Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: multivariate lpoly
From
Nick Cox <[email protected]>
To
"[email protected]" <[email protected]>
Subject
Re: st: multivariate lpoly
Date
Mon, 10 Mar 2014 18:19:50 +0000
I'd restrict the descriptor "multivariate (anything) regression" to
multiple responses.
Here you want multiple predictors: the official Stata offering
includes -mfp- and in addition there is a sibling spline family in
-mvrs- (SJ).
One difficulty with a local polynomial is quite how you report it,
except as a series of graphs and an estimated response for every
distinct observation.
Nick
[email protected]
On 10 March 2014 17:46, László Sándor <[email protected]> wrote:
> Hi all,
> I found some prior notes on the list about multivariate nonparametric
> regression, but no definitive conclusion. Is there something you would
> recommend now? Anything that scales as gracefully with the number of
> x's as possible? (The curse of dimensionality is always an issue with
> this kind of work.)
>
> Think of
> lpoly y x1 [x2 ...], generate(ybar)
> instead of the simple
> lpoly y x, generate(ybar)
>
> Some older code is available as kernreg.ado of Chuck Manski detailed
> in http://faculty.wcas.northwestern.edu/~cfm754/bounds_stata.pdf
>
> Thanks!
>
> Laszlo
>
> *
> * 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/
*
* 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/