CHESS

This is the main interface for CHESS.

class chess.CHESS(data: numpy.memmap, metric: str)

Clustered Hierarchical Entropy-Scaling Search Object.

build(stopping_criteria=None)

Clusters points recursively until stopping_criteria returns True.

Parameters

stopping_criteria – optional override to cluster.partitionable

compress(filename: str)

Compresses the clusters.

knn_search(query, k)

Searches the clusters for the k-nearest points to the query.

static load(filename)

Loads the CHESS object from the given file.

search(query, radius)

Searches the clusters for all points within radius of query.

write(filename)

Writes the CHESS object to the given filename.