How to add horizontal lines in different facets for 2x2 arrangements using ggplot2?

Question

I have a database that was plotted and separated by facets. The facets of the first row (row a) need a horizontal line at 0.5, while the facets of the second row (row b) need a line at 1. I have already partially achieved my goal following this example. However, the horizontal lines at 0.5 and 1 appear in all facets.

library(ggplot2)

#Data
values <- c(0.4, 0.6, 0.9, 1.1)
Column <- c("UW", "LW", "UW", "LW")
Row <- c("a", "a", "b", "b")
DF <- data.frame(Row, Column, values)
DF$Column <- factor(DF$Column,
                 levels = c("UW", "LW"))
DF$Row <- factor(DF$Row,
                 levels = c("a", "b"))

#Auxiliar DF
Target <- c("a", "b")
Lines <- c(0.5, 1)
Lines_in_plot <- data.frame(Target, Lines)
Lines_in_plot$Target <- factor(Lines_in_plot$Target)

#Plot
ggplot(data = DF, aes(y = values)) +
  geom_bar() +
  facet_grid(Row~Column,
             scales = "free") +
  geom_hline(data = Lines_in_plot,
             yintercept = Lines,
             linetype = "dashed",
             color = "red")

This MWE runs but displays the following warning message:

geom_hline(): Ignoring `data` because `yintercept` was provided.

What happens if you reduce the data=Lines_in_plot part from geom_hline — CALUM Polwart, Mar 16 '21 at 22:41

score 1 · Accepted Answer · answered Mar 16 '21 at 22:53

For the intercept to show up in specific panels, you'll need to have the Row referred to in facet_grid available as a variable inside Lines_in_plot. You'll also want to put yintercept inside aes so that ggplot knows to refer to the Lines_in_plot data for that yintercept.

...
#Auxiliar DF
Row <- c("a", "b")
Lines <- c(0.5, 1)
Lines_in_plot <- data.frame(Row, Lines)
Lines_in_plot$Row <- factor(Lines_in_plot$Target)

#Plot
ggplot(data = DF, aes(y = values)) +
  geom_bar() +
  facet_grid(Row~Column,
             scales = "free") +
  geom_hline(data = Lines_in_plot,
             aes(yintercept = Lines),
             linetype = "dashed",
             color = "red")

Ah! @Jon Spring has the same solution as me but types faster!! — CALUM Polwart, Mar 16 '21 at 22:57

score 1 · Answer 2 · answered Mar 16 '21 at 22:56

Here is your solution:


library(ggplot2)

#Data
values <- c(0.4, 0.6, 0.9, 1.1)
Column <- c("UW", "LW", "UW", "LW")
Row <- c("a", "a", "b", "b")
DF <- data.frame(Row, Column, values)
DF$Column <- factor(DF$Column,
                 levels = c("UW", "LW"))
DF$Row <- factor(DF$Row,
                 levels = c("a", "b"))

#Auxiliar DF
Row <- c("a", "b")
Lines <- c(0.5, 1)
Lines_in_plot <- data.frame(Row, Lines)
Lines_in_plot$Row <- factor(Lines_in_plot$Row)

#Plot
ggplot(data = DF, aes(y = values)) +
  geom_bar() +
  facet_grid(Row~Column,
             scales = "free") +
  geom_hline(data = Lines_in_plot,
             aes(yintercept = Lines),
             linetype = "dashed",
             color = "red")

Two changes:

Move y-intercept into the aesthetic
Rename target to Row to match the Facet so it knows what to do with them

How to add horizontal lines in different facets for 2x2 arrangements using ggplot2?

2 Answers2