Mapping the value of a continuous trait at only the nodes of a plotted phylogenetic tree using contMap

Inspired by some visualizations during our recent R phylogenetics workshop at UMass-Boston…

Had a terrific first day of our @NSF funded #PCMs in #Rstats workshop at @UMassBoston with @lukejharmon. Sorry for making our participants endure this obligatory @PrincetonUPress book photoshoot! pic.twitter.com/vYX75fSvP1

— Liam Revell (@phytools_liam) March 3, 2023

… as well as by similar plots in the recent scientific literature…

Exploring the impact of hyper-aggressive honeyeater species on local bird communities reveals the intertwined effects of historical and ecological factors in shaping community structure.#communitystructure #ecology #bioticinteractionshttps://t.co/5zpoX7VdMe pic.twitter.com/4EaNLi2JVB

— J. Biogeography (@JBiogeography) March 3, 2023

… I decided to add a simple nodes_only=TRUE option to the S3 plot.contMap method for the phytools "contMap" object class. Updated R code can be seen here.

Here’s how it works.

First, load phytools (updated from GitHub).

library(phytools)
packageVersion("phytools")

## [1] '1.5.11'

Now, load a dataset. Here, I’ll use the mammal tree & body mass data from Garland et al. (1992).

data(mammal.tree)
data(mammal.data)
lnBodyMass<-setNames(log(mammal.data$bodyMass),
	rownames(mammal.data))

We can next create a "contMap" object as follows.

Here is a typical visualization.

mammal.contMap<-contMap(mammal.tree,lnBodyMass,plot=FALSE)
mammal.contMap<-setMap(mammal.contMap,palette()[c(2,4)])
plot(mammal.contMap,fsize=c(0.7,0.9),ftype="i",
	leg.txt="log(body mass)")

Now let’s test the nodes_only=TRUE option.

plot(mammal.contMap,nodes_only=TRUE,fsize=c(0.7,0.9),
	ftype="i",leg.txt="log(body mass)")

So far so good.

The implementation works with trees of different types and orientation. Here is a “fan” style tree plot.

data(anoletree)
data(anole.data)
svl<-setNames(anole.data$SVL,rownames(anole.data))
anole.contMap<-contMap(anoletree,svl,plot=FALSE)
anole.contMap<-setMap(anole.contMap,viridisLite::viridis(10))
plot(anole.contMap,nodes_only=TRUE,fsize=c(0.7,0.8),
	ftype="i",leg.txt="log(SVL)",type="fan",legend=5,
	offset=2)