DAGMiningResult

DAGMiningResult(chains, diversity, parent_child_edges, n_components)

Result of DAG mining on an embedding space.

DAG mining finds hierarchical structures by using neighbor diversity as a generality signal. Points with diverse neighbors (connecting many topics) are considered “general”, while points with coherent neighbors (tight clusters) are “specific”.

Attributes

Name Type Description
chains List[DAGChain] List of extracted DAG chains
diversity np.ndarray Neighbor diversity score for each point
parent_child_edges List[Tuple[int, int, float, float]] List of (parent, child, similarity, gap) tuples
n_components int Number of connected components in the DAG

Methods

Name Description
get_chains_by_coherence Return chains with at least min_coherence.
get_chains_by_length Return chains with at least min_length nodes.
summary Return summary statistics.

get_chains_by_coherence

DAGMiningResult.get_chains_by_coherence(min_coherence=0.65)

Return chains with at least min_coherence.

get_chains_by_length

DAGMiningResult.get_chains_by_length(min_length=3)

Return chains with at least min_length nodes.

summary

DAGMiningResult.summary()

Return summary statistics.