sampling doesn't make sense above the resolution of the graph

[masenf]

Screen.Recording.2026-07-22.at.10.49.29.PM.mov

zoom in until reaching the saturation point of the window (number of points less than the scatter density threshold)

then zooming out from there leaves the saturation point and starts to show sampled points -- why are there now individual points in view.

not sure how complex it is to calculate the saturation point for a requested density window and only give points if they wouldn't totally blot out the chart 🤔 but a fixed sample rate at all zoom levels is really misrepresenting this dataset: examples/fastapi 100m point scatter