logo rkeops

RKeOps contains the R bindings for the cpp/cuda library KeOps. It provides standard R functions that can be used in any R (>=3) codes.

For a full documentation you may read:


Feel free to contact us for any bug report or feature request, you can also fill an issue report on GitHub.

KeOps, PyKeOps, KeOpsLab




The KeOps library provides seamless kernel operations on GPU, with auto-differentiation and without memory overflows.

With RKeOps, you can compute generic reductions of very large arrays whose entries are given by a mathematical formula. It combines a tiled reduction scheme with an automatic differentiation engine. It is perfectly suited to the computation of Kernel dot products and the associated gradients, even when the full kernel matrix does not fit into the GPU memory.

For more information (installation, usage), please visit https://www.kernel-operations.io/ (especially the section dedicated to RKeOps) and read the vignettes available in R with the command browseVignettes("rkeops") or on the CRAN.



Install from CRAN

Note: RKeOps is not avaible on CRAN yet, it should be soon.

{r install, eval=FALSE} install.packages("rkeops")

Install from sources

!! In most recent version of devtools, the args argument is not available anymore and it is not possible to use devtools::install_git. Please check next section to install from sources.

                      subdir = "rkeops", 
# not possible to use `devtools::intall_github()` because of the required submodule

Get sources and install from local repository

git clone --recurse-submodules="keops/lib/sequences" https://github.com/getkeops/keops
# or
git clone https://github.com/getkeops/keops
cd keops
git submodule update --init -- keops/lib/sequences
# other submodules are not necessary for RKeOps