Introduction to ggplot2

This is a powerful library, which has revolutionized the way graphics are plotted by incorporating a "grammar of graphics", which allows different features to combined and overlaid. This approach has inspired a ports to Python and other languages, such as Julia.

Let's start by loading Fisher's classic data set on three species of Iris.

library(ggplot2)
data(iris)
head(iris)

	Sepal.Length	Sepal.Width	Petal.Length	Petal.Width	Species
1	5.1	3.5	1.4	0.2	setosa
2	4.9	3	1.4	0.2	setosa
3	4.7	3.2	1.3	0.2	setosa
4	4.6	3.1	1.5	0.2	setosa
5	5	3.6	1.4	0.2	setosa
6	5.4	3.9	1.7	0.4	setosa

In this data set we have morphometric data on flower shape from three species. In ggplot, the main plotting command is ggplot, which takes a data frame and sets up an aesthetic, which is the mapping between the variables. This mapping can then be visualized by applying a geometry. E.g.,

ggplot(iris, aes(x=Sepal.Length, y=Sepal.Width))+geom_point()

In this case, we use set up a relationship between Sepal.Length and Sepal.Width using the aes function and then plotted it using a scatter plot (geom_point function in ggplot). We can easily supply additional aesthetics, such as species:

ggplot(iris, aes(x=Sepal.Length, y=Sepal.Width, color=Species))+geom_point()

In this example, we added an aesthetic that color-codes each species, though we could also use something like shape

ggplot(iris, aes(x=Sepal.Length, y=Sepal.Width, shape=Species))+geom_point()

You can find out more about available aesthetics by asking R directly:

?geom_point

geom_point {ggplot2}

R Documentation

Points, as for a scatterplot

Description

The point geom is used to create scatterplots.

Usage

geom_point(mapping = NULL, data = NULL, stat = "identity",
  position = "identity", na.rm = FALSE, ...)

Arguments

mapping	The aesthetic mapping, usually constructed with aes or aes_string. Only needs to be set at the layer level if you are overriding the plot defaults.
data	A layer specific dataset - only needed if you want to override the plot defaults.
stat	The statistical transformation to use on the data for this layer.
position	The position adjustment to use for overlapping points on this layer
na.rm	If FALSE (the default), removes missing values with a warning. If TRUE silently removes missing values.
...	other arguments passed on to layer. This can include aesthetics whose values you want to set, not map. See layer for more details.

Details

The scatterplot is useful for displaying the relationship between two continuous variables, although it can also be used with one continuous and one categorical variable, or two categorical variables. See geom_jitter for possibilities.

The bubblechart is a scatterplot with a third variable mapped to the size of points. There are no special names for scatterplots where another variable is mapped to point shape or colour, however.

The biggest potential problem with a scatterplot is overplotting: whenever you have more than a few points, points may be plotted on top of one another. This can severely distort the visual appearance of the plot. There is no one solution to this problem, but there are some techniques that can help. You can add additional information with stat_smooth, stat_quantile or stat_density2d. If you have few unique x values, geom_boxplot may also be useful. Alternatively, you can summarise the number of points at each location and display that in some way, using stat_sum. Another technique is to use transparent points, geom_point(alpha = 0.05).

Aesthetics

geom_point understands the following aesthetics (required aesthetics are in bold):

• x

• y

• alpha

• colour

• fill

• shape

• size

See Also

scale_size to see scale area of points, instead of radius, geom_jitter to jitter points to reduce (mild) overplotting

Examples

p <- ggplot(mtcars, aes(wt, mpg))
p + geom_point()

# Add aesthetic mappings
p + geom_point(aes(colour = qsec))
p + geom_point(aes(alpha = qsec))
p + geom_point(aes(colour = factor(cyl)))
p + geom_point(aes(shape = factor(cyl)))
p + geom_point(aes(size = qsec))

# Change scales
p + geom_point(aes(colour = cyl)) + scale_colour_gradient(low = "blue")
p + geom_point(aes(size = qsec)) + scale_size_area()
p + geom_point(aes(shape = factor(cyl))) + scale_shape(solid = FALSE)

# Set aesthetics to fixed value
p + geom_point(colour = "red", size = 3)
qplot(wt, mpg, data = mtcars, colour = I("red"), size = I(3))

# Varying alpha is useful for large datasets
d <- ggplot(diamonds, aes(carat, price))
d + geom_point(alpha = 1/10)
d + geom_point(alpha = 1/20)
d + geom_point(alpha = 1/100)

# You can create interesting shapes by layering multiple points of
# different sizes
p <- ggplot(mtcars, aes(mpg, wt))
p + geom_point(colour="grey50", size = 4) + geom_point(aes(colour = cyl))
p + aes(shape = factor(cyl)) +
  geom_point(aes(colour = factor(cyl)), size = 4) +
  geom_point(colour="grey90", size = 1.5)
p + geom_point(colour="black", size = 4.5) +
  geom_point(colour="pink", size = 4) +
  geom_point(aes(shape = factor(cyl)))

# These extra layers don't usually appear in the legend, but we can
# force their inclusion
p + geom_point(colour="black", size = 4.5, show_guide = TRUE) +
  geom_point(colour="pink", size = 4, show_guide = TRUE) +
  geom_point(aes(shape = factor(cyl)))

# Transparent points:
qplot(mpg, wt, data = mtcars, size = I(5), alpha = I(0.2))

# geom_point warns when missing values have been dropped from the data set
# and not plotted, you can turn this off by setting na.rm = TRUE
mtcars2 <- transform(mtcars, mpg = ifelse(runif(32) < 0.2, NA, mpg))
qplot(wt, mpg, data = mtcars2)
qplot(wt, mpg, data = mtcars2, na.rm = TRUE)

# Use qplot instead
qplot(wt, mpg, data = mtcars)
qplot(wt, mpg, data = mtcars, colour = factor(cyl))
qplot(wt, mpg, data = mtcars, colour = I("red"))

---

[Package ggplot2 version 1.0.1 ]

Geometries can also be changed, allowing the same data set to be visualized in different ways, by changing which variables you want to plot

ggplot(iris, aes(x=Species, y=Sepal.Length))+geom_point()

You can change geometries to make different types of plots

ggplot(iris, aes(x=Species, y=Sepal.Length))+geom_boxplot()

For instance, the boxplot can summarize data into quantiles and medians, which is helpful when you have too many points to visualize. In this manner you can play around with different types of visualizations and choose which one looks best.

Adding more elements to plots

Layers

Technically geom_point is a layer. You can add them together easily to make more complex plots

ggplot(iris, aes(x=Species, y=Sepal.Length))+geom_boxplot()+geom_jitter()

In this example, we added a scatterplot (slightly jittered to prevent overplotting) on top of a boxplot summarizing the data.

Facets

Facets are panels that allow you to subdivide the data

ggplot(iris, aes(x=Sepal.Width, y=Sepal.Length))+geom_point()+facet_grid(.~Species)

In this case, we split the plot into three components oriented along the x-axis What is the effect of ggplot(iris, aes(x=Sepal.Width, y=Sepal.Length))+geom_point()+facet_grid(Species~.)

Scales

The axes can be transformed to suit your needs, e

ggplot(iris, aes(x=Sepal.Width, y=Sepal.Length))+geom_point()+scale_x_continuous(limits=c(2,3),"x axis")+scale_y_reverse()

Warning message:
: Removed 67 rows containing missing values (geom_point).

Themes

You can change visual aspects of the plot (how text is rendered, what is colored using the

ggplot(iris, aes(x=Sepal.Length, y=Sepal.Width, color=Species))+geom_point()+theme_bw()

There are, of course, lots of other packages for plotting data in R, and they may make certain kinds of plots easier.

require(graphics)
mosaicplot(Titanic, main = "Survival on the Titanic")