Scatter Chart
Documentation Index
Fetch the complete documentation index at: /llms.txt. Use this file to discover all available pages before exploring further.
A composable scatter chart for relationships, clusters, bubbles, and quadrant analysis.
Overview
Use a scatter chart to inspect relationships between two numeric measures, reveal clusters, and spot outliers. Add bubble sizing when a third measure matters, or divide the plot into labeled quadrants for prioritization and portfolio analysis.
Anatomy
ScatterChart owns the rows, numeric x/y fields, optional grouping, loading state, and selection. Add one Scatter per series, then compose optional Grid, XAxis, YAxis, Legend, Tooltip, and Quadrants parts.
Accessibility
Provide a meaningful ariaLabel, keep tooltips enabled for exact values, and pair the canvas with a table or written summary when point-level data is essential. Bubble area and quadrant color should supplement labels and values rather than carry meaning alone.
Installation
Usage
import { ScatterChart, type ChartConfig } from "honestui/charts";<ScatterChart
data={data}
config={chartConfig}
xDataKey="seats"
yDataKey="adoption"
groupDataKey="segment"
pointNameDataKey="account"
>
<ScatterChart.Grid />
<ScatterChart.XAxis label="Licensed seats" />
<ScatterChart.YAxis label="Adoption" />
<ScatterChart.Legend />
<ScatterChart.Tooltip />
<ScatterChart.Scatter dataKey="enterprise" />
<ScatterChart.Scatter dataKey="startup" />
</ScatterChart>When groupDataKey is supplied, each Scatter renders rows whose group value matches its dataKey. Without grouping, a single series renders every row.
Bubble Chart
Set variant="bubble" and provide sizeDataKey to encode a third numeric measure by area. minSize and maxSize keep small values visible and large values from overwhelming the plot.
Quadrant Chart
Add Quadrants with numeric split points and optional labels. Honest UI renders quiet analytical regions behind the points while retaining exact x/y positioning.
Loading State
The loading state preserves the plot dimensions with neutral placeholder points and the shared chart loading treatment.