R help - Jul 2005 - Problem specifying "function" for "mle" operation

Hello fellow R users,

Below are two cases using the "mle" operation from the stats4 package.
In CASE 1 the code runs fine, in CASE 2 errors occur:

CASE 1

x, alpha_current, s, and n are vectors of the same length.

ll_beta<-function(b0=0,b1=0)
-sum(s*b0+s*b1*x+s*alpha_current-n*log(1+exp(b0+b1*x+alpha_current)))

fit_beta<-mle(ll_beta)

CASE 2

The error message is as follows => Error in validObject(.Object) : invalid
class "mle" object: invalid object for slot "fullcoef" in
class "mle": got class "list", should be or extend class
"numeric"

x,s,n, and alpha are vectors of the same length (5 dimensions in this specific
case).

ll_alpha<-function(alpha=c(0.5,0.5,0.5,0.5,0.5))
-sum(s*b0+s*b1*x+s*alpha-n*log(1+exp(b0+b1*x+alpha))-0.5*(alpha/sigma)^2)

fit_alpha<-mle(ll_alpha)

MY THEORY:

I think it has something to do with the fact that the function argument is a
vector in CASE 2. Hope this doesn't mean I have to re-write the function in
a way that it doesn't require vector inputs.....

Any suggestions on how to make CASE 2 work would be appreciated.

peace

Narcyz.


This message is intended for the addressee named and may con...{{dropped}}

"Narcyz Ghinea" <Narcyz.Ghinea at swsahs.nsw.gov.au> writes:
> MY THEORY:
> 
> I think it has something to do with the fact that the function
> argument is a vector in CASE 2. Hope this doesn't mean I have to
> re-write the function in a way that it doesn't require vector
> inputs.....
You do. The innards of mle has

    f <- function(p) {
        l <- as.list(p)
        names(l) <- nm
        l[n] <- fixed
        do.call("minuslogl", l)
    }

which converts the parameters to a list before calling the likelihood.
 
> Any suggestions on how to make CASE 2 work would be appreciated.
Well, either you conform or you patch mle to handle functions with
vector parameters....

-- 
   O__  ---- Peter Dalgaard             ??ster Farimagsgade 5, Entr.B
  c/ /'_ --- Dept. of Biostatistics     PO Box 2099, 1014 Cph. K
 (*) \(*) -- University of Copenhagen   Denmark          Ph:  (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)                  FAX: (+45) 35327907