Chad Scherrer

I don't think so, the pr targets that branch

OTOH no problem waiting if you prefer, it's pretty minor I think

BTW for `AffinePushfwd` (which should share more code with this than it currently does) I handled (1) above by defining e.g. ```julia @inline function logdensity_def(d::AffinePushfwd{(:λ,)}, x::AbstractArray) z = d.λ *...

> A `setinverse` construct convenient to have anyway, I think. Nice! I like this idea, and made a comment in that issue. > ChangesOfVariables can't depend on AD packages, that...

> can we solve that with better forwarding mechanisms in MB somehow? That's the hope - that's why this is an MB issue ;)

I think this problem of calling a function multiple time is specific to pushforwards, and doesn't come up in most cases. Here's the typically-called `logdensityof`: ```julia @inline function logdensityof(μ::AbstractMeasure, x)...

Ok, let's try that first. It seemed kind of hacky when I did it for affine transforms, because it's doing an end run around this whole nice system we set...

Oh, even better for the classification example: ```julia m = @model X,pool begin n = size(X, 1) # number of observations p = size(X, 2) # number of features k...

I agree this would be at least interesting, and possibly very useful. Somewhere in that interval ;) The challenge here isn't specific to SossMLJ, or even to Soss. the density...

Got this to work: ```julia julia> using Distributions julia> using Soss julia> import TransformVariables julia> struct Elastic{T} L1 :: T σ :: T end julia> function Distributions.logpdf(α::Elastic, x) ℓ1 =...