Tuesday, February 28, 2006

Re: RE: st: behavior of -areg-

Dear Rich,

Thanks a lot for taking your time to help out. This is certainly very good to know. I also noticed that -xtreg, fe- does not have the same problem as -areg-, i.e. with -xtreg, fe-, the same collinear variable drops out. Does -xtreg, fe- use a different precision?

Thanks again, -Radu

2006/2/28, Richard Gates <rgates@stata.com>: > In response to Radu's <raduban@gmail.com> query on why -areg- was not dropping > collinear variables, Scott <smerryman@kc.rr.com> made an interesting > demonstration on how machine precision can affect whether -_rmcoll- drops > variables: > > > You replicate this with the grunfeld dataset and the mvalue variable. ... > > I would like to add to this demonstration where I show the > dangers of having single precision variables (float) and not > promoting them to double precision with caution. > > Note that single precision will have only about 7 digits > of precision while double has about 16 corresponding to > 24 and 53 bits for the mantissia: > ...

* * 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/


Links to this post:

Create a Link

<< Home

This page is powered by Blogger. Isn't yours?