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

# Re: st: Residuals from sequence of regressions

 From DBernstein@lecg.com To statalist@hsphsun2.harvard.edu Subject Re: st: Residuals from sequence of regressions Date Mon, 18 Oct 2004 13:38:13 -0400

```Hello,

I'm new to replying to messages, so I hope I post this correctly.
Anyway, here is a solution to the residual question that should work:
If anyone has a more efficient solution, I would be interested in seeing
it.

levels year, local(Y)

gen residual=.

for each y of local Y {
reg  y  x1  x2  x3  if year==`y'
predict temp, residual
replace residual=temp if year==`y'
drop temp
}

Regards,

David J. Bernstein, Ph.D.
Managing Economist
LECG Inc.
dbernstein@lecg.com
Office:  215-546-4950 ext. 239
Fax:      215-546-3568

Subject: st: Residuals from sequence of regressions

Hi,

I'm relatively new to Stata, and have a seemingly simple question but,
despite consulting physical and on-line manuals, I have been unable to
answer it so assistance would be appreciated.

I have a large balanced panel of N observations over T years. I would like
to do a cross-sectional regression for each year, and record the residuals
for the regression for each of the N observations for that year. E.g. for
1960, I want each of the N observations to have a new variable which is the
residual from the 1960 cross-sectional regression across the N
observations.
Then I want the same for 1961 etc.

I know how to run individual regressions for each year, but I am unsure how
to store the residuals from the regression for each year. To run the
regressions, I type:

Sort year
By year: quietly reg y x1 x2 x3

Normally, after a regression you can type "predict res, residual", but I am
not sure how to incorporate it in the above framework. Stata doesn't seem
to
let me type more than one command after the "by" function. E.g. it won't
let
me type something like

By year: quietly reg y x1 x2 x3, predict res, residual

and typing "predict res, residual" at the end merely gives me the residuals
for the cross-sectional regression for the last year.

Any help would be gratefully received.

Thanks, Alex

*
*   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/

------------------------------

End of statalist-digest V4 #1765
********************************

*
*   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 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index