Multiple QVD loads with same column names

[JamGardner]

Hi, hoping for some help; 

I currently have multiple QVD's with different types of performance in each - as they are people performance tables they all have similar columns, apart from the results. i.e. Date, Organisation, Team Area, Team, Team Code. 

When I load these in the script editor I get multiple synthetic keys - which have worked in the past however when I add a more detailed qvd which has slightly different columns (which I try to rename in the load) i.e Team area is called Serving area, the number of these keys grows (+34) and then causes performance issues as the load seems to time out. 

The question I have is how to load these without causing the issues I am encountering. 

Thanks

[edwin]

you need to model your data in terms of facts and dimensions.  typically, all facts are concatenated and fields are conformed (meaning if two fields have the same meaning you want them to have a single field name: Account, customer Account, Account Number, etc.. can just be called Account).  if there are different granularities of data and you are confident they should not be concatenated then you either build a link table or link two tables using a synthesis of the keys and not include the individual keys in one of the tables:

if two tables has Account and SubAccount as common fields, create a new field called LINK = Account & '|' & SubAccount, then retain the individual fields on the side you are confindent has all the values.  of course you can also rename the same fields in one of the tables differently.  the link field will connect the two tables and the individual field (original and renamed) can be used as dimensions if needed (especially when comparing)

hope that helps.

[QFabian]

Hi, if all of your tables shared a significant part of the fields, you should put CONCATENATE between each load.

In that way, you are going to get just 1 table, with all of the records from the source tables loaded.

All common fields are going to works as common filters, and the different fields can be use in visualizations.

if you don't want to hace all the tables in 1 table, you can use qualify *, in order to force that every field have fdifferents names according on the table's names.

Here a link that you can check for a PDF about data modeling in qlik.

https://community.qlik.com/t5/Qlik-Chile/The-QlikView-Data-Model-Best-Practices/ta-p/1760708