Hello, I am in the process of producing clusters for a script, and one of the first operations in the cluster is to ‘match tree’ with another input which will always have data in it. Only when clustering, the ‘match tree’ returns the common error of:
“1. Input trees must have equal number of paths”
As an explanation, sometimes the primary ‘tree’ input doesn’t have any data in some branches.
For some reason, when the primary tree input is brought into a cluster, the branches with no data are removed entirely, which makes the ‘match tree’ component inside the cluster turn red with an error.
Outside of a cluster, this same operation returns no errors, as expected. (No branches are removed)
Oddly enough, despite the ‘match tree’ showing an error in the cluster, and the cluster showing no output when viewed through a panel inside the cluster-- outside of the cluster, the output is correct (introduces the missing branch again)
Attached are 3 screen shots of an example, which hopefully explains this.
The way we deal with this at Packhunt is by filling empty branches with null values, and then removing the null values within the cluster at the input side. This way you can still use clusters reliably and in a maintainable way.
It would be nice if this get’s solved, but this bug is very old from what I gather, so I am personally not counting on it.
I used a combine data component to introduce null values. Filling the empty branches worked beautifully as you described. It’s rare a work-around is so simple!
Regardless of the simplicity of this solution, I hope in the near future, this historic bug gets it’s legs back.
As you both seam to like / use clusters a lot - please check it.
and maybe add a
as Reply - this will pop the topic back to more attention - thanks
in the editor:
