Sum of most recent values across time

[APMarsh]

Hello,

I have tried various functions and have not been able to come up with a solution to this.

I have a series of reports that are submitted with a number of projects (reference number). I need to track the latest EFC across time at a reference number level.

I have tried numerous methods, one of which is:

var test3XX = 
SUMX(
  SUMMARIZE('Prog report data',
  'Prog report data'[Reference Number],"Reference number",
  CALCULATE(SUM('Prog report data'[Estimated Final Cost]),
    USERELATIONSHIP('Date'[Date],'Prog report data'[Lodgement date]),
    FILTER('Prog report data','Prog report data'[Lodgement date]=MAX('Prog report data'[Lodgement date])),
    FILTER(ALL('Date'[Date]),'Date'[Date]<=MAX('Date'[Date])),
    FILTER(ALL('Date'[Date]),'Date'[Date]<>BLANK())
    )),
  [Reference number]
)

The issue seems to be that all the functions only return ONE MAX/LATEST date - so when filtering on a project (reference) it is ok, however, when slicing and dicing at an aggregated level, it doesnt recognise the latest lodgement date for each reference number.

The data looks something like this (redacted).

Reference Number	Report Number	Lodgement date	Estimated Final Cost
ABC.0002.1819I.SXG	ABC.PR03.1819.SXG	09/04/2020	$                          111,111.11
ABC.0002.1819I.SXG	ABC.PR01.1819.SXG	11/02/2020	$                          111,111.11
ABC.0002.1819I.SXG	ABC.PR05.1819.SXG	12/06/2020	$                          111,111.11
ABC.0002.1819I.SXG	ABC.PR06.1819.SXG	17/07/2020	$                          111,111.11
ABC.0002.1819I.SXG	ABC.PR08.1819.SXG	15/09/2020	$                          111,111.11
ABC.0004.1819E.FGH	ABC.PR03.1819.FGH	01/07/2020	$                          286,729.76
ABC.0004.1819E.FGH	ABC.PR06.1819.FGH	09/11/2020	$                          268,972.07
ABC.0004.1819E.FGH	ABC.PR04.1819.FGH	18/09/2020	$                          268,972.07
ABC.0004.1819E.FGH	ABC.PR10.1819.FGH	10/12/2020	$                          268,972.07
ABC.0014.1819E.FGH	ABC.PR03.1819.FGH	01/07/2020	$                          397,962.54
ABC.0014.1819E.FGH	ABC.PR06.1819.FGH	09/11/2020	$                          326,679.50
ABC.0014.1819E.FGH	ABC.PR04.1819.FGH	18/09/2020	$                          324,312.77
ABC.0014.1819E.FGH	ABC.PR10.1819.FGH	10/12/2020	$                          326,679.50
XYZ.0001.1819I.SXG	XYZ.PR01.1819.SXG	05/12/2019	$                            23,961.11
XYZ.0001.1819I.SXG	XYZ.PR02.1819.SXG	28/01/2020	$                            23,961.11
XYZ.0001.1819I.SXG	XYZ.PR03.1819.SXG	25/02/2020	$                            23,961.11
XYZ.0001.1819I.SXG	XYZ.PR05.1819.SXG	11/03/2020	$                            23,961.11
XYZ.0001.1819I.SXG	XYZ.PR06.1819.SXG	13/05/2020	$                            23,961.11
XYZ.0002.1920A.LIP	XYZ.PR03.1920.LIP	09/06/2020	$                              5,112.99
XYZ.0002.1920A.LIP	XYZ.PR01.1920.LIP	13/05/2020	$                              5,112.99
XYZ.0003.1920K.FGH	XYZ.PR05.1920.FGH	10/12/2020	$                          455,466.64
XYZ.0003.1920K.FGH	XYZ.PR03.1920.FGH	04/11/2020	$                          530,471.67
XYZ.0003.1920K.FGH	XYZ.PR02.1920.FGH	23/09/2020	$                          460,416.67
XYZ.0003.1920K.FGH	XYZ.PR01.1920.FGH	21/08/2020	$                          460,416.67
XYZ.0005.1920K.FGH	XYZ.PR05.1920.FGH	10/12/2020	$                      1,400,645.44
XYZ.0005.1920K.FGH	XYZ.PR03.1920.FGH	04/11/2020	$                      1,481,465.83
XYZ.0005.1920K.FGH	XYZ.PR02.1920.FGH	23/09/2020	$                      1,355,555.56
XYZ.0005.1920K.FGH	XYZ.PR01.1920.FGH	21/08/2020	$                      1,355,555.56
XYZ.0006.1920K.FGH	XYZ.PR01.1920.FGH	21/08/2020	$                          144,166.67
XYZ.0006.1920K.FGH	XYZ.PR02.1920.FGH	23/09/2020	$                          144,166.67
XYZ.0006.1920K.FGH	XYZ.PR03.1920.FGH	04/11/2020	$                          144,166.67
XYZ.0006.1920K.FGH	XYZ.PR05.1920.FGH	10/12/2020	$                          144,166.67

