Skip to content
Prev 7087 / 20628 Next

Very large HPDintervals, mcmcsamp, lmer

Thierry,

I have two quick follow up questions for my model that I hope you can
can answer for me.
model<-lmer(TN ~ wsh*rip + (1|stream), data=all24)

First, the one I should know: how did you determine the number of
parameters? Watershed (4) + reach (2) +reach x watershed (1) = 7?

And how many parameters are in this model? 9?
model<-glmer(Y ~ wsh*rip + (1|stream) + (1|stream:rip) + (1|obs),
data=ept, family=binomial(link="logit"))

Second: is there a simple explanation for why 10 observations per
parameter is the rule of thumb?

Thank you

Colin

On Tue, Nov 22, 2011 at 12:59 AM, ONKELINX, Thierry
<Thierry.ONKELINX at inbo.be> wrote:

  
    
Message-ID: <CACu_zZmuebfCCEN-wFn6jk6yA8qTG1qR6eJM8TGSSquLyeV0uw@mail.gmail.com>
In-Reply-To: <AA818EAD2576BC488B4F623941DA742756FCFC78@inbomail.inbo.be>