How to set similarity function for cluster algorithms?

[Gang Su]

Hi developers,

I have been using sspace and I really like the available clustering algorithms in this library. I have encountered one difficulty - many of these algorithms have no input argument for similarity function. The hierarchical clustering has one but not for directkmeans, etc. It seems that the interface with the Properties as input is not always implemented by all clustering algorithms. And even I added them, it's not clear whether the clustering algorithm actually used them.

Without hacking the source code, is there anyway to figure out how the similarity parameter is set? Thank you very much.

Best wishes,

Gang Su