How to model various facts and conformed dimensions in a DMR Framework

[onsteinj]

Hi guys,

First post in this part of the forum. Hope you guys can help me out.

I am wondering how I should model my framework. I know of the best practice, using several layers:
1) Import            (importing tables from DB; source is starscheme dimensional DM's with conformed dimensions (Time and Currency) and Fact based dimensions)
2) Foundation      (defining relationships between facts and dimensions)
3) Consolidation   (makeup for better presentation, currently not used)
4) Dimensional     (DMR with the dimensions, facts, hierarchies and scopes)
5) Presentation     (Packages)

I am wondering how I should model, orden and present my Facts with their related dimensions? Each fact in a different Workspace? Each fact in a seperate Package? How do I make sure Cognos won't query 2 facts at the time?

Please some advice!?

Thank you in advance for your time.

Cheers,
J.

[MFGF]

Hi guys,

First post in this part of the forum. Hope you guys can help me out.

I am wondering how I should model my framework. I know of the best practice, using several layers:
1) Import            (importing tables from DB; source is starscheme dimensional DM's with conformed dimensions (Time and Currency) and Fact based dimensions)
2) Foundation      (defining relationships between facts and dimensions)
3) Consolidation   (makeup for better presentation, currently not used)
4) Dimensional     (DMR with the dimensions, facts, hierarchies and scopes)
5) Presentation     (Packages)

I am wondering how I should model, orden and present my Facts with their related dimensions? Each fact in a different Workspace? Each fact in a seperate Package? How do I make sure Cognos won't query 2 facts at the time?

Please some advice!?

Thank you in advance for your time.

Cheers,
J.

Hi,

The beauty of modelling according to best practice is that you don't need to worry about users querying multiple fact tables. If your modelling is correct, Cognos will generate a stitch query (query split) that interrogates each fact query subject and aggregates based on its individual level of granularity, and stitches together based on (coalesced) conformed dimension items. If your facts are held at different grains (eg Sales at a day level and Sales Targets at a Month level) you will need to add appropriate determinants to the relevant conformed dimension(s) (in this case the Date dimension) so that each fact will be aggregated correctly either side of the stitch.

Cheers!

MF.

[onsteinj]

Thank you!! I´ve forwarded your comments. That should do it!

p.s. How was your Grolsch?  ;)

Cheers,
J.

[MFGF]

p.s. How was your Grolsch?  ;)

Delicious! Thank you!! ;)