Skip to content
Prev 379083 / 398502 Next

[FORGED] Re: Quantile Density Contours

Regardless of how many dimensions you have for independent variables, the density is one-dimensional, and if you assume the density function has been determined (e.g. by kernel estimation or by a Gaussian copula) then if you integrate the density function along that dimension there will be unique slices of the multivariate input domain determined by those slices. They might in general be disjoint regions of the independent variable space, but that is what the contour function does.

I am not seeing your point, Bert, unless you are unwilling to assume a density function model?
On March 27, 2019 2:18:18 PM PDT, Bert Gunter <bgunter.4567 at gmail.com> wrote: