lime91/ebstatmax

simulations regarding the EBStatMax project

ebstatmax

Simulation-based computation of power and type-I error of statistical testing procedures for the EBStatMax project. Currently, nparLD and various GPC methods are implemented.

Contents

The main part of this project is the diacerein.R command line tool that provides various options to perform simulations. Its name alludes to the clinical trial by Vally et al. 2018, in which a drug called Diacerein was applied to treat Epidermolysis Bullosa. The provided tool is purely implemented in R. Internally, diacerein.R uses various functions and a global CONFIG object from the provided simUtils package.

The original study data is provided in the data/ subdirectory. It is used to compute rejection rates of the implemented statistical hypotheses tests. Various command line arguments of diacerein.R determine how this data is used and manipulated during simulations.

Install and Run

Requirements

In order to run the script, install R (the project was developed with version 4.2.1). Additionally, the following R packages are required:

• optparse (1.7.3)
• data.table (1.14.6)
• dplyr (1.0.10)
• nparLD (2.2)
• jsonlite (1.8.3)
• devtools (2.4.5)

The version number is the one used to develop this project.

Execute

Perform the following steps:

• Clone this repo and navigate into the repo root
• On Linux, type ./diacerein.R --help to familiarize yourself with the available program options. This requires an Rscript executable located in /usr/bin/. If you're using Windows (or if Rscript is located elsewhere), type Rscript diacerein.R --help. Make sure the Rscript command is detectable via the PATH environment variable.
• Each program invocation simulates either type-I error or power in a specific scenario. For instance, if you want to simulate the power of nparLD when normally-distributed effects are added at post-treatment time only to the Pruritus variable in the placebo group, then invoke the program with arguments -m nparld -e norm -t Pruritus. Thereby, the size of the random effects is determined in the config file simUtils/R/config.R.

Test

There are unittests available for the simUtils package. Execute all unittests with Rscript -e "devtools::test('./simUtils')".

Support and Copyright

For general questions contact the main developer Konstantin Emil Thiel. For questions regarding GPC and its implementation contact co-developer Johan Verbeeck.

Copyright (C) 2022 Konstantin Emil Thiel

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see https://www.gnu.org/licenses/.