Sorry for my poor English, I'm trying to improve it
My problem is that I have two tables like this:
and other tables with iduser...
If I have to do a simple chart with sum(cash_earning) by data_earning or sum(cash_spending) by data_spending I wouldn't have problem (attached charts)
My problem appears when I try to put this two charts together in one simple chart because if I chose "data_earning" from table 1, sum(cash_spending) don't work and viceversa.
Anyway I found a solution to solve and avoid cartessian product... I load the tables two times, once with the right name dimension: data_earning and data_spending and other renaming these two dimension as "date"
I use the new date dimension to put sum(cash_earning) and sum(cash_spending) together (join tables by iduser and new date dimension)
I'm happy because I resolve it but I think that perhaps isn't necessary to load two times the same data.
I hope you understand me