Visualizations and the grammar of graphics

Overview

• Identify the importance of graphics in communicating information
• Define the layered grammar of graphics
• Demonstrate how to use layered grammar of graphics to build Minard’s graph of Napoleon’s invasion of Russia
• Practice generating layered graphics using ggplot2

Before class

• Read chapters 1-4 from R for Data Science
  • Chapters 1, 2, and 4 are very short but set the stage for the next few weeks
  • Chapter 3 is crucial - you need to read this chapter and complete some of the exercises before coming to class. Exercise solutions can be found here.
• Read The grammar of graphics
  • If you have additional time, also read/skim A Layered Grammar of Graphics
• Install and setup software
  • Definitely have R and RStudio installed and operational for class today
  • If you haven’t gotten GitHub setup yet, that’s fine but you will need it to complete homework 1

Class materials

• Why visualize data?
• How to build a complicated, layered graphic

• Practice generating layered graphics using ggplot2

• Exploring Minard’s 1812 plot with ggplot2 - a much fancier (and more complex) version

Additional resources

Graphical design

• Tufte, Edward R. The Visual Display of Quantitative Information. Classic book on statistical graphics and visualization design.
• Healey, Kieran. Data Visualization: A Practical Guide. An applied introduction to graphical design with lots of applications in ggplot2 (and many code examples).

ggplot2

• ggplot2: Elegant Graphics for Data Analysis, 2nd Edition – Hadley Wickham. Excellent resource for learning the intricacies of ggplot2.
• Documentation for ggplot2
• R Graphics Cookbook, 2nd edition – Winston Chang. A practical guide with 150 examples to generate quality statistical graphics based on the data you wish to present.
• Why do we learn the ggplot2 graphics library and not the base graphics system? David Robinson explains it well in Don’t teach built-in plotting to beginners (teach ggplot2), and follows up with a longer defense of ggplot2 in Why I use ggplot2

Useful cheatsheets

• Data visualization with ggplot2 cheat sheet
• RStudio IDE Cheat Sheet - if you don’t know what all the buttons and panels do in RStudio, this is a great decoder

What you need to do after class

• Complete the first homework assignment
• Install the rcis library from GitHub. To install the package, run the command remotes::install_github("cis-ds/rcis") in the console. We will be using data from this package in class next week.