Is there any way to manipulate the titles of a ctree plot?

Question

Is there any way to change the title sizes of a ctree plot?

Use the following variables to quickly set up a ctree plot

a<-c(41, 45, 50, 50, 38, 42, 50, 43, 37, 22, 42, 48, 47, 48, 50, 47, 41, 50, 45, 45, 39, 45, 46, 48, 50, 47, 50, 21, 48, 50, 48, 48, 48, 46, 36, 38, 50, 39, 44, 44, 50, 49, 40, 48, 48, 45, 39, 40, 44, 39, 40, 44, 42, 39, 49, 50, 50, 48, 48, 47, 48, 47, 44, 41, 50, 47, 50, 41, 50, 44, 47, 50, 24, 40, 43, 37, 44, 32, 43, 42, 44, 38, 42, 45, 50, 47, 46, 43,
37, 47, 37, 45, 41, 50, 42, 32, 43, 48, 45, 45, 28, 44,38, 41, 45, 48, 48, 47 ,49, 16, 45, 50, 47, 50, 43, 49, 50)

X1<-c(NA,NA,NA,NA,NA,1,2,2,2,NA,2,2,2,2,2,2,2,NA,NA,2,2,2,2,NA,2,2,2,2,2,2,2,NA,NA,NA,NA,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,NA,2,2,2,2,2,2,2,2,2,2,NA,2,2,2,2,2,2,2,2,2,2,2,1,2,2,2,2,2,1,2,2,2,2,2,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,1,1,2,2,1,2,2,2)
X2<-c(NA,NA,NA,NA,NA,NA,2,2,2,NA,NA,2,2,2,2,2,2,NA,2,2,2,2,2,NA,2,2,2,2,2,2,2,NA,NA,NA,NA,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,NA,2,2,2,2,2,2,2,2,2,2,2,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,1,2,2,2,2,2,2,1,2,2,2,2,2,2,2)
X3<-c(NA,35,40,NA,10,NA,31,NA,14,NA,NA,15,17,NA,NA,16,10,15,14,39,17,35,14,14,22,10,15,0,34,23,13,35,32,2,14,10,14,10,10,10,40,10,13,13,10,10,10,13,13,25,10,35,NA,13,NA,10,40,0,0,20,40,10,14,40,10,10,10,10,13,10,8,NA,NA,14,NA,10,28,10,10,15,15,16,10,10,35,16,NA,NA,NA,NA,30,19,14,30,10,10,8,10,21,10,10,35,15,34,10,39,NA,10,10,6,16,10,10,10,10,34,10)
X4<-c(NA,NA,511,NA,NA,NA,NA,NA,849,NA,NA,NA,NA,1324,1181,832,1005,166,204,1253,529,317,294,NA,514,801,534,1319,272,315,572,96,666,236,842,980,290,843,904,528,27,366,540,560,659,107,63,20,1184,1052,214,46,139,310,872,891,651,687,434,1115,1289,455,764,938,1188,105,757,719,1236,982,710,NA,NA,632,NA,546,747,941,1257,99,133,61,249,NA,NA,1080,NA,645,19,107,486,1198,276,777,738,1073,539,1096,686,505,104,5,55,553,1023,1333,NA,NA,969,691,1227,1059,358,991,1019,NA,1216)

p<-cbind(X1,X2,X3,X4)

With the following you should then get the plot below

library(party)  
urp<-ctree(a~., data=data.frame(a,p))
plot(urp, main = "Broken Title")

How do I change the title size? I've tried the following which does nothing:

plot(urp, main = "Broken Title",cex = 1.5)
plot(urp, main = "Broken Title",cex.main = 1.5)

In fact, can I manipulate the title at all? font.main similarly does nothing. What about the titles "Node 2" and "Node 3." Is there likewise no way for me to manipulate them?

A similar question was asked here: https://stackoverflow.com/questions/18817522/ctree-changing-titles-of-inner-nodes

Answer 1

There is good news and bad news.

So the plot() function that's actually doing all the work there is party:::plot.BinaryTree. The help is available from ?plot.BinaryTree but the bad news is it doesn't have any easily accessible parameters for font formatting. However, the good news is that the function uses grid graphics to draw to the screen and you can update properties after you've created the plot.

So after you run

library(party)  
urp<-ctree(a~., data=data.frame(a,p))
plot(urp, main = "Broken Title")

You can run

for(gg in grid.ls(print=F)[[1]]) {
   if (grepl("text", gg)) {
       print(paste(gg, grid.get(gg)$label,sep=": "))
   }
}

to see all the text boxes on the plot. For example, I see

[1] "GRID.text.673: Broken Title"
[1] "GRID.text.677: X1"
[1] "GRID.text.678: p = 0.03"
[1] "GRID.text.680: 1"
[1] "GRID.text.682: phantom(0) <= 1"
[1] "GRID.text.684: phantom(0) > 1"
[1] "GRID.text.686: Node 2 (n = 8)"
[1] "GRID.text.697: Node 3 (n = 109)"

Here i see the node names, and the text they contain. Note that the node names are not the same from plot to plot and change everything you draw the same plot. But you can use this list to find the ones you want to change and update them. So if I wanted to make the main text bigger, I would run

grid.edit("GRID.text.673", gp=gpar(fontsize=20))

or If i wanted to italize the node labels i would run

grid.edit("GRID.text.686", gp=gpar(fontface=3))
grid.edit("GRID.text.697", gp=gpar(fontface=3))

and that gives