Generally I find the GetDist Python package based on this paper does a much better job producing accurate contours than this one, in particular for close-to-Gaussian distributions and especialy with boundaries. E.g. here's a comparison for 100 samples from a unit normal distribution (top row) or a unit normal truncated below 0 (bottom row; code at the end of this post):
I don't know if there's any motivation here from anyone to borrow / implement some ideas from that paper, but I figured I'd put it out there, it would be very nice to have a pure-Julia version.
Generally I find the GetDist Python package based on this paper does a much better job producing accurate contours than this one, in particular for close-to-Gaussian distributions and especialy with boundaries. E.g. here's a comparison for 100 samples from a unit normal distribution (top row) or a unit normal truncated below 0 (bottom row; code at the end of this post):
I don't know if there's any motivation here from anyone to borrow / implement some ideas from that paper, but I figured I'd put it out there, it would be very nice to have a pure-Julia version.