Dear Statalist readers,
does anyone know of an implementation of Huber's robust A-estimator
(described by Hampel/Ronchetti/Rousseeuw/Stahel (1986): Robust
Statistics, The approach based on influence functions, p. 113). I am
trying to do some robust standardization for non gaussian variables and
use Huber's 1-step estimator for the mean, but I need a robust estimate
for the variance. Does anyone has an idea?
Thanks,
Thomas
begin:vcard
n:M�hlmann;Thomas
tel;fax:0221-470-2305
tel;work:0221-470-2628
x-mozilla-html:FALSE
org:Universit�t zu K�ln;Bankseminar
adr:;;Albertus-Magnus-Platz;50923 K�ln;;;
version:2.1
email;internet:[email protected]
fn:Dipl.-Kfm. Thomas M�hlmann
end:vcard