Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: lock variable in model so that swaic cannot remove it


From   Nick Cox <njcoxstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: lock variable in model so that swaic cannot remove it
Date   Thu, 4 Oct 2012 12:25:51 +0100

Depends what you mean by "similar". I don't know anything based on
AIC, if that's what you mean, but other people will know much more
about that than I do, as I don't follow automated variable selection
programs closely.

I am a great fan of selecting variables by thinking about what makes
scientific sense and what seems to work well statistically, but that's
hard to automate, and not everyone has the luxury to do that.

Nick

On Thu, Oct 4, 2012 at 10:23 AM, Neophytos Stylianou
<neophytos.stylianou@postgrad.manchester.ac.uk> wrote:
> Thanks Nick
>
> Sorry for not having all details there. It is my first time in statalist. Ill make sure next time I will be more descriptive.
>
> By the way is there another way of getting an optimal model similar to swaic?
>
> Thanks,
> Neo
>
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nick Cox
> Sent: 04 October 2012 10:02
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: st: lock variable in model so that swaic cannot remove it
>
> -swaic- is a user-written command. Please remember that you are asked to explain where user-written commands you refer to come from. In this case I doubt that either author is a member of the list, and I know that Tom Steichen is very firmly retired.
>
> . search swaic
>
> Keyword search
>
>         Keywords:  swaic
>           Search:  (1) Official help files, FAQs, Examples, SJs, and STBs
>
> Search of official help files, FAQs, Examples, SJs, and STBs
>
>
> SJ-5-4  sg134_1 . . . . . . . . . . . . . . . . . .  Software update for swaic
>         (help swaic if installed) . . . . . . . . . Z. Wang and T. J. Steichen
>         Q4/05   SJ 5(4):607
>         has new varwidth option, now works with xi, and now works
>         after clogit
>
> STB-54  sg134 . . . . . Model selection using the Akaike information criterion
>         (help swaic if installed) . . . . . . . . . . . . . . . . . .  Z. Wang
>         3/00    pp.47--49; STB Reprints Vol 9, pp.335--337
>         command for stepwise model selection using the Akaike
>         information criterion (AIC)
>
> From a look at the help file, there appears to be no way to do what you want. -swaic- offers consenting adults a completely automated selection method. You can select forwards or backwards, but otherwise there is no way to interfere with small details.
>
> -nestreg- may offer a congenial alternative. -stepwise- itself also allowing locking.
>
> Nick
>
> On Thu, Oct 4, 2012 at 9:47 AM, Neophytos Stylianou <neophytos.stylianou@postgrad.manchester.ac.uk> wrote:
>
>> Can I please get some help with the swaic command.
>> I run a multiple logistic regression and then I perform the swaic, model command.
>> AIC takes a variable out that I definitely want to be kept in the final model no matter its significance or penalisation.
>> Is there a way of forcing swaic to keep a variable in the final model?
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


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