Measurement

giant.ufo.measurements:

class giant.ufo.measurements.Measurement

This ABC defines the interface for a measurement that is expected by the ExtendedKalmanFilter class for a measurement.

To implement a new measurement type, simply subclass this class, implement the abstract methods and properties (plus whatever you need for the methods/properties) and then use it to feed measurements into the EKF.

abstract property covariance: ndarray

Returns the covariance of the measurement as a numpy array

abstract property observed: Sequence | ndarray | Real

Returns the observed measurement.

abstract property observer_location: State

Returns the state of the observer at the time the measurement was captured in the base dynamics frame

abstract property time: datetime

Returns the time of the measurement as a python datetime object

abstract property identity: Hashable

The identity of this measurement.

This is used primarily by the tracker to link measurement back to observation ids

Summary of Methods

compare_residuals	This compares residuals computed using these measurement models.
compute_jacobian	Computes and returns the change in the predicted measurement given a change in the state.
predict	Predicts what the measurement should be given the current state.