Stat 705 HW 20 Due Friday April 11, 2008.
-----------------------------------------

Consider the R dataset "mtcars" which can be found 
in the base R package. It consists of measurements 
of miles per gallon (mpg) -- the response variable 
of interest -- and 10 other quantitative variables
on each of 32 domestic and foreign automobiles from 
1973-1974 as reported in "Motor Trend" magazine.

(a) Perform "forward" and "stepwise" automatic model 
selection to find the best 2 linear regression model 
fits with predictors chosen from among an Intercept, 
the 10 predictive variables in the dataset, and the 
pairwise interactions among the variables
     cyl     disp   hp    drat     wt    am
Do your "forward" model selections in R, with k=4, and 
your "stepwise" (with both forward and backward steps)
selections in SAS using SLENTRY=.1 and SLSTAY=.05.
(b) Compare your best model obtained from each of the 
two approaches by means of any graphical display that 
you think makes sense. Do not do any other model-
fitting steps than those specified in (a). In this part 
of the problem, assess the quality of your "best" fitted
models graphically. One graph for each model together 
with some concluding narrative is enough.