Hullo--I'm having an issue with something that I suspect can be easy--I'm making branching splines (vein-like) using find shortest path. I want the pscales to be biggest in the middle of the branches and smaller at the ends (like remapping a curveu on a single line). The closest I've come is shortest path-->fuse-->polypath-->edge transport from point 0 (which seems to default to a central point with branching), promoting @dist to min/max details, and fitting the dist between those to use as my 0-1 which I remap. This works great for the farthest points in the branches, but for the shorter branches the end pscales don't go to 0. I tried setting end points to @pscale = 0 and attrib blurring it, but that didn't do the trick. I assume this is the same trick to make branching/tree sweeps on L-systems, but it's eluding me. What am I missing? Here is the line alone and with spheres copied on points to visualize pscales.

Lots of different ways this could be done. Don't know what you actually have set up in your file or where you want to go with it, since that will determine best set up for your purposes.

But hopefully this file could give you something to work with.

Thanks for this! Unfortunately, since it's not procedural it doesn't really work for me. By some judicious attribute promotion and remapping of my find shortest path's cost attributes I was able to get what I needed.

since it's not procedural it doesn't really work for me.

It wouldn't be hard to modify it so that it can be. The example was to simply show you a method to start with.