Charts · Relationship
Scatter and bubble
Relationship charts are the right answer whenever two prepared measures might move together. Quadrants and regression lines are annotations; the chart never fits a model.
Scatter
Two prepared measures, one point per observation, axis crosshair on both axes.
Factor exposure vs forward return
Prepared buckets; hover reads both axes.
Bubble and connected scatter
Bubble encodes a third prepared value as area; a connected scatter is a line series with symbols, ordered by x.
Bubble
Symbol area is proportional to the prepared turnover field.
Connected scatter
Ordered by x, so the path reads as a trajectory.
Regression lines
The chart accepts the fit; it never computes it.
A fitted line is a second series whose data arrives already prepared by the caller. That keeps the chart honest: whatever the model did is visible in the data layer, not hidden inside a renderer.
<ScatterChart
data={points}
xKey="factor"
series={[
{ key: "forward", tone: "accent" },
{ key: "fitted", tone: "warning", type: "line", symbol: false },
]}
/>