Skip to contents

Robust linear regression using Huber M-estimation

Source: R/rlm.R
rlm.Huber.Rd

Fits a linear regression model using Huber M-estimation¹ via iteratively reweighted least squares.

Usage

rlm.Huber(X, y, k2 = 1.345, maxit = 100, tol.min = 1e-04, tol.target = 1e-04)

Arguments

X

numeric design matrix.

y

numeric response vector.

k2

tuning constant for Huber ψ-function. Default = 1.345.

maxit

maximum number of iterations. Default = 100.

tol.min

minimum convergence tolerance. Default = 0.0001.

tol.target

target convergence tolerance. Default = 0.0001.

Value

An object of classes "lm" and "rlm", broadly compatible with objects returned by MASS::rlm() with minor differences due to rounding.

References

  1. Huber, P.J., 1973. Robust regression: asymptotics, conjectures and Monte Carlo. The Annals of Statistics, pp. 799–821.

See also

Other rlm: AICR.Huber()