I’m working on a script that creates doors in a wall, but I need to get rid of the line segments in the wall that are of equal length as the doors.
Since there are multiple items to be matched for some branches ({8} in this example set), that solution doesn’t work:
exactly, because the matching would be like:
Member index works well in these cases, together with Cull Index:
Yeah, that’s what I thought, but the results were empty. So the number conversion was where the issue was at… Thanks!
yep, stuff being output by a Data Parameter is not a Number GH type, while output coming from a Length output is Number GH type, so they won’t match by default
many times I also found myself getting number inputs from external sources, and despite passing both through a Number Parameter before the Member Index, also a trim to 2 or 3 decimals was necessary 
Yeah, the decimals (and rounding) can be a pain.
not a big fan of juggling datatrees.
for my eyes a simple script does offer more readability:
Cull lines_02_tp.gh (15.5 KB)
private void RunScript(List<double> lTest, List<double> lRef, ref object isHit)
{
double tolerance = 0.001;
List<bool> returnHits = new List<bool>(lRef.Count);
foreach (double d in lTest)
{
bool found = lRef.Any(item => Math.Abs(item - d) <= tolerance);
returnHits.Add(found);
}
isHit = returnHits;
}
EDIT:
and a proper approach for comparing doubles would be to use a expression with some threshold / tolerance, retrieve a true / false and then cull with a pattern.
I left the expression part in above definition, but I was not motivated to get the datatrees in the right structure to perform the logic…
I’m very much a Python beginner, I would take this chance to mimic something like your nice and slender C# with some basic Python:
# x, y inputs both floats with List access
def same_length(a, b, tol):
if abs(a-b) <= tol:
return True
return False
a = []
for test_len in x:
res = True
for bad_len in y:
if same_length(test_len, bad_len, 0.001):
res = False
break
a.append(res)
how about:
def same_length(a, b, tol):
return abs(a-b) <= tol:
?
…
and just curious - how to implement the same logic without scripting, with pure vanilla gh-components ? (get the same true,false, … tree ?)
that is a bit painful 
my best guess is, given tree_A and tree_B:
graft tree_A in such a way each branch carries just one single item
in tree_B, duplicate branches times the amount of respective elements in ungrafted tree_A
then Mass Addition to OR everything together and get one single boolean per branch, then Trim Tree et voilà
visually:
Cull lines_Re_Re.gh (22.3 KB)
ok - somehow you proofed my fear…
I think it would be super strong feature, if a cluster could have the same input / access as a script component. I wrote this as a wish, but did not get any answers:
Propagate Ancestors component from TreeSloth would be my main choice here. Then again, doing it natively teaches you valuable lessons about DataTrees - like how to painfully crush your toes over and over again by hitting them against DataTrees.
I would always solve stuff like this with a script … but anyhow it seamed a nice challenge…
…here is my vanilla gh solution to find a cull pattern.
instead of multiply data I use cross reference (holistic) to get the combinations for the equality-test.
flattening the final pattern and unflatten with the original tree as guide seems more readable for me…
compare_lists_with_lists_01.gh (16.6 KB)
