*   R. Carter Hill, William E. Griffiths and Guay C. Lim,  
*   Principles of Econometrics, Fourth Edition, Wiley, 2011.

* Chapter 4.6  Log-Log Models (pages 156-7)  

READ (newbroiler.dat) / names

* The data set has missing values
SET SKIPMISS
STAT / ALL

* Use the GENR command to create Log transformed observations
GENR LQ=LOG(Q)
GENR LP=LOG(P)

* Equation estimation 
* The RSTAT option reports the goodness-of-fit measure: 
*     R-SQUARE BETWEEN ANTILOGS OBSERVED AND PREDICTED
OLS LQ LP / LOGLOG RSTAT PREDICT=YHAT

* Get predictions
GENR QHAT=EXP(YHAT)*EXP($SIG2/2)

SORT P QHAT
GRAPH QHAT P / LINEONLY

STOP