Calculate "Running sick days per employee (long time absences)"

[fitzi44]

Hi all

I'm currently facing the challenge to create a measure for all our employees which are categorized as "long time absences". The definition for a "long time absence" is, that an employee is "sick" for at least 30 working days running. If he is 30 and more days sick in a row, the period (each date in this period) is automatically defined and flagged as "long time absence":

Date	Employee	Sick	Running sick days per employee	Long time absence
02/02/2023	Employee A	0	0	0
03/02/2023	Employee A	0	0	0
06/02/2023	Employee A	0	0	0
07/02/2023	Employee A	0	0	0
08/02/2023	Employee A	0	0	0
09/02/2023	Employee A	0	0	0
10/02/2023	Employee A	0	0	0
13/02/2023	Employee A	0	0	0
14/02/2023	Employee A	0	0	0
15/02/2023	Employee A	0	0	0
16/02/2023	Employee A	0	0	0
20/02/2023	Employee A	0	0	0
21/02/2023	Employee A	0	0	0
22/02/2023	Employee A	0	0	0
23/02/2023	Employee A	0	0	0
24/02/2023	Employee A	0	0	0
27/02/2023	Employee A	0	0	0
28/02/2023	Employee A	0	0	0
01/03/2023	Employee A	0	0	0
02/03/2023	Employee A	0	0	0
03/03/2023	Employee A	0	0	0
06/03/2023	Employee A	1	1	0
07/03/2023	Employee A	1	2	0
08/03/2023	Employee A	1	3	0
09/03/2023	Employee A	1	4	0
10/03/2023	Employee A	1	5	0
13/03/2023	Employee A	1	6	0
14/03/2023	Employee A	1	7	0
15/03/2023	Employee A	1	8	0
16/03/2023	Employee A	1	9	0
17/03/2023	Employee A	1	10	0
20/03/2023	Employee A	1	11	0
21/03/2023	Employee A	0	0	0
22/03/2023	Employee A	0	0	0
23/03/2023	Employee A	0	0	0
27/03/2023	Employee A	0	0	0
28/03/2023	Employee A	0	0	0
29/03/2023	Employee A	0	0	0
03/04/2023	Employee A	1	1	0
04/04/2023	Employee A	1	2	0
05/04/2023	Employee A	1	3	0
06/04/2023	Employee A	1	4	0
11/04/2023	Employee A	1	5	0
12/04/2023	Employee A	1	6	0
13/04/2023	Employee A	1	7	0
14/04/2023	Employee A	1	8	0
17/04/2023	Employee A	1	9	0
18/04/2023	Employee A	1	10	0
19/04/2023	Employee A	1	11	0
20/04/2023	Employee A	1	12	0
21/04/2023	Employee A	1	13	0
24/04/2023	Employee A	1	14	0
25/04/2023	Employee A	1	15	0
26/04/2023	Employee A	1	16	0
27/04/2023	Employee A	1	17	0
28/04/2023	Employee A	1	18	0
01/05/2023	Employee A	0	0	0
02/05/2023	Employee A	1	1	1
03/05/2023	Employee A	1	2	1
04/05/2023	Employee A	1	3	1
05/05/2023	Employee A	1	4	1
08/05/2023	Employee A	1	5	1
09/05/2023	Employee A	1	6	1
10/05/2023	Employee A	1	7	1
11/05/2023	Employee A	1	8	1
12/05/2023	Employee A	1	9	1
15/05/2023	Employee A	1	10	1
16/05/2023	Employee A	1	11	1
17/05/2023	Employee A	1	12	1
19/05/2023	Employee A	1	13	1
22/05/2023	Employee A	1	14	1
23/05/2023	Employee A	1	15	1
24/05/2023	Employee A	1	16	1
25/05/2023	Employee A	1	17	1
26/05/2023	Employee A	1	18	1
30/05/2023	Employee A	1	19	1
31/05/2023	Employee A	1	20	1
01/06/2023	Employee A	1	21	1
02/06/2023	Employee A	1	22	1
05/06/2023	Employee A	1	23	1
06/06/2023	Employee A	1	24	1
07/06/2023	Employee A	1	25	1
08/06/2023	Employee A	1	26	1
09/06/2023	Employee A	1	27	1
12/06/2023	Employee A	1	28	1
13/06/2023	Employee A	1	29	1
14/06/2023	Employee A	1	30	1
15/06/2023	Employee A	1	31	1
16/06/2023	Employee A	1	32	1

I'm currently struggling with the task to calulcate the "Running sick days per employee" and therefore also with the calculation of the "Long time absence" flag in DAX.
Does anyone has some help for me?

