R help - Jul 2005 - integrate fails with errors

Hi all,
i'm new to R,
I need to modelize in R a statistic algorithm,
This algo use Weibull, normal law, linear regression, normalisation, root mean
square, to find eta and beta fitting the weibull model (to analyse few results)
and further when we will get more information apply bayes model .

the problem is when When i try to integrate it fails with errors.

by the way i like to integrate something like that :

beta0,eta0,n are initialized as single integer, temp is a 1 dimension array
containing 9 integer.

integrate(function(beta)
((beta/(eta0)^beta)^n)*prod(temp^(1-beta)*exp(-sum(temp^beta)/(eta^beta)))*(1/(sqrt(2*pi))*exp(((beta-beta0)^2,0,Inf)

R says : The longuest object isn't a multiple of the shortest object in
temp^beta , the same for temp^(1-beta).

I don't understand why R fails to take calculate each temp^beta then sum it
over again for each beta values.

don't know if my post is explicit my head is burned out with these things
today

thks.

////////////////////////////////////////////////////////////
// Webmail Oreka : http://www.oreka.com
////////////////////////////////////////////////////////////

On Tue, 19 Jul 2005, [iso-8859-1] herodote at oreka.com wrote:
>
> beta0,eta0,n are initialized as single integer, temp is a 1 dimension array
containing 9 integer.
>
> integrate(function(beta)
((beta/(eta0)^beta)^n)*prod(temp^(1-beta)*exp(-sum(temp^beta)/(eta^beta)))*(1/(sqrt(2*pi))*exp(((beta-beta0)^2,0,Inf)
>
> R says : The longuest object isn't a multiple of the shortest object in
temp^beta , the same for temp^(1-beta).
>
> I don't understand why R fails to take calculate each temp^beta then
sum it over again for each beta values.
>
The help page for integrate() says
     f: an R function taking a numeric first argument and returning a
           numeric vector of the same length.

Your f is not of that form.

 	-thomas