Parallelize 'SuperLearner' functions

Henrik Bengtsson

=

The futurize package allows you to easily turn sequential code into parallel code by piping the sequential code to the futurize() function. Easy!

TL;DR

library(futurize)
plan(multisession)
library(SuperLearner)

res <- CV.SuperLearner(Y = Y, X = X, SL.library = SL.library) |> futurize()

Introduction

This vignette demonstrates how to use this approach to parallelize SuperLearner functions such as CV.SuperLearner().

The SuperLearner package provides a framework for ensemble machine learning in R. The algorithm utilizes V-fold cross-validation to combine multiple prediction algorithms into a single ensemble predictor. Since cross-validation involves training many models independently, it is a perfect candidate for parallelization.

Example: Cross-Validated Super Learner

The CV.SuperLearner() function evaluates the cross-validated risk of the Super Learner ensemble. For example:

library(SuperLearner)

n <- 100
p <- 5
X <- as.data.frame(matrix(rnorm(n * p), n, p))
Y <- X[, 1] + X[, 2] + rnorm(n)
SL.library <- c("SL.glm", "SL.mean")

res <- CV.SuperLearner(Y = Y, X = X, V = 10, SL.library = SL.library)

Here CV.SuperLearner() evaluates sequentially. To run in parallel, pipe to futurize():

library(futurize)
library(SuperLearner)

res <- CV.SuperLearner(Y = Y, X = X, V = 10, SL.library = SL.library) |> futurize()

This will distribute the cross-validation fold evaluations across the available parallel workers, given that we have set up parallel workers, e.g.

plan(multisession)

The built-in multisession backend parallelizes on your local computer and works on all operating systems. There are other parallel backends to choose from, including alternatives to parallelize locally as well as distributed across remote machines, e.g.

plan(future.mirai::mirai_multisession)

and

plan(future.batchtools::batchtools_slurm)

Supported Functions

The following SuperLearner functions are supported by futurize():

• CV.SuperLearner() with seed = TRUE as the default