[Jihwan_Kim · Accepted solution]

Hi,

Please check the below picture and the attached pbix file.

Running sick: = 
VAR _list =
    ALL ( Data[Date] )
VAR _t =
    ADDCOLUMNS (
        FILTER ( ALL ( 'Calendar' ), 'Calendar'[Date] IN _list ),
        "@sicktotal", [Sick total:]
    )
VAR _prevsick =
    ADDCOLUMNS (
        _t,
        "@prevsick", MAXX ( OFFSET ( -1, _t, ORDERBY ( 'Calendar'[Date], ASC ) ), [@sicktotal] )
    )
VAR _condition =
    ADDCOLUMNS (
        _prevsick,
        "@condition",
            IF ( [@sicktotal] <> [@prevsick] || [@prevsick] == BLANK (), 1, 0 )
    )
VAR _group =
    ADDCOLUMNS (
        _condition,
        "@group",
            SUMX (
                FILTER ( _condition, 'Calendar'[Date] <= EARLIER ( 'Calendar'[Date] ) ),
                [@condition]
            )
    )
VAR _runningsick =
    ADDCOLUMNS (
        _group,
        "@runningsick",
            SUMX (
                FILTER (
                    _group,
                    'Calendar'[Date] <= EARLIER ( 'Calendar'[Date] )
                        && [@group] = EARLIER ( [@group] )
                ),
                [@sicktotal]
            )
    )
VAR _longtimeabsence =
    ADDCOLUMNS (
        _runningsick,
        "@longtimeabsence",
            IF (
                30
                    IN SUMMARIZE (
                        FILTER ( _runningsick, [@group] = EARLIER ( [@group] ) ),
                        [@runningsick]
                    ),
                1,
                0
            )
    )
RETURN
    MAXX (
        FILTER ( _longtimeabsence, 'Calendar'[Date] = MAX ( 'Calendar'[Date] ) ),
        [@runningsick]
    )

Long absence: = 
VAR _list =
    ALL ( Data[Date] )
VAR _t =
    ADDCOLUMNS (
        FILTER ( ALL ( 'Calendar' ), 'Calendar'[Date] IN _list ),
        "@sicktotal", [Sick total:]
    )
VAR _prevsick =
    ADDCOLUMNS (
        _t,
        "@prevsick", MAXX ( OFFSET ( -1, _t, ORDERBY ( 'Calendar'[Date], ASC ) ), [@sicktotal] )
    )
VAR _condition =
    ADDCOLUMNS (
        _prevsick,
        "@condition",
            IF ( [@sicktotal] <> [@prevsick] || [@prevsick] == BLANK (), 1, 0 )
    )
VAR _group =
    ADDCOLUMNS (
        _condition,
        "@group",
            SUMX (
                FILTER ( _condition, 'Calendar'[Date] <= EARLIER ( 'Calendar'[Date] ) ),
                [@condition]
            )
    )
VAR _runningsick =
    ADDCOLUMNS (
        _group,
        "@runningsick",
            SUMX (
                FILTER (
                    _group,
                    'Calendar'[Date] <= EARLIER ( 'Calendar'[Date] )
                        && [@group] = EARLIER ( [@group] )
                ),
                [@sicktotal]
            )
    )
VAR _longtimeabsence =
    ADDCOLUMNS (
        _runningsick,
        "@longtimeabsence",
            IF (
                30
                    IN SUMMARIZE (
                        FILTER ( _runningsick, [@group] = EARLIER ( [@group] ) ),
                        [@runningsick]
                    ),
                1,
                0
            )
    )
RETURN
    MAXX (
        FILTER ( _longtimeabsence, 'Calendar'[Date] = MAX ( 'Calendar'[Date] ) ),
        [@longtimeabsence]
    )

[fitzi44]

Thanks a lot @Jihwan_Kim 

Really great help here. You solution worked, however it was bot very performant in the end in the report, so we opted to implement the logic already on Database level.

kind regards and thanks

Andreas