Skip to main content

New to Qlik Sense

If you’re new to Qlik Sense, start with this Discussion Board and get up-to-speed quickly.

Announcements
QlikWorld 2023, a live, in-person thrill ride. April 17 - 20, 2023, in Las Vegas! REGISTER TODAY
cancel
Showing results for 
Search instead for 
Did you mean: 
StacyCui
Creator
Creator

Qlike Sense Data Modeling

Hi, I am a new in Qlik Sense. I meet an issue of data modeling in my work. Business Background: We have two order tables that are respective CY ORand PY OR. In business scope. We have district structure that Territory→Branch→Office. Each year, we adjusted the structure a little, like maybe office A belongs to Branch AA in last year, but this year, it maybe belongs to Branch BB. So, the district structure in CY and PY order table, it doesn't completely same. Q: When I design the modeling, Is it necessary to build a district structure table as a dimension table to connect with the fact tables. Besides, for PY district structure, we don't has a record. I have to distinct the structure in PY OR table to obtain. Kindly appreciate for your reply.

Labels (2)
1 Solution

Accepted Solutions
marcus_sommer
MVP & Luminary
MVP & Luminary

The officially recommended data-model is a star-scheme by merging (concatenating and/or mapping/joining) all facts within a single fact-table and n dimension-tables. And this fits quite well to your data because both order-tables contain the same - just another year - and could be simply concatenated. The district-information could be outsourced in a dimension-table - just linked per Office to plot the current structure against all data or per Office & '|' & Year to respect any changes in regard to the Year.

- Marcus

View solution in original post

4 Replies
marcus_sommer
MVP & Luminary
MVP & Luminary

The officially recommended data-model is a star-scheme by merging (concatenating and/or mapping/joining) all facts within a single fact-table and n dimension-tables. And this fits quite well to your data because both order-tables contain the same - just another year - and could be simply concatenated. The district-information could be outsourced in a dimension-table - just linked per Office to plot the current structure against all data or per Office & '|' & Year to respect any changes in regard to the Year.

- Marcus

StacyCui
Creator
Creator
Author

Hi, Marcus. I'm so glad to receive your reply. Your explanation is very useful for me. But I want to double check it. I draft it in an Excel. The right table is the district dimension table, is that right as you mentioned?

StacyCui_0-1660745876302.png

Usually, in our business, people would like to select each layer alone. So I divided them into 3 columns as a dimension table. I think it also works, is it right?

StacyCui_1-1660746117197.png

marcus_sommer
MVP & Luminary
MVP & Luminary

Yes, each layer alone is right. Means the district table contains in the end 5 fields: Territory, Branch, Office, FY + the key-field to the order-table of Office & FY which might be also a numeric by using an autonumber().

- Marcus

StacyCui
Creator
Creator
Author

Exactly. Thank you so much. I will try it in practice.