Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
 nagarajupinnibo
		
			nagarajupinnibo
		
		
		
		
		
		
		
		
	
			
		
		
			
					
		Hi ,
Is it possible to have multiple data models in single QVW application? If Yes, What scenarios we will use this.
Is it right thing to create multiple data models in single QVW.
Can you please explain this??
Thanks,
Nagaraju
 tresesco
		
			tresesco
		
		
		
		
		
		
		
		
	
			
		
		
			
					
		Couple of years ago the restriction(that too with Doc CALs only) was there, now you can have multiple data models in a single qvw.
 nagarajupinnibo
		
			nagarajupinnibo
		
		
		
		
		
		
		
		
	
			
		
		
			
					
		Hi tresesco,
Is it good practice to create multiple data models in Qlikview. What scenarios we will create these type of models.
Thanks,
Nagaraju
 
					
				
		
 danieloberbilli
		
			danieloberbilli
		
		
		
		
		
		
		
		
	
			
		
		
			
					
		As far as I know, there are limitations regarding the use of Document licenses. If your model moves too far away from the classic star schema (regarding a central fact table and dimension tables with significant smaller amount of distinct values) then it might be, that Document licenses do not work...but I never experienced something like that
 
					
				
		
YES POSSIBLE!
ON SCRIPT EDITOR!
USE SEPARATE TABS FOR EACH DATA MODEL
 tresesco
		
			tresesco
		
		
		
		
		
		
		
		
	
			
		
		
			
					
		In general, it should be avoided, because that could create confusion on front-end behavior (because of un-associated data model) and would be expensive in terms of script maintenance. However, you could face a situation where there would a small set of data that really is not part of main analysis data, in that case you might want to keep that part as a separate model.
 petter
		
			petter
		
		
		
		
		
		
		
		
	
			
		
		
			
					
		It is tecnical possible but seldom advisable since you can achieve the same effect with document chaining withou the disadvantages. Speed and size will be a challenge for a comprehensive multiple (non-associated) business data models.
Having data islands - one or a few minor tables is not uncommon. Mostly for UI covenience. The bulk data model is the business data model and should be one associated whole. If you have a problem getting the associations/links correctly it might be a sign that you either should split your app into two separate apps or spend even more time on your data modelling.
When it comes to the license terms a customer is specifically not allowed to have two or more business data models (larger) side-by-side with the Small Busines Edition of the server.
 petter
		
			petter
		
		
		
		
		
		
		
		
	
			
		
		
			
					
		Tabs will never separate your data model - tabs are purely a UI covenience
 
					
				
		
 jonathandienst
		
			jonathandienst
		
		
		
		
		
		
		
		
	
			
		
		
			
					
		Well is is possible ... but may not be desirable. You will have to design the model carefully to properly handle multiple fact tables with varying granularity and avoid undesirable circular loops and synthetic keys. You may land up with having duplicate copies of the dimension tables, which will be confusing for the users and it is more complex to encode expressions.
One of the great features of QV is its modular nature, and you lose many benefits of that if you try do to much in a single qvw file.
 
					
				
		
 datanibbler
		
			datanibbler
		
		
		
		
		
		
		
		
	
			
		
		
			
					
		Hi,
excuse me if this is a stupid question - but what exactly is the factor toggling whether to speak of one or of several data_models in a QlikView_app?
I know that several tables, which are usually linked in some way or other, form a data_model, of course. But:
Is every table that is not linked to any other a separate data_model? Then I have up to 6 or so in some apps.
If it may be that at some point, a Document_license stops working, then I guess it is relevant - I never had any issues with that up to now.
Or maybe everything starting from 2 linked tables?
Thanks a lot!
Best regards,
DataNibbler
