Average Distinct Count Over a set time period (ignoring date filter)

[marcritchie]

Hi there,

I have a report where I want to have a dynamic distinct count of users who have booked time to a project within the last month of a certain date. This is the information that I want to show with the Department Size being the dynamic count and the Capacity worked out from that:

The report has a filter on the reporting date (a Week Commencing date which is in a date table). The relationship is shown here:

 

 

What I want to show is basically the average distinct count of the Employee ID over the previous 4 weeks.

The problem I have is that the slicer causes the result of my measure to be fixed over time whereas I want to see over time the change in this average based on the previous 4 weeks of wherever it lies on the timeline.

As you can see here the Capacity (which uses the Calculated Department Size measure) remains constant over time (as I think it will always pick up the reporting date selected in the slicer). What I want is e.g. 5th December should work out the average over its previous 4 weeks (14th Nov-5th Dec), not the 4 weeks up to the reporting date. This chart isn't connected to the Reporting Date slicer but defaults to last 2 months.

 

This is where I am with the DAX measure:

Calculated Department Size = 
CALCULATE (
    DISTINCTCOUNT ( Timesheet[Employee ID] ),
    ALL ( 'Reporting Date'[Week Commencing] ),
    Timesheet[Booked Hours] > 0,
    'Reporting Date'[Week Commencing]
        > MAX ( 'Reporting Date'[Week Commencing] ) - 28,
    'Reporting Date'[Week Commencing] < MAX ( 'Reporting Date'[Week Commencing] )
)

I don't know whether one measure can do the job or if I need to set up one to cover the reporting date and one to cover the trend over time. 

Thanks,

Marc

[v-rongtiep-msft]

Hi @marcritchie ,

Please have a try.

Change all to allselected.

Calculated Department Size = 
CALCULATE (
    DISTINCTCOUNT ( Timesheet[Employee ID] ),
    ALLSELECTED ( 'Reporting Date'[Week Commencing] ),
    Timesheet[Booked Hours] > 0,
    'Reporting Date'[Week Commencing]
        > MAX ( 'Reporting Date'[Week Commencing] ) - 28,
    'Reporting Date'[Week Commencing] < MAX ( 'Reporting Date'[Week Commencing] )
)

If I have misunderstood your meaning, please provide your pbix file without privacy information.

Best Regards

Community Support Team _ Polly

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[amitchandak]

@marcritchie , In case you can have week rank in date or week yeat tbale

a new column

Week Rank = RANKX(all('Date'),'Date'[Year Week],,ASC,Dense) //YYYYWW format

CALCULATE(AverageX(values('Date'[Year Week]) ,calculate(DISTINCTCOUNT ( Timesheet[Employee ID] ))), FILTER(ALL('Date'),'Date'[Week Rank]>=max('Date'[Week Rank])-8 && 'Date'[Week Rank]<=max('Date'[Week Rank])))

Rolling 4 = CALCULATE(AverageX(values('Date'[Year Week]) ,calculate(DISTINCTCOUNT ( Timesheet[Employee ID] ))),DATESINPERIOD('Date'[Date ],MAX('Date'[Date ]),-28,DAY))