Package com.azure.cosmos.models
Class CosmosVectorIndexSpec
java.lang.Object
com.azure.cosmos.models.CosmosVectorIndexSpec
Vector Indexes spec for Azure CosmosDB service.
-
Constructor Summary
-
Method Summary
Modifier and TypeMethodDescriptionGets the indexing search list sizegetPath()
Gets path.Gets the quantization byte sizegetType()
Gets the vector index type for the vector indexsetIndexingSearchListSize
(Integer indexingSearchListSize) Sets the indexing search list sizeSets path.setQuantizationSizeInBytes
(Integer quantizationByteSize) Sets the quantization byte sizeSets the vector index type for the vector index
-
Constructor Details
-
CosmosVectorIndexSpec
public CosmosVectorIndexSpec()Constructor
-
-
Method Details
-
getPath
Gets path.- Returns:
- the path.
-
setPath
Sets path.- Parameters:
path
- the path.- Returns:
- the SpatialSpec.
-
getType
Gets the vector index type for the vector index- Returns:
- the vector index type
-
setType
Sets the vector index type for the vector index- Parameters:
type
- the vector index type- Returns:
- the VectorIndexSpec
-
getQuantizationSizeInBytes
Gets the quantization byte size- Returns:
- quantizationByteSize the number of bytes used in product quantization of the vectors. A larger value may result in better recall for vector searches at the expense of latency. This applies to index types DiskANN and quantizedFlat. The allowed range for this parameter is between 1 and 3.
-
setQuantizationSizeInBytes
Sets the quantization byte size- Parameters:
quantizationByteSize
- the number of bytes used in product quantization of the vectors. A larger value may result in better recall for vector searches at the expense of latency. This applies to index types DiskANN and quantizedFlat. The allowed range for this parameter is between 1 and min(Dimensions, 512). The default value would be 64.- Returns:
- CosmosVectorIndexSpec
-
getIndexingSearchListSize
Gets the indexing search list size- Returns:
- indexingSearchListSize which represents the size of the candidate list of approximate neighbors stored while building the DiskANN index as part of the optimization processes. The allowed range for this parameter is between 25 and 500.
-
setIndexingSearchListSize
Sets the indexing search list size- Parameters:
indexingSearchListSize
- indexingSearchListSize which represents the size of the candidate list of approximate neighbors stored while building the DiskANN index as part of the optimization processes. The allowed range for this parameter is between 25 and 500.- Returns:
- CosmosVectorIndexSpec
-