Forum Discussion

duesouth's avatar
duesouth
Helper I
1 year ago
Solved

Add Arrows to a Matrix

I've got a matrix for % attendance in a school - which is cumulative attendance week-on-week, as below:                                                                                          ...
Need Help
  • Elena_Kalina's avatar
    Elena_Kalina
    1 year ago

    Hi duesouth 

    1. First create a measure to calculate the difference:

    Attendance Change = 
VAR CurrentWeek = MAX('Date'[Date])
VAR CurrentAttendance = [Attendance Measure]

VAR PreviousWeek = 
    CALCULATE(
        MAX('Date'[Date]),
        FILTER(
            ALLSELECTED('Date'),
            'Date'[Date] < CurrentWeek
        )
    )

VAR PreviousAttendance = 
    CALCULATE(
        [Attendance Measure],
        FILTER(
            ALLSELECTED('Date'),
            'Date'[Date] = PreviousWeek
        )
    )

RETURN
    IF(
        NOT(ISBLANK(PreviousAttendance)),
        CurrentAttendance - PreviousAttendance,
        BLANK()
    )

    1. Then apply conditional formatting to your attendance values:

      • Select your matrix visual

      • Go to "Conditional formatting" → "Icons"

      • Set up rules like:

        • When value > 0.001 → Green up arrow

        • When value < -0.001 → Red down arrow

      • Base the formatting on the "Attendance Change" measure

    If this post helps, then please consider Accepting as solution to help the other members find it more quickly, don't forget to give a "Kudos" – I’d truly appreciate it! 

    Thank you.

     

Elena_Kalina's avatar
Elena_Kalina
Solution Sage
1 year ago

Hi duesouth 

  1. First create a measure to calculate the difference:

Attendance Change = 
VAR CurrentWeek = MAX('Date'[Date])
VAR CurrentAttendance = [Attendance Measure]

VAR PreviousWeek = 
    CALCULATE(
        MAX('Date'[Date]),
        FILTER(
            ALLSELECTED('Date'),
            'Date'[Date] < CurrentWeek
        )
    )

VAR PreviousAttendance = 
    CALCULATE(
        [Attendance Measure],
        FILTER(
            ALLSELECTED('Date'),
            'Date'[Date] = PreviousWeek
        )
    )

RETURN
    IF(
        NOT(ISBLANK(PreviousAttendance)),
        CurrentAttendance - PreviousAttendance,
        BLANK()
    )

  1. Then apply conditional formatting to your attendance values:

    • Select your matrix visual

    • Go to "Conditional formatting" → "Icons"

    • Set up rules like:

      • When value > 0.001 → Green up arrow

      • When value < -0.001 → Red down arrow

    • Base the formatting on the "Attendance Change" measure

If this post helps, then please consider Accepting as solution to help the other members find it more quickly, don't forget to give a "Kudos" – I’d truly appreciate it! 

Thank you.

 

  • Elena_Kalina's avatar
    Elena_Kalina
    Solution Sage
    1 year ago

    If the default arrow icons don't meet your needs, there are several alternative ways to implement custom trend indicators in Power BI:

    1. DAX Character Arrows
      You can directly include arrow symbols (▲▼↗↘→) in your DAX measures by concatenating them with your values. This approach uses Unicode characters and works across all Power BI visuals.

    2. JSON Formatting
      Power BI allows custom icon sets through JSON formatting rules. You can:

      • Create a JSON file defining your preferred icons

      • Specify exact colors and positioning

      • Apply different icon styles for various value ranges
        This method provides consistent styling across your report.

    3. SVG Icons
      For complete design control:

      • Embed SVG code directly in your measures

      • Customize every aspect (shape, color, size)

      • Use any vector graphic as an indicator
        Note: SVG rendering only works in web browsers, not the desktop app.

     

    If this post helps, then please consider Accepting as solution to help the other members find it more quickly, don't forget to give a "Kudos" – I’d truly appreciate it! 

    Thank you.

  • duesouth's avatar
    duesouth
    Helper I
    1 year ago

    Worked perfectly - thank you so much!