Static documentation

Data Visualisation Catalogue Gallery

Rendered markdown notes for the catalogue review, compiler mapping, implementation tracker, and chart method specs.

38 chart methods
10 with reference assets
43 documents

Histogram

Source: Data Visualisation Catalogue - Histogram

Histogram reference

Histogram anatomy

Why This Belongs In The Compiler Gallery

The catalogue defines histograms as distributions over continuous intervals. The compiler already models a bin transform and vertical rect bars, making this a direct fit and an important contrast with categorical bar charts.

Data Contract

Input can be raw observations or pre-binned intervals.

FieldTypePurpose
valuequantitativeContinuous measure to bin when starting from raw observations.
minquantitative, derivedInclusive lower bound for compiler-generated bins.
maxquantitative, derivedUpper bound for compiler-generated bins.
countquantitativeFrequency in the bin.
labeltextHuman-readable bin label.

Expected Visual

The faithful visual should show contiguous or nearly contiguous bars along a continuous x axis. Bar height encodes count/frequency. Unlike a bar chart, the horizontal position represents an interval, not a category. The catalogue reference uses vertical columns with simple guide axes.

ElementExpected treatment
BarsRects representing interval bins; gaps should be minimal.
X axisContinuous interval labels or tick marks.
Y axisCount or frequency.
SortingAlways by bin start, not by count.
LabelsOptional count labels; avoid treating bins as unordered categories.

Compiler Mapping

Compiler partMapping
Data sourceobservations raw table and bins output table.
Transformbin from raw values to count rows.
Scalesx: linear over bin bounds; y: linear.
Layersrect bins, line grid/baseline, text bin labels and optional count labels.
InteractionBin hover should expose interval, count, and contributing source rows.

Faithfulness Notes

The gallery histogram now renders compiler-generated bins with x0/x1 from a linear scale, which keeps the bar-chart/histogram distinction intact. Remaining polish is tick formatting for interval labels and optional density/frequency normalization.