Quantile Regression Outlier Diagnostics with K Left Out Analysis

Diagnostics methods for quantile regression models for detecting influential observations: robust distance methods for general quantile regression models; generalized Cook's distance and Q-function distance method for quantile regression models using aymmetric Laplace distribution. Reference of this method can be found in Luis E. Benites, Víctor H. Lachos, Filidor E. Vilca (2015) ; mean posterior probability and Kullback–Leibler divergence methods for Bayes quantile regression model. Reference of this method is Bruno Santos, Heleno Bolfarine (2016) .

R package for diagnosing quantile regression models

• quokar can be easily installed on Windows and Linux using command:

devtools::install_github("wenjingwang/quokar")

• To install quokar on Mac, you need to install a fortran compiler. You can install gfortran by command,

curl -O http://r.research.att.com/libs/gfortran-4.8.2-darwin13.tar.bz2
sudo tar fvxz gfortran-4.8.2-darwin13.tar.bz2 -C /