Equivalent of scipy.cluster.vq.kmeans

### Description

The RAFT project (https://github.com/rapidsai/raft) has several building blocks for different algorithms, which include the k-means and single-linkage clustering algorithms, which could be exposed directly to cupy. I'm hoping we might be able to expose this in the form of a `pylibraft` artifact, which could expose either the cython or the Python in a similar way to `pylibcugraph`.

Review the RAFT k-means and single-linkage building blocks and compare how pylibcugraph exposes similar functionality. Define how a pylibraft artifact would expose these algorithms to CuPy through Python or Cython; done means the requested clustering functionality is available through that integration.

Written by the indexing model from the issue text.