Forum Discussion
Best Way for work with Multiple Fact Tables
itchyeyeballs Thanks for your replys, are very helpful, So I am going continue investigating about this, because I think there must be some modeling technique to get a star schema, even in complex models.
Try concatening them in just 1 big fact table. I used this technique in a model with 5 fact tables and it worked great.
You can add a column with an identifier for each concatened fact table so when you use a formula you can filter in DAX only the rows that met the condition for the fact table you want.
Tell me if it works for you.
Cya
alexanderg i have to admit when I first read your question( a long time ago), it sounds very strange, I did not get it, now that I have worked a little bit with Qlik, i see where you are coming from, Tabular handle multiple fact tables very gracefully , no need for concatenation or the Link table technique.
thanks smizgier for your suggestion, at the moment I have two alternatives for this situation, one is that you mention, concatenate tables into one, but I think this technique is most useful when table structures are similar, the other option I am considering is to create a central table containing the fields in common dimension, allowing the joining of two or more fact tables against a common set of dimensions. To be more specific this technique consists of:
1- In the query editor get the fact tables and concatenate the common key dimension fields into one compuond key.
2- create a new aditional table by appending the distinct values from the fact tables. this will be the central table that contain the new compound key with the key dimension fields.3- in the model, you have to link each fact table and each dimension table to the new central table
it seems that this works...
That's like creating a linked, right? it shud works.
Hope it works
