TL;DR
Thefuturize package makes it extremely simple to parallelize your existing map-reduce calls, but also a growing set of domain-specific calls. All you need to know is that there is a single function calledfuturize() that will take care of everything, e.g.
Thefuturize() function parallelizes viafutureverse, meaning your code can take advantage of anysupported future backends, whether it be parallelization on your local computer, across multiple computers, in the cloud, or on a high-performance compute (HPC) cluster. Thefuturize package has only one hard dependency - thefuture package. All other dependencies are optional “buy-in” dependencies as shown in the below tables.
Supported calls
Supported map-reduce packages
Thefuturize package supports transpilation of functions from multiple packages. The tables below summarize the supported map-reduce and domain-specific functions, respectively. To programmatically see which packages are currently supported, use:
To see which functions are supported for a specific package, use:
futurize_supported_functions("caret")| Package | Functions | Requires |
|---|---|---|
| base | lapply(),sapply(),tapply(),vapply(),mapply(),.mapply(),Map(),eapply(),apply(),by(),replicate(),Filter() | future.apply |
| stats | kernapply() | future.apply |
| purrr | map() and variants,map2() and variants,pmap() and variants,imap() and variants,modify(),modify_if(),modify_at(),map_if(),map_at(),invoke_map() | furrr |
| crossmap | xmap() and variants,xwalk(),map_vec(),map2_vec(),pmap_vec(),imap_vec() | (itself) |
| foreach | %do%, e.g. foreach() %do% { },times() %do% { } | doFuture |
| plyr | aaply() and variants,ddply() and variants,llply() and variants,mlply() and variants | doFuture |
| pbapply | pblapply(),pbsapply() and variants,pbby(),pbreplicate() andpbwalk() | (itself) |
| BiocParallel | bplapply(),bpmapply(),bpvec(),bpiterate(),bpaggregate() | doFuture |
Table: Map-reduce functions currently supported byfuturize() for parallel transpilation.
Here are some examples:
library(futurize)plan(multisession)xs<-1:10ys<-lapply(xs,sqrt)|>futurize()xs<-1:10ys<-purrr::map(xs,sqrt)|>futurize()xs<-1:10ys<-crossmap::xmap_dbl(xs,~.y*.x)|>futurize()library(foreach)xs<-1:10ys<-foreach(x=xs)%do%{sqrt(x)}|>futurize()xs<-1:10ys<-plyr::llply(xs,sqrt)|>futurize()xs<-1:10ys<-pbapply::pblapply(xs,sqrt)|>futurize()xs<-1:10ys<-BiocParallel::bplapply(xs,sqrt)|>futurize()and
Supported domain-specific packages
You can also futurize calls from a growing set of domain-specific packages (e.g. boot,fwb,caret,glmnet,lme4,mgcv, andtm) that have optional built-in support for parallelization.
| Package | Functions | Requires |
|---|---|---|
| boot | boot(),censboot(),tsboot() | future |
| caret | bag(),gafs(),nearZeroVar(),rfe(),safs(),sbf(),train() | doFuture |
| fwb | fwb(),vcovFWB() | (itself) |
| glmnet | cv.glmnet() | doFuture |
| lme4 | allFit(),bootMer() | future |
| mgcv | bam(),predict.bam() | future |
| strucchange | breakpoints() | doFuture |
| tm | TermDocumentMatrix(),tm_index(),tm_map() | future |
Table: Domain-specific functions currently supported byfuturize() for parallel transpilation.
Here are some examples:
ctrl<-caret::trainControl(method="cv", number=10)model<-caret::train(Species~., data=iris, method="rf", trControl=ctrl)|>futurize()ratio<-function(d,w)sum(d$x*w)/sum(d$u*w)b<-boot::boot(boot::city,ratio, R=999)|>futurize()f<-fwb::fwb(boot::city,ratio, R=999)|>futurize()cv<-glmnet::cv.glmnet(x,y)|>futurize()m<-lme4::allFit(models)|>futurize()b<-mgcv::bam(y~s(x0, bs=bs)+s(x1, bs=bs), data=dat)|>futurize()bp<-strucchange::breakpoints(Nile~1)|>futurize()m<-tm::tm_map(crude,content_transformer(tolower))|>futurize()