Rate This Document

Findability

Accuracy

Completeness

Readability

Parametrized Constructor

API Definition

class KBest(dim: int, R: int, L: int, A: int, graph_opt_iter:int, metric: string, init_builder_type: string, index_type: string, numa_enabled: bool, numa_nodes_number: int)

Function

Initializes the build of graph search.

Parameters

Parameter	Data Type	Description	Value Range
dim	int	Feature dimension.	[1, 2999]
R	int	Number of neighboring nodes.	[11, 499]
L	int	Size of the candidate node list during the graph build.	[11, 1999]
A	int	Angle threshold during the pruning of graph build.	[11, 360]
graph_opt_iter	int	Number of rounds for graph index self-iteration.	[0, 30]
metric	Character string	Distance measurement indicator.	L2 or IP. L2 indicates the Euclidean distance. IP indicates the inner product distance.
init_builder_type	Character string	Algorithm for building the k-nearest neighbor graph.	RNNDescent and NNDescent indicate two different algorithms for building the k-nearest neighbor graph. RNNDescent is recommended.
index_type	Character string	Neighbor selection strategy.	HNSW, NSG, TSDG, and SSG indicate the four different neighbor selection strategies.
numa_enabled	Boolean	Indicates whether to enable NUMA optimization.	-
numa_nodes_number	int	Number of used NUMA nodes.	≥ 1

Parent topic: Python