Range Slicer

One click each, free: Q1 to Q4, middle 50%, above or below the median, top or bottom 10%, without outliers (Tukey's fences), above the mean, and Equal groups (5 gives quintiles, 10 deciles). Percentiles are computed like numpy's default, linear interpolation over all the rows the Measure counts, and tested against numpy. Presets > Default selection opens the page on any of them or on a custom range.

Free presets are applied as numbers: a top 10% chosen today stays the same range after a refresh. Distribution rules (premium) > Keep rules live turns a rule into a live one that is re-resolved whenever the data or the other filters change, and adds mean ± 1σ or 2σ, P5 – P95, P1 – P99, top or bottom N% and above a target.

Put Customer in Entity (measure per entity) and Total revenue in Measure. The slider then ranges over revenue per customer, and the filter keeps the customers in range as a plain list on the Customer column, so every visual on the page respects it and the filter pane shows it. When a year or region slicer changes their revenue, the slicer re-lists the customers that are now in range.

That replaces the parameter table, flag measure and visual-level filter many reports carry. Measure per entity is free for up to 30,000 entities, one data window; beyond that, up to 150,000, is Premium. With Premium, Distribution analysis (premium) > Top entities listed names the largest customers the range kept and dropped, with Other, and What the range dropped shows them as a report page tooltip.

Yes, with Premium. Several ranges (Shift + drag) lets a reader hold several bands at once, typed as "10-50; >200" too, Exclude mode turns the selection into bands to leave out, and Named buckets turns "Budget: <40; Standard: 40-120; Premium: 120+" into chips the reader clicks, one or several. Power BI's range filter takes two conditions, so several ranges are written as the list of values inside them and re-listed when the data changes.

Second numeric field (premium) adds its own slider, with both ranges filtering together, and Joint density of the two fields draws a 2-D density where dragging a rectangle sets both ranges. Split by (premium) stacks the histogram by a segment in its real names, and Comparison (premium) draws a measure such as last year's count as an outline over the bars.

Slider > Open at the ends is on by default: a handle at the minimum or maximum writes no bound there, so rows above today's highest price stay selected after tomorrow's refresh. By default the histogram and the slider's ends follow the other filters; Slider > Bounds > Fixed (widest seen) keeps them. A range that keeps no loaded value is never written, because an empty filter would blank the report.

Bookmarks and sync slicers work, cross-highlighting from other visuals shows on the bars, and the Apply button makes heavy models wait until the reader is done. Typed bounds accept the reader's own separators, 1.5k, 2M, 12.5% and currency, and are never snapped; dragged handles snap to the step, to values in the data or to bin edges as set in Slider > Handles snap.