s-news
[Top] [All Lists]

Re: [S] cox proportional models

To: Chung-chi Chang <sph1606@utsph.sph.uth.tmc.edu>
Subject: Re: [S] cox proportional models
From: Prof Brian D Ripley <ripley@stats.ox.ac.uk>
Date: Tue, 23 Feb 1999 21:17:00 +0000 (GMT)
Cc: "'s-news@wubios.wustl.edu'" <s-news@wubios.wustl.edu>
In-reply-to: <01BE5F3D.DF8A9E40@sph39214.sph.uth.tmc.edu>
Sender: owner-s-news@wubios.wustl.edu
On Tue, 23 Feb 1999, Chung-chi Chang wrote:

> I tried to do the estimation from cox proportional hazards models 
> by specifying the baseline intensity follows "Weibull" forms.
> I don't see "coxph" codes in s-plus can do that, can't it? Or I have 
> to use "survreg" to find the scale parameter of the Weibull on 
> baseline intensity function?

The Cox proportional hazards model (note `Cox' is capitalized: he is a
living person) assumes a non-parametric baseline hazard. The Weibull model
is also a proportional hazard model and can be fitted by survreg, but you
will need to transform the coefficients to PH form: see Venables & Ripley
for worked examples. The coefficients will be different when you assume
a Weibull or non-parametric baseline hazard.

For information, Sir David Cox is (more or less) on record as saying that
he gave too little emphasis to parametric survival models in that
record-cited 1972 paper.

-- 
Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu.  To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message:  unsubscribe s-news

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