phylo.heatmap with missing values as NA

I just posted an update to phylo.heatmap that addresses a user request that the function permit missing values as NA. I did this first by just adding a bunch of na.rm=TRUE tags; however there remained a problem that in each instance that there was an NA in the data matrix to be plotted, the cell would be colored white. This is fine the color white is completely outside the range of plotted colors; however deceptively misleading if it is included, or at the margin, of this range!

My solution was to first identify the cells with NAs, and then draw a 'strikeout' line through each plotted cell, with white as the background. Here's my code to do so:

if(any(is.na(X))){
    ii<-which(is.na(X),arr.ind=TRUE)
    x.na<-seq(START,END,by=(END-START)/(ncol(X)-1))[ii[,2]]
    y.na<-seq(0,1,by=1/(nrow(X)-1))[ii[,1]]
    for(i in 1:length(x.na)){
        xx<-x.na[i]+c(1/2,-1/2)*(END-START)/(ncol(X)-1)
        yy<-y.na[i]+c(-1/2,1/2)*1/(nrow(X)-1)
        lines(xx,yy)
    }
}

I also added the option, in response to two different user comments (1, 2) to turn off the plotted points at the end of each tip of the tree (even though I like that feature).

Here's a demo:

library(phytools)
colors<-colorRampPalette(colors=c("white","black"))(100)
phylo.heatmap(tree,X,colors=colors,standardize=TRUE,lwd=3,
    pts=FALSE)

This same argument, pts, can also be used in dotTree and in the S3 plotting method for objects of class "cophylo".

A version of phytools with these updates can be installed using devtools:

library(devtools)
install_github("liamrevell/phytools")

The data & tree were simulated as follows:

tree<-rtree(n=20)
X<-fastBM(tree,n=12)
X[sample(1:(ncol(X)*nrow(X)),3)]<-NA ## random missing data