Do not input private or sensitive data. View Qlik Privacy & Cookie Policy.
Skip to main content

Announcements
Qlik Open Lakehouse is Now Generally Available! Discover the key highlights and partner resources here.
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

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

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

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.