[R] package MASS - MLE of negative binomial distributions
pdalgd at gmail.com
Thu Dec 10 17:19:15 CET 2015
The author usually points out that this is support software for a book, which you should probably consult.
Generally speaking, theta in glm.nb plays the same role as the variance in least squares, and you can and cannot do essentially the same things with it.
If you want to do more than that, you might want to look at generic MLE routines, like the one in stats4, or Ben Bolker's packages.
On 10 Dec 2015, at 17:04 , Rik Verdonck <rik.verdonck at bio.kuleuven.be> wrote:
> That makes sense indeed, but it also raises further questions to me if you don't mind:
> - Is there a good reason why the default assumption would be a constant overdispersion? Is it because you lose fewer degrees of freedom which would give you more power in comparing models?
> - Is there a way to tell glm.nb to fit the thetas separately per group?
> - Is there a way to fix theta (i.e. feed it a value) and only make an estimate of the expected value?
> Many thanks!
> Van: peter dalgaard [pdalgd at gmail.com]
> Verzonden: donderdag 10 december 2015 16:53
> Aan: Rik Verdonck
> CC: r-help at r-project.org
> Onderwerp: Re: [R] package MASS - MLE of negative binomial distributions
> glm.nb fits 6 mean parameters plus 1 theta. 6 x fitdistr fits two parameters each.
> On 10 Dec 2015, at 16:21 , Rik Verdonck <Rik.Verdonck at bio.kuleuven.be> wrote:
>> Dear list,
>> I have a question about the exact estimate of the maximum likelihood for a negative binomial fit. I'm trying to approach this in two different ways: the first one is a fit using the glm.nb method, and the second one is a fit using the fitdistr function for each condition separately, where I add up all log likelihoods. These two methods do not yield the same values for the log likelihood of the fit. They do yield the same log likelihood if all data are one group (no summation), so I assume I'm doing something wrong when I sum up log likelihoods. Am I not "allowed" to do this?
>> Example code:
>> for(group in levels(groups))
>> NBfit<-fitdistr(x[groups==group],"Negative Binomial")
>> Many thanks!
>> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
> Peter Dalgaard, Professor,
> Center for Statistics, Copenhagen Business School
> Solbjerg Plads 3, 2000 Frederiksberg, Denmark
> Phone: (+45)38153501
> Office: A 4.23
> Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Office: A 4.23
Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com
More information about the R-help