The dimensionality of vectors.
The depth of exploration at add time.
The depth of exploration at add time.
The value to set.
The depth of exploration of the search.
The depth of exploration of the search.
The value to set.
The type of index.
returns a boolean that indicates whether training is required.
Whether training is required.
Argument of the metric type.
The metric of the index.
The number of verctors currently indexed.
Add n vectors of dimension d to the index using the provided labels.
Input matrix, size n * d
Vector identifiers
Add n vectors of dimension d to the index with ID's.
Input matrix, size n * d
Vector identifiers
Merge the current index with another IndexHNSW instance.
The other IndexHNSW instance to merge from.
Query n vectors of dimension d to the index. return at most k vectors. If there are not enough results for a query, the result array is padded with -1s.
Input vectors to search, size n * d.
The number of nearest neighbors to search for.
Output of the search result.
Static fromStatic fromStatic readGenerated using TypeDoc
IndexHNSW Index. The Hierarchical Navigable Small World indexing method is based on a graph built on the indexed vectors.
Param
The dimensionality of index.
Param
The number of neighbors used in the graph (defaults to 32).
Param
Metric type (defaults to L2).