This repository presents the code written for the paper by Hurault et al. (2020), Impact of environmental factors in predicting daily severity scores of atopic dermatitis (preprint). The code is written in R language for statistical computing.
The dataset that was used in the study originates from the paper "Short-term effects of weather and air pollution on atopic dermatitis symptoms in children: A panel study in Korea" by Kim et al. (2017) and can be requested from the Zenodo database.
With the dataset file named 2017Plos_breakdown.csv, the data can be loaded and process with the functions in 001_data_import.R.
Data exploration is performed in 002_Data_exploration.R.
Model validation is conducted in 003_LR_model.R for the mixed effect logistic regression model that was previously published here by Kim et al.(2017).; and in 004_MOLR_model.R for the mixed effect ordinal logistic regression model.
The project also contains visualisation scripts:
005_plot_coef.R: the script used to produce the plots presenting the coefficients associated to environmental factors and the difference of performance between models without environmental covariates and models with environmental covariates.006_plot_lear_curve.R: the script used to plot the learning curves.