create_searcher

API Definition

def create_searcher(db: numpy.ndarray, scann_config: str, training_threads: int, docids: numpy.ndarray, kmopt_pq_iter: int, kmopt_pq_sample: int, kmopt_pq_init: int, kmopt_ivf_iter: int, kmopt_ivf_sample: int, kmopt_ivf_init: int, filter_thr: float, filter_type: int) -> scann.scann_ops.ScannSearcher

Function

Builds an index based on the input parameters and returns the ScannSearcher (filter_thr and filter_type are KScaNN-specific parameters).

Parameters

Parameter	Data Type	Description	Value Range
db	numpy.ndarray	Base library vector.	The value cannot be null.
scann_config	Character string	Configuration file required for creating the index, containing all configuration parameters.	-
training_threads	int	Number of threads used for index construction.	[0, MAXINT]. The default value is 0.
docids	numpy.ndarray	Document ID.	The default value is None.
kmopt_pq_iter	int	Number of K-means iterations in PQ.	[0, MAXINT]. The default value is 0.
kmopt_pq_sample	int	Sample size for PQ K-means clustering.	[0, MAXINT]. The default value is 0.
kmopt_pq_init	int	K-means initialization type in PQ.	{0, 1, 2, 3}. The default value is 0. 1: initialization based on the average distance. 2: initialization based on the K-means++ algorithm. 3: random initialization. 0: Do not enable K-means optimization in PQ.
kmopt_ivf_iter	int	Number of K-means iterations in IVF.	[0, MAXINT]. The default value is 0.
kmopt_ivf_sample	int	Sample size for IVF K-means clustering.	[0, MAXINT]. The default value is 0.
kmopt_ivf_init	int	IVF K-means initialization type.	{0, 1, 2, 3}. The default value is 0. 1: initialization based on the average distance. 2: initialization based on the K-means++ algorithm. 3: random initialization. 0: Do not enable K-means optimization in IVF.
filter_thr	float	Filter threshold.	[0, 1]. The default value is 0.
filter_type	int	Filter type.	The value is 0 or 1. The default value is 0. 0: filtering based on the number of zero elements in the vector. 1: filtering based on the deviation of elements from the mean value.

Return Value

Data Type	Description
scann.scann_ops.ScannSearcher	The build starts and the built ScannSearcher is returned.

Parent topic: Python APIs