I am trying to build a cube from a number of tables that conform pretty
closely to a "star schema".

There are a number of dimension tables in my star schema to represent
different dimension types.

One of the dimension tables can contain members for up to 8 discrete
dimensions - all of similar type - indexed 0-n.
The dimension table has an dimension index field that can take values from
0-n, it also has an member id field that is unique within a dimension (i.e.
for a given dimension index). I want to relate this member id field to
columns in my FACT table. Somehow I need to specify that where the
dimension index is 0 it's member id field is related to the first column of
the FACT table, when it is 1 the relationship is to the second column of the
FACT table etc. It is not obvious to me how to do this using the cube
editor. Indeed it may not be possible. It strikes me as a very simple and
common thing to want to do so thought it was worth putting a post here.

There will never be more that once instance of a dimension for all my other
dimension types so the relationship from the member id to the fact table is
straight forward in those cases.

Any suggestions would be greatly appreciated.

You could create a view for each discrete dimension contained in your base
table.
Then you can use the views as dimension sources and in cube editor as if
they were regular dimension tables.
HTH,
Brian
www.geocities.com/brianaltmann/olap.html

"Andy" wrote:

Thanks Brian,

That was exactly what I needed.
I had a hunch that there was a really simple solution that I was not seeing.

Andy

"Brian Altmann" <findme@thesignaturewebsite> wrote