Bubble chart in ggplot2

A bubble chart is a type of scatter plot where the points are scaled based on a variable.

Sample data

In order to create a bubble plot in ggplot2 you will need a data frame containing the coordinates of the points (x and y) and a variable to map the sizes of the bubbles (size). In addition, you can have a variable representing groups or labels.

# Sample data
x <- c(2, 5, 7, 3, 6, 1, 9, 2)
y <- c(2.2, 2, 1, 2, 1, 4, 1, 6)
size <- c(100, 30, 50, 250, 120, 140, 80, 36)
group <- c("A", "A", "A", "B", "C", "B", "D", "B")

# Data frame
df <- data.frame(x = x, y = y, size = size, group = group)

Bubble plots in ggplot2 with `geom_point` and `scale_size` or `scale_size_area`

In order to create a basic bubble chart in ggplot2 you will need to pass the coordinates and the variable that represents sizes to the aes function and then use the geom_point function.

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

ggplot(df, aes(x = x, y = y, size = size)) +
  geom_point()

Bubble chart in ggplot2

Resizing the bubbles

However, the previous plot can be improved making use of the scale_size function, as the default bubbles are small. This function scales the size of the bubbles to a specific range of sizes. Note that you can provide a name that will be shown in the legend of the chart.

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

ggplot(df, aes(x = x, y = y, size = size)) +
  geom_point() +
  scale_size(name = "Size", range = c(3, 15))

Bubble plot in ggplot2 with scale_size

scale_size_area

An alternative function to create bubble graphs is scale_size_area that ensures that a value of zero is mapped to a size of 0 and instead of a range a maximum size must be specified with max_size. This is more useful when the size represents a count of observations.

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

ggplot(df, aes(x = x, y = y, size = size)) +
  geom_point() +
  scale_size_area(max_size = 10)

ggplot2 bubble graph

Labelling each bubble

With the geom_text or geom_label functions you can add texts to the bubbles. Recall that the functions provide the nudge_x and nudge_y arguments to customize the position of the texts or labels.

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

ggplot(df, aes(x = x, y = y, size = size)) +
  geom_point() +
  geom_text(label = group, nudge_x = 0.5, color = "red") +
  scale_size(name = "Size", range = c(3, 15))

Label the bubbles of a ggplot2 bubble plot

Turn off the legend or change its position

It is possible to remove the legend or change its position with the legend.position argument of the theme function.

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

ggplot(df, aes(x = x, y = y, size = size)) +
  geom_point() +
  scale_size(name = "Size", range = c(3, 15)) +
  theme(legend.position = "none")

Remove or change the legend of a bubble chart in ggplot

Bubble colors

Change the color of the bubbles in ggplot2

The default black color of the bubbles can be customized through the color argument of the geom_point function, specifying a new color.

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

ggplot(df, aes(x = x, y = y, size = size)) +
  geom_point(color = "red") +
  scale_size(name = "Size", range = c(3, 15))

Fill and border color of the ggplot2 bubbles

Fill and border color of the bubbles

Note that there are some symbols that can handle both a fill color and a border color, such as the symbol 21.

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

ggplot(df, aes(x = x, y = y, size = size)) +
  geom_point(fill = "cornflowerblue", color = "black", shape = 21) +
  scale_size(name = "Size", range = c(3, 15))

Transparency of the bubbles of a bubble chart in R

Color transparency

You can also control the level of transparency of the colors with alpha, which ranges from 0 (transparent) to 1 (opaque).

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

ggplot(df, aes(x = x, y = y, size = size)) +
  geom_point(color = "red", alpha = 0.5) +
  scale_size(name = "Size", range = c(3, 15))

Bubble chart in R with color by group

Color based on a categorical variable

If you have a categorical variable that represents groups you can pass the name of that variable to color inside aes, so each group of bubbles will be filled with a different color. Use scale_color_manual or an equivalent function if you want to customize the default colors.

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

ggplot(df, aes(x = x, y = y, size = size, color = group)) +
  geom_point() +
  scale_size(name = "Size", range = c(1, 10))

Color of the bubbles based on a continuous variable in ggplot2

Color based on a continuous variable

The color of the bubbles can also be mapped to a continuous variable. In the following example we are using the y variable and hence a gradient color will be applied based on that variable. Use scale_color_gradient, scale_color_gradient2 or scale_color_gradientn to customize the color palette.

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

ggplot(df, aes(x = x, y = y, size = size, color = y)) +
  geom_point() +
  scale_size(name = "Size", range = c(1, 10))