donostah {robust} | R Documentation |
Compute a robust estimate of location and scale using the Donoho-Stahel projection-bsed estimator.
donostah(x, control)
x |
a numeric matrix containing the data. |
control |
a list of control parameters. The utility function covRob.control creates a list of the control parameters and their default values. See details for the required control parameters and their default values. |
This function is called by the high-level function covRob
when the Stahel-Donoho estimator is specified (via the optional argument estim = "donostah"
). It may also be of interest to power users who want to compute a Stahel-Donoho estimate with a minimum of fuss.
nresamp = "auto"
nresamp
may not be reached if too many of the size p
subsamples, chosen out of the observed vectors, are in a hyperplane. If nresamp = 0
all subsamples are taken. The default nresamp = "auto"
is calculated to provide a breakdown point of eps
with probability prob
.maxres = "auto"
maxres = "auto"
is 2 times nresamp
.random.sample = FALSE
FALSE
then the random seed is set so that the estimate is reproducible.center = TRUE
center = TRUE
then a robust estimate of the center is computed; if center = FALSE
then no centering takes place and the center is taken to be the zero vector.tune = 0.95
prob = 0.99
nresamp
when nresamp = "auto"
.eps = 0.5
nresamp
when nresamp = "auto"
.a list with the following components:
cov |
a numeric matrix containing the Stahel-Donoho estimate of the covariance/correlation matrix. |
center |
a numeric vector containing the Stahel-Donoho estimate of the location vector. |
dist |
a numeric vector containing the squared Mahalanobis distances. |
R. A. Maronna and V. J. Yohai (1995) The Behavior of the Stahel-Donoho Robust Multivariate Estimator. Journal of the American Statistical Association 90 (429), 330–341.
data(woodmod.dat) X <- as.matrix(woodmod.dat) ds.control <- covRob.control("donostah") donostah(X, ds.control)