Hi Emannuel (and list),
Just wanted to let you know that after some testing with my own cases,
the new drop.tip appears to fix the error, so the issue is closed from
my perspective.
Cheers,
-Dave
On Tue, Jul 28, 2015 at 12:25 PM, Emmanuel Paradis
<emmanuel.para...@ird.fr> wrote:
> Hi David,
>
> I've conducted more tests and the bug was really in drop.tip and sometimes
> apparent without calling ladderize. The fix seems to work well. I have built
> ape 3.3-0.6 which is available on ape-package.ird.fr as source package.
>
> Best,
>
> Emmanuel
>
>
> Le 25/07/2015 19:33, Emmanuel Paradis a écrit :
>>
>> David,
>>
>> I was thinking about some simple tests to check that the function still
>> works correctly in other situations. It can happen that when fixing a
>> bug which occurs in a special situation in a piece of code, the code
>> then fails in other (possibly simpler) situations. In the present case,
>> the fix is the part of the code that renumbers the nodes which is thus
>> used in all cases (even if there are no node labels). I'll conduct some
>> tests later, but in the mean time if you see something strange, just
>> tell me.
>>
>> I attach the new source file.
>>
>> Best,
>>
>> Emmanuel
>>
>> Le 25/07/2015 18:54, David Bapst a écrit :
>>>
>>> Hi Emmanuel,
>>>
>>> Thank you for the fix! And, yes, I realize, its probably one of ape's
>>> most widely used functions. Perhaps what we need is a function that
>>> tests whether there is a mismatch in the node.labels, across trees
>>> that might have different sets of taxa, which will help in the future
>>> to alert us if this bug appears again. Of course, such a function
>>> would need to avoid using drop.tip entirely. I've looked and I'm not
>>> aware of any function across the entire Phylogenetics taskview that
>>> fulfills this criteria.
>>>
>>> So, I admit that's a tough order. I've made a particularly ugly first
>>> stab at such a thing, mostly depending on prop.part. Its rather
>>> clumsy, as I originally wrote it as code to fix node.labels, rather
>>> than find the mismatches. Unfortunately, at present, it misidentifies
>>> root-ward losses of clades, due to dropping of stem/out-group tips, as
>>> a mismatch between node labels. Perhaps someone else will see a
>>> cleaner, more accurate solution.
>>>
>>>
>>> ###########################################################################
>>>
>>>
>>> library(ape)
>>>
>>> testNodeLabels<-function(tree1,tree2,printMisMatch=TRUE,plot=FALSE){
>>> nlab1<-tree1$node.label
>>> if(!is.null(tree2$node.label)){
>>> nlab2<-tree2$node.label
>>> }else{
>>> nlab2<-rep(NA,Nnode(tree2))
>>> }
>>> #if tree2 has any taxa not in tree1, stop
>>> noMatch<-sapply(tree2$tip.label,function(x)
>>> !any(x==tree1$tip.label))
>>> if(any(noMatch)){
>>> stop("tree2 has taxa not found in tree1 and thus is
>>> uncomparable")
>>> }
>>> #
>>> desc1<-lapply(prop.part(tree1),function(x) tree1$tip.label[x])
>>> desc2<-lapply(prop.part(tree2),function(x) tree2$tip.label[x])
>>> #need to remove taxa not shared by one without using drop.tip
>>> taxa<-c(tree1$tip.label,tree2$tip.label)
>>> shared<-taxa[sapply(taxa,function(x)
>>> any(x==tree1$tip.label) & any(x==tree2$tip.label))]
