I often use Smoother+Overlay to check how independent categorical variables affect dependent continuous variables in a time trend plot. But one issue I often encounter is when different categories have different volumes of data, the smoother lines are equally as big even if one category has 200 data points and another category has 4 data points. This can create busy plots where it's hard to distinguish the important signals. (And if I'm checking a large number of different independent variables in the Overlay role, it is tedious to exclude small-data-volume groups for each variable.)
Is there any way to do something similar to: make the Smoother line thicker if there's a high data volume, and thinner if there's less data? (Even BETTER would be if the line dynamically changed thickness depending on the deltaX between two points: e.g., if one category has two points 1 day apart and another category has two point 2 days apart, the smoother line for the first category would be ~twice as thick.)
Here's the type of "busy graph" I'm dealing with: variable thickness lines would help me determine at a glance which signals are important and which ones aren't.
