Statalist


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

st: RE: cubic fit ci graph?


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: cubic fit ci graph?
Date   Mon, 13 Apr 2009 17:38:00 +0100

No. 

You could try finding the code for -qfit- and cloning it. 

On the other hand, you could ponder why StataCorp stopped at -qfit- and why no user-programmer has apparently filled this gap. 

While quadratics are quite often useful, at least locally, it seems rare that cubics are good overall fits to data. Once you have the idea that the data may wiggle and waggle a bit, it's more fruitful to consider other kinds of fit. -lowess-, -lpoly-, -fracpoly- and -rcspline- (the last user-written from SSC) are some alternatives likely to be much more useful. However, -rcspline- requires Stata 10. 

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

Chris Witte

I'm using version 9.2.  Does anyone know of a graphing command similar to _lfitci_ and _qfitci_, but that can handle the cubic power of "xvar"?


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   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