roblm.control {roblm} | R Documentation |
Tuning parameters for the MM-regression estimator and the associated S-estimator
roblm.control(M = 2000, Nres = NA, seed = 99, fixed = FALSE, tuning.chi = 1.54764, tuning.psi = 4.685061, compute.roboot = FALSE, compute.rd = TRUE, max.it = 50, groups = 5, n.group = 400, k.fast.s = 1)
M |
Number of bootstrap samples. This is
used when compute.roboot = TRUE |
Nres |
Number of re-sampling candidates to be used to find the initial S-estimator. This parameter is currently set to 500, which works well in most situations (see References below). User-choice capability will be added in future releases |
seed |
Random seed for the re-samples used in obtaining candiates for the initial S-estimator. |
fixed |
If FALSE the explanatory variables are
treated as random variables.
Used when compute.roboot = TRUE |
tuning.chi |
Tuning constant for the S-estimator. The choice 1.54764 yields a 50% breakdown estimator. |
max.it |
Maximum number of IRWLS iterations |
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. |
compute.roboot |
If TRUE standard errors are computed using the Robust Bootstrap of Salibian-Barrera and Zamar (2002). |
compute.rd |
If TRUE robust distances (based on the MCD robust covariance matrix) are computed for the robust diagnostic plots. This may take some time to finish, specially for large data sets. |
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. |
Matias Salibian-Barrera
Rousseeu and Yohai (1984); Yohai (1987); Salibian-Barrera and Zamar (2002); Salibian-Barrera and Yohai (2005)