A small hint for those who are struggeling with poor quality when clustering.
In some case the output of a clustering might look as in the following image:
This is not, what you might expect.
If you set the following parameter to 1, the result would be more likely what you expect:
CompactnessLevel=1
This is the result with the changed CompactnessLevel:
In this particular example the underlying planning locations were given with very heterogeneous activities between 1 and >100.
This made it challenging for the algorithm to satisfy compactness AND balance.
This made it challenging for the algorithm to satisfy compactness AND balance.