Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: multiple rolling regressions


From   "Erasmo Giambona" <e.giambona@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: multiple rolling regressions
Date   Tue, 13 May 2008 12:46:23 -0400

Hi Austin,
This code works except that it produces missing values for d starting
from firm number 5918, despite the fact that y and x are availables. I
thought it was a memory problem, so I set the memory to 850m, but I
still get missing values. Does it mean that I have to run the code for
different subsamples?
Thanks.
Erasmo

On Fri, May 9, 2008 at 2:04 PM, Austin Nichols <austinnichols@gmail.com> wrote:
> Erasmo--this problem seems constantly to have a "new" wrinkle.  Try this:
>
> clear
> input firm year y x
>      1 1980 20 23
>      1 1981 22 34
>      1 1982 20 19
>      1 1984 22 34
>      1 1985 20 19
>      1 1986 28 34
>      1 1987 30 19
>      1 2004 33 48
>      1 2005 30 45
>      1 2006 35 50
>      2 1980 15 23
>      2 1981 55 34
>      2 1982 29 19
>      2 1983 15 23
>      2 1984 55 34
>      2 1985 29 19
>      2 1986 28 34
>      2 1987 30 19
> end
> su year, meanonly
> loc ny=r(max)-r(min)+1
> loc fy=r(min)
> tempfile allyrs
> preserve
> drop _all
> set obs `ny'
> g year=`fy'+_n-1
> save `allyrs'
> restore
> append using `allyrs'
> fillin firm year
> drop if mi(firm)
> tsset firm year
> sort firm year
> g double d=.
> qui forv i=3/`=_N-2' {
>  loc f=`i'-2
>  loc l=`i'+2
>  if firm[`f']==firm[`i'] & firm[`l']==firm[`i'] {
>  cap reg y year in `f'/`l' if firm==firm[`i']
>  if _rc==0 {
>  if e(N)>2 {
>  predict double ye if e(sample), resid
>  reg x year in `f'/`l' if firm==firm[`i']
>  predict double xe if e(sample), resid
>  reg ye xe
>  replace d=_b[xe] in `l'
>  drop ye xe
>  }
>  }
>  }
> }
> li, noo sepby(firm)
>
> On Fri, May 9, 2008 at 11:58 AM, Nick Cox <n.j.cox@durham.ac.uk> wrote:
> > The -fillin- does not work as you hoped because _no_ observations exist
> > anywhere for years 1988-2003. -fillin- rectangularises, rather than
> > "fills in"!
> >
> > You'd need to fix that by hand by explicitly adding some observations
> > with those years but missings otherwise. However, for the example data
> > given here, that sounds futile if I understand what is being tried, as
> > your gap is too big to be bridged, anyway.
> >
> > Erasmo Giambona
> >
> > Sorry Austin. Here are the codes and outputs for my questions 1 and 2.
>
> *
> *   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/
>
*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index