3 Why Git and GitHub

3.1 What is version control, Git, and Github?

Version control is the practice of tracking and managing changes to (statistical) code and other files.
Git is a version control system. It tracks what is changed in a file, when and by whom and synchronizes the changes to a central server so that multiple contributors can manage changes to the same set of files (Wilson et al., 2017).
GitHub is a hosting service on the web for Git repositories.

Reference: Wilson G, et al. Good enough practices in scientific computing. PLoS Comp Bio. 2017.

Version control eliminates the need to send code or outputs (graphics, reports) via email or share folders between collaborators. With a version control system, everyone has access to the most recent set of files.

Have you ever tried to reproduce an analysis you did 3 years ago?
Have you ever tried to reproduce someone else’s analysis?
- Their links don’t work on your computer
- You may or may not have access to the data

Because everyone has access to the same files, a project’s workflow can be set up to ensures that the analyses are reproducible for everyone.
In R, this can be as simple as hitting the “knit” button to run the analyses on anyone’s computer – no need to update the file pathways, no need to download new versions of the code or data.

With Github, you can easily view changes made to statistical code. So if you are working together, it is easy to tell what lines of code were changed, alongside downstream changes to reports or data visualizations as a result of the change to the analysis.

You can use Git to roll back to a previous version of a file at any point. To do this, you would search for the commit ID of the version you want to revert to and then use the git checkout {commit ID} command to revert to that version. This is useful if your team decides that a change made to the code was not beneficial or wants to revert back to a different strategy that was used previously.