
A Future for R: Available Future Backends
Henrik Bengtsson
Source:vignettes/future-2b-backend.md.rsp
future-2b-backend.Rmd
The future package comes with built-in future backends that leverage the parallel package part of R itself. In addition to these backends, others exist in package extensions, e.g. future.callr, future.mirai, and future.batchtools. Below is an overview of the most common backends that you as an end-user can chose from.
Package / Backend | Features | How futures are evaluated |
---|---|---|
futuresequential
|
📶 ♻️ |
sequentially and in the current R process; default |
futuremultisession
|
📶 ♻️ |
parallelly via background R sessions on current machine |
futurecluster
|
📶 ♻️* |
parallelly in external R sessions on current, local, and/or remote machines |
futuremulticore
|
📶 ♻️ |
(not recommended) parallelly via forked R processes on current machine; not with GUIs like RStudio; not on Windows |
future.callrcallr
|
📶(next) ♻️(next) |
parallelly via transient callr background R sessions on current machine; all memory is returned when as each future is resolved |
future.miraimirai_multisession
|
📶(next) ♻️(next) |
parallelly via mirai background R sessions on current machine; low latency |
future.miraimirai_cluster
|
♻️(next) |
parallelly via mirai daemons running locally or remotely |
future.batchtoolsbatchtools_lsf batchtools_openlava batchtools_sge batchtools_slurm batchtools_torque
|
📶(soon) ♻️(soon) |
parallelly on HPC job schedulers (Load Sharing Facility [LSF], OpenLava, TORQUE/PBS, Son/Sun/Oracle/Univa Grid Engine [SGE], Slurm) via batchtools; for long-running tasks; high latency |
📶: futures relay progress updates in real-time, e.g. progressr
♻️:
futures are interruptible and restartable; * disabled by default
(next): next release; (soon): in a near-future release