0

I'm struggling with choosing proper package for parallel asynchronous execution of scripts by external programs with R (R script also provides inputs and aggregates outputs), I know about four popular packages : future, doSNOW, parallel and foreach but which to choose for this task ? I have about 30 independent tasks and need to run 5 of them in parallel on local machine on Windows

HenrikB
  • 6,132
  • 31
  • 34
Qbik
  • 5,885
  • 14
  • 62
  • 93
  • this might be of some help: https://stackoverflow.com/questions/41057849/run-asynchronous-function-in-r – tushaR Jan 23 '19 at 14:32
  • Author of the future framework here: if you're familiar with `lapply()` and friends, I recommend that you use the [future.apply](https://cran.r-project.org/package=future.apply) package that implements `future_lapply()` etc. See https://www.jottr.org/2018/06/23/future.apply_1.0.0/ for an introduction. It should provide the smallest mental threshold to get over. The performance of the different approaches (future.lapply, foreach, ...) are approximately the same because they all build on the same parallel backends internally. – HenrikB Jan 24 '19 at 00:32

0 Answers0