pbdMPI (version 0.5-1)

global sort: Global Quick Sort for Distributed Vectors or Matrices

Description

This function globally sorts distributed data for all ranks.

Usage

comm.sort(x, decreasing = FALSE, na.last = NA,
          comm = .pbd_env$SPMD.CT$comm,
          status = .pbd_env$SPMD.CT$status)

Value

The returns are the same size of x but in global sorting order.

Arguments

x

a vector.

decreasing

logical. Should the sort order be increasing or decreasing?

na.last

for controlling the treatment of NAs. If TRUE, missing values in the data are put last; if FALSE, they are put first; if NA, they are removed.

comm

a communicator number.

status

a status number.

Warning

All ranks may not have a NULL x.

Author

Wei-Chen Chen , George Ostrouchov, Drew Schmidt, Pragneshkumar Patel, and Hao Yu.

Details

The distributed quick sort is implemented for this functions.

References

Programming with Big Data in R Website: https://pbdr.org/

Examples

if (FALSE) {
### Save code in a file "demo.r" and run with 2 processors by
### SHELL> mpiexec -np 2 Rscript demo.r

spmd.code <- "
### Initialize
suppressMessages(library(pbdMPI, quietly = TRUE))

.comm.size <- comm.size()
.comm.rank <- comm.rank()

### Examples.
comm.set.seed(123456, diff = TRUE)
x <- c(rnorm(5 + .comm.rank * 2), NA)
# x <- sample(1:5, 5 + .comm.rank * 2, replace = TRUE)
comm.end.seed()

if(.comm.rank == 1){
  x <- NULL    ### Test for NULL or 0 vector
}

y <- allgather(x)
comm.print(y)

y <- comm.sort(x)
y <- allgather(y)
comm.print(y)

### Finish.
finalize()
"
# execmpi(spmd.code, nranks = 2L)
}