This post explains how to avoid overlapped
text with automatic positioning in ggplot2
plots using the ggrepel package.
This post showcases
the key features of ggrepel and provides a
set of graph examples using the package.
{ggrepel}
The ggrepel package in R is an extension of the ggplot2
package, designed to simplify the process of avoiding overlapped
texts in plots.
It offers 2 main functions: geom_text_repel() and
geom_label_repel()
✍️ author → Kamil Slowikowski
📘 documentation → github
⭐️ more than 1000 stars on github
To get started with ggrepel, you can install it directly
from CRAN using the install.packages function:
The ggrepel package allows you to display labels using a
single geom and invert the functions geom_text_repel() and
geom_label_repel() without constraint:
Here’s a basic example:
library(ggplot2)
library(ggrepel)
data(iris)
ggplot(iris, aes(Sepal.Width, Sepal.Length, label=Species, color=Species)) +
geom_text_repel()
You can labels on individual data points with the
geom_label_repel() function. All points will not be
displayed if there is not enough room for them, but you can force it
with the max.overlaps argument (check below).
Example:
library(ggplot2)
library(ggrepel)
data(iris)
ggplot(iris, aes(Sepal.Width, Sepal.Length, label=Species, color=Species)) +
geom_label_repel() +
geom_point() +
theme_classic(base_size = 16) +
theme(legend.position = "none") 
The easiest way to display only some labels is to create a new column on your dataframe with a non-empty value only for the observations you’re interested in.
Example:
# Create a new columns with the label only for row 2,3 and 14
mtcars$car = ""
idx_to_label = c(2, 20, 14)
mtcars$car[idx_to_label] = rownames(mtcars)[idx_to_label]
# Display the result
ggplot(mtcars, aes(wt, mpg, label = car)) +
geom_text_repel() +
geom_point(color = ifelse(mtcars$car == "", "grey50", "red"))
The geoms provided by ggrepel try do avoid
as much as possible overlapping, with a maximum of 10 by default.
However, you can change this value if overlapping is not a problem for
you with the max.overlaps argument!
Example:
library(ggplot2)
library(ggrepel)
data(iris)
ggplot(iris, aes(Sepal.Width, Sepal.Length, label=Species, color=Species)) +
geom_text_repel(max.overlaps=nrow(iris)) + # ensure all data points are displayed
geom_point() +
theme_classic(base_size = 10) +
theme(legend.position = "none") 
Parallel coordinates plot are easy to create with
ggrepel combined with the geom_segement()
function from ggplot2
Example:
library(hrbrthemes)
df = data.frame(x1 = 1,
y1 = rnorm(10),
x2 = 2,
y2 = rnorm(10),
lab = state.name[1:10])
ggplot(df, aes(x1, y1, xend = x2, yend = y2, label = lab, col = lab)) +
geom_segment(size = 0.5) +
guides(color = "none") + # remove legend
theme_bw() + # change background color and overall theme
geom_text_repel(nudge_x = -0.2, direction = "y", hjust = "right") +
geom_text_repel(aes(x2, y2), nudge_x = 0.1, direction = "y", hjust = "left")
Thanks to the color, bg.color and
bg.r, you can change colors of your labels.
Example:
mtcars$car = rownames(mtcars)
ggplot(mtcars, aes(wt, mpg, label = car)) +
geom_point(color = "red") +
geom_text_repel(color = "white", # text color
bg.color = "grey30", # shadow color
bg.r = 0.12) # shadow radius
