February 25, 2014 at 1:36 am
Hi folks,
It was difficult to select the appropriate title for my question.
Background: The business analyst specify the following report (this is the simplified version):
Also to simplify sopose that I have two "Fact Tables"
- Purchase Orders:
- Inventory:
The challenge: I'm trying to build a cube in order to create a powerpivot report that looks like in the specs.
The drawback: the purchase orders table and the inventory table have both the Article and Country dimension. The problem arises when I want to include the time dimension in the pivot, because the inventory is not related to the time dimension.
Alternative 1: we added a dummy time id to the inventory table and then we could hide columns in the report with NULL values
Alternative 2: include the quantity or sold articles in the purchase order pro month as a calculated member
I'm not satisfy with this solutions and I think I'm missing something here from the theory.
Any comment or suggestion about how to tackle this issue will be appreciated.
Kind Regards,
February 25, 2014 at 8:11 am
Inventory measures are usually modeled based on:
1. A periodic snapshot fact table at a daily or monthly date grain and sometimes.....
2. An accumulating snapshot done for the current month.
A semi-additive measure such as LastNonEmpty or LastChild is usually used to measure the inventory on hand at the end of each period.
One measure group would use the purchase orders fact table and....
One measure group would use the inventory fact table.
Are you SURE you do not have any historical data regarding inventory (no auditing history of previous quantities on hand?)
February 26, 2014 at 8:43 am
i would suggest to use date dimension as a confirmed dimension with 2 fact tables. Create a dateid key in inventory fact table and use date dimension as confirmed dimension(shared dimension)
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply