[R] autokorrelationresistente Kovarianz in R und S-plus

[Achim Zeileis]

Carsten,

the language on this list is English.

On Wed, 2 Jun 2004 11:17:04 +0200  Carsten.Colombier at efv.admin.ch wrote:

> Liebes r-help-Team,
> 
> ich bin gerade an meiner NDK-Abschlussarbeit und wollte anfragen, ob
> Sie wissen, ob R und S-plus eine Kovarianz für  robuste MM-Schätzer
> zur Verfügung stellen, die gegen Autokorrelation resistent ist.

This can be done with a combination of the function rlm() for fitting
robust linear models in MASS and the functions from the package sandwich
for computing HAC (heteroskedasticity and autocorrelation consistent)
covariance matrices. One possibility would be to compute a kernel HAC
estimate with automatic bandwidth selection (Andrews, 1991, Econometric)
which is implemented in kernHAC. So you first fit the model 
  fm <- rlm(yourmodel, method = "MM")
using rlm() and then can estimate the corresponding HAC covariance
matrix of the coefficients by
  kernHAC(fm)

hth,
Z

> Können Sie mir da weiterhelfen?
> 
> Vielen Dank,
> Carsten Colombier
> 
> 
> Dr. Carsten Colombier
> Economist
> Swiss Federal Finance Administration FFA
> Group of Economic Advisers
> Bundesgasse 3
> CH-3003 Bern
> Switzerland
> 
> email  carsten.colombier at efv.admin.ch
> phone 0041-31-3226332
> fax     0041-31-3230833
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://www.stat.math.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
> http://www.R-project.org/posting-guide.html
>