[R] uneven list to matrix

[Gabor Grothendieck]

OK.  One other thought. The R merge command has a sort= argument
that you can try out.  See ?merge


On 8/24/07, Christopher Marcum <cmarcum at uci.edu> wrote:
> Hi Gabor,
>
> My apologies. Both solutions work just fine on large lists (n=1000,
> n[[i]]>=500). A memory problem on my machine caused the error and
> fail-to-sort. Thank you!
>
> PS - The zoo method is slightly faster.
>
> Best,
> Chris
>
> Gabor Grothendieck wrote:
> > On 8/24/07, Christopher Marcum <cmarcum at uci.edu> wrote:
> >> Hi Gabor,
> >>
> >> Thank you. The native solution works just fine, though there is an
> >> interesting side effect, namely, that with very large lists the rows of
> >> the output become scrambled though the corresponding columns are
> >> correctly
> >> sorted. The zoo package solution does not work on large lists: there is
> >> an
> >> error:
> >>
> >> Error in order(na.last, decreasing, ...) :
> >>        argument 1 is not a vector
> >
> > They both work on the example data.  Please provide reproducible
> > examples to illustrate your comments if you would like a response.
> >
> >>
> >> Gabor Grothendieck wrote:
> >> > Here are two solutions.  The first repeatedly uses merge and the
> >> > second creates a zoo object from each alph component whose time
> >> > index consists of the row labels and uses zoo's multiway merge to
> >> > merge them.
> >> >
> >> > # test data
> >> > m <- matrix(1:5, 5, dimnames = list(LETTERS[1:5], NULL))
> >> > alph <- list(m[1:4,,drop=F], m[c(1,3,4),,drop=F], m[c(1,4,5),,drop=F])
> >> > alph
> >> >
> >> > # solution 1
> >> > out <- alph[[1]]
> >> > for(i in 2:length(alph)) {
> >> >       out <- merge(out, alph[[i]], by = 0, all = TRUE)
> >> >       row.names(out) <- out[[1]]
> >> >       out <- out[-1]
> >> > }
> >> > matrix(as.matrix(out), nrow(out), dimnames=list(rownames(out),NULL))
> >> >
> >> > # solution 2
> >> > library(zoo)
> >> > z <- do.call(merge, lapply(alph, function(x) zoo(c(x), rownames(x))))
> >> > matrix(coredata(z), nrow(z), dimnames=list(time(z),NULL))
> >> >
> >> >
> >> > On 8/23/07, Christopher Marcum <cmarcum at uci.edu> wrote:
> >> >> Hello,
> >> >>
> >> >> I am sure I am not the only person with this problem.
> >> >>
> >> >> I have a list with n elements, each consisting of a single column
> >> matrix
> >> >> with different row lengths. Each row has a name ranging from A to E.
> >> >> Here
> >> >> is an example:
> >> >>
> >> >> alph[[1]]
> >> >> A 1
> >> >> B 2
> >> >> C 3
> >> >> D 4
> >> >>
> >> >> alph[[2]]
> >> >> A 1
> >> >> C 3
> >> >> D 4
> >> >>
> >> >> alph[[3]]
> >> >> A 1
> >> >> D 4
> >> >> E 5
> >> >>
> >> >>
> >> >> I would like to create a matrix from the elements in the list with n
> >> >> columns such that the row names are preserved and NAs are inserted
> >> into
> >> >> the cells where the uneven lists do not match up based on their row
> >> >> names.
> >> >> Here is an example of the desired output:
> >> >>
> >> >> newmatrix
> >> >>  [,1]  [,2]  [,3]
> >> >> A  1     1     1
> >> >> B  2     NA    NA
> >> >> C  3     3     NA
> >> >> D  4     4     4
> >> >> E  NA    NA    5
> >> >>
> >> >> Any suggestions?
> >> >> I have tried
> >> >> do.call("cbind",list)
> >> >> I also thought I was on the right track when I tried converting each
> >> >> element into a vector and then running this loop (which ultimately
> >> >> failed):
> >> >>
> >> >> newmat<-matrix(NA,ncol=3,nrow=5)
> >> >> colnames(newmatrix)<-c(A:E)
> >> >> for(j in 1:3){
> >> >> for(i in 1:5){
> >> >> for(k in 1:length(list[[i]])){
> >> >> if(is.na(match(colnames(newmatrix),names(alph[[i]])))[j]==TRUE){
> >> >> newmatrix[i,j]<-NA}
> >> >> else newmatrix[i,j]<-alph[[i]][k]}}}
> >> >>
> >> >> Thanks,
> >> >> Chris
> >> >> UCI Sociology
> >> >>
> >> >> ______________________________________________
> >> >> R-help at stat.math.ethz.ch mailing list
> >> >> https://stat.ethz.ch/mailman/listinfo/r-help
> >> >> PLEASE do read the posting guide
> >> >> http://www.R-project.org/posting-guide.html
> >> >> and provide commented, minimal, self-contained, reproducible code.
> >> >>
> >> >
> >>
> >>
> >>
> >
>
>
>