*
* Example starting on page 79
*
open data food.dat
data(format=free,org=columns) 1 40 food income
*
linreg food
# constant income
*
* The instruction PRJ computes fitted values. For a specific set of explanatory
* variables, you use the option XVECT=||x1,...,xn||. Note that you include the 1
* for the CONSTANT, which in RATS is treated like any other explanatory variable.
*
* After PRJ, the fitted value is in %prjfit and the standard error of prediction
* is in %prjstderr.
*
prj(xvect=||1.0,20.0||)
*
* DISPLAY (disp or dis for short) displays computed values, possibly with
* descriptive strings. If you want to control the number of digits displayed, you
* can use "pictures" before the values. For instance, if you want to show %prjfit
* with four digits right of the decimal, you would add *.#### before the %PRJFIT
* in the next instruction.
*
disp "Predicted Value at income=20" %prjfit
*
* The %invttest function gives two-tailed critical values, so to get the .95
* percentage confidence band, you go from - the .05 critical value to + the .05
* critical value.
*
disp "Confidence Interval" %prjfit-%invttest(.05,%ndf)*%prjstderr %prjfit+%invttest(.05,%ndf)*%prjstderr