Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: dfbeta for survival time data


From   Ricardo Ovaldia <ovaldia@yahoo.com>
To   statalist@hsphsun2.harvard.edu
Subject   RE: st: dfbeta for survival time data
Date   Fri, 7 Oct 2005 11:33:46 -0700 (PDT)

--- Steinar Fossedal <Steinar.Fossedal@nhh.no> wrote:

> Are you sure about that? It seems to me it isn't (or
> I am doing
> something terribly wrong). -dfbeta- and -predict-
> with option
> dfbeta(varname) generate 'last estimates not found'

You are correct, I also get "last estimates not
found". Unfortunately I do not have my [ST] manual
with me to verify, but I thought that is what it said
when I read it this morning. I could be wrong.

That said, in the book "An introduction to survival
data analysis with Stata", by Cleves, Gould and
Gutierrez, on pages 194 and 195, they show an easy way
to do this:

1. Run stcox and save the efficient score residual for
all the covariates.
2. make a matrix of these residuals get the V-C matrix
from the regression.
3. Multiply the two matrices.
4. the element of this last matrix are the dfbetas.

example:
	. stcox var1 var2 var3, esr(esr*)
	. mkmat esr1 esr2 esr3
 	. mat V=e(V)
        . mat Inf=esr*V 

Ricardo

Ricardo Ovaldia, MS
Statistician 
Oklahoma City, OK


		
__________________________________ 
Yahoo! Mail - PC Magazine Editors' Choice 2005 
http://mail.yahoo.com
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index