[R] Understanding R code

[Rolf Turner]

On 21/08/2009, at 3:30 PM, David Winsemius wrote:

>
> On Aug 20, 2009, at 4:31 PM, kfcnhl wrote:
>
>>
>> What is
>> 1. par.ests <- optimfit$par
>
> Just a guess... the parameter estimates from an optimization  
> procedure?
>
>> 2. fisher <- hessb(negloglik, par.ests, maxvalue=maxima);
>
> Probably the Fisher Information matrix.
>
>> 3. varcov <- solve(fisher);
>
> I seem to remember being told that the inverse of the Fisher
> Information matrix gives you the variance covariance matrix.
>
>> 4. par.ses <- sqrt(diag(varcov));
>
> And the square roots of the diagonal elements of the variance-
> covariance matrix would be the standard errors of the parameter
> estimates.
>
> Is the pop quiz over?

	You might also have asked him ``Would you like coffee with that?'' :-)

		cheers,

			Rolf Turner

######################################################################
Attention:\ This e-mail message is privileged and confid...{{dropped:9}}