Skip to contents

EczemaPred is a R package implementing models to serve as building blocks for predicting the evolution of eczema severity, and provides a set of generic functions to manipulate these models. The models are implemented in the probabilistic programming language Stan.

EczemaPred was first introduced in Hurault et al. (2022), “EczemaPred: A computational framework for personalised prediction of eczema severity dynamics”, published in Clinical and Translational Allergy. The analysis code of this research article is available here.


The package requires RStan and C++ toolchain, which can be installed by following these instructions.

Then, the package can be installed by typing the following commands in R:


Or to install a specific version, for example the initial release (v0.1.0):


The package can take a few minutes to install as the models needs to be compiled (but no compilation will be required when using the package). Many warnings may be displayed during compilation but they can be safely ignored.

NB: EczemaPred requires HuraultMisc, my personal function library, to work.


The package is loaded with:


If you are working on a local, multicore CPU with excess RAM, you may want to call options(mc.cores = parallel::detectCores()) to run Stan on multiple cores on parallel.

The list of functions and datasets is available on the package website or by typing help(package = "EczemaPred"). Examples on how the package can be used are provided in vignettes (long form documentation).

Basic knowledge of Bayesian modelling with Stan and the package rstan is required to analyse models’ outputs. The Stan documentation is available here.

NB: While the purpose of the package is to abstract the implementation to the user, the R code and Stan code can be accessed in the R/ and inst/stan directories, respectively.


The open source version of EczemaPred is licensed under the GPL v3 license, which can be seen in the LICENSE file. A closed source version of EczemaPred is also available without the restrictions of the GPL v3 license with a software usage agreement from Imperial College London. For more information, please contact Diana Yin.