This folder contains the following files:

- BART_FinalModel.RData: an RData file that contains the final BART model created and presented in the article.
- BinaryOutcome_CVSL_Analyses.R: an R script used to evaluation the cross validation Super Learner for the binary outcome of stunted growth. The script gives the coefficients per fold across all algorithms and the evaluation measures summarized across the folds per algorithm.
- BinaryOutcome_FinalModel.R: an R script used to fit the final BART model to the data and produce the evaluation measures and variable importance and partial dependence plots for this model. 
- ContinuousOutcome_CVSL_Analyses.R: an R script used to evaluation the cross validation Super Learner for the continuous outcome of height-for-age z-score (HAZ). The script gives the coefficients per fold across all algorithms and the evaluation measures summarized across the folds per algorithm.
- ContinuousOutcome_FinalModel.R: an R script used to fit the final stepwise linear regression model to the data and produce the evaluation measures.
- CV Super Learner_both outcomes.R: an R script used to create the cross validation Super Learner and single run of the Super Learner models for the binary outcome of stunted growth and the continuous outcome of HAZ.
- HAZ_all_NNLS.RData: an RData file that contains the Super Learner models created for the continuous outcome of HAZ.
- HAZind_all_AUC.RData: an RData file that contains the Super Learner models created for the binary outcome of stunting.
- PROVIDE_Data_All.RData: an RData file with the data used for analyses with correct structure and class types per variable. The numerical variables have been standardized. This file includes both the binary and continuous outcomes along with all predictors.
- Sen_PP_ROC.R: an R script of a function that calculates and outputs the evaluation measures of AUC, BS adjusted, and Minimum of Precision and Recall for the binary outcome models.
- Supplementary Tables and Analyses: a PDF file of additional tables, plots, and analyses. The additional tables and plot include numerical summaries of the predictor variables and a correlation plot for the quantitative predictors. Another table includes brief descriptions of all algorithms used in the Super Learner library. The additional analyses include the results for the continuous outcome HAZ.