Integrator provides erroneous results after an update of the integration problem via callbacks #1061
Comments
|
Does updating packages not make a difference? |
I just updated it but the issue is still present ! |
|
Because discrete callbacks resolve immediately after continuous callbacks, so the DiscreteCallback would be true on the same time as the ContinuousCallback affect!. Is that what you wanted or did you want it to take a step with the new dynamics first? Because if that's the case, then |
|
I first wrote this example while having a wrong understanding of the interpolar's capabilities. You already told me on Julia discourse that modifying the problem won't cause the interpolator to be updated during that integration step. That initial wrong understanding is the reason why the example is ill-conceived. |
Describe the bug 🐞
Simply put, I am trying to implement a saturation feature on ODEs' output by detecting a threshhold value via a continuous callback.
Then via a discrete callback I can check if yes or no I can deactivate the saturation. The problem is that once the continuous callback triggers and the problem is modified (i.e. activate the saturation by flipping a boolean state) two strange behaviours happen :
Expected behavior
affectfunction of the continuous callback. However an error is thrown when the problem is modified.Minimal Reproducible Example 👇
Error & Stacktrace⚠️
Environment :
using Pkg; Pkg.status()using Pkg; Pkg.status(; mode = PKGMODE_MANIFEST)versioninfo()The text was updated successfully, but these errors were encountered: