Forum Discussion
weird behaviour (cross join?) in very simple model - table visualization
The data is not entirely clear to me, but your weird results are due to the id column in fact1.
You'll see that the result of the sum of fact2 is repeated for each id from fact1. To solve this you could simply remove the id column from your table, but I'm not sure what the id in fact1 is meant to represent.
If you were to remove the id column you would get the sum of both fact1 and fact2 by category.
Hi Jog, removing id from the visual avoids the behaviour, but I need the table to show the data at the lowest level of detail (fact1).
My question here is why if I drag the column "reseller_sales" (making an implicit sum measure) to the visual works perfect and If I I use the measure fact2_reseller =sum(fact2[reseller_sales]) does not work?
thanks!
Hey Shoemaker,
Alright, gotcha.
First I would like to pose the question if your results are really 'correct' with an implicit measure. 😉
Not saying the results are 'wrong' per se, but at the very least confusing for an end user. Even if your users are okay with it now.. will new users be okay with it in the future? Or the same users with bad memory in the future. 😉 Combining two fact tables of different granularities in one model is generally a bad idea/practice. It'll set you up for trouble.
As for the reason the implicit measure 'works' and explicit doesn't. I'm not 100% sure, because I'm not entirely positive how an implicit measure gets calculated. But I suspect it has something to do with context transition. The explicit measure below returns the result you'd expect without any bidirectional relationships and other shenanigans.