Consensus clustering (consensus)

Note

This module is still under development and its interface may change at any time.

Orange.clustering.consensus.consensus_matrix(data, n, subset_fn, cluster_fn)

Compute a consensus matrix.

Parameters:

• data – data set to work on
• n – number of different clusterings.
• subset_fn – a function that takes a data table and returns a new data table with a subset of instances or attributes (or both). Subset function needs to preserve meta attribute “index”.
• cluster_fn – a function that takes a data set and returns a list of examples in the clusters.