardless of which set-up procedure you follow below.

Option 1: miniconda (python distribution) and conda (python version manager, environment manager, package manager, and package dependency manager)

• Note: mamba, a faster alternative to conda, is now installed as a module on the HPC, so the above instructions can be followed using 'mamba' instead of 'conda' if you prefer.

Option 2: pip (python package manager) and venv (python environment manager)

Local set-up

Prereqs

All commands below should be run in a bash shell (Windows users can use the 'mingw64' terminal that is included when installing git).

1. Clone this repository: create a 'ProjectAeon' directory in your home directory, clone this repository there, and cd into the cloned directory:

mkdir ~/ProjectAeon
cd ~/ProjectAeon
git clone https://github.com/SainsburyWellcomeCentre/aeon_mecha
cd aeon_mecha

Set-up

Ensure you stay in the ~/ProjectAeon/aeon_mecha directory for the rest of the set-up instructions, regardless of which set-up procedure you follow below.

Option 1: miniconda (python distribution) and conda (python version manager, environment manager, package manager, and package dependency manager)

• Note: mambaforge and mamba can be used as faster, drop-in replacements for 'miniconda' and 'conda', respectively. You can set up the Aeon environment using them, following roughly the same instructions as above. See here for more info.

Option 2: pip (python package manager) and venv (python environment manager)

Repository Contents

• .github/workflows/ : GitHub actions workflows for building the environment and running tests
• aeon/ : Source code for the Aeon Python package
  • aeon/dj_pipeline: Source code for the Aeon DataJoint MySQL database pipeline
  • aeon/io: Source code for loading raw data
  • aeon/processing: Source code for processing raw data
  • aeon/qc: Source code for quality control of raw data
  • aeon/schema: Examples of 'experiment schemas': variables that can be used to load raw data from particular experiments
• docker/ : Dockerfiles for building Docker images for the Aeon DataJoint MySQL database pipeline.
• docs/ : Documentation for the Aeon project
  • docs/devs/ : Documentation for developers
  • docs/env_setup/ : Documentation for setting up the Aeon Python environment
  • docs/examples/ : Aeon usecase examples
  • docs/using_hpc_jupyterhub.md : Instructions for using Jupyter notebooks to access Aeon data via SWC's HPC
  • docs/using_online_dashboard.md : Instructions for connecting to Aeon's online dashboard
• env_config/ : Configuration files for the Aeon Python environment
• tests/ : Unit and integration tests
  • tests/data : Data used by tests

Citation Policy

If you use this software, please cite it as below:

Sainsbury Wellcome Centre Foraging Behaviour Working Group. (2023). Aeon: An open-source platform to study the neural basis of ethological behaviours over naturalistic timescales, https://doi.org/10.5281/zenodo.8411157