Change the color of specific gridline ggplot2 R script

0
votes

I have been trying to plot my 2-way interactive boxplots with ggplot2 and I was quite successful so far. Since I want the viewers to be able to compare the results based on them being negative and positive, I figured it would be neat to change the color of gridline at y=0; I searched a lot and I could not find any similar questions online. I even tried to manipulate the theme() feature myself.

I realized that the the minor gridlines cannot have labels and in case minor and major gridlines overlap, the line would inherit the theme features for the major lines. So, I could transform y=0 into a minor gridline without any labels or I could have y=0 as a major line but then the rest would be minor and thus, no labels are shown. (I obviously don't want all of the gridlines to be darker than y=0 because then it would not be significantly obvious.

The chart below shows the first scenario:

y=0 as a minor line but no labels

p + scale_y_continuous(expand = c(0, 0), breaks = c(-40, 40, -20, 20), minor_breaks = c(0))  
  +theme(panel.grid.minor.y = element_line(color="#54545B"))

And This chart shows the second scenario:

y=0 as major line but then the rest are minors and thus, no labels!

p + scale_y_continuous(expand = c(0, 0), breaks = seq(-40, 40, 40), minor_breaks = c(-20, 20)) 
  +theme(panel.grid.major.y = element_line(color="#54545B"))

Did anyone have the same issue? Is there any good solution to do this without using Photoshop (manually manipulating the charts!)?

2
rggplot2boxplotrscript
I'd use geom_hline(yintercept = 0) instead. – neilfws

2 Answers

1
votes

I'd use geom_hline(yintercept = 0). Specify it before geom_boxplot so as it is in the layer underneath the boxes.

Example data:

set.seed(123)
data.frame(x = rep(c("A", "B"), each = 10), 
           y = sample(-100:100, 20, replace = TRUE)) %>% 
ggplot(aes(x, y)) + 
  geom_hline(yintercept = 0) + 
  geom_boxplot()

enter image description here

0
votes

In addition to neilfws's response, add your colour code to the color argument in geom_hline(). I've edited neilfw's code to include this:

set.seed(123)
data.frame(x = rep(c("A", "B"), each = 10), 
           y = sample(-100:100, 20, replace = TRUE)) %>% 
  ggplot(aes(x, y)) + 
  geom_hline(yintercept = 0,color="#54545B") + 
  geom_boxplot()