lmrob.control {robustbase}R Documentation

Tuning parameters for lmrob

Description

Tuning parameters for, lmrob, the MM-regression estimator and the associated S-estimator.

Usage

lmrob.control(seed = 37, Nres = 500,
        tuning.chi = 1.54764, bb = 0.5, tuning.psi = 4.685061,
        max.it = 50, groups = 5, n.group = 400, k.fast.s = 1,
        compute.rd = TRUE)

Arguments

seed random seed for the re-samples used in obtaining candiates for the initial S-estimator. The default, 37 used to be frozen in the underlying C code.
Nres number of re-sampling candidates to be used to find the initial S-estimator. Currently defaults to 500 which works well in most situations (see References below). User-choice capability will be added in future releases.
tuning.chi tuning constant for the S-estimator. The default, 1.54764, yields a 50% breakdown estimator.
bb expected value under the normal model of the "chi" function with tuning constant equal to tuning.chi. This is used to compute the S-estimator.
tuning.psi tuning constant for the re-descending M-estimator. The choice 4.685061 yields an estimator with asymptotic efficiency of 95% for normal errors.
max.it integer specifying the maximum number of IRWLS iterations.
groups This parameter is for the fast-S algorithm. Number of random subsets to use when the data set is large.
n.group This parameter is for the fast-S algorithm. Size of each of the groups above.
k.fast.s This parameter is for the fast-S algorithm. Number of local improvement steps for each re-sampling candidate.
compute.rd logical indicating if robust distances (based on the MCD robust covariance estimator covMcd) are to be computed for the robust diagnostic plots. This may take some time to finish, particularly for large data sets.

Author(s)

Matias Salibian-Barrera

See Also

lmrob, also for references and examples.


[Package robustbase version 0.1-5 Index]