[R] apply for nested lists

[Gabor Grothendieck]

Alexandre Sanchez Pla <asanchez <at> ub.edu> writes:

: 
: Hi,
: 
: I am working with lists whose terms are lists whose terms are lists. 
Although 
: the real ones contain locuslink identifiers and GO annotations (I work with 
the 
: Bioconductor GO) package, I have prepared an simplified example of what I 
have 
: and what I would like to do with it:
: 
: Imagine I have a list such as:
: 
: tst.list<-list("1"=list("1A"=list(ID="1A",VAL=172),"1B"=list
: (ID="1B",VAL=134),"1C"=list(ID="1C",VAL=0)),"2"=list("2A"=NA),"3"=list
("3A"=list
: (ID="3A",VAL=33),"3B"=list(ID="3B",VAL=2)))
: 
: I would like, for instance, to be able to extract some values such as the 
: content of the "VAL" field, which may sometimes not be available.
: I may do it using a nested for such as:
: 
: x<-character(0)
: for (i in 1:length(tst.list)){
:     if (!is.na(tst.list[[i]][[1]][[1]])){
:         for (j in 1:length(tst.list[[i]]))
:             {x<-c(x,tst.list[[i]][[j]]$VAL)}}
:     else    
:             {x<-c(x, NA)}}
: 
: which gives me what I need
: 
: > x
: [1] "172" "134" "0"   NA  "33"  "2"  
: 
: According to most R documents this may be done more efficiently using apply 
: instructions, but I have failed in my temptatives to obtain the same
: 
: Thanks for any help.
: 

Try this:

f <- function(x) if (is.null(x$VAL)) NA else x$VAL
unlist(lapply(tst.list, lapply, f))