Skip to main content
Submitted

Clone dimensions within and between semantic models

  • December 17, 2019
  • 6 replies
  • 1 view

Forum|alt.badge.img

When creating a tabular model, I might use the same DWH view twice. E.g. when creation a supply chain model, there should be a 'shipping location' and a 'receiving location' with both dimensions built on the same DWH view 'locations'. Once one of both dimensions is built in a tabular model, it would save time when one could just copy/paste the dimension and just change the name (besides the name, both dimensions are identical).

6 replies

Forum|alt.badge.img
  • Known Participant
  • January 7, 2020

This is certainly a feature that would save us a lot of time creating role-playing dimensions - just think about the date-dimension with a lot of attributes and hierarchies.


Forum|alt.badge.img
  • Known Participant
  • January 9, 2020

Agree that this is needed. Even if you are working in Tabular and following their recommendation to use only one date dimension per model, any time you use Type II SCD history you'll need to create role-playing copies of those tables as well if you want to be able to analyze on both current and historical data. Adding these tables by hand - often to multiple models - can get really laborious, especially if you are renaming fields. 

In general, additional automation in the semantic layer is a common and popular request.  I've compiled some of these suggestions below. 

Copy Models: 

https://support.timextender.com/hc/en-us/community/posts/360041497911-Clone-Semantic-Model

Copy Endpoints: 

https://support.timextender.com/hc/en-us/community/posts/360017509026-Semantic-Model-Copy-Endpoints

Copy Tables (very similar request to this one): 

https://support.timextender.com/hc/en-us/community/posts/360047240851-Copy-table-structure-between-semantic-models

Copy Measures: 

https://support.timextender.com/hc/en-us/community/posts/360040073711-Copy-Paste-Measures-in-Semantic-Layer

Semantic-Layer Snippets: 

https://support.timextender.com/hc/en-us/community/posts/360055585071-Semantic-DAX-Snippets

https://support.timextender.com/hc/en-us/community/posts/360042754271-Snippets-for-semantic-models

 


Forum|alt.badge.img
  • Known Participant
  • February 16, 2020

Hi,

Since Discovery Hub is an automation tool it should not stick to the structure of SSAS Tabular Model where you define a dimension al over again when you want to setup e.g. a role playing dimension. 
Discovery Hub should have a feature in which you can define the dimension and then allow the user to add that dimension to the tabular model (inherit or copy - more or less like in OLAP Multi-dim).  Such a feature would allow us to work much faster in Tabular Model.

Best regards,

Peter


Forum|alt.badge.img
  • Known Participant
  • February 17, 2020

Hi Peter, 

I actually love this idea.  It feels far more "automated" to me, and I'd be very happy to see something like this implemented.


Forum|alt.badge.img
  • Known Participant
  • February 28, 2020

Agree with this idea since Peter and I have posted this suggestion numerous times already over the past years (from at least 2014). You should be able to copy / paste any node for that matter.


In the end it is just metadata that is being moved around in the projectrepository => this would truly match their credo of being a real data warehouse AUTOMATION tool.

 

If you export the project to XML file.

  • Copy/past all OLAP related stuff
  • replace the primary keys by a newly generated GUID (can do this in whatever tool you like)
  • import the XML again
  • you now have a copy of your olap server or cube or dimensions

This shows it is actually not that difficult to add this functionality in the TX application.

 

Please also check Visual Studio which has the option to read the metadata from an existing OLAP database and redeploy it to another (also a way to copy an existing database / cube)


  • Participating Frequently
  • February 12, 2021

Thanks for your product feedback.

We have passed your input to our R&D department for further evaluation.

The roadmap of our product is focused on a split in ODX, MSW, Semantic Layer and Customer Portal.

With that split we will see new features and functions appear that might exceed your expectations or make the request obsolete.