Register now to learn Fabric in free live sessions led by the best Microsoft experts. From Apr 16 to May 9, in English and Spanish.
Hello,
I am currently working on a report and decided i really need a star scheme. Why? Because everywhere i read it says its the best structure. However i have a little issue with my data setup where id like some help from you experts out here.
The report is about systems(applications). There is a dataset connecting all other data sets. So that has to be the Fact table i believe. This one would look like this:
Portal |
Application ID |
Applications Title |
Supportgroup |
Created Date |
Modified Date |
The created date and modified date are the dates the record has been added or changed. for example (Created: 01-01-2017, Modified: 04-05-2018)
Now i have a couple of dim tables. these are connections to different systems where i can collect changes, incidents, issues, costs etc about that defined application. An example of these tables look like this:
Changes |
Change ID |
CreatedDate |
ReportDate |
CreatedThisMonth |
DeletedThisMonth |
Leadtime |
Supportgroup |
Impact |
etc. |
Incidents |
Incident ID |
CreatedDate |
ReportDate |
CreatedThisMonth |
DeletedThisMonth |
Leadtime |
Supportgroup |
Risk |
etc. |
Issues |
Issue ID |
Original estimate |
Work completed |
Work left |
CreatedDate |
ReportDate |
Iscurrent |
Applications Title |
etc. |
As you can see i collored where i can connect them. In Changes and incidents there is a creation date, a report date and some boolians deciden wether or not it was created/deleted that month. On Issues its different. There is also a created date but the status of each issue is measured by day, not by month.
Last step i would like to add a date tabel to the whole structure to be able to for example ask for everything between may 2019 and now. How would i go about connecting this date table to the current structure? here is the date table example:
Date |
Date |
Month |
Week |
etc. |
Thanks in advance.
Quality over Quantity
Solved! Go to Solution.
Hi @EricHulshof ,
I think in your case you are mixing up Dimension tables and Fact tables. Fact tables often represent events or transactions. Dim tables describe objects tied to those events or transactions. In your case, Applications are usually a Dimension table as they describe objects (Applications). The table Changes can be construed as a dimension table (every change is an object in itself) but is much more logically a Fact table (changes are every accumulating and they happen in a chronological order, and are tied to for example Applications (by Application ID). The same goes for Incidents and Issues, these are typically Fact tables.
Now think of it this way; you have multiple Fact tables (center of the Star model) that just happen to share similar Dimension tables (you don't need seperate Dimension table for Application for Incident or Changes table).
I hope this makes sense and helps you forward. It seems you have a complicated model and you will need to use a lot of USERELATIONSHIP() because on your date table you will create multiple inactive relationships with your fact tables. For example, you can create an active relation on CreatedOn and inactive relations on ReportDate.
Let me know if you have any questions 🙂
Kind regards
Djerro123
-------------------------------
If this answered your question, please mark it as the Solution. This also helps others to find what they are looking for.
Keep those thumbs up coming! 🙂
Proud to be a Super User!
Hi @EricHulshof ,
I think in your case you are mixing up Dimension tables and Fact tables. Fact tables often represent events or transactions. Dim tables describe objects tied to those events or transactions. In your case, Applications are usually a Dimension table as they describe objects (Applications). The table Changes can be construed as a dimension table (every change is an object in itself) but is much more logically a Fact table (changes are every accumulating and they happen in a chronological order, and are tied to for example Applications (by Application ID). The same goes for Incidents and Issues, these are typically Fact tables.
Now think of it this way; you have multiple Fact tables (center of the Star model) that just happen to share similar Dimension tables (you don't need seperate Dimension table for Application for Incident or Changes table).
I hope this makes sense and helps you forward. It seems you have a complicated model and you will need to use a lot of USERELATIONSHIP() because on your date table you will create multiple inactive relationships with your fact tables. For example, you can create an active relation on CreatedOn and inactive relations on ReportDate.
Let me know if you have any questions 🙂
Kind regards
Djerro123
-------------------------------
If this answered your question, please mark it as the Solution. This also helps others to find what they are looking for.
Keep those thumbs up coming! 🙂
Proud to be a Super User!
This actually makes alot of sence! Thanks alot, here, take this kudo 🙂
Quality over Quantity
Covering the world! 9:00-10:30 AM Sydney, 4:00-5:30 PM CET (Paris/Berlin), 7:00-8:30 PM Mexico City
Check out the April 2024 Power BI update to learn about new features.
User | Count |
---|---|
113 | |
100 | |
78 | |
76 | |
52 |
User | Count |
---|---|
146 | |
109 | |
106 | |
88 | |
61 |