s-news
[Top] [All Lists]

Summary "coxph"

To: s-news@lists.biostat.wustl.edu
Subject: Summary "coxph"
From: Mika Nakazono <mika@msi.co.jp>
Date: Fri, 25 Jul 2003 10:59:26 +0900
Importance: Normal
Hi,

I have received two email about my question.
Thanks to Terry Therneau and John Loewy!

My original question is;
****************************************************************************
****
I have a question about "coxph".
I'm using S-PLUS 6.0 for Windows on Windows XP.

Is there any method or option which is equivalent to "TIES=DISCRETE"
option for PROC PHREG on SAS?
**********************************************************

Terry Therneau's answer is;
****************************************************************************
****
  The "exact partial likelihood" (as labeled in Cox's original paper) is
the appropriate form to use for discrete data (as discussed in Kalbfliesch
and
Prentice).
  SAS calls this ties=discrete, Splus calls it ties=exact.
**********************************************************

John Loewy's answer is;
****************************************************************************
****
I believe the options are "Breslow" "Efron" and "exact". I think the one
that corresponds to the SAS option is "exact". The problem is if you have a
relatively large dataset with lots of ties the program will not run or take
a very long time to run.
**********************************************************

Thank you very much.
Mika

Mika Nakazono
Mathematical Systems Inc.
Japan




--------------------------------------------------------------------
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>
  • Summary "coxph", Mika Nakazono <=