>>> #get ndesc for desc2
>>> ndesc2<-sapply(desc2,length)
>>> #reorder desc2 and nlab2
>>> desc2<-desc2[order(ndesc2)]
>>> nlab2<-nlab2[order(ndesc2)]
>>> result<-TRUE
>>> for(i in 1:length(desc1)){
>>> target<-desc1[[i]]
>>> targetName<-nlab1[i]
>>> sharedDesc<-target[sapply(target,function(x)
>>> any(x==shared))]
>>> if(length(sharedDesc)>1){
>>> #find matches in 2
>>> matches<-sapply(desc2,function(x)
>>> all(sapply(sharedDesc,function(z) any(z==x))))
>>> }else{
>>> matches<-FALSE
>>> }
>>> #get richest match - if there is no match, get NA
>>> matchClade<-which(matches)[1]
>>> if(!is.na(matchClade)){
>>> matchName<-identical(nlab1[i],nlab2[matchClade])
>>> if(!matchName){
>>> result<-FALSE
>>> if(printMisMatch){
>>> warning(paste("\n A node with descendants:",
>>> paste0(sharedDesc,collapse=', '),
>>> "\n is labeled:",nlab1[i],
>>> "in tree1 but labeled:",
>>> nlab2[matchClade],"in tree2 \n"))
>>> }
>>> }
>>> }
>>> }
>>> #
>>> if(plot){
>>> #plot it
>>> layout(1:2)
>>> plot(tree1,show.tip.label=TRUE,use.edge.length=FALSE)
>>> nodelabels(tree1$node.label)
>>> plot(tree2,show.tip.label=TRUE,use.edge.length=FALSE)
>>> nodelabels(tree2$node.label)
>>> layout(1)
>>> }
>>> return(result)
>>> }
>>>
>>> set.seed(1)
>>> tree<-rtree(10)
>>> tree$node.label<-rep(NA,Nnode(tree))
>>> tree$node.label[1]<-"ROOT"
>>> tree$node.label[5]<-"HELLO"
>>> tree$node.label[8]<-"NOPE"
>>>
>>> tree1<-drop.tip(tree,"t2")
>>> tree2<-ladderize(tree)
>>> tree3<-drop.tip(tree2,"t2")
>>>
>>> testNodeLabels(tree,tree1)
>>> testNodeLabels(tree,tree2)
>>> testNodeLabels(tree,tree3)
>>>
>>> testNodeLabels(tree,tree3,plot=TRUE)
>>>
>>> ################################################
>>>
>>> Additionally, could you send the full revised source file? I have to
>>> admit, I always get lost looking for ape's newest source files on the
>>> ape website.
>>>
>>> Cheers,
>>> -Dave
>>>
>>> On Sat, Jul 25, 2015 at 6:20 AM, Emmanuel Paradis
>>> <emmanuel.para...@ird.fr> wrote:
>>>>
>>>> Hi David,
>>>>
>>>> Here is a fix for drop.tip (line numbers refer to the source file
>>>> drop.tip.R):
>>>>
>>>> 229,231c229,230
>>>> < ## executed from right to left, so newNb is modified before
>>>> phy$edge:
>>>> < phy$edge[sndcol, 2] <- newNb[phy$edge[sndcol, 2]] <-
>>>> < (n + 2):(n + phy$Nnode)
>>>> ---
>>>>>
>>>>> newNb[sort(phy$edge[sndcol, 2])] <- (n + 2):(n + phy$Nnode)
>>>>> phy$edge[sndcol, 2] <- newNb[phy$edge[sndcol, 2]]
>>>>
>>>>
>>>> Since this function is widely used, this requires more tests to validate
>>>> this fix.
>>>>
>>>> Best,
>>>>
>>>> Emmanuel
>>>>
>>>>
>>>> Le 18/07/2015 06:42, David Bapst a écrit :
>>>>>
>>>>>
>>>>> Hello all,
>>>>>
>>>>> Recently I noticed a complex function of mine that does some tree
>>>>> transformations was randomly scrambling node.label elements.
>>>>>
>>>>> In the course of doing so, I found this old email (below) from Rebecca
>>>>> Best in 2012, which outlined an issue that occurred when a ladderized
>>>>> tree had tips dropped. It appears that bug has reappeared in more
>>>>> recent version of ape.
>>>>>
>>>>> Here's a slightly modified version of her example code, which appears
>>>>> to still replicate a node label shuffling in ape v3.3.
