DataJoint Element for Pose Estimation with DeepLabCut

DataJoint Element for markerless pose estimation with DeepLabCut. DataJoint Elements collectively standardize and automate data collection and analysis for neuroscience experiments. Each Element is a modular pipeline for data storage and processing with corresponding database tables that can be combined with other Elements to assemble a fully functional pipeline. This repository also provides a tutorial environment and notebooks to learn the pipeline.

Experiment Flowchart

Data Pipeline Diagram

Getting Started

• Please fork this repository.

• Clone the repository to your computer.

  git clone https://github.com/<enter_github_username>/element-deeplabcut

• Install with pip:

  pip install -e .

• Interactive tutorial on GitHub Codespaces

• Documentation

Support

• If you need help getting started or run into any errors, please open a GitHub Issue or contact our team by email at [email protected].

Interactive Tutorial

• The easiest way to learn about DataJoint Elements is to use the tutorial notebooks within the included interactive environment configured using Dev Container.

Launch Environment

Here are some options that provide a great experience:

• (recommended) Cloud-based Environment

  • Launch using GitHub Codespaces using the + option which will Create codespace on main in the codebase repository on your fork with default options. For more control, see the ... where you may create New with options....
  • Build time for a codespace is a few minutes. This is done infrequently and cached for convenience.
  • Start time for a codespace is less than 1 minute. This will pull the built codespace from cache when you need it.
  • Tip: Each month, GitHub renews a free-tier quota of compute and storage. Typically we run into the storage limits before anything else since Codespaces consume storage while stopped. It is best to delete Codespaces when not actively in use and recreate when needed. We'll soon be creating prebuilds to avoid larger build times. Once any portion of your quota is reached, you will need to wait for it to be reset at the end of your cycle or add billing info to your GitHub account to handle overages.
  • Tip: GitHub auto names the codespace but you can rename the codespace so that it is easier to identify later.

• Local Environment

  Note: Access to example data is currently limited to MacOS and Linux due to the s3fs utility. Windows users are recommended to use the above environment.

  • Install Git
  • Install Docker
  • Install VSCode
  • Install the VSCode Dev Containers extension
  • git clone the codebase repository and open it in VSCode
  • Use the Dev Containers extension to Reopen in Container (More info is in the Getting started included with the extension.)

You will know your environment has finished loading once you either see a terminal open related to Running postStartCommand with a final message of Done or the README.md is opened in Preview.

Once the environment has launched, please run the following command in the terminal:

MYSQL_VER=8.0 docker compose -f docker-compose-db.yaml up --build -d

Instructions

1. We recommend you start by navigating to the notebooks directory on the left panel and go through the tutorial.ipynb Jupyter notebook. Execute the cells in the notebook to begin your walk through of the tutorial.

2. Once you are done, see the options available to you in the menu in the bottom-left corner. For example, in Codespace you will have an option to Stop Current Codespace but when running Dev Container on your own machine the equivalent option is Reopen folder locally. By default, GitHub will also automatically stop the Codespace after 30 minutes of inactivity. Once the Codespace is no longer being used, we recommend deleting the Codespace.