This repository contains the code for simulations testing the performance of our R package ensembleGGM and an illustrative application to ovarian cancer data.
Ensemble learning can be computationally expensive, and so most of the code in this repository is designed to be run in a large-scale computing environment with multiple cores. Our simulations were conducted primarily on the Massachusetts Green High Performance Computing Cluster (https://www.mghpcc.org/).
To allow the user to pilot these simulations and applications locally and allocate resources depending on what they have available, we have used the config R package along with configuration files Simulations/config.yml and Applications/config.yml to encode pilot versions of the scripts that include a small number of folds (2-3) and a small number of cores (1-2). We recommend using K=10 folds in general. The number of cores can be adjusted as desired based on the available hardware.
Please use GitHub issues to report general questions. For individual-specific questions, email Kate Hoff Shutta at kshutta at hsph.harvard.edu.