András Rusu

I think `lyapunovspectrum` does return the values after the warnings, it's just that if your T is large, it takes a long time to print the same warning so many...

There is a `check_error(integrator)` in SciMLBase which looks at the success fields, but I couldn't get it to work with integrator made from `ds`. ``` function chaotictest!(du,u, p, t) k1,...

Can we extend `check_error` by simply omitting the checks that use fields that the default `SimpleATsit5` doesn't have? I mean: ``` function check_error(integrator::SimpleDiffEq.SimpleATsit5Integrator) # This implementation is intended to be...

Hi! I've modified this pull request. Now it's much cleaner with the suggested `successful_step` functions. And also, I hope it works for all simple integrators and tests for all retcodes,...

My problem is that I have to use a stiff solver. This leaves me two possibilities: using autodifferentiation, which at the moment, doesn't work or don't use autodifferentiation, which works...

> because I can still reproduce this issue on main Yes, it's still there in the 3.0 release as well

This might be a separate issue, but since it is related, I'll mention it here. Why one get different number of states in the outcome space when using `probabilities` vs....

> I suppose it is reasonable to think convergence could be achieved faster if p_est was used. I'm not sure what the runtime difference would be in practice, but it...