navlie.types.MeasurementModel

class navlie.types.MeasurementModel

Bases: ABC

Abstract measurement model base class, used to implement measurement models of the form

\[\mathbf{y} = \mathbf{g}(\mathcal{X}) + \mathbf{v}\]

where \(\mathbf{v} \sim \mathcal{N}(\mathbf{0}, \mathbf{R})\). To implement a measurement model, you must inherit from this class and implement the evaluate method, which must return a numpy array. You must also specify covariance matrix \(\mathbf{R}\) by implementing the covariance method.

abstract evaluate(x: State) → ndarray

Evaluates the measurement model \(\mathbf{g}(\mathcal{X})\).

abstract covariance(x: State) → ndarray

Returns the covariance \(\mathbf{R}\) associated with additive Gaussian noise.

jacobian(x: State) → ndarray

Evaluates the measurement model Jacobian with respect to the state.

\[\mathbf{G} = \frac{D \mathbf{g}(\mathcal{X})}{D \mathcal{X}}\]

evaluate_with_jacobian(x: State) → Tuple[ndarray, ndarray]

Evaluates the measurement model and simultaneously returns the Jacobian as its second output argument. This is useful to override for performance reasons when the model evaluation and Jacobian have a lot of common calculations, and it is more efficient to calculate them in the same function call.

jacobian_fd(x: State, step_size=1e-06)

Calculates the model jacobian with finite difference.

sqrt_information(x: State)

previous

navlie.types.Measurement

next

navlie.types.ProcessModel