ggplot background highlight between two treatments

Question

I want the grey background highlight to line up with the two vertical dashed lines that I've inserted into my plot. I can get the highlight to line up with "nd" and "pc", but I'd like the background to align between the two time points (pre/post treatment).

I've tried identifying the border of the highlight as x= 1.5 and 4.5, like I've done with the dashed lines, but I get "Error: Discrete value supplied to continuous scale" as a result.

Data:

> dput(LipCon)
structure(list(ChillTime = structure(c(1L, 1L, 1L, 1L, 2L, 2L, 
2L, 2L, 3L, 3L, 3L, 3L, 4L, 4L, 4L, 4L, 5L, 5L, 5L, 5L), .Label = c("nd", 
"6", "13", "24", "pc"), class = "factor"), Lipid = c(30.85703644, 
21.91554596, 25.19641351, 56.05474457, 43.02224726, 31.93075251, 
21.50358848, 28.74947619, 31.81816769, 30.68972065, 26.63482725, 
39.6305118, 29.90226926, 24.28663997, 23.10808485, 30.8010717, 
23.78336938, 18.92581619, 24.73146066, 22.48963284)), .Names = c("ChillTime", 
"Lipid"), row.names = c(NA, -20L), class = "data.frame")

My current code is as follows:

ggplot(LipCon, aes(x = ChillTime, y = Lipid)) +
  theme_bw() +
  labs(x = 'Time (weeks)', y = 'Lipid Content (%)') +
  ggtitle("Lipid Content") +
  theme(plot.title = element_text(hjust = 0.5, face='bold')) +
  geom_rect(aes(xmin = 'nd', xmax = 'pc', ymin=-Inf, ymax=Inf), alpha=0.6, fill="grey90") +
  geom_boxplot() +
  geom_vline(aes(xintercept=1.5), linetype="dashed") +
  geom_vline(aes(xintercept=4.5), linetype="dashed")

Current plot

Without data or a plot, I have no idea what you're seeing... — CPak, Jun 20 '17 at 18:05
Please read [How to make a great reproducible example in R?](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) — M--, Jun 20 '17 at 18:08
I think your question is very much related to this one: https://stackoverflow.com/questions/32642856/how-do-i-use-geom-rect-with-discrete-axis-values or this one: https://stackoverflow.com/questions/16816700/ymin-and-ymax-for-discrete-axis-in-geom-area — Steven M. Mortimer, Jun 20 '17 at 18:18

Steven M. Mortimer · Accepted Answer · 2017-06-20T21:11:49.107

Do not write it like this with the aes:

geom_rect(aes(xmin = 1.5, xmax = 4.5, ymin=-Inf, ymax=Inf), alpha=0.6, fill="grey90")

Write this instead:

geom_rect(xmin = 1.5, xmax = 4.5, ymin=-Inf, ymax=Inf, alpha=0.6, fill="grey90")

The use of aes forces your rectangle to conform to the scales that's plotting LipCon, where the x-axis is discrete for boxplots. Omitting the aes part frees you from this constraint so you can plot on the x-axis using scalars to refer to exact places on the axis.

Complete Code

ggplot(LipCon, aes(x = ChillTime, y = Lipid)) +
    theme_bw() +
    labs(x = 'Time (weeks)', y = 'Lipid Content (%)') +
    ggtitle("Lipid Content") +
    theme(plot.title = element_text(hjust = 0.5, face='bold')) +
    geom_rect(xmin = 1.5, xmax = 4.5, ymin=-Inf, ymax=Inf, alpha=0.6, fill="grey90") +
    geom_boxplot() +
    geom_vline(aes(xintercept=1.5), linetype="dashed") +
    geom_vline(aes(xintercept=4.5), linetype="dashed")

Happy to help, and welcome to Stack Overflow. If this answer or any other one solved your issue, please mark it as accepted. — Steven M. Mortimer, Jun 20 '17 at 18:40

score 0 · Answer 2 · answered Jun 20 '17 at 18:46

Use annotate with geom = "rect" to add a single rectangle to a plot instead of many rectangles on top of each other.

To avoid the issue of creating a continuous scale instead of a discrete scale when you use the rectangle layer as your first plotting layer, you can use geom_blank as the first geom layer.

ggplot(LipCon, aes(x = ChillTime, y = Lipid)) +
     geom_blank() +
     theme_bw() +
     labs(x = 'Time (weeks)', y = 'Lipid Content (%)') +
     ggtitle("Lipid Content") +
     theme(plot.title = element_text(hjust = 0.5, face='bold')) +
     annotate(geom = "rect", xmin = 1.5, xmax = 4.5, ymin = -Inf, ymax = Inf, alpha = 0.6, fill = "grey90") +
     geom_boxplot() +
     geom_vline(aes(xintercept=1.5), linetype="dashed") +
     geom_vline(aes(xintercept=4.5), linetype="dashed")

ggplot background highlight between two treatments

2 Answers2