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

RE: st: Residual and influence diagnostics for conditional logistic regression


From   "Dupont, William" <william.dupont@Vanderbilt.Edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: Residual and influence diagnostics for conditional logistic regression
Date   Thu, 20 Oct 2005 11:45:26 -0500

Nick

I am a little confused as to who can and should do what.  Perhaps I am
using "keyword" incorrectly.  However, it is currently possible to use
-findit- to find user-written programs that have not been published in
STB or SJ by searching for "keywords".  In fact, brief experimentation
suggests that -findit- can find any phrase that is given in a program's
help file.

For example I submitted a program to calculate restricted cubic splines
to the RePEc data base.  To date, this program has not been described in
either STB or SJ.  However,
-findit cubic splines- finds the program without difficulty.  My help
file contains the phrase -implicitly specifies k knots located at xvar-
.  The command

-findit implicitly specifies k knots located at xvar-

immediately finds this program.  Hence, I infer that as long as authors
put appropriate keywords or phrases in their help file, users will be
able to find their programs using -findit-.

Incidentally, Ronna Cong does not mention conditional logistic
regression in her help file.  She does mention "Delta-Beta influence
statistics" and -findit- finds her program using this phrase.

I guess what I should have said was to urge all authors who contribute
programs to the Stata community to include lots of helpful keywords and
phrases in their documentation.  If I am correct that -findit- searches
all contributed help files for matches, then this fact is of value for
the Stata community to know.

Bill

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nick Cox
Sent: Wednesday, October 19, 2005 3:56 PM
To: statalist@hsphsun2.harvard.edu
Subject: RE: st: Residual and influence diagnostics for conditional
logistic regression

The Statalist FAQ advises that 

"User-written programs accessible from the StataCorp web site have
essentially the same status as those published on Statalist. Their
posting there is basically a matter of convenience to users and is not
an official endorsement by StataCorp." 

Strictly speaking, I believe that this applies even to those programs
that happen to have been written by people who were at the time of
publication employees of StataCorp, in this case Ronna Cong. 

As a consequence, programs such as -clfit- on
http://www.stata.com/users/ are not represented by any keywords at all,
in so far as they are not included in entries in the *.key files,
explained in the help for -keyfiles-, which are what -search- [narrow
sense] examines. 

Only if you use -findit- or -search, all- will you find -clfit-, but not
because of its keywords, but because of write-ups at that website that
are detected by an internet search. 

This is not to disagree with Bill on
the merits of -clfit-. But it is clear that there is no way that
user-written programs, other than those published via the STB or SJ, can
systematically and practically be included in the .key files. 

Nick
n.j.cox@durham.ac.uk 

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu]On Behalf Of Dupont, 
> William
> Sent: 19 October 2005 20:19
> To: statalist@hsphsun2.harvard.edu
> Cc: john.moran@nwahs.sa.gov.au
> Subject: RE: st: Residual and influence diagnostics for conditional 
> logistic regression
> 
> 
> John
> 
> Many thanks.  That is exactly what I was looking for.
> 
> Unfortunately it is a little hard to find.  Searches for "conditional 
> logistic regression", "residuals conditional logistic", "diagnostic 
> conditional logistic", "diagnostic logistic", etc. all failed to find 
> this useful program.  I would suggest that StataCorp expand the 
> keywords for this program in order to make it easier to find.
> 
> With best wishes.
> 
> Bill
> 
> 
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Moran, John
> (NWAHS)
> Sent: Monday, October 17, 2005 7:25 PM
> To: statalist@hsphsun2.harvard.edu
> Subject: RE: st: Residual and influence diagnostics for conditional 
> logistic regression
> 
> See "clfit" on the StataCorp users site
> 
> E-mail: john.moran@nwahs.sa.gov.au
> 
> 
> 
> *
> *   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/
> 
> *
> *   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/
> 

*
*   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/

*
*   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