ggplot2 tips: distinction between aesthetics and attributes

Aesthetics are defined inside aes() in ggplot syntax and attributes are outside the aes().

e.g. ggplot(data, aes(x, y, color=var1) + geom_point(size=6)

We typically understand aesthetics as how something looks, color, size etc.
But in ggplot’s world how things look is just an attribute.

I will create an imaginary data frame to apply those concepts.

library(ggplot2)
points <- 500
# Defining the Golden Angle
angle <- pi*(3-sqrt(5))
t <- (1:points) * angle
x <- sin(t/2)
y <-cos(t/2)
z <- rep(c(1,2,3,4,5,6,7,8,9,10), times=50)
w <- rep(c(1,2), times=250)
df <- data.frame(t, x, y, z, w)
# Have a look at the data
head(df)
##           t          x           y z w
## 1  2.399963  0.9320324  0.36237489 1 1
## 2  4.799926  0.6754903 -0.73736888 2 2
## 3  7.199890 -0.4424710 -0.89678282 3 1
## 4  9.599853 -0.9961710  0.08742572 4 2
## 5 11.999816 -0.2795038  0.96014460 5 1
## 6 14.399779  0.7936008  0.60843886 6 2

The dataframe we created has 3 numeric (t, x, y) variables and 2 discrete variables (z, w). With ggplot2 I can map any of the variables on my plot by defining them inside the aes().

# Make a scatter plot of points of a spiral
p <- ggplot(df, aes(x*t, y*t))
p + geom_point()

Example use of an aesthetics

By defining col=factor(z) inside aes(), I can map z to colors. So now the graph shows x, y and also values z.

# Make a scatter plot of points in a spiral
p <- ggplot(df, aes(x*t, y*t, col=factor(z)))
p + geom_point()

Each different color now represents different values of z.

Example use of an attribute

Attribute is how somethings looks. e.g. you can the points bigger by defining size=4. But it does not give any extra information about data.

# Make a scatter plot of points in a spiral
p <- ggplot(df, aes(x*t, y*t, col=factor(z)))
p + geom_point(size = 4)

Use shape as an attribute

Same goes here. I am changing how something looks like. The data point shape change to 24 which defines a empty triangle. But nothing is mapped onto it. It is just an attribute.

# Make a scatter plot of points in a spiral
p <- ggplot(df, aes(x*t, y*t, color=factor(z)))
p + geom_point(shape=24, size=4)

Here, xt, yt and factor(z) is mapped on to our graph.

Using shape as an aesthetics

By defining shape and color inside aes() I can map w and z to my plot as well.

points <- 500
# Defining the Golden Angle
angle <- pi*(3-sqrt(5))
t <- (1:points) * angle
x <- sin(t)
y <-cos(t)
z <- rep(c(1,2,3,4,5,6,7,8,9,10), times=50)
w <- rep(c(1,2), times=250)
df <- data.frame(t, x, y, z, w)

p <- ggplot(df, aes(x*t, y*t, shape=factor(w), color=factor(z)))
p + geom_point(size=3)

Spirals look nice and we got some basics of ggplot. Now let’s use it to create a pattern designer, with Shiny. Many patterns in Nature can be explained by mathematical terms, Shapes of sunflowers, dandelions or snowflakes etc.

I will tell the rest of the story in the next update. Now you can play with the app to create your patterns!

Until next time!
Serdar