HNSW¶
-
class
hubness.neighbors.
HNSW
(n_candidates: int = 5, metric: str = 'euclidean', method: str = 'hnsw', post_processing: int = 2, n_jobs: int = 1, verbose: int = 0)¶ Bases:
hubness.neighbors.approximate_neighbors.ApproximateNearestNeighbor
Attributes Summary
Methods Summary
fit
(X[, y])Setup the HNSW index.
kneighbors
(X[, n_candidates, return_distance])Attributes Documentation
-
valid_metrics
= ['euclidean', 'l2', 'minkowski', 'cosine', 'cosinesimil']¶
Methods Documentation
-
fit
(X, y=None) → hubness.neighbors.hnsw.HNSW¶ Setup the HNSW index.
-
kneighbors
(X: numpy.ndarray, n_candidates: Optional[int] = None, return_distance: bool = True)¶
-