Pareto chart in Power BI: why the DAX workaround keeps breaking
Power BI has no native Pareto chart. The usual recipe is a combo chart with a cumulative-percentage measure: rank the categories with RANKX, sum every value at or above the current rank, divide by the grand total. It works on the first dataset and then it breaks in one of three ways. Two categories tie and the ranking skips a number, so the cumulative line jumps. A slicer changes the ranking context and the line no longer adds to 100. Or somebody sorts the axis by a different field and the bars are no longer descending, which is the one thing a Pareto chart must be.
SmartVisuals Pareto Chart does the whole calculation inside the visual. Map a category and a measure and the bars are ranked by contribution, the cumulative line is drawn, the threshold is marked and the vital few are coloured, in every filter state, with no measure to maintain. Ties are broken deterministically, the sort is stable in Power BI Desktop and in the Service, and duplicate category rows are aggregated before ranking.
