You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Points 2 is half-addressed since it does not completely remove type checks in the code to handle the inputs, instead of using dispatch -- but i think it is good as is :)
Point 3 is not addressed, discretize still lives in the ReachSets module -- i don't have any particular suggestion here. We could move this file upwards to the root src/discretize.jl..
Point 5 is interesting and not addressed in Revise discretize #503; in general one can try to invert the matrix P and obtain Phi1 and Phi2 without exponentiation. I don't have an intuition about the outcome but i propose to make a new issue for experiments.
After solving #41, we may want to do some refactorings:
Outsource the file to a different module? (It is used in(outdated)ReachSets
andProperties
.)exp(δ*Anorm) - 1. - δ*Anorm
). Avoid this.Add an option for trying to invert the matrix in the interpolation method (instead of the 3n×3n matrix computations).see Add an option to discretization for computing the matrix inverse #514The text was updated successfully, but these errors were encountered: