a retake on parallel:pvec

pvec.R

pvec2<-function(v, FUN, ..., mc.set.seed = TRUE, mc.silent = FALSE, 
                mc.cores = getOption("mc.cores", 2L), mc.cleanup = TRUE
                ,combineFUN = `c`
                ) {
### AUTHOR: [email protected]
### PURPOSE: an improvement(?) on parallel:pvec which
###  (1) does not require v to be a vector, rather, v must
###      implement `[`, `length`.  Thus, i.e. BioConductor List (including
###      GRangesList) is supported.
###  (2) uses `parallel:splitIndices` to compute indices into v
###  (3) takes an optional <combineFUN>, defaulting to `c`, which
###      combines list of results from each job.
###  (4) is backwards compatible (thus I've not renamed v as, e.g., X)
###  
### Offered in reply to:
### http://thread.gmane.org/gmane.science.biology.informatics.conductor/43660
###
### TODO: > offer as contrib to parallel package (and removing "parallel:::")
###       > let each job compute its own indices based on job sequence
###         number instead of using splitIndices, thus spreading this work
###         around and esp. removing need to serialize long vectors of
###         indices between processes.
  cores <- as.integer(mc.cores)
  n<-length(v)
  cores<-min(cores,n)
  if (cores < 1L) 
    stop("'mc.cores' must be >= 1")
  if (cores == 1L) 
    return(FUN(v, ...))
  if (mc.set.seed) 
    mc.reset.stream()
  si<-splitIndices(n,cores)
  jobs <- NULL
  cleanup <- function() {
    if (length(jobs) && mc.cleanup) {
      mccollect(parallel:::children(jobs), FALSE)
      parallel:::mckill(parallel:::children(jobs), if (is.integer(mc.cleanup)) 
                        mc.cleanup
                        else 15L)
      mccollect(parallel:::children(jobs))
    }
    if (length(jobs)) {
      mccollect(parallel:::children(jobs), FALSE)
    }
  }
  on.exit(cleanup())
  FUN <- match.fun(FUN)
  jobs <- Map(function(si) {mcparallel(
                             FUN(v[si],... )
                             ,mc.set.seed = mc.set.seed
                             ,silent = mc.silent)}
              ,si
              )
  res <- mccollect(jobs)
  names(res) <- NULL
  if(!missing(combineFUN)) {
    combineFUN <- match.fun(combineFUN)
    res <- do.call(combineFUN, res)
  }
  if (length(res) != n) 
    warning("some results may be missing, folded or caused an error")
  res
}