calibrain.BMN#

calibrain.BMN(L, y, noise_var, n_orient=1, max_iter=1000, tol=1e-06, init_gamma=None, verbose=False, normalization=False, logger=None, **kwargs)[source]#

BMN estimate with optional sLORETA normalization.

Return type:: Dict[str, Any]

Supports#

n_orient = 1 -> fixed (EEG or MEG) n_orient = 2 -> reduced free MEG n_orient = 3 -> free EEG

Notes

posterior_mean and posterior_cov are returned in the original coefficient space.
gamma is the learned common scalar hyperparameter in the internal optimization parameterization, so it should be treated mainly as a diagnostic quantity, especially when normalization=True.