Suppose I have several quantiles of my distribution of interest. (Say I've run a quantile regression at each of several quantiles. Now I want a 'simple', smooth parametric approximation for the distribution underlying those quantiles.) It would be great to have a function that estimates the best-fitting sgt distribution for these quantiles, optionally allowing for fitting only a subset of the sgt parameters while holding the others fixed.