** Airline Example in SAS **

*(1) Create the log transformed data and 
the monthly dummy varaibles **; 

Data Air; set sashelp.air; 
 LAIR = log(air); 
 month = month(date); 
 array M(11); 
 do i=1 to 11;  M(i) = (month=i); end; 
 drop i; 

*(2) Print and plot the data **; 
 Proc print noobs data=air; 
   
 axis1 label=(angle=90 font=centb h=1.2); 
 Proc gplot data=air; 
 plot (air lair)*date/vaxis=axis1; 
 plot air*date air*date air*date/overlay vaxis=axis1; 
 plot lair*date lair*date lair*date/overlay vaxis=axis1;
 symbol1 v=none i=join c=black; 
 symbol2 v=none i=smooth10 c=red; 
 symbol3 v=none i=smooth60 c=blue;
run;  

* (3) LAIR seems to have the better statistical behavior. 
Analyze LAIR

   * (A) Using REG; 
PROC REG; MODEL LAIR=DATE M1-M11; run; 
   * (B) Using GLM; 
PROC GLM; CLASS MONTH; 
    MODEL LAIR = DATE MONTH/SOLUTION; 
run;