More on splitting vertical line colors by the mapped states on daughter edges

I just posted an update to the phytools function splitEdgeColor which allows a user plotting an object of class "simmap" in which the mapped character changes state exactly at one more nodes, to split the color of the vertical edges in a square phylogram to split colors according to the differing states of the two daughter edges (if they differ).

This update tells R what to do in the case of multifurcating nodes. What I decided is that the line segment between each daughter edge should be the color of the daughter edge north of that edge for daughters north of the parent edge; and vice versa for daughters south of the parent edge.

To clarify what that means, we can try the following 'toy' example:

library(devtools)
install_github("liamrevell/phytools",quiet=TRUE)

library(phytools)
tree<-read.newick(text="((A:1.0,B:1.0,C:1.0,D:1.0):1.0,E:2.0);")
tree<-paintBranches(tree,1,"b","a")
tree<-paintBranches(tree,3,"c","a")
tree<-paintBranches(tree,5,"b")
colors<-setNames(c("black","blue","red"),c("a","b","c"))
plot(tree,colors,split.vertical=TRUE,lwd=3)

Here are some other examples of how it looks:

set.seed(100)
tree<-rtree(n=26,tip.label=LETTERS)
tree$edge.length[(tree$edge.length<0.25)*(tree$edge[,2]>Ntip(tree))==1]<-0
di2multi(tree)->tree
b<-sample(tree$edge[,2],10)
c<-sample(setdiff(tree$edge[,2],b),10)
tree<-paintBranches(tree,b,"b","a")
tree<-paintBranches(tree,c,"c")
plot(tree,colors,split.vertical=TRUE)

Or when we have used paintSubTree:

tree<-rtree(n=26,tip.label=LETTERS)
tree$edge.length[(tree$edge.length<0.25)*(tree$edge[,2]>Ntip(tree))==1]<-0
di2multi(tree)->tree
plotTree(tree,node.numbers=TRUE)

tree<-paintSubTree(tree,33,"b","a",stem=TRUE)
tree<-paintSubTree(tree,42,"c","a",stem=TRUE)
tree<-paintSubTree(tree,41,"b",stem=TRUE)
plot(tree,colors,split.vertical=TRUE)

That's it.