A better approach replicating 'Curves in Curves'

If we’re going for pure performance… Here’s a shot using Impala to speed parts of it up. First finding the neighborhood of all nearby curves (curves with a centrepoint within the ‘radius’ of the largest point in the set) and then doing containment testing. All of this is parallel and comes out to just over a hundred milliseconds all told… it’s the boundary surfacing to preview at the end that takes the time :slight_smile:


curveInCurve_2.gh (100.3 KB)