LocallyLinearizedInformationOperator

class probnum.diffeq.odefilter.information_operators.LocallyLinearizedInformationOperator(information_operator, forward_implementation='sqrt', backward_implementation='sqrt')

Bases: ApproximateInformationOperator

Approximate information operators based on local linearization.

Methods Summary

__call__(t, x)	Call self as a function.
as_transition([noise_fun])	Return an approximate transition.
jacobian(t, x)

Methods Documentation

Parameters

• information_operator (InformationOperator) –

• forward_implementation (Optional[str]) –

• backward_implementation (Optional[str]) –

__call__(t, x)

Call self as a function.

as_transition(noise_fun=None)

Return an approximate transition.

In this case, an EKF component.

Parameters

noise_fun (Optional[RandomVariable]) –

jacobian(t, x)