R ggplot2 facet_wrap reordering subfigure and set different colors for each ID label

Question

I am trying to reorder the subfigures order in ggplot facet_wrap and specify a certain color for their subfigure text label.

Here is my dataset:

  ID <- rep(c('ABC123','DEF456','GHI789','JKL012'),each = 10)
  Vref <- c((runif(10,1,2)),(runif(10,3,5)),(runif(10,6,9)),(runif(10,0,2)))
  Time <- rep(c(1:10),4)
  df <- data.frame(ID,Vref,Time)
  ggplot(df) + geom_point(aes(x=Time, y=Vref)) + facet_wrap(~ID, nrow = 2)

Now the order of subfigure is ABC123 --> DEF456 --> GHI789 --> JKL012. All text labels are in black.

My question is:

How to change the order to GHI789 --> ABC123 --> JKL012 --> DEF456 ? How to specify the color in red, green, yellow, and blue ?

Thanks.

score 1 · Answer 1 · edited May 23 '17 at 11:50

1

I recommend checking out these already answered questions:

How to change the order of facet labels in ggplot (custom facet wrap labels)

facet label font size

The ordering can be achieved through a factor, and the coloring of the facet strip is done through a theme as in the following (making the colour orange):

ggplot(df) + geom_point(aes(x=Time, y=Vref)) + facet_wrap(~ID, nrow = 2) +
theme(strip.text.x = element_text(size = 8, colour = "orange"))

For colouring each facet strip individually, I don't know of a clean way to do this. I found a similar question that discusss using custom graphics objects (grobs) to modify the background colours of the strips. Here is the link:

ggplot2: facet_wrap strip color based on variable in data set

edited May 23 '17 at 11:50

Community

1
1

answered Mar 08 '16 at 05:24

Daniel D.

134
1
4
7

Thanks for the link! About the issues of the coloring the facet strip, the theme setting is to change the element_text all in one color. However, I want to specify the colors for each individual subfigures. – Kuo-Hsien Chang Mar 08 '16 at 15:35
Thanks for clarifying. I missed that earlier. Anyway, I've updated my answer though only to point to a similar SO question. I'm not aware of a clean "aes" way to do this. – Daniel D. Mar 08 '16 at 18:29

Sandy Muspratt · Accepted Answer · 2018-07-23T23:06:30.057

One approach to specifying colours for the strip texts is to use editing functions from the grid package. Usually grid.ls() and other editing functions see just one grob, but grid.force() makes all the grobs in the ggplot visible to grid's editing functions.

library(grid)
library(ggplot2)

ID <- rep(c('ABC123','DEF456','GHI789','JKL012'),each = 10)
Vref <- c((runif(10,1,2)),(runif(10,3,5)),(runif(10,6,9)),(runif(10,0,2)))
Time <- rep(c(1:10),4)
df <- data.frame(ID,Vref,Time)

# Change order of strip texts
df$ID = factor(df$ID, levels = c("GHI789", "ABC123", "JKL012", "DEF456"))

p = ggplot(df) + geom_point(aes(x=Time, y=Vref)) + facet_wrap(~ID, nrow = 2) 

gp <- ggplotGrob(p)  # Get ggplot grob

# Get the names of the grobs
# grid.force makes all the grobs visible to grid's editing functions
names.grobs <- grid.ls(grid.force(gp))$name

# Inspect the list. The required grobs' names begin with GRID.text,
# but so do other text grobs - to do with the axes.
# Therefore, in the edit, use gPath to limit the editing to GRID.text in the strip,
# and not the GRID.text in the axes.

strip.text <- names.grobs[which(grepl("GRID.text", names.grobs))]

# Set up the colours
colour <- c("yellow", "blue", "red", "green")

# The edit
for(i in 1:4) gp = editGrob(grid.force(gp), 
                     gPath("GRID.titleGrob", strip.text[i]), 
                     grep = TRUE,    
                     gp = gpar(col = colour[i]))

# Draw it 
grid.newpage()
grid.draw(gp)

Very useful post. It helps me a lot. I believe it will also help others. — Kuo-Hsien Chang, Mar 08 '16 at 19:53

R ggplot2 facet_wrap reordering subfigure and set different colors for each ID label

2 Answers2

Linked