#Double for loop in r code#

At that point, the replica task will be executed: itвЂ™ll first queue a replica of itself, and will then enlist in the processing of the loop. The loop will be processed serially until the underlying scheduler decides to spare a thread to process the queued replica. When that task is executed, itвЂ™ll first queue a replica of itself, and will then enlist in the processing of the loop at this point, itвЂ™s the only task processing the loop. Instead, Parallel.For begins by creating just one task. Parallel.For doesnвЂ™t just queue MaxDegreeOfParallelism tasks and block waiting for them all to complete that would be a viable implementation if we could assume that the parallel loop is the only thing doing work on the box, but we canвЂ™t assume that, in part because of the question that spawned this blog post. To understand why, it helps to have a high-level understanding for how Parallel.For is implemented. Let me start by waving my hands a bit and saying that, for the most part, you donвЂ™t need to worry about such things, in that Parallel.For was designed to accommodate this usage pattern. Second, theyвЂ™re concerned that the outer loopвЂ™s threads will end up blocking waiting for the inner parallel loops to finish.

#Double for loop in r code#

if their code was running on an eight-core machine, theyвЂ™re concerned that the outer loop would run eight of the nested loops in parallel, each of which would use eight threads, such that there would be 64 threads all churning on the inner loop. First, theyвЂ™re concerned that each nested loop will assume it вЂњowns the machineвЂќ and will thus try to use all of the cores for itself, e.g. Typically when folks ask this question, theyвЂ™re concerned about one of two things. Every now and then, I get this question: вЂњis it ok to use nested Parallel.For loops?вЂќ The short answer is вЂњyes.вЂќ As is often the case, the longer answer is, well, longer.