>>>>>
>>>>> #############################
>>>>> require(ape)
>>>>>
>>>>> #read tree
>>>>> mytree<-read.tree()
>>>>> ((D,(E,G)1)1,((H,J)0.8,(K,(((L,M)0.5,(N,O)0.6)1,(P,(Q,R)1)1)0.7)1)1);
>>>>>
>>>>> #ladderize tree
>>>>> mytree.lad<-ladderize(mytree)
>>>>>
>>>>> #node labels display on both trees correctly
>>>>> layout(1:2)
>>>>> plot(mytree,show.node.label=TRUE)
>>>>> plot(mytree.lad,show.node.label=TRUE)
>>>>>
>>>>> #drop tips from both trees
>>>>> drop.mytree<-drop.tip(mytree,c("L","D","G"))
>>>>> drop.mytree.lad<-drop.tip(mytree.lad,c("L","D","G"))
>>>>>
>>>>> #plot both trees, node labels are incorrect for ladderized tree
>>>>> dev.new()
>>>>> layout(1:2)
>>>>> plot(drop.mytree,show.node.label=TRUE)
>>>>> plot(drop.mytree.lad,show.node.label=TRUE)
>>>>> #############################
>>>>>
>>>>> Although I'm still in the process of dismantling my own issue, so I am
>>>>> not 100% certain, I strongly believe this is the culprit in my case,
>>>>> as my script partly drops.tips from an input tree (that happens to
>>>>> always be ladderized).
>>>>>
>>>>> Cheers,
>>>>> -Dave
>>>>>
>>>>>
>>>>> ---------- Forwarded message ----------
>>>>> From: Rebecca Best <rjb...@ucdavis.edu>
>>>>> Date: Tue, Oct 2, 2012 at 12:26 AM
>>>>> Subject: [R-sig-phylo] ladderize + drop.tip = shuffled node labels
>>>>> To: r-sig-phylo@r-project.org
>>>>>
>>>>>
>>>>> Hi all
>>>>>
>>>>> I have been plotting some pruned trees recently, and have run into a
>>>>> problem using drop.tip() after ladderize(). If you ladderize() and
>>>>> then drop tips from the ladderized tree, then at least in my case the
>>>>> node labels are no longer correct. This may be an unlikely sequence of
>>>>> commands, but I thought I'd post this in case it is an easy fix, or it
>>>>> helps anyone else avoid issues.
>>>>> Thanks!
>>>>>
>>>>> Rebecca
>>>>>
>>>>> ##
>>>>>
>>>>> require(ape)
>>>>>
>>>>> #read tree
>>>>>
>>>>> mytree<-read.tree()
>>>>> ((D,(E,G)1)1,((H,J)0.8,(K,(((L,M)0.5,(N,O)0.6)1,(P,(Q,R)1)1)0.7)1)1);
>>>>>
>>>>> #ladderize tree
>>>>>
>>>>> mytree.lad<-ladderize(mytree)
>>>>>
>>>>> #node labels display on both trees correctly
>>>>>
>>>>> plot(mytree,show.node.label=TRUE)
>>>>> plot(mytree.lad,show.node.label=TRUE)
>>>>>
>>>>> #drop tips from both trees
>>>>>
>>>>> drop.mytree<-drop.tip(mytree,c("L","D","G"))
>>>>> drop.mytree.lad<-drop.tip(mytree.lad,c("L","D","G"))
>>>>>
>>>>> #plot both trees, node labels are incorrect for ladderized tree
>>>>>
>>>>> plot(drop.mytree,show.node.label=TRUE)
>>>>> plot(drop.mytree.lad,show.node.label=TRUE)
>>>>>
>>>>> _______________________________________________
>>>>> R-sig-phylo mailing list
>>>>> R-sig-phylo@r-project.org
>>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
>>>>>
>>>>>
>>>>
>>>
>>>
>>>
>
--
David W. Bapst, PhD
Adjunct Asst. Professor, Geology and Geol. Eng.
South Dakota School of Mines and Technology
501 E. St. Joseph
Rapid City, SD 57701
http://webpages.sdsmt.edu/~dbapst/
http://cran.r-project.org/web/packages/paleotree/index.html
_______________________________________________
R-sig-phylo mailing list - R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/
