(Not sure if this is a bug) Disagreement between KeplerianPropagator and NumericalPropagator
This is difficult to explain...
When I use a KeplerianPropagator, in all three modes, after a particular duration of propagation, the InitialState and Date within that propagator instance are unchanged, remaining as the very first initial state.
When I use a NumericalPropagator(the DormandPrince853Integrator one), in all three modes, after a propagation for some duration, the InitialState and Date within that propagator are CHANGED to be the the final value at the final date.
I mean that they are all propagators and should be supposed to perform consistently, so that users do not need to pay attention to their differences. So I suppose you want to know about this small differences in handling interior data. It might be a problem if a numerical propagator use would like to get the very first initial condition, in which situation the kepler one is correct but the DP853 one is not.
Is it possible I have misunderstood something? Or misused the classes you provided to us? Thanks a lot.
P.S. I did not test other analytical propagators yet. I thought they are designed with similar prototype.
(from redmine: issue id 251, created on 2016-10-07, closed on 2017-07-27)
- Uploads: