huberWeightLS: Huber's function

Description Usage Arguments Details Value Author(s) References Examples

Description

This function provides the usual Huber's weight function in Robust estimation context. See Huber (1981) for details. Let ρ(x) be the usual Huber's function, this function is ρ'(x)/x, where ρ'(x) is the derivative of ρ(x).

Usage

1

Arguments

data

The input vector or matrix

k

The parameter to control robustness, the default value is 1.345, as suggested in many Robust Statistics textbooks

Details

See details of the huber weights in iterative least squares in the references

Value

the weight vector or matrix for Huber-type robust regression

Author(s)

Lingsong Zhang (lingsong@purdue.edu) and Chao Pan (panc@purdue.edu)

References

Huber, P. J. (1981). Robust statistics. Wiley series in probability and mathematical statistics.

Examples

1
2
3
4
5
#generate a t distrbution matrix
x<-matrix(rt(100, 1), nrow=10)

#generate the huber weight matrix with k=1.345
y=huberWeightLS(x, k=1.345)

Example output



RobRSVD documentation built on May 2, 2019, 9:39 a.m.