Data flow: How to distribute a value on different entities

Alessandro Carosso

Hi everyone,

I have to distribute one value per customer and month on a cube per customer, month, item and agent, based on the turnover cube, also with customer, month, item and agent. On these 2 last cubes, "customer", "article" and "agent" are in sparsity. I made a dataflow with the following steps:
a: value per customer / month to be distributed
b: turnover cube with "references" function for customer month.
c: turnover cube
d: (value distributed per month, customer, agent, article) result of c * (a / b).

 

Question 1) Is the approach correct?
Question 2) if it is correct, why are only some customers calculated?

Previous Member

Hi Alessandro,

 

what do you mean by "references function"? From reading your case i am assuming that you used the "total by" feature so my answer is based on this on:

 

Good news first: You don't have to worry about your sparsity since you are using the same sparsity in C and D and your are doing a multiplication you only get values for your sparse combinations anyway.

 

I think the total by feature is not the right feature to use in your case. You usually use the total by if you want to show the aggregated values of a parent entity to one element of your child entity (e.g. the values for the year 2016 when looking at April 16). In your case you want an aggregation over the whole entities and not an aggregation into a parent entity.

 

Therefore i suggest that you create a data flow writing the values of your turnover cube into a cube with article and agent as dimensions and use this cube in your dataflow instead of the one with the total by feature.

 

Please let me know if this works for you.

 

Mark