class probnum.diffeq.odefilter.information_operators.ApproximateInformationOperator(information_operator)

Bases: probnum.diffeq.odefilter.information_operators.InformationOperator, abc.ABC

Approximate information operators.

An approximate information operator is a version of an information operator that differs from its non-approximated operator in two ways:

  1. When it is transformed into a transition, the output is an approximate transition such as an EKF component.

  2. The Jacobian might be different to the Jacobian of the original version.

Approximate information operators are returned by approximation strategies such as EK0 and EK1. For instance, the EK0 changes the Jacobian of the information operator (in the sense that it sets the Jacobian of the ODE vector field to zero).

Methods Summary

__call__(t, x)

Call self as a function.

as_transition([measurement_cov_fun, …])

jacobian(t, x)

Methods Documentation

__call__(t, x)[source]

Call self as a function.

abstract as_transition(measurement_cov_fun=None, measurement_cov_cholesky_fun=None)[source]
jacobian(t, x)[source]