Poisson data in Lotka-Volterra ODE example?

[ElOceanografo]

A question/suggestion on this example: https://github.com/TuringLang/docs/blob/main/tutorials/bayesian-differential-equations/index.qmd

It currently has the data normally distributed around the ODE solution. However, given that this is an ecological model, one would never actually observe negative densities or counts of animals in the wild. (The existence of atto-foxes is problematic enough without positing the existence of theoretical anti-foxes...) Would it make sense to make the data and observation distributions Poisson counts instead? I just tried it out and the model still converges easily.

[penelopeysm]

@ElOceanografo I enjoyed the atto-foxes!

Would you be able to do a quick PR for this?

[ElOceanografo]

Sure, no problem

[ElOceanografo]

I've updated everything except for the final example model fitlv_sensealg. Unless I'm missing something, it is identical to the first one, just using Zygote instead of ForwardDiff? Happy to update it, but I'm not sure exactly what the point of the example is supposed to be...

[penelopeysm]

Hmmmm, I did some git trawling and that example used to have a sensealg specified, but it was removed here: #316

I agree it looks a bit weird now. There is still some point in keeping this because I think SciMLSensitivity only works with specific AD backends, but it needs to be explained much better imo.

Feel free to ignore it, and I can open a separate issue!

[penelopeysm]

Closed by #614