This must be a very old bug since it’s in R7. Beyond frustrating 
I have a tree where the last branch is empty. Outside the cluster, all three outputs of TStat (Tree Statistics) show the empty branch.
cluster_trees_2024Aug6a.gh (9.6 KB)
Inside the cluster, TStat does not show the empty branch. In addition, the cluster output is visible inside but not outside 
Shoddy work and a waste of time. 
P.S. Same in R8.
@Joseph_Oster - can you provide a sample?
– Dale
OOPS! So sorry, not used to starting threads. Added above.
By the way, Prune Tree fixed the problem, outside or inside the cluster - so I added it inside.
cluster_trees_2024Aug6b.gh (10.2 KB) Fixed.
I found another mystery yesterday on a different project and wonder if it’s related? I’ll try to simplify this but please bear with me, it’s complicated.
These two versions are identical except for the yellow group added to “fix” this weird behavior.
This is supposed to work (without the yellow group) but as you can see, some of the surfaces are not colored as expected by the ColorBa cluster, which uses TStat (Tree Statistics).
At first I thought it was because ColorBa needed Match Tree internally, but as you can see, it works the same without it. This is odd but I have a vague memory that this is intentional?
I disabled the ‘Tree/List Viewer’ tool because its output only adds to the mystery, but it too uses TStat (Tree Statistics). Try enabling it, you will see.
Hard to believe I’ve never noticed these problems because I’ve been using these tools for years.
P.S. Like my first post in this thread, I discovered this flaw in R7 but it works the same in R8.
This thread reports and discusses the same issues:
I am not sure if I understand correctly, and maybe you already found the solution here, but isn’t this just an issue with the mismatching of trees due to the member-index component generating a branch per input-member at the index output?
I don’t know what you mean? Without code to examine, I won’t look into it further.
Having studied it back in August, as reported, it’s hard to believe any “just an issue” theory.
My appologies, Here is the ,gh model:
sort by angle_bugfinding.gh (67.3 KB)
Just for clarity’s sake, let me rephrase my question a bit better:
Can you explain why the algorithm is supposed to work without the yellow group? I ask this because I can’t see the purpose here of having two mismatched tree structures at the input of the list item component.
I think this is the same:
