AIC.lmr: Robust version of AIC for an MM-estimated linear model

In harrysouthworth/margarita: Support functions for texmex

Description

Robust version of AIC for an MM-estimated linear model

Usage

## S3 method for class 'lmr'
AIC(object, scale, ..., k = 2)

Arguments

object	An object of class 'lmr'.
scale	An optional scale parameter. If not provided, it is taken from object.
...	Not used.
k	The penalty per parameter to be used.

Details

The robust AIC correponds to equation (16) of Tharmaratnam and Claeskens. The 'penalty' term is not a simple function of the number of parameters in the model, so increaseing k does not necessarily result in simpler models being chosen.

References

Tharmaratnam, K. and Claeskens, G. A comparison of robust versions of the AIC based on M, S and MM-estimators, Technical Report KBI 1014, Katholieke Universiteit Leuven, 2010. https://lirias.kuleuven.be/bitstream/123456789/274771/1/KBI_1014.pdf

harrysouthworth/margarita documentation built on Aug. 19, 2021, 5 a.m.