bfork

R library for spawning and managing multiple processes (basic process forking for R). A paper thin wrapper around fork() and waitpid().

The package consists of three major functions fork, wait, and waitpid.

Example 1

    # destination for plots
    dest <- "/path/to/dest/"

    # example function to be forked
    fn <- function() {
        Sys.sleep(1) # simulate some work
        img_name <- paste("child", i, sep="-")
        png(paste(dest, img_name, ".png", sep=""))
        plot(0,0, xlim=c(0,2), ylim=c(0,2), type="n") # create canvas
        text(1,1, paste("Child [", i, "]: Hello, World!", sep="")) # write text to canvas
        dev.off()
    }
    
    starttime <- Sys.time()

    # fork the child processes
    for(i in 1:10) {
        fork(fn)
    }
    
    # wait for all processes to finish
    wait()

    endtime <- Sys.time()

    print(paste("Total time taken to run 10 1s processes: ",
                round(endtime - starttime, 2), "s", sep=""))
    # Output: (Time taken may vary)
    #   Total time taken to run 10 1s processes: 1.68s
    #   Ten images at `dest` with "Child [<child number>]: Hello, World!" in the center

Example 2

    # example functions to be forked
    fn1 <- function() {
        Sys.sleep(2) # simulate some work
        print("Multiple")
    }

    fn2 <- function() {
        Sys.sleep(5) # simulate more work
        print("Processes")
    }

    print("Hello")

    # fork the processes
    pid1 = fork(fn2)
    pid2 = fork(fn1)

    # wait for both functions to finish
    pid1 = waitpid(pid1)
    pid2 = waitpid(pid2)
    # Output:
    #   Hello
    #   Multiple
    #   Processes

Note:: While the above example does use print it is not advised to do so when forking processes

Why bfork

Because time is important, so why waste it.

Who is bfork written for

If you have a R script that calls the same function that has a long runtime multiple times which also sends its output to a separate file, bfork is for you. E.g. if you're like me and you make a lot of plots in loops (See Example 1)