Rolling total of calculated measure

[pereir20]

I have developed measures to identify % variances between years, however, I am attempting to "sum" these together to create a rolling total. I was able to successfully create a formula to do this, however, when I published the report I received the error "visual has exceeded the available resources". I am trying to identify a more streamlined method of calculating the rolling total.

Example: The % increase between 2016 and 2018 is 8.3% and the price between 2017 and 2018 is 7.9%. I would like the 'Price Δ Aggregate' field show the running total of 'Price Δ % (Aggregate)' as shown below. 

Delivery Year	Price Δ % (Aggregate)	Price Δ Aggregate
2016	0.00%	0.00%
2017	8.30%	8.30%
2018	7.90%	16.20%

I was able to use the below measure to calculate the correct 'Price Δ Aggregate', however, I am receiving the resource error. Is there a more efficient way to perform this calculation?

Price Δ Aggregate = CALCULATE(SUMX(ADDCOLUMNS(SUMMARIZE(Data,Data[Delivery Year]),"Priceagg", [Price Δ % (Aggregate)]),[priceagg]),FILTER(ALLSELECTED(Data),Data[Delivery Year]<=MAX(Data[Delivery Year])))

[OwenAuger · Accepted solution]

@pereir20

I would try simplifying your measure down to something like this:

Price Δ Aggregate =
SUMX (
    FILTER (
        ALLSELECTED ( Data[Delivery Year] ),
        Data[Delivery Year] <= MAX ( Data[Delivery Year] )
    ),
    [Price Δ % (Aggregate)]
)

Filtering (a subset of) the entire fact table with FILTER ( ALLSELECTED ( Data ) ... ) is probably contributing to the performance problems. You can get away with just filtering Delivery Year.

Does this still return the correct result, and is there any performance improvement?

Regards,

Owen

[pereir20]

Thanks Owen, this solution is much more efficient reducing the processing time dramatically. I appreciate your help!