Use the DISJCLUS macro.

1. Issue the following commands to SAS (with appropriate
   substitutions for "U:\").

LIBNAME GFLib 'U:\dm\sasdata' ; 

LIBNAME RCLib 'U:\dm\RC' ;

2. Apply the DISJCLUS macro.  How to fill in the blanks 
is described on pages 126-129 of the textbook. 

   a. I used the following to obtain {GFLIB.DIABET140.rtf},
      which provides plots that may suggest an appropriate 
      number of clusters.
 
	GFLib.diabet1		yes
	3 			yes
 	x1 x2 x3 x4 x5
	id			[blank]
	C:\dm\RC\		C:\dm\RC\
	40			word

   b. I used the following to obtain {GFLIB.DIABET141.rtf},
      which explicitly identifies the observations assigned 
      to each cluster and performs a canonical discriminant 
      analysis treating the cluster assignments as values 
      of a pseudo-response.
 
	GFLib.diabet1		[blank]
	3 			yes
 	x1 x2 x3 x4 x5
	id			[blank]
	C:\dm\RC\		C:\dm\RC\
	41			word