Archives of the TeradataForum

Message Posted: Wed, 16 Jun 2004 @ 15:55:04 GMT

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

If you already know the data contains only numbers and letters, then the following will select all rows with pure numeric values:

sel col1 where upper(col1) (cs) = lower(col1);

This works by converting the alpha characters to uppercase (left side) and lowercase (right side) and then doing a casespecific comparison. If the two sides are not equal (for a particular row), then you know column1 does not contain pure numerics (e.g. digits 0 through 9).

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