Class KnnVectorPropertyDescriptor<T>

Inheritance
object
DescriptorBase<KnnVectorPropertyDescriptor<T>, IKnnVectorProperty>
PropertyDescriptorBase<KnnVectorPropertyDescriptor<T>, IKnnVectorProperty, T>
CorePropertyDescriptorBase<KnnVectorPropertyDescriptor<T>, IKnnVectorProperty, T>
DocValuesPropertyDescriptorBase<KnnVectorPropertyDescriptor<T>, IKnnVectorProperty, T>
KnnVectorPropertyDescriptor<T>
Implements
Inherited Members
Namespace: OpenSearch.Client
Assembly: OpenSearch.Client.dll
Syntax
public class KnnVectorPropertyDescriptor<T> : DocValuesPropertyDescriptorBase<KnnVectorPropertyDescriptor<T>, IKnnVectorProperty, T>, IDescriptor, IKnnVectorProperty, IDocValuesProperty, ICoreProperty, IProperty, IFieldMapping where T : class
Type Parameters
Name Description
T

Constructors

| Edit this page View Source

KnnVectorPropertyDescriptor()

Declaration
public KnnVectorPropertyDescriptor()

Methods

| Edit this page View Source

CompressionLevel(string)

The compression level applied to the vector field (for example, 1x, 2x, 4x). Higher compression reduces memory footprint at the cost of some search accuracy.

Declaration
public KnnVectorPropertyDescriptor<T> CompressionLevel(string compressionLevel)
Parameters
Type Name Description
string compressionLevel
Returns
Type Description
KnnVectorPropertyDescriptor<T>
| Edit this page View Source

DataType(string)

The data type of the vector elements. Defaults to float. Set to byte to store each dimension as an 8-bit signed integer ([-128, 127]), or binary for binary vectors, reducing memory and storage requirements.

Declaration
public KnnVectorPropertyDescriptor<T> DataType(string dataType)
Parameters
Type Name Description
string dataType
Returns
Type Description
KnnVectorPropertyDescriptor<T>
| Edit this page View Source

Dimension(int?)

The dimension of the vector.

Declaration
public KnnVectorPropertyDescriptor<T> Dimension(int? dimension)
Parameters
Type Name Description
int? dimension
Returns
Type Description
KnnVectorPropertyDescriptor<T>
| Edit this page View Source

Method(Func<KnnMethodDescriptor, IKnnMethod>)

The method to use when the underlying Approximate k-NN algorithm does not require training.

Declaration
public KnnVectorPropertyDescriptor<T> Method(Func<KnnMethodDescriptor, IKnnMethod> selector)
Parameters
Type Name Description
Func<KnnMethodDescriptor, IKnnMethod> selector
Returns
Type Description
KnnVectorPropertyDescriptor<T>
| Edit this page View Source

Mode(string)

The mode that determines the default configuration of the vector field, balancing between low latency and low cost (for example, in_memory or on_disk).

Declaration
public KnnVectorPropertyDescriptor<T> Mode(string mode)
Parameters
Type Name Description
string mode
Returns
Type Description
KnnVectorPropertyDescriptor<T>
| Edit this page View Source

ModelId(string)

The model to use when the underlying Approximate k-NN algorithm requires a training step.

Declaration
public KnnVectorPropertyDescriptor<T> ModelId(string modelId)
Parameters
Type Name Description
string modelId
Returns
Type Description
KnnVectorPropertyDescriptor<T>
| Edit this page View Source

SpaceType(string)

The vector space used to calculate the distance between vectors. Can be specified at the top level of the field, as an alternative to specifying it on the Method.

Declaration
public KnnVectorPropertyDescriptor<T> SpaceType(string spaceType)
Parameters
Type Name Description
string spaceType
Returns
Type Description
KnnVectorPropertyDescriptor<T>

Implements

Extension Methods