Forum Discussion

azs's avatar
azs
New Member
5 years ago
Solved

DAX Measure Average Count over Time

Hi,

 

I am working with work order counts. Any work order is having a time stemp - opened, due to and closed - depending on "due to" and "closed" it can be evaluated, if the job was running overdue or is closed in time.

I am trying to get a measure, which is giving just the average number of overdue jobs over time.

 

for example:

 

A;B;C;D are opened on monday and due to by tuesday.

 

Monday: opened A;B;C;D --> overdue count 0

Tuesday: A is due to and closed as well; B&C&D are due to but not finished yet --> overdue count 3

Wednesday: B&C&D are closed --> overdue count 0

 

So the Average Overdue count these 3 days is "1".

 

I defined Overdue-Count as:

 

CALCULATE(
    COUNTA('Order'[Due to]),
    ALLSELECTED(Calender[Date])
 
and put on the filters "not closed" and "Due to"<"TODAY()".
 
This is working for current overdue jobs but it does not give me a historical average evaluation. I tried somehow with AverageX but still i don't now how to replace the "TODAY" with the time period, for which the overdue evaluation has to be done - "Calculate the overdues from every day by comparing "due to<=current date" AND "closed is no", then add up overdue counts per day and calculate the average over all time up to today".
 
thanks for any hints,
best regards,
azs
Need Help
  • Anonymous's avatar
    Anonymous
    5 years ago

    Hi azs 

    Due to I don't know your data model, I build a sample to have a test.

    Here I have two open days 2021/05/20 and 2021/05/24.

    2021/05/20 Part:

    2021/05/20: opened A;B;C;D --> overdue count 0

    2021/05/21: C is due to and closed as well; A&B&D are due to but not finished yet --> overdue count 3

    2021/05/22: A&B are overdue but closed; D is overdue but not closed --> overdue count 1

    2021/05/23: D is overdue but closed --> overdue count 0

    So the Averge count is (0+3+1+0)/ 4 = 1

    2021/05/20 Part is the same like yours.

    Measures:

    Diff Day = DATEDIFF(MAX('Table'[Due to]),MAX('Table'[Close]),DAY)
    Avg = 
VAR _Overdue = SUMX('Table',[Diff Day])
VAR _AllDaycount = DATEDIFF(MAX('Table'[Open]),MAX('Table'[Close]),DAY) +1
VAR _Avg = DIVIDE(_Overdue,_AllDaycount)
Return
_Avg

    Result is as below.

     

    Best Regards,

    Rico Zhou

     

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

2 Replies

  • amitchandak's avatar
    amitchandak
    Super User
    5 years ago

    azs , Something like this


    COUNTA('Order'[Due to]),filter('Order', 'Order'[Status] ="not closed" && 'Order'[due to] <today()))

     

    or

     


    COUNTA('Order'[Due to]),filter('Order', 'Order'[Status] ="not closed" && 'Order'[due to] <today()), ALLSELECTED(Calender[Date]))

     

    or

     


    COUNTA('Order'[Due to]),filter('Order', 'Order'[Status] ="not closed" && 'Order'[due to] <today()), ALL(Calender[Date]))

  • Anonymous's avatar
    Anonymous
    Not applicable
    5 years ago

    Hi azs 

    Due to I don't know your data model, I build a sample to have a test.

    Here I have two open days 2021/05/20 and 2021/05/24.

    2021/05/20 Part:

    2021/05/20: opened A;B;C;D --> overdue count 0

    2021/05/21: C is due to and closed as well; A&B&D are due to but not finished yet --> overdue count 3

    2021/05/22: A&B are overdue but closed; D is overdue but not closed --> overdue count 1

    2021/05/23: D is overdue but closed --> overdue count 0

    So the Averge count is (0+3+1+0)/ 4 = 1

    2021/05/20 Part is the same like yours.

    Measures:

    Diff Day = DATEDIFF(MAX('Table'[Due to]),MAX('Table'[Close]),DAY)
    Avg = 
VAR _Overdue = SUMX('Table',[Diff Day])
VAR _AllDaycount = DATEDIFF(MAX('Table'[Open]),MAX('Table'[Close]),DAY) +1
VAR _Avg = DIVIDE(_Overdue,_AllDaycount)
Return
_Avg

    Result is as below.

     

    Best Regards,

    Rico Zhou

     

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