11-04-2014 05:19 AM
I want to build a kind of DWH for tables that form the basis for a large variety of VA reports. These tables will have, e.g. all bills, all members, all providers, etc. Based on the reports I will define filters for the bits of data that I need. Often, I will want to combine information from different tables (e.g. total of bills from all members in a given city). I am wondering what the best way to so is. In particular:
- If I create a star schema or other query in the data builder, my impression is that a new table in memory is created. This practically multiplies the size of the bills table (the biggest I have) with the number of reports I want to base o it (for all different combinations of data). Is there a way to have a kind of 'view', i.e. a data query that doesn't result in a table in memory (but, of course metadata does get created), but only when it gets called (at the moment of the creation of a report) and then doesn't exist anymore once the visualisation is made?
- If I create one big mastertable, it means that all data of a given member or provider is repeated on all lines. I think there is no way around that if I want one data source?
- If I have the separate tables for bills, members and providers, it means that I cannot combine the city of a member with the amount on a bill, without a data query, right?
All input will be very helpful!
11-19-2014 10:43 AM
It seems now that it is best to use Star Schemas. The only annoyabnce could be that between fact tables and dimension tables one can only use one linking variable, not a combination.