Archives of the TeradataForum

Message Posted: Wed, 09 Oct 2002 @ 13:26:26 GMT

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

I completely concur. When I read the question first, I saw two possible situations:

1) the data has been standardized and cleansed prior to the load. In this case comparison for *identity* is straight forward and there are many ways of optimizing the comparison check.

2) the data has not been standardized nor cleansed prior to the load. In this case I would not use SQL. Reason is, it is not powerful enough to do a thorough job. A better idea is to extract the data and run it through specialized tools (Vality, Trillium etc.) They do address cleansing and matching efficiently and also allow cleanup of some info.

Cheers

Thomas

Thomas Buehlmann

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