Installation Instructions

General

This guide navigates through some quirks and considerations during the installation of the rosace and cmdstanr packages, particularly those related to compiler configurations on different operating systems. While installations on the latest MacOS and Linux with Ubuntu have been smooth, we’ve noted a few issues for installing on the Linux CentOS distribution related to C++ configurations. Alternatively, you may want to use a Docker container with the image we provided to bypass those issues.

Installing cmdstanr

Before installing rosace, you will need to install cmdstanr first using the code below.

install.packages("cmdstanr", repos = c("https://mc-stan.org/r-packages/", getOption("repos")))

# Use cmdstanr to install CmdStan, this requires a working C++ toolchain and compiler
library(cmdstanr)
install_cmdstan(cores = 2)

# Check that installation is successful by checking the CmdStan version
cmdstan_version()

If you are downloading on a Linux platform and encounter errors, it is likely due to requirements for a C++ complier since stan is built in C++. According to stan-dev, stan is tested on Linux with g++ 5 (a compiler for C++). Ensure that your system is equipped with a gcc version > 5 which includes g++. You can check the current version in the terminal using:

gcc --version

If gcc is not installed or if your version is below the required, you’ll need to install or upgrade it, following guidelines specific to your Linux distribution. You may check the distribution information in the terminal using:

lsb_release -a

Installing impute

If you encounter xx is not available for this version of R, or xx is not available for rosace during the automatic installation, you may try download it manually. For example, the package impute is not available on CRAN and can be downloaded using the following:

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("impute")

Using Docker

If you prefer to use a Docker container to run the analysis, we’ve provided a Docker image for rosace. Please follow the instructions to install Docker on your system. Once Docker is installed, you can pull the rosace Docker image with the following command:

docker pull roseraosh/rosace:latest

To run the Docker container and access its shell environment, use a command similar to the following, which creates a container and enters an interactive shell environment

docker run --rm --platform linux/amd64 -it --entrypoint bash --name rosacecontainer roseraosh/rosace

You may also add an argument to mount your local directories or data files to the container. Replace /PATH/TO/DATADIR with the actual path to your data directory on your local system:

docker run --rm --platform linux/amd64 -it -v /PATH/TO/DATADIR:/home/rosace/data --entrypoint bash --name rosacecontainer roseraosh/rosace

Once you are in the container, you may invoke R with:

R

And start the analysis by loading the rosace package:

library("rosace")