Discussion Board for collaboration related to QlikView App Development.
Hi QlikView Experts,
I am stuck in the data modelling process. I have four fact tables. All share the same two dimensions (e.g. Dim1 and Dim2).
Moreover two fact tables share two dimensions (e.g. Dim3 and Dim4) which doesn't exist in the other two fact tables.
Following the two approaches I've tested. The first one works (because I want to do a selection for "Special_ID" in Fact1) but QlikView creates a synthetic key and I don't know how to avoid this.
For the second approach I've used generic keys, but can't do selections for the "Special_ID" :-(.
Should I live with the synthetic key from the first approach or do you have any ideas how best to solve this.
Many thanks in advance!
You may want to check out a couple of articles written by hic.
Fact Table with Mixed Granularity
I would probably concatenate all four fact tables into one long table. This would result in a solution that is more efficient than one with a link table.
Thank you both for your responses,
but after concatenating the tables some of my calculations within the diagramms don't work anymore ;-(.
So can we say that link table architecture is a Star Schema ?