Moritz Schauer

Yes, that's nicest because it incorporates reality into specification.

This is fixed for the general `PDMat` constructor which does not check positive definitiness: ``` julia> L = SMatrix{2,2}(1,2,0,0);rand(MvNormal(PDMats.PDMat(L*L', Base.LinAlg.Cholesky(L, :L)))) 2-element Array{Float64,1}: 1.35396 2.70793 ``` Is this now officially...

Update: So for the graphs protocol, the vertices must be integers: https://juliagraphs.org/Graphs.jl/dev/core_functions/interface/#Base.eltype-Tuple{AbstractGraph} Do `MultilayerGraphs` have an O(1) map from their vertex representation to the Integers and back?

People starting to use `sum(x)/length(x)` could be an indicator of friction caused by moving `mean` to the periphery

Speaking for the serious statisticians: we have no interest in developing the package in a direction where it stops being practical

That's cool! Would make sense to open an issue and think about for which functions this would be a good idea to soften landing for newcomers? For example for `I`...

I am torn, `compound` is more familiar to people, but somehow it obscures the compositional nature even more and I wouldn't know how to name the other compositions which are...

Sorry, one correction: If I think about it, this this is not a *density* but a likelihood, so we would like to represent the function class, not the class of...

Sometimes there is a conjugate distribution from a different class of distributions than the prior. But *here* Bayesian inversion is not only conjugate but also fusion and pullback through linear...

@filtron