hilo(<dist_default>) does not work for multivariate distributions

[mitchelloharawild]

Should <hilo> objects allow for matrix (or even non-numeric) inputs for lower/upper? Or should hilo(<dist_default>) restructure inputs to provide multiple hilo objects?

I'm currently inclined to generalise the object class, as intervals of variables in multivariate distributions should be tightly coupled. Similarly, you might want intervals of ordered categories.

library(distributional)
dist <- dist_multivariate_normal(mu = list(c(1,2)), sigma = list(matrix(c(4,2,2,3), ncol=2)))
hilo(dist)
#> Error: `lower` must be a vector with type <double>.
#> Instead, it has type <double[,2]>.

^{Created on 2021-11-19 by the reprex package (v2.0.0)}