Packages

For this post, we need to install and load the ggbeeswarm package.

We can install it from CRAN using install.packages("ggbeeswarm"). Then, we can load it:

# install.packages("ggbeeswarm")
library(ggbeeswarm)
library(ggplot2)

Dataset

Since beeswarm plots are made to visualize individual data points, we need a dataset that contains numerical values. Here, we’ll use the iris dataset, which is a built-in dataset in R.

We can easily load it:

data(iris)

Beeswarm with ggplot2

The package comes with 2 main functions:

• geom_beeswarm(): to create a beeswarm plot with ggplot2
• geom_quasirandom(): to create a quasirandom plot, which is a mix between a beeswarm and a jitter plot.

In this post, we’ll focus on the geom_beeswarm() function.

ggplot(iris,aes(y=Sepal.Length,x='')) +
  geom_beeswarm()

Flip the plot

We can super easily flip the plot by changing the x and y aesthetics:

ggplot(iris,aes(x=Sepal.Length,y='')) +
  geom_beeswarm()

Custom theme

We can change the color of the points and the overall theme of the plot using the theme() function:

ggplot(iris,aes(y=Sepal.Length,x='')) +
  geom_beeswarm(color='blue') +
  theme_minimal()

Change the position behavior

By default, the geom_beeswarm() function will use the swarm method to position the points. We can change this behavior using the method argument. Here are the available options:

• swarm: the default method
• compactswarm: similar to swarm, but the points are more compact
• center: the points are centered on the x-axis
• hex: the points are positioned in hexagons
• square: the points are positioned in squares

Here’s an example with center:

ggplot(iris,aes(y=Sepal.Length,x='')) +
  geom_beeswarm(method='center')

Going further

This post explains how to create and customize a beeswarm plot with ggbeeswarm.

You might also be interested in how to create a grouped beeswarm plot with ggplot2.

Related chart types

---

Violin

Density

Histogram

Boxplot

Ridgeline