This PR implements automatic Jacobian computation for the LinearizedSystem/MeasurementModel classes.
The main changes are:
getSystemJacobian(...), getMeasurementJacobian(...) and getNoiseJacobian(...) functions in the LinearizedSystem/MeasurementModel classes are no longer pure virtual. If they are not overloaded, a central finite difference approximation of order two is used to calculate the respective Jacobian.
The observe(...) function in the MeasurementModelBase class, no longer only takes the state as an input, but also a noise sample. This is done to be consistent with the implementation of the propagate(...) function in the SystemModelBase class, as well as allowing the automatic Jacobian computation w.r.t. the measurement noise.
Additional documentation for the newly introduced implementations and some spelling mistake fixes.
This PR implements automatic Jacobian computation for the
LinearizedSystem/MeasurementModelclasses. The main changes are:getSystemJacobian(...),getMeasurementJacobian(...)andgetNoiseJacobian(...)functions in theLinearizedSystem/MeasurementModelclasses are no longer pure virtual. If they are not overloaded, a central finite difference approximation of order two is used to calculate the respective Jacobian.observe(...)function in theMeasurementModelBaseclass, no longer only takes the state as an input, but also a noise sample. This is done to be consistent with the implementation of thepropagate(...)function in theSystemModelBaseclass, as well as allowing the automatic Jacobian computation w.r.t. the measurement noise.