Sebastian Persson

> Observables could depend on species, parameters, and variables. What if an unknown symbol is encountered in the expression? We could either default to creating it as a parameter, or...

As metadata in catalyst - or downstream packages?

Does this PR also account for the interface to PEtab.jl?

Thanks for the fast reply! I agree with you on this. Error from failed optimization should be caught, while errors from the model should be handled appropriately by the modeling...

As an update to this issue, for another ODE model with a bad startguess I encountered an Optim.jl related error (see below). This error is rare, but can happen for...

Sorry for late reply (missed this in my notifications). Hopefully this attached manifest file makes it run [Manifest.zip](https://github.com/user-attachments/files/16893204/Manifest.zip). If this does not work I can try to create a smaller...

Thanks for the fast reply, I managed to get it working! Providing a custom gradient to `AutoMala` is a bit involved, would it be possible to add a custom gradient...

Some additional thoughts (I maintain [PEtab.jl](https://github.com/sebapersson/PEtab.jl) which allows users to set up likelihood parameter estimation problems for ODE models, and currently a key functionality lacking in the modeling pipeline is...

Providing a custom gradient is covered by providing `OptimizationProblem` - as when creating it (or rather the `OptimizationFunction`) the user can set the gradient (for example in PEtab we do...

A bit late to the game, but I agree with Paul here that we should avoid nesting. I find the nesting to be confusing, while the long format is more...