Diagnostic statistics and visualization for quantile regression

This project aims to extend diagnostic statistics in the R package quokar. Currently in this package we have several methods such as absolute residual and robust distance, general Cook's distance, Q-function distance , mean posterior probability and Kullback-Leibler divergence to detect outliers in the framework of quantile regression models. Recently, the research on sensitivity analysis of quantile regression has attracted more and more attention. Improved methods have been introduced in the literature, which have not yet been implemented in R. This project aims to extend diagnostic statistics in quokar. It will provide users with the much needed methodology to diagnose outliers in quantile regression, thereby reducing the risk of a detrimental estimation impact that outliers may have.