Moritz Schauer

*Correction:* They do count or estimate the number of moves for each CPDAG $M_t$ and can use this later to account for the chain $c_t$ visiting CPDAGs with a large...

Let's have a chat on https://julialang.zulipchat.com about it if you like.

You are reading the code! I don't remember anymore but I think you are right. Let's remove the reference to stable and keep the issue open if someone wants to...

> store the to-be-deleted edges temporarily and remove them at the d = d+1 line This might be a very good idea because removing single edges is linear in the...

I gave a shot at the stable skeleton algorithm in #101

Didn't merge this right away because I feel there should be a third-party package doing this the right way.

If you have SDEs (a noisy relaxation of your ODE), you can do multiple shooting in MCMC

``` julia> d = censored(Normal(0, 1), 0, Inf) Censored(Normal{Float64}(μ=0.0, σ=1.0); lower=0.0, upper=Inf) julia> pdf(d, 0.0001) 0.3989422784067213 julia> pdf(d, 0) 0.5 julia> pdf(d, -0.0001) 0.0 ``` I think a digestible statement...

What do you mean with the last statement? Is this good or bad?

If you are interested, can you read https://github.com/JuliaStats/Distributions.jl/issues/1468 which also refers to this discussion: https://github.com/JuliaMath/DensityInterface.jl/pull/4#issuecomment-959970063 and see if that somehow relates to your problem?