Hi, I suppose the clusterer.exemplars_ returns points in the order of clusterer.labels_, this is, the first exemplar refers to label 0, the second refers to points assinged to label 1, ...

I would like to double check this information, since I didn't find it on the documentation.

I have a clusterer object which is returning exemplars far different than the points associated with the each label, so I wonder if the exemplars_ are shown in a differente order than the labels_.

Thanks and congrats on the work done on hdbscan and umap.