如何为图中的不同区域设置不同的 ggplot2 背景

Question

I have this empty plot defined by two lineal equations defining three regions in a plot.

lin.func <- function(int, slope, x) {
  y = int + slope*x
}

ggplot(data.frame(x = c(32, 40)), aes(x)) + 
  stat_function(fun = lin.func, args = list(int = 78.8, slope = -1.76), 
                colour="black", size = 1) +
  stat_function(fun = lin.func, args = list(int = 105.3, slope = -2.38), 
                colour="black", size = 1)+
  theme_bw() + 
  scale_x_continuous(limits=c(28,44), breaks = c(seq(min(28),max(44), length.out = 5))) +
  scale_y_continuous(limits=c(10,25), breaks = c(seq(min(10),max(25), length.out = 4))) 

The plot looks like this:

What I need is top have a green background in the central diagonal strip, a red in the leftmost region, and blue on the rightmost region.

Is that possible to do? I was looking around and could find any solution.

Another question would be how to make the axis start right at the numbers I set in scale_x_continuous and not a little bit less and far than that..

Thanks so much in advance

Answer 1

I have to do this from time to time and it can be somewhat tedious, but the best way I've found is to define polygons. In This case, you would need to define three polygons, each identified by a unique id. See my example of the first polygon below. If you need to do this multiple times, it may be easier to define a function that creates polygons based on their vertices. I also switched to coord_cartesian because it allows you to define the polygons based on points that are not visible on the plot. For the middle background, I would suggest defining it by three points: the intersection of the two lines and two other points, each far off your graph on one line. I hope this gives you some kind of start. Good luck.

polygons <- data.frame(x= c(10, 10, 20, 40), y = c(5, 25, 78.8-1.76*c(20,40)), group = c(rep(1,4)))

ggplot(data.frame(x = c(32, 40)), aes(x)) + 
  geom_polygon(data = polygons, aes(x = x, y=y, group = group, fill= factor(group))) + 
  stat_function(fun = lin.func, args = list(int = 78.8, slope = -1.76), 
                colour="black", size = 1) +
  stat_function(fun = lin.func, args = list(int = 105.3, slope = -2.38), 
                colour="black", size = 1)+
  theme_bw() + 
  coord_cartesian(x = c(28,44), y = c(10,25))
  

Answer 2

If you want to colour the area to the left, between and to the right of two curves (or two lines, as in your case), you can use geom_ribbon with properly set xmin, xmax, ymin and ymax columns. Add to that a column with a factor of three levels (left, between, right) and set the aesthetics to colour by that column.

Answer 3

There are several ways to define the limits, some more efficient than others, but I think drawing rectangles in the background is probably easiest?

To remove the padding around the edges, use expand = c(0,0)

lin.func <- function(int, slope, x) {
  y = int + slope*x
}
miny = 10
maxy = 25
x1 = c(28,31)
x2 = c(31,40)
x3 = c(40,44)

ggplot(data.frame(x = c(32, 40)), aes(x)) + 
    geom_rect(aes(xmin=x1[1], xmax=x1[2], ymin=miny, ymax=maxy), fill='#BB0000', color=NA, alpha=0.2) +
    geom_rect(aes(xmin=x2[1], xmax=x2[2], ymin=miny, ymax=maxy), fill='#00BB00', color=NA, alpha=0.2) +
    geom_rect(aes(xmin=x3[1], xmax=x3[2], ymin=miny, ymax=maxy), fill='#0000BB', color=NA, alpha=0.2) +
  stat_function(fun = lin.func, args = list(int = 78.8, slope = -1.76), 
                colour="black", size = 1) +
  stat_function(fun = lin.func, args = list(int = 105.3, slope = -2.38), 
                colour="black", size = 1)+
  theme_bw() + 
  scale_x_continuous(expand = c(0, 0),limits=c(28,44), breaks = c(seq(min(28),max(44), length.out = 5))) +
  scale_y_continuous(expand = c(0, 0),limits=c(miny,maxy), breaks = c(seq(min(miny),max(maxy), length.out = 4)))