navlie.datagen.DataGenerator

class navlie.datagen.DataGenerator(process_model: ProcessModel, input_func: Callable[[float], ndarray], input_covariance: ndarray, input_freq: float, meas_model_list: List[MeasurementModel] | None = None, meas_freq_list: float | List[float] | None = None, meas_offset_list: float | List[float] | None = None)

Bases: object

A class used for data generation given a process model, and as many measurement models as you want. Frequencies and offsets of each measurement can also be specified.

Parameters:

• process_model (ProcessModel) – Process model to be used for ground truth trajectory simulation

• input_func (Callable[[float], np.ndarray]) – A function that returns the input value to be used. The function must accept a timestamp as a float from the data generator.

• input_covariance (np.ndarray or Callable[[float], np.ndarray].) – Covariance used for noise generation to be applied to input values. Either provided as a static value or as a function returning the time-varying Q

• input_freq (float) – Frequency if the input.

• meas_model_list (List[MeasurementModel], optional) – MeasurementModel of measurements to be generated, by default []

• meas_freq_list (float or List[float], optional) – If a measurement model list is provided, also provide a list of frequencies associated with each measurement. If a single frequency is provided as a float, it will be used for all measurements.

• meas_offset_list (float or List[float], optional) – If a measurement model list is provided, you can also provide a list of initial offsets associated with each measurement. If no offsets provided, the initial offset is assumed to be 0. If a single offset is provided as a float, it will be used for all measurements.

add_measurement_model(model: MeasurementModel, freq: float, offset: float = 0.0)

Adds a new measurement model to the existing list of measurement models.

Parameters:

• model (MeasurementModel) – The MeasurementModel of measurements to be generated.

• freq (float) – The frequency associated with these measurements.

• offset (float, optional) – The initial offset associated with these measurements, by default 0.0.

generate(x0: State, start: float, stop: float, noise=False)

Generates data by integrating the process model using noiseless input values, and using the states computed to generate measuements.

Parameters:

• x0 (State) – Initial state

• start (float) – Starting timestamp

• stop (float) – Ending timestamp

• noise (bool, optional) – Whether to add noise to inputs/measurements, by default False

Returns:

• List[State] – Ground truth states

• List[Input] – Inputs, possibly noisy if requested.

• List[Measurement] – Measurements, possibly noisy if requested.

previous

navlie.datagen.generate_measurement

next

navlie.filters