Uses of Class
org.apache.lucene.util.quantization.ScalarQuantizer
Packages that use ScalarQuantizer
Package
Description
HNSW vector helper classes.
Lucene 9.9 file format.
Provides quantization methods for scaling vector values to smaller data types and possibly fewer
dimensions
-
Uses of ScalarQuantizer in org.apache.lucene.codecs.hnsw
Methods in org.apache.lucene.codecs.hnsw with parameters of type ScalarQuantizerModifier and TypeMethodDescriptionstatic float
ScalarQuantizedVectorScorer.quantizeQuery
(float[] query, byte[] quantizedQuery, VectorSimilarityFunction similarityFunction, ScalarQuantizer scalarQuantizer) Constructors in org.apache.lucene.codecs.hnsw with parameters of type ScalarQuantizerModifierConstructorDescriptionScalarQuantizedRandomVectorScorerSupplier
(VectorSimilarityFunction similarityFunction, ScalarQuantizer scalarQuantizer, RandomAccessQuantizedByteVectorValues values) -
Uses of ScalarQuantizer in org.apache.lucene.codecs.lucene99
Fields in org.apache.lucene.codecs.lucene99 declared as ScalarQuantizerModifier and TypeFieldDescriptionprotected final ScalarQuantizer
OffHeapQuantizedByteVectorValues.scalarQuantizer
Methods in org.apache.lucene.codecs.lucene99 that return ScalarQuantizerModifier and TypeMethodDescriptionLucene99HnswVectorsReader.getQuantizationState
(String field) Lucene99ScalarQuantizedVectorsReader.getQuantizationState
(String fieldName) OffHeapQuantizedByteVectorValues.getScalarQuantizer()
static ScalarQuantizer
Lucene99ScalarQuantizedVectorsWriter.mergeAndRecalculateQuantiles
(MergeState mergeState, FieldInfo fieldInfo, Float confidenceInterval, byte bits) Merges the quantiles of the segments and recalculates the quantiles if necessary.Methods in org.apache.lucene.codecs.lucene99 with parameters of type ScalarQuantizerModifier and TypeMethodDescriptionOffHeapQuantizedByteVectorValues.load
(OrdToDocDISIReaderConfiguration configuration, int dimension, int size, ScalarQuantizer scalarQuantizer, VectorSimilarityFunction similarityFunction, FlatVectorsScorer vectorsScorer, boolean compress, long quantizedVectorDataOffset, long quantizedVectorDataLength, IndexInput vectorData) Constructors in org.apache.lucene.codecs.lucene99 with parameters of type ScalarQuantizerModifierConstructorDescriptionDenseOffHeapVectorValues
(int dimension, int size, ScalarQuantizer scalarQuantizer, boolean compress, VectorSimilarityFunction similarityFunction, FlatVectorsScorer vectorsScorer, IndexInput slice) -
Uses of ScalarQuantizer in org.apache.lucene.util.quantization
Methods in org.apache.lucene.util.quantization that return ScalarQuantizerModifier and TypeMethodDescriptionstatic ScalarQuantizer
ScalarQuantizer.fromVectors
(FloatVectorValues floatVectorValues, float confidenceInterval, int totalVectorCount, byte bits) This will read the float vector values and calculate the quantiles.static ScalarQuantizer
ScalarQuantizer.fromVectorsAutoInterval
(FloatVectorValues floatVectorValues, VectorSimilarityFunction function, int totalVectorCount, byte bits) QuantizedVectorsReader.getQuantizationState
(String fieldName) RandomAccessQuantizedByteVectorValues.getScalarQuantizer()
Methods in org.apache.lucene.util.quantization with parameters of type ScalarQuantizerModifier and TypeMethodDescriptionfloat
ScalarQuantizer.recalculateCorrectiveOffset
(byte[] quantizedVector, ScalarQuantizer oldQuantizer, VectorSimilarityFunction similarityFunction) Recalculate the old score corrective value given new current quantiles