Grammer of Graphics

Lecture 3

Dr. Elijah Meyer + Konnie Huang

Duke University
STA 199 - Fall 2022

September 5th, 2022

Checklist

– Have a GitHub account and have it connected with R (Lab 0)

– Have checked the GitHub username sheet

– Keeping up with the Prepare link on course website

– Go to the course GitHub org and find your ae-02-starwars (repo name will be suffixed with your GitHub name). – Clone the repo in your container, open the Quarto document in the repo

Goals for today

– Think about what to do (and not to do) with visualizations

– Understand the fundamentals of ggplot

– Build appropriate visualizations

– More practice with R

Announcements

– Turn in AE’s via GitHub (will go over today)

– Will turn in Lab 1 via GitHub (Thursday)

– Release HW 1 Wednesday (Turned in on Gradescope: Being set up now)

Prepare Reading Questions

https://app.sli.do/event/g6fnKCDuZ1sw8NKiZTu3Rb

Reading Highlights

– Sometimes you’ll run the code and nothing happens. Check the left-hand of your console: if it’s a +, it means that R doesn’t think you’ve typed a complete expression and it’s waiting for you to finish it.

Examining data visualization

Discuss the following for the visualization.

– What is the visualization trying to show?

– What is effective, i.e. what is done well?

– What is ineffective, i.e. what could be improved?

– What are you curious about after looking at the visualization?

04:00

Application exercise

`ae-02-starwars`

Go to the course GitHub org and find your ae-02-starwars (repo name will be suffixed with your GitHub name).
Clone the repo in your container, open the Quarto document in the repo, and follow along and complete the exercises.
Render, commit, and push your edits by the AE deadline – 3 days from today (Thursday by 11:59p ET).

Exercises 2.3.1

Match the variables to plots

– 1 categorical variable, 1 quantitative variable (often ordinal)

– 2 quantitative variables

– 1 quantitative variable

04:00

– geom_histogram

– geom_point

– geom_bar

Recap of AE

– Construct plots with ggplot().

– Layers of ggplots are separated by +s.

– The formula is (almost) always as follows:

ggplot(DATA, aes(x = X-VAR, y = Y-VAR, ...)) +
  geom_XXX()

– Aesthetic attributes of a geometries (color, size, transparency, etc.) can be mapped to variables in the data or set by the user.

– Use facet_wrap() when faceting (creating small multiples) by one variable and facet_grid() when faceting by two variables.