At 14:23 25/03/99 -0500, HERMANS1 wrote:

>Offhand I'd say that the SAS System already has the capability that you want >(even if it does require use of the SQL procedure that some may consider >unaesthetic). You can specify a physical ordering of the observations in the >dataset by either creating an ordered view of a dataset in SQL and SETting the >view in a data step, or by defining the output of a datastep as a view (say, >Xvw, and sorting the dataset in a SQL statement (say, ....SELECT * FROM Xvw >ORDER BY UCASE(var), LENGTH(var), ABS(numvar). For those inclined to data step >programming can regard these methods as special SORT procedures that sort on >functions of variables. Sig

Sig, yes, as I've responded to others, those are all options - and it wouldn't surprise me if performance was comparable in all cases - but I'm not sure that alters what I said. As we often discuss here, the fact that a particular functionality can be achieved 'in other ways' is not necessarily an excuse for not allowing a user-friendly syntax - is that not the point of any 'higher level' computer language? A vast amount of the SAS semantics and syntax could be eliminated without making it impossible for us to find a way of coding the functionality we want - but, just as we seem to like having the (technically redundant) SQRT() function, I think it wouldn't be unreasonable to be able to sort 'BY UPCASE(var)'. Indeed, for all I care, the compiler might interpret such an instruction in terms of a 'SQL solution'; that would be fine by me!

Kind Regards

John

---------------------------------------------------------------- Dr John Whittington, Voice: +44 (0) 1296 730225 Mediscience Services Fax: +44 (0) 1296 738893 Twyford Manor, Twyford, E-mail: medisci@powernet.com Buckingham MK18 4EL, UK mediscience@compuserve.com ----------------------------------------------------------------