allowed to create J-QPD distributions for a full set of samples at once, rather than an individual one at a time (only for J-QPD versions, not the extended ones)
switched to original J-QPDs instead of extended ones for the quantile regression chain (faster and more convenient due to the above point)