Re: Type 2 SCD Question
- From: "Jeje" <willgart@xxxxxxxxxxx>
- Date: Sat, 10 Feb 2007 00:04:01 -0500
Simply group by the name of the sales person.
if you group all the sales for Bob you'll have all the sales for Bob for all the territories where Bob was.
or you can group by the natural key (not the SCD surrogate key) if Bob's name changes over the time (his natural key don't change but the name can)
in a datawarehouse database you certainly have the right indexes to do this calculation quickly.
if this type of queries is common for you, try to use indexed views or cubes.
"JP" <JP@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message news:A5BF1138-6F60-4276-AC55-14ADD36B7AD5@xxxxxxxxxxxxxxxx
In designing a data warehouse I understand the importance of the.
dimension/fact relationship. I also understand why you would need type 2
dimensions.
The example I have seen over and over again demonstrates when a sales person
changes sales territories. Here you would need a type 2 dimension so that
the person's sales figures relate to the territory.
What happens when I want to get total sales for a person irrespective of the
sales terriroty? (Basically all sales attributed to one sales person). If a
sales person never left a sales region you could sum up all sales. But, if a
sales person left a sales region there would be fact records relating to the
person with more than one surrogate key. (When the sales person changes
territories a new record is added to the sales person dimension and the new
surrogate key is used to map to the facts).
- Prev by Date: Loading data
- Next by Date: Re: Office Web Components & Analysis Services
- Previous by thread: RE: Type 2 SCD Question
- Next by thread: Re: Publishing cubes in cognos
- Index(es):
Relevant Pages
|
|
