Re: Pathological schema examples

> vicenti_at_ix.netcom.com (Richard Vicenti) writes:
> I am compiling a list of "challenging" schemas that a data =
> warehouse project might run into and have to cleanse.
>
> Examples include:
>
> Compound key from 2 tables
> Sort by Price WHERE IF Units < 1000 THEN Regular.Price ELSE =
> Discount.Price
> Richard Vicenti

<snip>

> vicenti.ix_at_netcom.com
>

>>>>
Richard,,

I was in the role of Data Warehouse Technical Architect for the last year, and I'm just finishing up some Standards & Guidelines for Data Warehousing here at my client's site.

I'm a little confused, though.. are you asking for challenging cases where the source data must be converted/transformed/integrated into the warehouse (what you refer to as 'cleanse')? Or are you asking about how to handle efficient query/analysis of data already in the warehouse (the use of your 'sort' example leads me to believe this).

If you're really serious about compiling a list, gimme an e-mail clarifying your requirements...

PS If you're looking for stock 'problem'/'solutions', I may not be able to help you. Many of the identified data problems required analysis by the Data Administrator and Source Data Manager; in other words, it required a case by case decision...

Gary Wong (gary.wong_at_cyberstore.ca)
Information Technology Consultant
(604) 360-1699 (Voice)
(604) 384-1178 (Fax) Received on Sun Sep 17 1995 - 00:00:00 CEST