How do I change the outline colours of a boxplot with ggplot?

2
votes

My data has values which are all the same so they are coming up as just a line in a box plot. This, however, means that I can't tell the difference between groups as the fill doesn't show up. How can I change the outlines of the boxplot to be a specific colour.

Note: I do not want all the outline colours to be the same colour, as in the next line of code:

library(dplyr)
library(ggplot2)

diamonds %>% 
      filter(clarity %in% c("I1","SI2")) %>% 
    ggplot(aes(x= color, y= price, fill = clarity))+
      geom_boxplot(colour = "blue")+
      scale_fill_manual(name= "Clarity", values = c("grey40", "lightskyblue"))+
      facet_wrap(~cut)

Instead, I would like all the plots for I1 (filled with grey40) to be outlined in black while the plots for SI2 (filled with lightskyblue) to be outlined in blue.

The following don't seem to work

geom_boxplot(colour = c("black","blue"))+

OR

scale_color_identity(c("black", "blue"))+

OR

scale_color_manual(values = c("black", "blue"))+
1
rggplot2boxplot
Add colour = clarity to aes() and use scale_color_manual(values = c("black", "blue")). – Andrey Kolyadin
And remove colour = "blue", because that would override the aes mapping. In addition, you probably want to name the color scale Clarity, too, because otherwise you get two legends. – lukeA
Perfect, thank you. – mckisa

1 Answers

3
votes

You have to:

  1. Add color = clarity to aesthetics
  2. Add scale_color_manual to ggplot object with wanted colors
  3. Name scale_color_manual the same way as scale_fill_manual to get single combined legend

Code :

library(dplyr)
library(ggplot2)
diamonds %>% 
    filter(clarity %in% c("I1","SI2")) %>% 
    ggplot(aes(x= color, y= price, fill = clarity, color = clarity))+
        geom_boxplot()+
        scale_fill_manual(name= "Clarity", values = c("grey40", "lightskyblue"))+
        scale_color_manual(name = "Clarity", values = c("black", "blue"))+
        facet_wrap( ~ cut)

Plot:

enter image description here