The Grammar of Graphics

At its simplest: A statistical graphic is a mapping of variables in a

• data set to
• aes()thetic attributies of
• geom_etric objects.

These are, in my view, the most important components to think about.

Famous graphical illustration by Minard of Napolean's march to and retreat from Moscow in 1812

6 dimensions (variables) of information on a 2 dimensional page:

Top Map:

Bottom Graph:

Load this in R:

simple <- 
  data.frame(
    A = c(2,1,4,9),
    B = c(3,2,5,10),
    C = c(4,1,15,80),
    D = c("a", "a", "b", "b")
  )

Run each of these individually:

ggplot(data=simple, aes(x=A, y=B)) + geom_point()

ggplot(data=simple, aes(x=A, y=B, color=D)) + geom_point()

ggplot(data=simple, aes(x=A, y=B, shape=D)) + geom_point()

Quotes from paper:

• “Iteratively update a plot.” i.e. build it in a modular (piece-by-piece) fashion instead of in one immutable (unchangeable) piece
• “Giving us a framework to think about graphics, and hopefully shortening the distance from mind to paper.”
• “Encourages the use of graphics customized to a particular problem rather than relying on generic named graphics.” Ex: terms like “scatterplot”, “histogram”, “coxcomb” plot

Note the key words: iterative, framework, and customized.

• Steep learning curve, just like learning a new grammar
• It teaches you grammar, but not how to write poetry. i.e. it sets up a framework for making graphics, but it doesn't teach you how to make good graphics.
• “Beyond that, it seems difficult to see how to do much more algorithmically, and we need to turn to education and training.”
• “all plots are static and separate.” i.e. non-interactive.

Packages for interactive graphics include

• ggvis: in development by Wickham
• plotly: from the creators of http://www.plot.ly
• shiny: later in the course. Install the shiny package -> File -> New File -> Shiny Web App… -> Single File -> Click on “Rup App”

and many, many more…