x3cflux.compute_perturbed_measurements

class x3cflux.compute_perturbed_measurements(simulator, params: ndarray, measurement_standard_deviations: Tuple[List[ndarray], List[float]] = None, num_samples: int = 1, random_seed: int = None)

Computes noisy measurements from simulated ones.

Noise simulation is based on given absolute errors. Errors are assumed to be normally distributed. Constraints are considered automatically and Markov Chain Monte Carlo algorithms are used to draw appropriate random numbers. The current version does not work for parameter measurements that contain multiple parameters and does not consider special error models.

Parameters:

• simulator – simulator for noise-free measurements

• params – parameters to simulate noise-free measurements from

• measurement_standard_deviations – standard deviations, defaults to the given standard deviations from the simulator

• num_samples – number of noisy measurements to generate

• random_seed – for deterministic measurement generation

Returns:

noisy measurement data