statsmodels.robust.norms.TukeyBiweight.get_tuning#

classmethod TukeyBiweight.get_tuning(bp=None, eff=None)[source]#

Tuning parameter for given breakdown point or efficiency

This currently only returns values from a table.

Parameters:
bpfloat in [0.05, 0.10, 0.15, …, 0.5], optional

Required breakdown point Either bp or eff has to be specified, but not both.

efffloat or None, optional

Required asymptotic efficiency. Either bp or eff has to be specified, but not both.

Returns:
tuple of float

The output depends on whether using bp or eff. If using bp, returns (tuning parameter, efficiency, scale bias) where c is the tuning. If using eff, (tuning parameter, breakdown point).