Samuel Brand

@JonasIsensee I've had the same issue, and also want to thank you for your work on this package. Good luck dealing with this issue.

Hi guys, We ran into the same issue around developing an epidemiological modelling/inference package (see [here](https://github.com/CDCgov/Rt-without-renewal/pull/190)). Personally, I'd support `prefix=true` default.

I also want to say thanks for creating Pathfinder! Just want to bump this, because as of (I think) last month Turing seems to have made it easier to interface...

@sethaxen That sounds great. For models with lots of parameters this will be super handy.

Thanks @sethaxen ! I'm currently working on [Epi inference project/package-to-be](https://github.com/CDCgov/Rt-without-renewal/tree/main) with @seabbs and @zsusswein so this is a good chance to try it out.

Hi @sethaxen, Following up from #93 . For our example, pathfinder fails [here](https://github.com/CDCgov/Rt-without-renewal/blob/da0057c1dec3a62c45b72e28c11a92e510160a24/EpiAware/docs/src/examples/getting_started.jl#L295) with a `MethodError` on a `promote_rule` > MethodError: promote_rule(::Type{IrrationalConstants.Log2π}, ::Type{ForwardDiff.Dual{ForwardDiff.Tag{OptimizationReverseDiffExt.OptimizationReverseDiffTag, Float64}, Float64, 12}}) is ambiguous. Its a...

> Its a bit of a complicated example so if that doesn't make a lot of sense; I can try and find a more minimal fail. Does this work fine?...

> Can you post the full stack trace? Here you go: > ERROR: MethodError: promote_rule(::Type{IrrationalConstants.Log2π}, ::Type{ForwardDiff.Dual{ForwardDiff.Tag{OptimizationReverseDiffExt.OptimizationReverseDiffTag, Float64}, Float64, 12}}) is ambiguous. > > Candidates: promote_rule(::Type{S}, ::Type{T}) where {S

> Can Turing sample fine with HMC (as in, compute the gradient of the log-density without erroring?) Yes, the workflow is that pathfinder is part of initialisation for a HMC/NUTS...

Thanks for looking into this!