This package implements a blockwise-majorization-decent algorithm for group quantile regression (GSQR). It computes efficiently the solution paths of the group-Lasso penalized quantile regression, the group-MCP penalized quantile regression, the group-SCAD penalized quantile regression and theirs approximations.
I was wondering if it would be difficult by any chance to also support sparse group quantile regression, by analogy to sparse group LASSO?