Total Outstanding measure stops when visual filter is applied.

[Anonymous]

I suspect this issue is some sort of conflict between the Total Outstanding Measure and the Visual Filter i am trying to apply.

As the attached pbix file shows in the Weekly Incident Outstanding tab the line continues on through all subsiquent weeks 

Under the Aged Incidents tab i have added a visual filter to show only tickets which are older than 30 days yet the blue line stops in mid air at Week 24. 

In an attempt to diagnose I mocked up this table quickly which implies there are no tickets over 30 days for weeks 25, 26 and 27.

Sample Data

Even so you would think total outstanding would continue counting down from 390 to 355 then 327 and finally 304.

Im aware the Weekly Outstanding visual has line continuing on the same number past week 27 (this is a result of removing/hiding data for the sample the production file is Ok.

Thanks in advanced 

Outstanding Aged Incidents over 30 Days  

[HotChilli]

Nested measures can be hard to deal with (and debug!) so let's start with something obvious.  The calendar and fault tables have cross-filter direction of both so that needs to be changed so that the calendar table filters the faults and not the other way around.

That makes a change to the line (it also exposes that using a YEAR slicer and ISOWEEK in the visual is returning results that you don't want for ISOWEEK 53 - that's because the ISOWEEK 53 results really belong to the previous year)

I'm not entirely happy with the extra filter clauses in some of these measures e.g. 

Arrived = CALCULATE(COUNTROWS(Faults),Faults[DateCreatedKey])

so maybe you should also look at those.

- also is there a reason to have so many records in the date table (220 years!) ?

[Anonymous]

@HotChilli 

Apolgies for the long gap in replies.

I've always disliked the Calendar in this report. Its one from the datawarehouse and used across the board (this is my only project), I was under the assumption it would be a bit cleverer and only show the Years and ISOWEEK of the appliable to the data in the Faults (FACT) table. Clearly not

Removing Cross Filter from the Calender>> Faults, seems to flatten the line on the graphs, indercating that removing the filter has somehow borken the calcualation 

In the meantime i will have a look at simpliying the calculation (even though its returning the correct result) 

Thanks