Does the current k-means implementation work fine for large distributed Dask arrays? Raul is also looking into the possibility of alternative clustering refinement steps, e.g. with a hierarchical approach. These could be exemplified in one of the tutorials, or added to the package.
Does the current k-means implementation work fine for large distributed Dask arrays? Raul is also looking into the possibility of alternative clustering refinement steps, e.g. with a hierarchical approach. These could be exemplified in one of the tutorials, or added to the package.