Currently they are two different classes. Equations of motion are often more naturally written in CT form But a lot of other addons/attributes/logging info are more naturally written in DT form (eg. KF, logging state information after a DT update in the gui
Current pipeline is CTsys = sys() DTsys = sys.discretize() where discretize is a method in ContinuousDynamicalSystem class.
Maybe best would be to have a single class, with a dt attribute.
f could have two behaviors, depending on whether self.dt is None or some value?
(we would still only have to implement the continuous-time _f method)