deneb-visuals

Add the visual with the pbir CLI using visualType deneb7E15AEF80B9E4D4F8E12924291ECE89A, bind fields to the single dataset role, then inject a Vega or Vega-Lite spec file with pbir visuals deneb. Validate the report afterward with pbir validate.

Vega-Lite is preferred for new Deneb visuals because it is more concise and covers most chart types. Use Vega only when you need signals, event streams, force layouts, or advanced cross-filtering with brush or lasso selection.

Set enableSelection to true in the vega object properties and handle the __selected__ field (on, off, neutral) in your encode blocks. Simple mode supports up to 250 data points; advanced mode uses pbiCrossFilterApply and pbiCrossFilterClear signals in Vega only.

Escaping depends on context: standalone spec files use JSON double quotes, but inside PBIR visual.json the spec is a single-quoted literal, so field names with spaces need doubled single quotes like datum[''Order Lines'']. Field names must match the nativeQueryRef display names.

Use the pbiColor(index) function for theme colors by index, with an optional shade parameter from -1 to 1, or schemes like pbiColorNominal, pbiColorLinear, and pbiColorDivergent for categorical and gradient palettes instead of hardcoded hex values.

Deneb defaults to a 10,000 row limit, which can be raised via dataLimit.override, though the model's own row cap still applies. For better performance, aggregate in DAX first and use canvas render mode for many marks before raising limits.