in a parent child dimension, can a child belong to more than one parent?
Has anyone had any experience with such a situation?


TIA

--
Message posted via http://www.sqlmonster.com

No, a child can not belong to more than one parent.

But I heard this kind of requirement sometimes, so I'd like to hear some
experience from others too.

Ohjoo Kwon
www.olapforum.com


"rama via SQLMonster.com" <forum (AT) SQLMonster (DOT) com> wrote

A new feature SQL 2005 Analysis Services is full support of many-to-many
dimensions.
--
Dave Wickert [MSFT]
dwickert (AT) online (DOT) microsoft.com
Program Manager
BI SystemsTeam
SQL BI Product Unit (Analysis Services)
--
This posting is provided "AS IS" with no warranties, and confers no rights.

"Ohjoo Kwon" <ojkwon (AT) olap (DOT) co.kr> wrote

i tried using the many to many feature, but that results in the 2 m:n
attributes being in separate dims. So we cannot create a user hierarchy for
the user to use on the UI

Also, this feature seems to roll up child aggrgates to ALL parents. I would
like to aggregate a child to any one parent only based on the fact details.

tia

--
Message posted via http://www.sqlmonster.com

If there are level1, level2 and M:N relationship between the two levels, I
understand AS 2005 can solve it with many-to-many dimension as following.

MainFact -> lev1 <-- Fact12(M:N) -> lev2
Of course, it does not support user hierarchy having M:N relationship among
levels. Actually important is that it automatically avoids data duplication
at aggregation level.

Here, I wonder how to solve next case.

If there are many levles having M:N relationship between levels.

MainFact->lev1<-Fact12(M:N)->lev2<-Fact23(M:N)->lev3<-FactMN(M:N)->levN

Does this make sense?

Ohjoo


"Dave Wickert [MSFT]" <dwickert (AT) online (DOT) microsoft.com> wrote

makes sense.

There might be an impact on the performance though - AS will have to query
all those joins...

--
Message posted via http://www.sqlmonster.com

If then, is it possible to set the latter case at Dimension Usage view of
cube builder? I'd like to know how to set continuous many-to-many
dimensions.


"rama via SQLMonster.com" <forum (AT) SQLMonster (DOT) com> wrote

These structures are only scanned went processing dimensions. Processing the
core fact tables are not impacted by a M:M -- remember we do processing of
dimensions *separately* from partition processing. These joins are dimension
only.
--
Dave Wickert [MSFT]
dwickert (AT) online (DOT) microsoft.com
Program Manager
BI SystemsTeam
SQL BI Product Unit (Analysis Services)
--
This posting is provided "AS IS" with no warranties, and confers no rights.

"rama via SQLMonster.com" <forum (AT) SQLMonster (DOT) com> wrote