TM1 Forum

Hello,

for a cube I chose to have the four dimensions: Customer, Year, Facts, Product. One customer in the customer dimension belongs to a special sector.
In the analysis some of the questions to answer are:
- How much sales revenue per sector?
- How much sales revenue per country?

Would you create two more dimensions, one for country and one for sector? Or would you create two attributes (country and sector) in the customer dimension?
What are the advantages/disadvantages?

Thanks a lot!

Toto

There are a number of options and the answer lies in how do you want to be able to analyze data.

1) Put a Sector and Country attribute on the Customer dimension. This is not a very good option because you would only be able to filter subsets of customers off this and it wouldn't provide subtotals and totals in the cube. You would be relegating the cube, at least in this respect, to more of a relational database and lose some of the OLAP functionality.

2) Build multiple hierarchies in the Customer dimension. You could have a By Sector tree and also a By Country tree.

3) Create a separate Sector dimension and Country dimension.

What you have to ask yourself, or your customer, is how do you want to be able to analyze the data and how is that data organized. If you choose option 2 then you won't be able to see the intersection of Country and Sector because they are in the same dimenson. The only way to see that would be option 3.

Hello Tomok,

thanks a lot for your detailed answer, it helped me a lot!

Regards,

Toto

Another consideration is does a single customer trade in multiple countries or is there scope for that to happen in the future? If so the seperate dims would be the best way forward.