This Github repository contains materials for the R sessions of the 2023 Masters of Machine Learning & Data Science program boot camp, including lecture notes, slides, exercises, and recommended resources as you continue to develop your skills.
| Date | Session | Content | Lecture notes | Slides | Exercises |
|---|---|---|---|---|---|
| Mon, 9/11 | R1 & R2 | Basic syntax and data structures; Reading and writing files; Data manipulation and visualization in base R Click here for a ZIP file containing Day 1 materials only. |
Lecture notes | Slides | Exercises and Answers |
| Tues, 9/12 | R3 & R4 | RMarkdown; Loops, conditionals, and functions; R and Github | Lecture notes | Slides | Exercises and Answers |
| Thu, 9/14 | R5 & R6 | Advanced manipulation: reshaping, merging, dplyr, and data.table | Lecture notes, pt 1 Lecture notes, pt 2 |
Slides, pt 1 Slides, pt 2 |
Exercises and Answers |
| Mon, 9/18 | R7 | Data visualization with ggplot | Lecture notes | Slides | Exercises |
| Mon, 9/18 | R8 | Final exercise | Exercise |
Here are some resources for learning R:
- Library of free learning resources, from NUIT RCS
- Compilation of resources often used by social scientists, from Emily Farris
learnrinteractive tutorials, from RStudio- Dataquest (coursework license available from NUIT)
NUIT's command line workshop includes some exercises and a well-curated list of resources. I'll add some commentary on what resources I think are useful for data analysts who need a working understanding of Bash, shells, and Unix.
Software Carpentry's Unix Shell course is a useful and matter-of-fact introduction. It probably won't convince you just how broadly useful the shell really is, though.
Learn Enough Command Line to be Dangerous, by Michael Hartl, is an excellent and realistic introduction with good exercises. I wish it had existed when I started learning this stuff. My only quibble is that the author is a macOS proselytizer, which I find unhelpful and out of step with the current landscape--Bash is for Windows users too!
Once you're comfortable with that, you can follow it up with Learn Enough Text Editor to be Dangerous. This might not be that exciting, but practicing this stuff will make you faster and more productive.
If you'd like a really accessible intro to the nuts and bolts of how all this stuff actually works, I like Julia Evans' work. She writes a blog as well as ingenious comics that teach Linux and Bash.
If you liked the DataCamp introductory R course, they also offer a free shell course.
Software Carpentry's Version Control with Git course is what we followed along with on day 2. Like their Bash course, it doesn't really introduce you to a real-world workflow, but it does help you understand the basic mechanics in a straightforward way.
Michael Hartl also wrote Learn Enough Git to be Dangerous. Just like his Bash and Text Editor tutorials, this is great for developing a practical understanding of the parts of Git that you really need to know.
NUIT also has a Git resource list.
The R materials used for R1-R5 are based on the Intro to R workshop from NUIT Research Computing Services, created by Christina Maimone. They have been expanded and modified by Kumar Ramanathan and Richard Morel. The materials on data.table used in R6 were originally developed by Ali Ehlen. The materials on ggplot2 used in R7 were originally developed by Kumar Ramanathan. Richard Morel, Ali Ehlen, and Kumar Ramanathan all jointly developed the overarching sequence of the sessions as well as the synthetic final exercise in R8.