Simulation studies play a crucial role in the development and assessment
of statistical methods. However, writing portable and scalable
simulation code is challenging. With the help of
Nextflow, simulatr decouples the
specification of a simulation from the execution of a simulation,
enabling users to develop a simulation study on their laptop and then
seamlessly run the simulation at-scale on a distributed computing
platform like a high-performance computing cluster or cloud computing
service.
Note: This package is experimental and under active development. Use with caution.
You can install the development version from GitHub with:
# install.packages("devtools")
devtools::install_github("katsevich-lab/simulatr")
We thank Jacob Bien for developing the simulator
package, which served as an
inspiration for simulatr. The primary difference is that simulatr is
specifically designed for seamless scaling via distributed computing
platforms.