Analysis Code for the Bayesian Level-Set (BALLET) Clustering Article

This repository contains code necessary to reproduce results from the Bayesian Level-Set Clustering article by David Buch, Miheer Dewaskar, and David Dunson.

The repository is organized into four top-level directories: data, R, scripts, and src.

1. When cloned, the data folder will be nearly empty, containing only a copy of the t-SNE embedded RNA sequencing data found on Renesh Bedre's website and the raw cosmological sky survey data, subsampled from the Edinburgh-Durham Southern Galaxy Catalogue, provided to us via personal correspondence from Woncheol Jang. The various other datasets which appear in our article are simulated by the scripts in this repository, especially scripts/data_processing/toy_data_processing.R.

2. The top-level directory R contains R function definitions and other helper code that is used in the course of our BALLET data analyses.

3. Similar to the R directory, the src directory contains code for functions which are used to perform the data analyses. However, the functions defined in src are written in C++, and bound to R functions using the Rcpp package.

Together, the R and src directories contain a complete collection of functions needed to carry out a BALLET analysis of a new dataset, should a user be interested. We may use these two folders as the foundation for an open-source BALLET software package in R.

4. The scripts directory contains the main R scripts which use the functions defined in R and src to carry out BALLET analyses and produce the output figures and tables seen in the main article and supplement. For ease of reading, the scripts in this top-level directory are grouped into four sub-folders: data_processing, illustrations, toy_challenge_analysis, and sky_survey_analysis.