I don't usually write to myself - but earlier this evening I wrote:

>Tim, I'm a bit confused by this reference to 'combinations' of class levels. I though >what the CLASS statement did was to create a design matrix of dummy variables - >effectively one additional variable (per observation) for each level of the class >variable. In the context of PROC MEANS/SUMMARY (but not things like TABULATE), I don't >quite see where 'combinations' (which I read, perhaps incorrectly, as relating to >'crosses') of class levels come into it. I'm obvioulsy missing something!

Having now actually done the thinking I should have done before, I now realise what it being said. I was, of course, thinking of a situation in which there was only a single CLASS variable. If there are two or more such variables, then it is obviously true that all 'combinations' of different levels of each of the 'nested' class variables *do* have to be considered - the theoretical 200 million limit on combinations may therefore be nothing other than the limit on the number of variables that can be handled in a SAS step (not dataset).

Regards John John

----------------------------------------------------------- Dr John Whittington, Voice: +44 1296 730225 Mediscience Services Fax: +44 1296 738893 Twyford Manor, Twyford, E-mail: johnw@mag-net.co.uk Buckingham MK18 4EL, UK CompuServe: 100517,3677 -----------------------------------------------------------