I have a challenging problem That I'm trying to implement as an olap
solution. I'm running MSAS 2005 and looking for any suggestions or options on
how best to accomplish this task.

I have a fact table that stores percentages called "Fact A". I also have
another fact table that stores activity called "Fact B" I would like to
proivde the ability for an end user to define a percentage range that would
be applied as a filter to the first fact table.

so in the client tool, a user would select 50% and 100% as dimensions. the
results returned are the data in "Fact B" that also meets the criteria of
"Fact A"

Here is some example data to help visualize

Fact A:
User Percentage
-------------------
User1 40%
User2 55%
User3 75%

Fact B

User Minutes
------------------
User1 10 Minutes
User2 20 Minutes
User3 20 Minutes

Given the above scenario(50% - 100%) The results from "Fact B" are:

User2 20 Minutes
User3 20 Minutes

User 1 is excluded because his percentage(40%) does not fall into the
selected range.

I was wondering if this could be implemented using Dimension writeback?

Is there is another approach that would work? I want to avoid having to
write a "Custom" Report.

Any Ideas?

Dear Jason,

This is a quick note to let you know that I'm performing research on this
issue and may need a more time.
I'll give you a reply as soon as possible. If it's convenient for you,
could you please mail me (changliw (AT) microsoft (DOT) com) your mail address so that
I can timely response to you?

Sincerely,
Charles Wang
Microsoft Online Community Support

================================================== ====
When responding to posts, please "Reply to Group" via
your newsreader so that others may learn and benefit
from this issue.
================================================== ====
This posting is provided "AS IS" with no warranties, and confers no rights.
================================================== ====

try to use calculations and the new scope feature.

-- only leaf level of the percentage dimension, do nothing at the all level
scope(measures.minutes,
percentagedimension.percentageattribute.percentage attribute.members);
--(or leaves(percentagedimension))
if (measures.minutes / (measures.minutes, percentagedimension.allmember)) <>
percentagedimension.currentmember.membervalue then this =null;
end scope;

make sure that you have setup the value column in the dimension with a value
between 0.0 and 1.0 which is the value to use in the comparison.
I hope this will help you.

"Jason" <Jason (AT) discussions (DOT) microsoft.com> wrote

Thanks for the advice I will try to implement a prototype today and see if
this is going to work.


"Jeje" wrote:

I suggest you to consider also a many-to-many relationship design.
It could be not so intuitive, but I already used it in similar
scenarios.
I am finishing a paper I hope to publish very soon about the
many-to-many relationships design. Let's contact me (through the
Contact link on my blog) if you are interested in a preview (I am in
the reviewing phase).

Marco Russo
http://www.sqljunkies.com/weblog/sqlbi
http://www.sqlbi.eu

Jason wrote:

Dear Jason,
Thanks for your email.

Your requirement is not related with Dimension Writeback.
Essentially, my idea is the same as Willgart now, using a calculation and
scope.
If this issue persists or if you have any other special requirement, please
feel free to let me know.
I'm very glad for further research.

Sincerely,
Charles Wang
Microsoft Online Community Support

Hi,

I am interested in this issue. Would you mind letting me know the result of
the suggestions? If you need further assistance, feel free to let me know.
I will be more than happy to be of assistance.

Have a great day!

Charles Wang
Microsoft Online Community Support

================================================== ====
When responding to posts, please "Reply to Group" via
your newsreader so that others may learn and benefit
from this issue.
================================================== ====
This posting is provided "AS IS" with no warranties, and confers no rights.
================================================== ====