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 at the end of May, and its replacement, statalist.org is already up and running.


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

st: RE: Plotting nonparametric reg. line alng with linear prediction line


From   Nick Cox <n.j.cox@durham.ac.uk>
To   "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu>
Subject   st: RE: Plotting nonparametric reg. line alng with linear prediction line
Date   Wed, 26 Oct 2011 13:45:02 +0100

Sounds like -lpoly-. 

As someone said, "The great thing about standards is that there are so many to choose from." I've met people who regard lowess, fractional polynomials and cubic splines as the standards. I suppose local linear regression looks standard from one mathematical-statistical viewpoint; it's just a pity that it often works rather poorly in practice. 

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

Ivica Rubil

Thanks Nick.
To be more specific, i just want the standard nonparametric stuff like
Nadaraya-Watson or local linear regression. How would either of these
fit in what you sent:
scatter y x || lfit y x || line ynpfitted x, sort

Nick Cox

-rcspline- from SSC is an example of a scatterplot smoother. (Deep
down, it is just another linear regression, but it is flexible in the
way that most nonparametric enthusiasts seek.) You could combine that
with a linear fit in this way.

. sysuse auto
(1978 Automobile Data)

. rcspline mpg weight, addplot(lfit mpg weight)

. rcspline mpg weight, lcolor(orange) addplot(lfit mpg weight,
lcolor(blue))

The line colors are changed in the last example, because with default
defaults (-scheme s2color-) the lines come out as red and green, a
graphic combination better avoided because it is problematic for many
readers. (I use -set scheme s1color-, in my profile.do.)

Nick

On Wed, Oct 26, 2011 at 8:26 AM, Nick Cox <njcoxstata@gmail.com> wrote:
> As there are many nonparametric regression methods, you may need to be
> more specific, but
>
> scatter y x || lfit y x || line ynpfitted x, sort
>
> is a general recipe where -ynpfitted- is the prediction or smoothing
> from your nonparametric method.
>
> Nick
>
> On Wed, Oct 26, 2011 at 8:02 AM, Ivica Rubil <irubil@eizg.hr> wrote:
>
>> is there anybody who knows how to plot the scatter points, the linear
>> prediction line and a nonparametric regression line on together on
one
>> graph. I plotted scattered points and the linear proediction line,
but I
>> cannot add the nonparametric regression line. Any suggestions?

*
*   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   |   Site index