<>
*****
sysuse auto, clear
reg price weight rep tr
g y=_b[_cons]+_b[weight]*weight+_b[rep78]*rep+_b[trunk]*trunk
*compare to built-in solution
predict pred
compare pred y
*****
HTH
Martin
_______________________
----- Original Message -----
From: "Chris Witte" <[email protected]>
To: <[email protected]>
Sent: Monday, March 30, 2009 9:55 PM
Subject: st: using coefficients from model
I am using stata 9.2. I would like to know how to call the coefficients
from the estimation results matrix e(b) and use them in an equation. For
example, if my matrix e(b) has values for "_cons" and coefficient "test", I
would like to calculate the answer to an equation y = [_cons] + x*[test].
How can I do this?
The reason why I want to do this is because I am trying to find the value
_fracpoly_ adjusts its estimates by. If I use this output from my
_fracpoly_ estimation:
study1twsg Coef. Std. Err. t P>t [95% Conf. Interval]
Itemp__1 15093.12 4454.121 3.39 0.012 4560.802 25625.45
Itemp__2 -10666.18 3136.84 -3.40 0.011 -18083.63 -3248.734
Itemp__3 1945.114 569.9302 3.41 0.011 597.4436 3292.785
_cons .0053322 .0008244 6.47 0.000 .0033828 .0072815
Deviance: -113.24. Best powers of temp among 164 models fit: -2 -2 -2.
and I plug in these coefficients with my 'x' having a value of 25, my
linear regression equation for predicting y is:
y=.0053322+(15093.12*25^-2)+(-10666.18*25^-2*log(25))+1945.114*25^-2*(log(25))^2
However, this equation does not yield the same value as _predict_ (or
_fracpred_), and this is because fracpoly secretly adjusts its constant.
If I can calculate the answer to this equation and put it into a variable,
I can then find the difference between this variable and the value that
_fracpred_ gives.
Thank you for your help. Hopefully this isn't too confusing.
Chris
*
* 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/
