soft_clustering_tuebingen

tangles.analysis.soft_clustering_tuebingen(tangle_matrix: ndarray, feat_ids: ndarray, feat_sys) → ndarray

Implements soft clustering as described in the Tuebingen Paper.

For each tangle, a corresponding cluster is calculated.

Parameters

tangle_matrixnp.ndarray

A (tangles x features)-matrix encoding whether a tangle contains a certain feature or its inverse. Contains 1 if the tangle contains the feature and otherwise 0.

feat_idsnp.ndarray

The feature ids corresponding to the columns of the tangle_matrix.

feat_sysFeatureSystem

A feature system.

Returns

clustering_matrixnp.ndarray

A (point x tangles)-matrix encoding the clustering score of each point in the dataset for each tangle. The value represents the proportion of the point contained within the cluster.