Alex Arslan

Done, but I still don't see CI status. Is the repository configured such that CI is not run on PRs from forks?

Looks like CI fails on Julia nightly with multiple threads, but not in a way that could plausibly be related to the changes here, so I think we're okay.

Seems reasonable to me, though my knowledge of LibPQ and Decimals is pretty limited. I do not, however, have write access to this repository, so my review here doesn't actually...

@devmotion, perhaps you're thinking of #1875? As @PaulSoderlind noted, there's no integration here.

CI failures on nightly are unrelated.

Ugh, type inference issue on 1.3 and I can't test on 1.3 locally 😑

The first here is using `fit!(model; REML=true, σ=1.0)` and the second does not specify `σ`: ```julia-repl julia> countmap(round.(exp.(getindex.(coef.(fts), 2)); digits=3)) Dict{Float64, Int64} with 1 entry: 0.928 => 500 julia> countmap(round.(exp.(getindex.(coef.(fts0),...

We even have QuadGK as a dependency already, so adding a fallback to integration wouldn't even require another dependency.

There's a convenient derivation [here](https://stats.stackexchange.com/a/511377) that relates the moments of the truncated log-normal to the moment generating function for the truncated normal: the $n^\text{th}$ moment of the log-normal distribution with...

Ah yes, thank you! Fixed.