s-news
[Top] [All Lists]

Re: mle functions or optimization routines

To: <s-news@lists.biostat.wustl.edu>
Subject: Re: mle functions or optimization routines
From: "nikolay" <niskrev@umich.edu>
Date: Sat, 24 Jan 2004 12:23:36 -0500
References: <3A822319EB35174CA3714066D590DCD504AF75E7@usrymx25.merck.com> <4011BA05.6000201@avaya.com>
Hello Everyone,
I was wondering if someone could help me in using S-Plus (6, Windows) to
compute the means of functions of Normal random variables.
Another question, somewhat related to the first one concerns a theorem that
I found (Theorem B.20 in Linear Regression, J. Grob, Springer), which says
that the E{f(z'z)z} = mE{f(x)}, where z~N(m,I) is n-dimensional vector, and
x is chi-square with n+2 d.f. and non-centrality parameter m'm/2. There is
restriction on n in the statement of the theorem, but I was wondering if it
would hold when f(y) = 1/y, and n = 1, i.e. when z is a scalar ?

Thanks for you help

Nikolay
----- Original Message ----- 
From: "Ping Zhang" <pingzhang@avaya.com>
To: <s-news@lists.biostat.wustl.edu>
Sent: Friday, January 23, 2004 7:19 PM
Subject: [S] mle functions or optimization routines


> In Splus, there is a function called nlminb() that can be used to
> solve arbitrary optimization problems and hence solving for MLE estimates.
> Does R have anything similar?
> Ping Zhnag
> Avaya Labs
>
>
> --------------------------------------------------------------------
> This message was distributed by s-news@lists.biostat.wustl.edu.  To
> unsubscribe send e-mail to s-news-request@lists.biostat.wustl.edu with
> the BODY of the message:  unsubscribe s-news
>


<Prev in Thread] Current Thread [Next in Thread>