Column names must be dynamic i.e. we should be able to change the column names in a Table/Matrix
fbcideas_migusr
6 years agoNew Member
'@Sebastian, you can achieve your required output today by putting the year attribute on the columns then changing the measure expression to the following = calculate(sum(value); FILTER(VALUES('Date'[Year]), 'Date'[Year] = year(today () )-2)) This measure will only return a value against the current year, it will return blanks against any other year which Power BI will automatically filter out.
Recent ideas
File level commit - Need the ability to edit the commit description
I tried the recently announce file level commit for Fabric items. One notable issue I have with the feature (testing this with a Fabric warehouse, I'm assuming this would hold true for other items): ...DataBard20 minutes agoMost Valuable ProfessionalNew30Views0likes2CommentsRename notebook cell name “SparkR (R)” to “R” and “PySpark (Python)” to “Python”
It's a simple idea. In Fabric notebooks, the cell language for R is labeled “SparkR (R)”. Since Spark 4.0, the SparkR package is deprecated, and the Fabric Runtime 2.0 documentation itself notes th...lorin_s1 hour agoNew MemberNew9Views0likes0CommentsAuto Display Units Should Dynamically Change Granularity
When auto is selected for display units, the granularity should automatically change to best suite the size of the data. For example, we often have charts that have numerical ranges from the milli...acason2 hours agoNew MemberNeeds Votes349Views26likes5Comments