statsmodels.robust.resistant_linear_model.RLMDetS.fit

RLMDetS.fit(h, maxiter=100, maxiter_step=5, start_params_extra=None)

Estimate the model using a deterministic S-estimator.

Parameters:

hint

Size of the initial sets used to compute the starting values for the S-estimator.

maxiterint, optional

Maximum number of iterations used for the final fit starting from the best starting value. Default is 100.

maxiter_stepint, optional

Maximum number of iterations used when fitting each starting value while searching for the best starting value. Default is 5.

start_params_extralist[ndarray] or None, optional

Additional starting parameter arrays to consider together with the starting sets generated internally. Default is None, which means no extra starting values are used.

Returns:

statsmodels.robust.robust_linear_model.RLMResults

Results instance corresponding to the best starting value.