/hbds5018_git

Version control lecture for WCM's HBDS 5018: Data Science I (R + Python)

Primary LanguageCSS

Getting Git Good for HBDS 5018

Version control (a lá git) has been standard practice in software engineering for decades, and its use has started to become expected in data science. In order to collaborate effectively, it is imperative to use version control tools alongside collaborative programming platforms like GitHub and Bitbucket.

A few "good enough" practices

  • Commit often; commits serve as "anchor points" in the history of a project. You can return to any point in the project history for which there is a commit.
  • Always work on new branches when adding features or developing a new line of work. Name the branches based on the work that they aim to contribute.
  • After the work on a branch is completed, create a pull request to the master (or main) branch of the repository as appropriate.
  • I'd recommend reviewing the "git flow" branching model. There are other models as well.

n.b., These are opinionated takes of mine, and many may disagree with them. As you learn more about version control best practices, you may find alternative guidance that better suit your, and your team's, work style.

Resources