Computational tools for data visualization, Bayesian Deming regression, and software documentation.

This dissertation consists of three chapters that, while unified under the broad context of the development of “computational tools”, are generally distinct. In the first chapter, we propose the ggdensity package for visualizing “highest density regions” in R by extending the ggplot2 package for data visualization. In the second chapter, we develop Bayesian models for hierarchical Deming regression data, providing an implementation with the BayesDeming R package. We also apply the proposed models to example data, with an application to a chemistry, manufacturing, and controls project at a major US pharmaceutical company. Finally, we propose the tldr package, providing a system of “example-forward” documentation in the R console. Included in the discussion is the accompanying tldrDocs package, providing documentation for popular base R functions.