[amitchandak]

@APMarsh , Try a measure like

calculate(sum(Table[Estimated Final Cost]), filter(Table,Table[Lodgement date] = calculate(max(Table[Lodgement date]), allexpcept(Table,table[Reference Number]))))

[APMarsh]

Hi Amitchandak,

Thanks for the suggestions. I did try this function as well - and it doesnt get the result i need. It needs to be able to sum across time (similar to a cumulative...but only summing the latest value). 

For example:

The sum on Wed night is 38.

The sum on Thurs night = 2+7+20+3+1+6+0 = 39

When i use any of the below code it works ok when i filter on a single reference number, however when I filter at the aggregated level, it doesnt work.

var test2 = 
CALCULATE(SUM('Prog report data - CUBE'[Estimated Final Cost]),
    FILTER('Prog report data - CUBE','Prog report data - CUBE'[PR lodgement date]=
    CALCULATE(MAX('Prog report data - CUBE'[PR lodgement date]),
    USERELATIONSHIP('Date'[Date],'Prog report data - CUBE'[PR lodgement date]),
    FILTER(ALL('Date'[Date]),'Date'[Date]<=MAX('Date'[Date])),
    FILTER(ALL('Date'[Date]),'Date'[Date]<>BLANK())
    )))

var test3 = 
SUMX(
    SUMMARIZE('Prog report data - CUBE',
    'Prog report data - CUBE'[Submission Number],"Sub NumberX",
    CALCULATE(SUM('Prog report data - CUBE'[Estimated Final Cost]),
        //FILTER('Prog report data - CUBE','Prog report data - CUBE'[PR lodgement date]=MAX('Prog report data - CUBE'[PR lodgement date]))
        USERELATIONSHIP('Date'[Date],'Prog report data - CUBE'[PR lodgement date]),        
        FILTER(ALL('Date'[Date]),'Date'[Date]<=MAX('Date'[Date])),
        FILTER(ALL('Date'[Date]),'Date'[Date]<>BLANK())
        )
    ),[Sub NumberX]
)

var test4 = 
CALCULATE(test3,
    //FILTER('Prog report data - CUBE','Prog report data - CUBE'[PR lodgement date]=MAX('Prog report data - CUBE'[PR lodgement date]))
    USERELATIONSHIP('Date'[Date],'Prog report data - CUBE'[PR lodgement date]),        
    FILTER(ALL('Date'[Date]),'Date'[Date]<=MAX('Date'[Date])),
    FILTER(ALL('Date'[Date]),'Date'[Date]<>BLANK())
)

var test5 = 
CALCULATE(
    SUM('Prog report data - CUBE'[Estimated Final Cost]),
    FILTER('Prog report data - CUBE','Prog report data - CUBE'[PR lodgement date]=CALCULATE(MAX('Prog report data - CUBE'[PR lodgement date]),
        ALLEXCEPT('Prog report data - CUBE','Prog report data - CUBE'[Submission Number])))
)

 Plotting over time doesnt work at aggregated level:

[amitchandak]

@APMarsh , Try a measure like

Measure =
VAR __id = MAX ('Table'[Reference Number] )
VAR __date = CALCULATE ( MAX('Table'[Lodgement date] ), ALLSELECTED ('Table' ), 'Table'[Reference Number] = __id )
CALCULATE ( Min ('Table'[Estimated Final Cost] ), VALUES ('Table'[Reference Number] ),'Table'[Reference Number] = __id,'Table'[Lodgement date ] = __date )

[APMarsh]

Hi @amitchandak ,

Have you had any further thoughts about this one. I still haven't managed to find a solution. 

[APMarsh]

Thanks Amitchandak,

I tried the below measure but it seems to give an uexpected result. 

var max1X = MAX('Prog report data - CUBE'[Submission Number])

var dateXX = CALCULATE(MAX('Prog report data - CUBE'[PR lodgement date]),ALLSELECTED('Prog report data - CUBE'),'Prog report data - CUBE'[Submission Number]=max1X)

return

CALCULATE(MIN('Prog report data - CUBE'[Estimated Final Cost]),VALUES('Prog report data - CUBE'[Submission Number]),'Prog report data - CUBE'[Submission Number]=max1X,'Prog report data - CUBE'[PR lodgement date]=dateXX)

I will keep trying and let you know when i get it.