In the future will you want a report of "Customers by Country" or any other
reports for which Country is a dimension? If so, then set up a separate
Country dimension. If not, and "Country" is just an attribute of "Customer"
(like "Age" is an attribute of "Customer"), then you're fine leaving it in
the Customer dimension. Personally I'd advise creating a separate Country
dimension, however, since most businesses like their data sliced and diced
by Country, by State/Territory, and by just about anything else you can
think up.
[quoted text, click to view] "Nestor" <n3570r@yahoo.com> wrote in message
news:op.tx23xvju83lx0t@nestor.gateway.2wire.net...
> Hi all,
>
> Some questions on identification of Slowly Changing Dimensions. When
> exactly do we determine if a dimension is classified as SCD?
>
> For example, if I have a Customer Dimension table
> Customer Key Name Country
> 1001 John USA
>
> If John changed his state to Australia, we can justify this as a SCD Type
> II (for historical recording) and insert the table so that we have
>
> Customer Dimension
> Customer Key Name Country
> 1001 John USA
> 1002 John Australia
>
> and in the fact table we can have something like
>
> Fact Table
> factid Customer Key Field1
> 1 1001 $10
> 2 1002 $10
>
> and there we have our SCD.
>
> However, if we normalize original dimension table into
>
> Customer Dimension
> Customer Key Name
> 1001 John
>
> Country Dimension
> Country Key Country
> 2001 USA
> 2002 Australia
>
> Then the fact table will be
> FactID CustomerKey CountryKey Field1
> 1 1001 2001 $10
> 2 1001 2002 $10
>
> and they both solve the same problem. The question is, when do you split
> the attributes out as a separate dimensional table and when do you make it
> a SCD? Are they any best practise for this?
>
> Cheers
>
>
>
>