Introduction

Torsten is a fork of Stan containing additional Stan language functions to facilitate implementation of pharmacometric models. It plays a role analogous to prediction for population pharmacokinetics (PREDPP) for NONMEM® models. The simplest way to install Torsten is to use the install_torsten() function from the bbr.bayes package.

This page demonstrates how to install Torsten using the install_torsten() function.

Set up

Load required packages and set file paths to your model and figure directories.

library(bbr)
library(bbr.bayes)
library(here)

here() starts at /data/expo4-torsten-poppk

Basic installation

The install_torsten() function attempts to download and install the latest release of Torsten. The following statement installs Torsten as a folder named “Torsten” in the project folder. By default, the installation will not overwrite a previous installation of the same Torsten version at that location. You can overwrite a previous installation by adding overwrite = TRUE. You may want to do that in cases where you want to change the C++ compiler options via the cpp_options argument.

install_torsten(dir = here("Torsten"))

Enabling multi-threading

One strategy for reducing computation time of Stan/Torsten analyses is to use the reduce_sum Stan function to allow within-chain parallel computation via multi-threading. That requires specific compiler options when Torsten is installed. The following statement installs Torsten using the compiler options needed for multi-threading.

install_torsten(dir = here("Torsten"), overwrite = TRUE,
                cpp_options = list(stan_threads = TRUE))