Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hello all,
I would like to understand how my set analysis could have a weird effect.
I have a Pivot table with country as dimensions and a count of a measure as expression.
And with this count as expression, there are values for each country ( normal values, everything is ok )
But what was needed as expression was a Sum of this same values, so i changed it.
But now, there is only one country ( country is the dimension ) who has value and all other ones are null.
I don't have identifiers in my set analysis about country and i tried with a identifier <Country= > to have all countries in case but nothing changes.
There is only one country ( one value of dimension ) who has the whole value when i use Sum
And when i use count, all dimensions have values...
Maybe i didn't explain very well, sorry 🙂
But i would to know if you had some ideas where this problem could come from
thanks
Maybe the measure isn't always numeric and then sum() could not return a valid result.
- Marcus
Thanks for response
but can't be that it's really always numeric
Please provide more information - the used expression, some screenshots from the object + data + data-model.
- Marcus
count({$<Year={"$(=max(Year))"},[expatriation country]=, [Code of contract]={'A}>} EndMonth*FLAG_NUMB*alt([Int. Soc.],0)/100)
Sum({$<Year={"$(=max(Year))"},[expatriation country]=, [Code of contract]={'A}>} EndMonth*FLAG_NUMB*alt([Int. Soc.],0)/100)
The values are numeric (EndMonth*FLAG_NUMB*alt([Int. Soc.],0)/100))
And the pivot table : As you see, in this case i put both set analysis, the count and the sum.
I suggest to check all possible values from:
EndMonth*FLAG_NUMB*alt([Int. Soc.],0)/100
again which values are really there. Especially the alt() function may mostly return the default-value of zero and then the entire multiplication/division will return zero. For the count() is this of course a valid value which could be counted.
- Marcus