Install data.table with multi-threading on Apple M1

Always get ready for the most powerful data wrangling package in R lang.

R is my favorite programming language when it comes to data analytics. And one of the reason I feel it is so powerful is solely due to this single package called data.table. It does not have quite the equal amount of attention to its major alternative, the dplyr family. I’ve spent years wondering why.

The elegance in API design and the unparalleled performance still won’t make it one of the most popular data wrangling package in R lang.

Anyway I’ve stick to it for so many years. And I’m sure there are so many more to come. :)

Ok the topic today is about using data.table in macOS, which can be a trouble due to a problem of the lacking of built-in OpenMP support in Apple’s machine. If we simply install data.table from CRAN, it will warn you that it is running on single-thread mode. That is not cool. So here is a very quick fix for that.

• Install OpenMP runtime

brew install libomp

• Update ~/.R/Makevars

This file instructs R for additional compiler arguments. Run the following one-off commands to update it:

OMP_PATH=$(brew --prefix libomp)

mkdir -p ~/.R
cat <<EOT >> ~/.R/Makevars
CPPFLAGS += -Xclang -fopenmp -I ${OMP_PATH}/include
LDFLAGS += -lomp -L ${OMP_PATH}/lib
EOT

• Install data.table from source

install.packages("data.table", type = "source", repos = "https://Rdatatable.gitlab.io/data.table")

Done. That should do the trick.

I’m using the native aarch64 build for R. Here is my sessionInfo:

R version 4.1.3 (2022-03-10)
Platform: aarch64-apple-darwin20 (64-bit)
Running under: macOS Monterey 12.2

Matrix products: default
BLAS:   /Library/Frameworks/R.framework/Versions/4.1-arm64/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.1-arm64/Resources/lib/libRlapack.dylib

locale:
[1] C/UTF-8/C/C/C/C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

loaded via a namespace (and not attached):
[1] compiler_4.1.3

Check it out:

library(data.table)
getDTthreads()  # check how many threads in use

Now we’re talking!