• What is version control?

• Introduction to Git and GitHub

• How Git and GitHub can be used for reproducible data analysis

• Note: This aim of this presentation is to introduce concepts and will be non-technical. If you would like technical support, please get in touch. (Contact details are at the end of the slides.)

Version control is the practice of tracking and managing changes to files.

Does this look familiar?

├── stats-publication
│   ├── publication-analysis-code.R
│   ├── publication-analysis-code-v2.R
│   ├── publication-analysis-code-v2 NEW METHODOLOGY.R
│   ├── publication-analysis-code-v2 NEW METHODOLOGY AB-changes.R
│   ├── publication-analysis-code-final.R
│   ├── publication-analysis-code-final April 2023.R

Git is a free and open source software for version control.

To use:

• Install Git on your computer

• Initiate Git in a project folder (also called a repository)

• Record any changes you make to files (these records are called ‘commits’)

• Undo changes and revert to previous version of files (if required)

• Collaborate with others on the same project using branches

Git is used via a terminal-like tool called Git Bash, or via RStudio.

Using Git means you don’t need to save multiple copies of the same file to retain older versions. This information is stored by Git.

Commits contain information on:

• what change was made,
• when the change was made,
• why the change was made, and
• who made the change.

Git tracks changes to the content of files, not just the file as a whole. This means the information above can be recorded for changes as small as one character on one line of code.

A version controlled code repository will usually contain files for one project and can contain:

• Code (e.g. R scripts)
• Documentation (e.g. README)
• Configuration files
• …but NOT DATA!

Data can be stored within your code repository, but it should not be tracked by Git. To ensure data files (and any other files containing sensitive information) are not tracked, a .gitignore file can be used to exclude them. Alternatively, data can be stored outside of your code repository.

More information on using Git safely can be found in the Duck Book.

GitHub is a web interface for hosting version controlled code. GitHub is owned by Microsoft but is mostly free to use, although some extra features are available for a fee. Rivals also exist, e.g. GitLab, Bitbucket.

GitHub can be used to:

• Make code publicly available (although repositories can also be private)

• Facilitate code review (using ‘pull requests’)

• Manage projects using tools such as issue tracking

• Navigate Git history and view previous versions of files

• View other people’s code and collaborate

https://github.com/DataScienceScotland

• Git can be used without GitHub

• GitHub is often used as the main copy of a code repository (or ‘remote’). Analysts or developers can take a copy (or ‘clone’) of the repository from GitHub to work on locally.

• Use Git locally to track changes and regularly ‘push’ to GitHub

• Use GitHub to facilitate code review and merging of branches

• Preferable to lots of copies of the same file with various names!

• Reproducible Analytical Pipelines (RAP)

  • Reproducible: You can rerun your code as it was at any point in time.

  • Auditable: You have a record of when changes were made and why.

  • Transparent: Code is publicly available on GitHub and available for others to review or reuse.

  • Good quality: Code review is built into the GitHub workflow.

Alice Byers
RAP Developer
Data Innovation Team, Scottish Government

• Email - alice.hannah@gov.scot

• GitHub - alice-hannah