Forum Discussion
New composite model - DAX help - calculation cost * countrows()
Hi Amitchandak,Thanks for the suggestion.
So I seem to have found the "solution" which is more like a work around but doesn't really matter as the whole project is already a workaround 🙂
So it seems that with the new composite model sumx works fine only and only when you don't put columns from the dim into the visual.
As soon as you add any columns of the dim into the visual with the sumx measure it throws an error "An unexpected error occurred (file 'xldqas.cpp', line 510, function 'DAXDefManager-..."
I gess Microsoft will fix it at some point we're only about a month since the release of this new massive feature so can't complain at all 🙂
benfedit , One is a column and one is the measure. What are trying a new column or new measure
Measure using a common dim
total cost= Sumx(Values(Dim[dim]) , Min([Cost]) * countrows(Fact))
a new column, but that will work when both tables are import mode
total cost= Cost * countrows(Fact)
Hi Amitchandak,Thanks for the suggestion.
So I seem to have found the "solution" which is more like a work around but doesn't really matter as the whole project is already a workaround 🙂
So it seems that with the new composite model sumx works fine only and only when you don't put columns from the dim into the visual.
As soon as you add any columns of the dim into the visual with the sumx measure it throws an error "An unexpected error occurred (file 'xldqas.cpp', line 510, function 'DAXDefManager-..."
I gess Microsoft will fix it at some point we're only about a month since the release of this new massive feature so can't complain at all 🙂
I'm having this same issue right now, thanks for bringing it up!! I'm trying to use averagex, it works when shown as an aggregated number, but the second I try to bring in the columns from dimension table in the composite model it gives me that same error.
In the meantime, I've been able to get around it be avoiding the averagex and simply dividing the measure by a count of the dimension table rows, although this gives slightly different numbers in the aggregate, but I'll look into that.
