Archives of the TeradataForum

Message Posted: Tue, 26 Dec 2006 @ 21:34:11 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Hi Victor,

Yes, I must have missed your comment/question. I've experienced many "lags" receiving mail from this list. Or maybe too much early eggnog?

The math function to calculate standard deviation for two columns looks fine to me, but what if you have 10 columns, for 100? I was trying to "broaden" the response.

The main thing I was trying to illustrate is that SAS data step functions are not the same Teradata aggregate functions. SAS functions "execute" across each row (or observation or record). The proper syntax for the SAS std deviation function is "STD(v1,v2,...vn)".

It's not unusual for SAS programmers to work with highly de-normalized tables (as SAS datasets), often with hundreds of columns. You probably would never see something like that in any formal database, especially Teradata. It takes a little while for old fashioned SAS programmers (like me) to "re-think" traditional SAS techniques into SQL-based equivalents.

Bob

	<Prev	Next>		<<First	<Prev	Next>	Last>>