Forum Discussion
sam_hoccane
4 years agoHelper I
Multiplication with Previous Values
Excel File: Link Hello, Could you please help me write the following DAX Querry? I want to create a measure to calculate values highlighted blue in the predicted-CO2 emission column using the...
- 4 years ago
sam_hoccane
Your wishes are my command 😉 Here is the workable solution exactly as you wish
https://we.tl/t-bo914kjbfJPredicted_Measure_4 = VAR CurrentEmission = SUM ( Data[Emission] ) VAR CurrentMonth = SELECTEDVALUE ( data[Year_Month] ) VAR CurrentForcast = SELECTEDVALUE ( Data[forcast] ) VAR PreviousMonths = FILTER ( ALLSELECTED ( Data ), data[Year_Month] < CurrentMonth ) VAR PreviousN_Months = FILTER ( PreviousMonths, Data[forcast] = "N" ) VAR PreviousN_Month = MAXX ( PreviousN_Months, data[Year_Month] ) VAR PreviousN_Emission = SUMX ( FILTER ( PreviousN_Months, data[Year_Month] = PreviousN_Month ), Data[Emission] ) VAR YearMonthOnly = DISTINCT ( SELECTCOLUMNS ( PreviousMonths, "@YearMonth", Data[Year_Month] ) ) VAR NumberOfY_Months = COUNTROWS ( FILTER ( YearMonthOnly, [@YearMonth] > PreviousN_Month && [@YearMonth] <= CurrentMonth ) ) + 1 VAR Result = IF ( CurrentForcast = "N", CurrentEmission, PreviousN_Emission * POWER ( 1.15, NumberOfY_Months ) ) RETURN Result
Dhacd
4 years agoResolver III
Hi sam_hoccane,
Please find another way to achieve the desired solution below,
PredictedEmissions =
Var TableN = filter('Table','Table'[forecast]="N")
Var MaxMonth = CALCULATE(MAX('Table'[Month]),TableN)
Var LastEmission = CALCULATE(MAX('Table'[CO2 Emission]),'Table'[Month]=MaxMonth,all('Table'))
Var Difference =datediff(MaxMonth,'Table'[Month],MONTH)
Var Predicted = (LastEmission *power(1.15,Difference))
return if( 'Table'[forecast]="Y",Predicted,'Table'[CO2 Emission])
Screenshot Attached
If this post helps, then please consider accepting it as the solution to help the other members find it more quickly.
Regards,
Atma.
tamerj1
4 years agoCommunity Champion
Dhacd
Good job!
A few issues here.
We don't know whether "N" and "Y" cycle overtime. Therefore the maximum Month with "N" might be after the first "Y" row.
If a month reading is missing from the data (not taken or not recorded for any reason) the DATEDIFF will not provide a correct result. COUNTROWS seems to be more reliable in this case.