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: RE: Choosing the bandwidth in a local linear regression
From
Jorge Eduardo Pérez Pérez <[email protected]>
To
"[email protected]" <[email protected]>
Subject
Re: st: RE: Choosing the bandwidth in a local linear regression
Date
Fri, 28 Oct 2011 14:17:40 -0400
The package "Estimation of quantile treatment effects with Stata" by
Frölich and Melly described in
http://www.stata-journal.com/article.html?article=st0203
includes an implementation of local linear regression, -locreg- with
bandwidth choice via leave one out cross validation.
To install it, type -net install st0203.pkg-.
_______________________
Jorge Eduardo Pérez Pérez
On Fri, Oct 28, 2011 at 1:36 PM, Nick Cox <[email protected]> wrote:
> -locpoly- is user-written, as you are asked to explain, but it has long been superseded by the official command -lpoly- (which is a modification). I don't see any support for least-squares cross-validation, so it seems that you need to write your own code to do this in Stata.
>
> Nick
> [email protected]
>
> Lina C
>
> I´m doing a matching estimation using the following command. I would
> like to calculate the bandwidth using least squares cross-validation,
> however I´m not sure whether this option or command is available in
> stata or whether I need to calculate it by my own. I would appreciate
> any help, many thanks,
>
> locpoly y d x, gen(yhat) at(d) biweight
>
> *
> * 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/
>
>
>
*
* 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/