BSquare: An R package for Bayesian simultaneous quantile regression

BSquare in an R package to conduct Bayesian quantile regression for continuous, discrete, and censored data. Quantile regression provides a comprehensive analysis of the relationship between covariates and a response. In quantile regression, by specifying different covariate effects at different quantile levels we allow covariates to affect not only the center of the distribution, but also its spread and the magnitude of extreme events. Unlike most approaches to quantile regression, such as those implemented in package quantreg and bayesQR, BSquare analyzes all quantile levels simultaneously. Therefore, this approach can borrow strength across nearby quantile levels to reduce uncertainty in the estimated quantile function, which can be advantageous when the covariate effects are similar across quantile levels. BSquare takes a model-based approach to quantile regression, which we briefly review in Section 1; for thorough descriptions of the model and its properties we refer t