CovNASest  R Documentation 
Computes SEstimates of multivariate location and scatter based on Tukey's biweight function for incomplete data using a fast algorithm similar to the one proposed by SalibianBarrera and Yohai (2006) for the case of regression. Alternativley, the Ruppert's SURREAL algorithm, bisquare or Rocke type estimation can be used.
CovNASest(x, bdp = 0.5, arp = 0.1, eps = 1e5, maxiter = 120,
nsamp = 500, impMeth = c("norm" , "seq", "rseq"), seed = NULL,
trace = FALSE, tolSolve = 1e13,
scalefn,
method = c("sfast", "surreal", "bisquare", "rocke", "suser", "sdet"), control,
t0, S0, initcontrol)
x 
a matrix or data frame. 
bdp 
a numeric value specifying the required
breakdown point. Allowed values are between

arp 
a numeric value specifying the asympthotic
rejection point (for the Rocke type S estimates),
i.e. the fraction of points receiving zero
weight (see Rocke (1996)). Default is 
eps 
a numeric value specifying the
relative precision of the solution of the Sestimate
(bisquare and Rocke type). Default is to 
maxiter 
maximum number of iterations allowed
in the computation of the Sestimate (bisquare and Rocke type).
Default is 
nsamp 
the number of random subsets considered. Default is 
impMeth 
select imputation method to use  choose one of "norm" , "seq" or "rseq". The default is "norm" 
seed 
starting value for random generator. Default is 
trace 
whether to print intermediate results. Default is 
tolSolve 
numeric tolerance to be used for inversion
( 
scalefn 

method 
Which algorithm to use: 'sfast'=FASTS, 'surreal'=SURREAL, 'bisquare', 'rocke' or 'sdet', which will invoke the deterministic algorihm of Hubert et al. (2012). 
control 
a control object (S4) of class 
t0 
optional initial HBDP estimate for the center 
S0 
optional initial HBDP estimate for the covariance matrix 
initcontrol 
optional control object to be used for computing the initial HBDP estimates 
Computes biweight multivariate Sestimator of location and scatter. The computation will be performed by one of the following algorithms:
An algorithm similar to the one proposed by SalibianBarrera and Yohai (2006) for the case of regression
Ruppert's SURREAL algorithm when method
is set to 'surreal'
Bisquare SEstimate with method
set to 'bisquare'
Rocke type SEstimate with method
set to 'rocke'
.
An S4 object of class CovNASest
which is a subclass of the
virtual class CovNARobust
.
Valentin Todorov valentin.todorov@chello.at, Matias SalibianBarrera matias@stat.ubc.ca and Victor Yohai vyohai@dm.uba.ar. See also the code from Kristel Joossens, K.U. Leuven, Belgium and Ella Roelant, Ghent University, Belgium.
M. SalibianBarrera and V. Yohai (2006) A fast algorithm for Sregression estimates, Journal of Computational and Graphical Statistics, 15, 414–427.
R. A. Maronna, D. Martin and V. Yohai (2006). Robust Statistics: Theory and Methods. Wiley, New York.
Todorov V & Filzmoser P (2009), An Object Oriented Framework for Robust Multivariate Analysis. Journal of Statistical Software, 32(3), 1–47. <doi:10.18637/jss.v032.i03>.
library(rrcov)
data(bush10)
CovNASest(bush10)
## the following four statements are equivalent
c0 < CovNASest(bush10)
c1 < CovNASest(bush10, bdp = 0.25)
c2 < CovNASest(bush10, control = CovControlSest(bdp = 0.25))
c3 < CovNASest(bush10, control = new("CovControlSest", bdp = 0.25))
## direct specification overrides control one:
c4 < CovNASest(bush10, bdp = 0.40,
control = CovControlSest(bdp = 0.25))
c1
summary(c1)
## Use the SURREAL algorithm of Ruppert
cr < CovNASest(bush10, method="surreal")
cr
## Use Bisquare estimation
cr < CovNASest(bush10, method="bisquare")
cr
## Use Rocke type estimation
cr < CovNASest(bush10, method="rocke")
cr
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.