x3cflux.compute_full_parameter_covariance

class x3cflux.compute_full_parameter_covariance(simulator, mle: ndarray)

Bases:

Compute the asymptotic covariance matrix for all parameters.

The matrix is based on the asymptotic free parameter covariance matrix from maximum likelihood theory. The algorithm is stable with regard to non-identifiable parameters.

Parameters:

• simulator (Simulator) – Labeling simulator used to evaluate the Hessian/Fisher information matrix.

• mle (numpy.ndarray) – Maximum likelihood estimator of the free parameters.

Returns:

• numpy.ndarray – Covariance matrix estimate.

• numpy.ndarray – Indices of non-identifiable parameters.