Version control

Can you use cloud services?

Check whether you can use cloud services like GitHub and GitLab before you upload any code.

This includes whether the repo needs to be private (at least to begin with) and what the license terms need to be.

Git

Use Git to version control to track and manage changes to code.

Tip

For collaborative data science work, we recommend committing frequently and following the principles of trunk-based development.

GitHub

The JGI has an GitHub organisation that you can use to create collaborative repositories.

You can get access from one of the other data scientists in the team. Whilst at the University you can also access additional education benefits on GitHub.

Organisation or personal repo?

Using an organisation repo is recommended for larger or long-running projects. This is helpful for continuity if there is a change the data scientist working on the project.

For smaller projects or Ask-JGI work, a repo under your personal profile is fine.

Data

You can also version control data but this is more unusual (raw or source data is generally considered read-only).

One data version control tool is DVC.