title: Working with RStudio, Git, GitHub output: html_document: toc: true
toc_depth: 4
Instructions used successfully in STAT 545 in 2014 and 2015.
Where we know the OS, here's the breakdown:
half Mac, just under half Windows (various flavours), a dash of Linux
2014 | 2015 | |
---|---|---|
Mac | 16 (41%) | 36 (51%) |
Windows 7 | 9 (23%) | 13 (19%) |
Windows 8 | 12 (31%) | 9 (13%) |
Windows 10 | 0 (0%) | 7 (10%) |
Linux | 2 (5%) | 5 (7%) |
Level 1: Installation
- Register a free GitHub account. If eligible, check the offers at GitHub Education.
- Install or upgrade R and RStudio. Get current, people.
- Install Git
- Install a Git client. Optional but recommended.
Level 2: Git and GitHub configuration
Level 3: RStudio and Git and GitHub Extravaganza
- Test connection between RStudio and GitHub
- Tell RStudio where to find Git
- GitHub credential caching
- Take possession of your STAT 545 repo
Level 666: Troubleshooting
Resources
We practice what we preach! This entire website is created with the tools and workflows described above. Go ahead and watch us work.
Long-term, you should understand more about what you are doing. Rote clicking in RStudio may be a short-term survival method but won't work for long.
trygit is to (command line) Git as swirl is to R. Learn by doing, in small bites.
The book Pro Git is fantastic and comprehensive.
GitHub's own training materials may be helpful. They also point to many other resources
Find a powerful Git client if you'd like to minimize your usage of Git from the command line.
Hadley Wickham's book R Packages has an excellent chapter on the use of Git, GitHub, and RStudio in R package development. He covers more advanced usage, such as commit best practices, issues, branching, and pull requests.
It's not you, it's Git! If you're not crying already, these fictional-but-realistic Git man pages should do the trick: git-man-page-generator.