Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!

Reply
fturmo
Frequent Visitor

Power BI Desktop: Variables in Measures

Dear all,

 

I am trying to create a measure that calculates the differences in hours between two columns taking into account the weekends and non worked days.

 

I am using the follwoing formula for the measure:

 

Measure Avg Time = 

var tabla=CALCULATETABLE(Table1)
var startdate= SUMX(tabla;Table1[Start Date])
var enddate=SUMX(tabla;Table1[End Date])
var Firstdayend =DATE(YEAR(startdate);MONTH(startdate);DAY(startdate)+1)
var LastdayStart = DATE(YEAR(enddate);MONTH(enddate);DAY(enddate))
var LastdayEnd = DATE(YEAR(enddate);MONTH(enddate);DAY(enddate)+1)
 
var secondsfirstday= SUMX(tabla;
                IF(Firstdayend<=startdate;
                               0;
                               DATEDIFF(startdate;Firstdayend;SECOND)
                ))
var secondslastday= SUMX(tabla;
                IF(enddate>=LastdayStart && enddate<=LastdayEnd;
                               DATEDIFF(LastdayStart;enddate;SECOND);
                               IF(enddate>LastdayEnd;
                                               DATEDIFF(LastdayStart;LastdayEnd;SECOND);
                                               0
                               )
                ))
var secondsmidday=SUMX(tabla; 
                (CALCULATE(DISTINCTCOUNT('Calendar'[Date Calculations]);FILTER('Calendar';'Calendar'[Date Calculations]>startdate 
                &&'Calendar'[Date Calculations]<enddate && 'Calendar'[Is Weekday]=TRUE()))-1)*3600*24
                )
return
 
SUMX(tabla;(secondsfirstday+secondslastday+secondsmidday)/3600)

The result of that formula is OK for each individual member but i don't know how to calculate the aggregate (average) for all the rows in the table.

 

Currently i have the following result:

 

Power BI Result.png

 And here you have the original Table:

 

As you can see this measure works fine (without the sumx) as a calculated column, but i want to apply this measure in a large database and it is not possible to use a calculated column because it is too slow and i have errors when i try to calculate it.

 

Original Table.png

 

 Thanks,

 

Francesc.

1 ACCEPTED SOLUTION
v-ljerr-msft
Employee
Employee

Hi @fturmo,

The result of that formula is OK for each individual member but i don't know how to calculate the aggregate (average) for all the rows in the table.

In this scenario, you may need two measures to work this out. First measure to calculate the time difference for the individual column which is similar to the measure you provided above.. The formula below is for your reference.

Measure Time For Individual Member =
VAR startdate =
    MAX ( Table1[Start Date] )
VAR enddate =
    MAX ( Table1[End Date] )
VAR Firstdayend =
    DATE ( YEAR ( startdate ), MONTH ( startdate ), DAY ( startdate ) + 1 )
VAR LastdayStart =
    DATE ( YEAR ( enddate ), MONTH ( enddate ), DAY ( enddate ) )
VAR LastdayEnd =
    DATE ( YEAR ( enddate ), MONTH ( enddate ), DAY ( enddate ) + 1 )
VAR secondsfirstday =
    IF ( Firstdayend <= startdate, 0, DATEDIFF ( startdate, Firstdayend, SECOND ) )
VAR secondslastday =
    IF (
        enddate >= LastdayStart
            && enddate <= LastdayEnd,
        DATEDIFF ( LastdayStart, enddate, SECOND ),
        IF ( enddate > LastdayEnd, DATEDIFF ( LastdayStart, LastdayEnd, SECOND ), 0 )
    )
VAR secondsmidday =
    (
        CALCULATE (
            DISTINCTCOUNT ( 'Calendar'[Date Calculations] ),
            FILTER (
                'Calendar',
                'Calendar'[Date Calculations] > startdate
                    && 'Calendar'[Date Calculations] < enddate
                    && 'Calendar'[Is Weekday] = TRUE ()
            )
        )
            - 1
    )
        * 3600
        * 24
RETURN
    ( secondsfirstday + secondslastday
        + secondsmidday )
        / 3600

Then use a new measure to sum each individual time difference together for the table.

Measure Avg Time = SUMX ( Table1, [Measure Time For Individual Member] )

Regards

View solution in original post

2 REPLIES 2
v-ljerr-msft
Employee
Employee

Hi @fturmo,

The result of that formula is OK for each individual member but i don't know how to calculate the aggregate (average) for all the rows in the table.

In this scenario, you may need two measures to work this out. First measure to calculate the time difference for the individual column which is similar to the measure you provided above.. The formula below is for your reference.

Measure Time For Individual Member =
VAR startdate =
    MAX ( Table1[Start Date] )
VAR enddate =
    MAX ( Table1[End Date] )
VAR Firstdayend =
    DATE ( YEAR ( startdate ), MONTH ( startdate ), DAY ( startdate ) + 1 )
VAR LastdayStart =
    DATE ( YEAR ( enddate ), MONTH ( enddate ), DAY ( enddate ) )
VAR LastdayEnd =
    DATE ( YEAR ( enddate ), MONTH ( enddate ), DAY ( enddate ) + 1 )
VAR secondsfirstday =
    IF ( Firstdayend <= startdate, 0, DATEDIFF ( startdate, Firstdayend, SECOND ) )
VAR secondslastday =
    IF (
        enddate >= LastdayStart
            && enddate <= LastdayEnd,
        DATEDIFF ( LastdayStart, enddate, SECOND ),
        IF ( enddate > LastdayEnd, DATEDIFF ( LastdayStart, LastdayEnd, SECOND ), 0 )
    )
VAR secondsmidday =
    (
        CALCULATE (
            DISTINCTCOUNT ( 'Calendar'[Date Calculations] ),
            FILTER (
                'Calendar',
                'Calendar'[Date Calculations] > startdate
                    && 'Calendar'[Date Calculations] < enddate
                    && 'Calendar'[Is Weekday] = TRUE ()
            )
        )
            - 1
    )
        * 3600
        * 24
RETURN
    ( secondsfirstday + secondslastday
        + secondsmidday )
        / 3600

Then use a new measure to sum each individual time difference together for the table.

Measure Avg Time = SUMX ( Table1, [Measure Time For Individual Member] )

Regards

That works perfectly.

 

Thanks Jerry for your Response.

 

Francesc.

Helpful resources

Announcements
April AMA free

Microsoft Fabric AMA Livestream

Join us Tuesday, April 09, 9:00 – 10:00 AM PST for a live, expert-led Q&A session on all things Microsoft Fabric!

March Fabric Community Update

Fabric Community Update - March 2024

Find out what's new and trending in the Fabric Community.