Partitioning method is a clustering algorithm, which is sensible to initial partitions (values of k), initial values and input sequence.

Enables parallelization of a query, as sourced by a custom partitioner that is responsible for splitting the input sequence into partitions.