Skip to content

Editing raw grid objects from a ggplot

Mara Averick edited this page Nov 2, 2018 · 7 revisions

Note: The ggplot2 wiki is no longer maintained, please use the ggplot2 website instead!

Editing raw Grid objects from a ggplot

  1. save the plot as a grob
  2. locate the particular grob in the gTree
  3. call grid.edit() with the gPath constructed from the information of step #2

A simple example goes like this, (adpated from r-help, http://finzi.psych.upenn.edu/Rhelp08/2009-May/199976.html)

  library(ggplot2)
  p <- # minimal example
  qplot(0,0, geom="blank")+ annotate("text",0,0,label="alpha")
  p
 
  g <- # store the plot as a grob
  ggplotGrob(p)
 
  # structure of the grob
  # grid.ls(g) # rather large!
 
  # find a particular grob in the gTree
  # grid.ls(grob=F, view=T)
  grid.ls(getGrob(g,"texts", grep = T))
 
  grid.edit(gPath("texts.gTree", "GRID.text"), grep=TRUE,
              gp=gpar(col="pink"))
 
  grid.edit(gPath("texts.gTree", "GRID.text"), grep=TRUE,
             label=expression(alpha^2), gp=gpar(col="red"))

Another example (removing either the x or y major/minor gridlines):

data  <- data.frame(Name = c("A","B","C","D","E","F","G","H"), y = c(100,200,300,400,200,300,400,350))
# you have to print before grob-ing
ggplot(data,aes(x=Name,y=y)) + geom_bar()

# locate the lines
 g <- # store the plot as a grob
  ggplotGrob(p)

grid.ls(g)

#notice 
#...
#panel-3-3
#    grill.gTree.383
#        panel.background.rect.374
 #       panel.grid.minor.y.polyline.376
  #      panel.grid.minor.x.polyline.378
   #     panel.grid.major.y.polyline.380
    #    panel.grid.major.x.polyline.382

grid.remove(gPath("GRID.gTree","layout","panel","grill.gTree","panel.grid.minor.x.polyline"),grep=T)
grid.remove(gPath("GRID.gTree","layout","panel","grill.gTree","panel.grid.major.x.polyline"),grep=T)
grid.remove(gPath("GRID.gTree","layout","panel","grill.gTree","panel.grid.major.y.polyline"),grep=T)

Note that ggsave() seems to discard these edits, so wrap with png()/pdf() and dev.off()

Note: The ggplot2 wiki is no longer maintained, please use the ggplot2 website instead!

Clone this wiki locally