Statalist The Stata Listserver


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

Re: st: Re: orthogonal estimation


From   Marcello Pagano <[email protected]>
To   [email protected]
Subject   Re: st: Re: orthogonal estimation
Date   Tue, 12 Dec 2006 09:24:22 -0500

If you wish to pursue this line of thinking outside the confines of Stata software, what you call "orthogonal regression" was studied by R.J. Adcock (1877,1878) and C.H. Kummell (1879) and Pearson (1901) under the label of what became known as "errors-in-variables" regression. Your perpendicular distance would presumably be considered when the errors in the dependent variable and independent variable have the same variance.

Adcock, R.J. (1877). Note on the method of least squares. Analyst 4 , 183-184.

Adcock, R.J. (1878). A problem in least squares. Analyst 5 , 53-54.

Kummell, C.H. (1879). Reduction of observed equations which contain more than one observed quantity. Analyst, 6, 97-105.

Pearson, K. (1901). On lines and planes of closest fit to systems of points in space. Philosophical Magazine, 2, 559-572.

m.p.

Kit Baum wrote:

Alvaro asks

Is there any program in STATA to perform orthogonal regressions? That
is, those which minimize the perpendicular distance between the
observations and the fitted line?


What you describe is known as limited-information maximum likelihood (LIML) regression. One of its features is that it is invariant to the choice of 'dependent variable'. OLS y on x minimizes residuals in y units; x on y minimizes residuals in x units. LIML y on x minimizes residuals perpendicular to the fitted line.

ssc describe ivreg2


Kit Baum, Boston College Economics
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html


On Dec 12, 2006, at 2:33 AM, statalist-digest wrote:


I have a set of observations, and for each observation I have  one
event date and n visit dates. How would I go about to get the date for
the last visit that occured before my event?
I'm guessing to use something like egen lastvisitdate=
rmin(eventdate-visitdate*), bu a) I'm not allowed to do that, and b) I
don't know how to only return positive values.
*
*   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 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index