Fix Turing 0.40 compatibility: use AbstractSciMLProblem type

[ChrisRackauckas-Claude]

Summary

• Fixed turing_inference function signature to accept SciMLBase.AbstractSciMLProblem instead of DiffEqBase.DEProblem
• This resolves CI failures with Turing 0.40 where SteadyStateProblem is now a subtype of AbstractNonlinearProblem rather than AbstractDEProblem

Root Cause

In Turing 0.40 / recent SciML ecosystem changes, SteadyStateProblem moved from being a subtype of AbstractDEProblem to AbstractNonlinearProblem. The turing_inference function was typed to only accept DEProblem, causing a MethodError when used with SteadyStateProblem.

Solution

Changed the type signature from:

function turing_inference(prob::DiffEqBase.DEProblem, ...)

to:

function turing_inference(prob::SciMLBase.AbstractSciMLProblem, ...)

This accepts both AbstractDEProblem (ODE/SDE/etc) and AbstractNonlinearProblem (including SteadyStateProblem).

Related

This supersedes the compat-only change in #358 by including the necessary code fix.

🤖 Generated with Claude Code

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Import/add SciMLBase for new type annotation

The new annotation SciMLBase.AbstractSciMLProblem relies on the SciMLBase module being in scope, but this package neither imports SciMLBase nor lists it as a dependency. In a clean environment, SciMLBase will be undefined in DiffEqBayes, so loading the module or parsing this method triggers UndefVarError: SciMLBase not defined. To avoid a load-time failure, add SciMLBase to [deps] and using SciMLBase (or switch to a type that is already in scope).

Useful? React with 👍 / 👎.