cancel
Showing results for
Did you mean:
Helper I

## Insufficient memory after creating new calculated colom

Hi,

I have just made these 2 coloms but when entering the code for the last colom PBI gives me an error term saying I have insuffient memory.

I have all other programs closed and I had hoped my laptop has enough memory (4GB) to do most things in PBI.

Now I was hoping that someone could help me out to rewrite my code so that it takes less memory or some other tips to help me out.

PotentialBenefit =
VAR Hours= (Fct_TV[RemainingWorkHours] + Fct_TV[TotalHoursCumulative])
RETURN
IF([BenefitByPerson] = RELATED(Dim_Benefit[MaxBenefit]) , ,
IF( Adv >= RELATED(Dim_Benefit[MaxBenefit]) , RELATED(Dim_Benefit[MaxBenefit])-[BenefitByPerson],
IF(Hours < RELATED(Dim_Benefit[Hourlimit]) , ,
))))

RemainingWorkHours =
VAR Today= Today()
VAR CALC = (261-DATEDIFF(CONCATENATE("01-01-",Fct_TV[YEAR]),(Today-7),Day)*0.715)*8
RETURN
IF( CALC <0,0,CALC )

The code works on a dataset with less data.
The code PotentialBenefit lets me know how many subsidy when can get when a person works for us the remainder of the year.
However there are limit to this "MaxBenefit" and a person most work over more than a certain number of hours "Hourlimit" with Dim_Benefit containing about 30 different rows.
In FcT_TV it says who has worked x number of hours "TotalHoursCumulative" amongst many other items and also a calculation on how much subsidy we can already get for a person taken based on the last time we have data from him. The Calculation  is called "BenefitByPerson"

BenefitByPerson =
CALCULATE( AVERAGE(Fct_TV[AchievedBenefit]),
FILTER(
Fct_TV,
Fct_TV[PeriodID] = CALCULATE( MAX( Fct_TV[PeriodID] ),
FILTER(
Fct_TV,
Fct_TV[CandidateID] = EARLIER( Fct_TV[CandidateID]) )
)))

I tried to translate the code to English to make it more readable.

Thanks to everyone trying to help in advance.
Super User

Use DAX Studio (install it if you don't have it yet) and then look at the produced query to see where your cartesian products are greatest (hig numbers of records). Make adjustments to your queries accordingly.