No problem. I just wasn’t sure that I was understanding your question correctly.
You also need to be careful of this:
Since 51 or any number 5*n+1 will result in an ending branch with a list length of one, the Nurb Curve component will get upset because it cannot create a curve with just a point. The easiest solution is probably to multiply the slider by the number you divide by, and make sure that the curve degree is at least one less than the multiplication factor:
Yeah. All of these can be found with a right-click at most in-/outputs, although they do also exist as components themselves.
What I showed at the end is also is not the most straight-forward way of doing things.
As you probably know, you can use the list of rational numbers as indices. They will be rounded by an input that takes them. I was just showing what was happening explicitly. Also, culling and then finding set difference was to show you the data trees of selected and unselected points. You can use the List Item component to get the selected points. Probably better, since I am not sure if the Populate Geometry component create unique values.
If you simplify and graft your tree structure–here at its earliest creation point at “Sorties Rond Point”:
your branch hierarchy reflects: {major street #; roundabout #; roundabout exit#}
You can use this data structure as a guide for the Match Tree component with the selected points list to reestablish order:
For this component, the content of the lists, both in value type and the list length, are irrelevant. Simply the number of lists (or the number of tree branches) must be the same.
You can then use this snippet:
to address a specific point, which simplified as a cluster:
looks much cleaner.
II-3_pavillonaire.gh (40.6 KB)