[Phylobase-devl] patch for nodeId

[François Michonneau]

Hello -

I realized that nodeId wasn't returning some nodes ID if the tree wasn't
fully dichotomous. I propose the following patch to fix this. It
slightly changes the order in which the nodes are returned compared to
the old version. However, it didn't in the code a place where it
matters. Does anybody have another suggestion to fix the problem or have
an opinion about the order nodeId needs to return the nodes ID?

Cheers,
  -- François

BEFORE
> nodeId(geospiza, "internal")
 [1] 16 17 18 19 20 21 22 23 24 25 26 27 15
> nodeId(geospiza, "all")
 [1] 16 17 18 19 20 21 22 23 24  1  2  3  4  5  6 25  7 26 27  8  9 10
11 12 13
[26] 14 15

__________________________________________________________________________


AFTER
> nodeId(geospiza, "internal")
 [1] 15 16 17 18 19 20 21 22 23 24 25 26 27
> nodeId(geospiza, "tip")
 [1]  1  2  3  4  5  6  7  8  9 10 11 12 13 14
> nodeId(geospiza, "all")
 [1] 15 16 17 18 19 20 21 22 23 24 25 26 27  1  2  3  4  5  6  7  8  9
10 11 12

___________________________________________________________________________

NEW VERSION

setMethod("nodeId", "phylo4",
function(x,which=c("internal","tip","allnode")) {
  which <- match.arg(which)
  tipNid <- x at edge[x at edge[,2]<=nTips(x),2]
  allNid <- unique(as.vector(x at edge))
  intNid <- allNid[! allNid %in% tipNid]
  nid <- switch(which,
                internal = intNid,
                tip = tipNid,
                allnode = allNid)

  #nid <- switch(which,
  #              internal=x at edge[x at edge[,2]>nTips(x),2],
  #              tip = x at edge[x at edge[,2]<=nTips(x),2],
  #              allnode = x at edge[,2])

  return(nid[!is.na(nid)])
})