[R] Multivariate Maximum Likelihood Estimation

[Konrad BLOCHER]

Thanks,

I managed to run gls, but my problem isn't solved :)

1. I do not know the autocorrelation between errors and its estimates
change for different methods. Therefore I wasn't including it in the
corAR1 function.

2. GLS yields counteraintuitive results with one of the variables (its
sign) and I need to make sure that it is either only the case for Least
Squares, or for the data (disregarding the influence of different
estimation techniques) - thus a need for multivariate MLE.

Cheers,

KB


> It seems like you didn't look at the examples in the helpfiles. See ?gls
> and ?corAR1.
>
>
> ------------------------------------------------------------------------
> ----
> ir. Thierry Onkelinx
> Instituut voor natuur- en bosonderzoek / Research Institute for Nature
> and Forest
> Cel biometrie, methodologie en kwaliteitszorg / Section biometrics,
> methodology and quality assurance
> Gaverstraat 4
> 9500 Geraardsbergen
> Belgium
> tel. + 32 54/436 185
> Thierry.Onkelinx at inbo.be
> www.inbo.be
>
> Do not put your faith in what statistics say until you have carefully
> considered what they do not say.  ~William W. Watt
> A statistical analysis, properly conducted, is a delicate dissection of
> uncertainties, a surgery of suppositions. ~M.J.Moroney
>
> -----Oorspronkelijk bericht-----
> Van: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
> Namens Konrad BLOCHER
> Verzonden: woensdag 6 februari 2008 16:42
> Aan: r-help at r-project.org
> Onderwerp: Re: [R] Multivariate Maximum Likelihood Estimation
>
> OK I got gls running but this is the error message that I got:
>
>> gls_results <- gls(LnWRPK ~ LnWGDP+LnWYIELD,correlation = corAR1)
>
> Error in switch(mode(x), "NULL" = structure(NULL, class = "formula"),  :
>   invalid formula
>
> Google didn't give any results unfortunately.
>
> Help! :)
>
> Thanks,
>
> KB
>
>> Have you tried gls()?
>>
>>
>>
> ------------------------------------------------------------------------
> ----
>> ir. Thierry Onkelinx
>> Instituut voor natuur- en bosonderzoek / Research Institute for Nature
> and Forest
>> Cel biometrie, methodologie en kwaliteitszorg / Section biometrics,
> methodology and quality assurance
>> Gaverstraat 4
>> 9500 Geraardsbergen
>> Belgium
>> tel. + 32 54/436 185
>> Thierry.Onkelinx at inbo.be
>> www.inbo.be
>>
>> Do not put your faith in what statistics say until you have carefully
> considered what they do not say.  ~William W. Watt
>> A statistical analysis, properly conducted, is a delicate dissection
> of
> uncertainties, a surgery of suppositions. ~M.J.Moroney
>>
>> -----Oorspronkelijk bericht-----
>> Van: r-help-bounces at r-project.org
> [mailto:r-help-bounces at r-project.org]
> Namens Konrad BLOCHER
>> Verzonden: woensdag 6 februari 2008 12:12
>> Aan: r-help at r-project.org
>> Onderwerp: [R] Multivariate Maximum Likelihood Estimation
>>
>> Hi,
>>
>> I am trying to perform Maximum Likelihood estimation of a Multivariate
> model (2 independent variables + intercept) with autocorrelated errors
> of
>> 1st order (ar(1)).
>>
>> Does R have a function for that? I could only find an univariate
> option
> (ar.mle function) and when writing my own I find that it is pretty
> memory-consuming (and sometimes wrong) so there must be a better way.
>>
>> Thanks,
>>
>> KB
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>