Use float instead of double for query vectors. (#46004)

Currently, when using script_score functions like cosineSimilarity, the query
vector is treated as an array of doubles. Since the stored document vectors use
floats, it seems like the least surprising behavior for the query vectors to
also be float arrays.

In addition to improving consistency, this change may help with some
optimizations we have been considering around vector dot product.

Author: jtibshirani

x-pack/plugin/src/test/resources/rest-api-spec/test/vectors/15_dense_vector_l1l2.yml

@@ -65,7 +65,7 @@ setup:
 
   - match: {hits.hits.1._id: "2"}
   - gte: {hits.hits.1._score: 12.29}
-  - lte: {hits.hits.1._score: 12.30}
+  - lte: {hits.hits.1._score: 12.31}
 
   - match: {hits.hits.2._id: "3"}
   - gte: {hits.hits.2._score: 0.00}

x-pack/plugin/src/test/resources/rest-api-spec/test/vectors/35_sparse_vector_l1l2.yml

@@ -63,7 +63,7 @@ setup:
 
   - match: {hits.hits.1._id: "2"}
   - gte: {hits.hits.1._score: 12.29}
-  - lte: {hits.hits.1._score: 12.30}
+  - lte: {hits.hits.1._score: 12.31}
 
   - match: {hits.hits.2._id: "3"}
   - gte: {hits.hits.2._score: 0.00}

x-pack/plugin/vectors/src/main/java/org/elasticsearch/xpack/vectors/mapper/VectorEncoderDecoder.java

@@ -130,7 +130,7 @@ public void swap(int i, int j) {
      * @param values - values for the sparse query vector
      * @param n - number of dimensions
      */
-    public static void sortSparseDimsDoubleValues(int[] dims, double[] values, int n) {
+    public static void sortSparseDimsFloatValues(int[] dims, float[] values, int n) {
         new InPlaceMergeSorter() {
             @Override
             public int compare(int i, int j) {
@@ -143,7 +143,7 @@ public void swap(int i, int j) {
                 dims[i] = dims[j];
                 dims[j] = tempDim;
 
-                double tempValue = values[j];
+                float tempValue = values[j];
                 values[j] = values[i];
                 values[i] = tempValue;
             }

x-pack/plugin/vectors/src/main/java/org/elasticsearch/xpack/vectors/query/ScoreScriptUtils.java

@@ -14,7 +14,7 @@
 import java.util.List;
 import java.util.Map;
 
-import static org.elasticsearch.xpack.vectors.mapper.VectorEncoderDecoder.sortSparseDimsDoubleValues;
+import static org.elasticsearch.xpack.vectors.mapper.VectorEncoderDecoder.sortSparseDimsFloatValues;
 
 public class ScoreScriptUtils {
 
@@ -37,7 +37,7 @@ public static double l1norm(List<Number> queryVector, VectorScriptDocValues.Dens
         Iterator<Number> queryVectorIter = queryVector.iterator();
         double l1norm = 0;
         for (int dim = 0; dim < docVector.length; dim++){
-            l1norm += Math.abs(queryVectorIter.next().doubleValue() - docVector[dim]);
+            l1norm += Math.abs(queryVectorIter.next().floatValue() - docVector[dim]);
         }
         return l1norm;
     }
@@ -59,7 +59,7 @@ public static double l2norm(List<Number> queryVector, VectorScriptDocValues.Dens
         Iterator<Number> queryVectorIter = queryVector.iterator();
         double l2norm = 0;
         for (int dim = 0; dim < docVector.length; dim++){
-            double diff = queryVectorIter.next().doubleValue() - docVector[dim];
+            double diff = queryVectorIter.next().floatValue() - docVector[dim];
             l2norm += diff * diff;
         }
         return Math.sqrt(l2norm);
@@ -97,11 +97,11 @@ public static final class CosineSimilarity {
         // calculate queryVectorMagnitude once per query execution
         public CosineSimilarity(List<Number> queryVector) {
             this.queryVector = queryVector;
-            double doubleValue;
+
             double dotProduct = 0;
             for (Number value : queryVector) {
-                doubleValue = value.doubleValue();
-                dotProduct += doubleValue * doubleValue;
+                float floatValue = value.floatValue();
+                dotProduct += floatValue * floatValue;
             }
             this.queryVectorMagnitude = Math.sqrt(dotProduct);
         }
@@ -130,7 +130,7 @@ private static double intDotProduct(List<Number> v1, float[] v2){
         double v1v2DotProduct = 0;
         Iterator<Number> v1Iter = v1.iterator();
         for (int dim = 0; dim < v2.length; dim++) {
-            v1v2DotProduct += v1Iter.next().doubleValue() * v2[dim];
+            v1v2DotProduct += v1Iter.next().floatValue() * v2[dim];
         }
         return v1v2DotProduct;
     }
@@ -139,15 +139,15 @@ private static double intDotProduct(List<Number> v1, float[] v2){
     //**************FUNCTIONS FOR SPARSE VECTORS
 
     public static class VectorSparseFunctions {
-        final double[] queryValues;
+        final float[] queryValues;
         final int[] queryDims;
 
         // prepare queryVector once per script execution
         // queryVector represents a map of dimensions to values
         public VectorSparseFunctions(Map<String, Number> queryVector) {
             //break vector into two arrays dims and values
             int n = queryVector.size();
-            queryValues = new double[n];
+            queryValues = new float[n];
             queryDims = new int[n];
             int i = 0;
             for (Map.Entry<String, Number> dimValue : queryVector.entrySet()) {
@@ -156,11 +156,11 @@ public VectorSparseFunctions(Map<String, Number> queryVector) {
                 } catch (final NumberFormatException e) {
                     throw new IllegalArgumentException("Failed to parse a query vector dimension, it must be an integer!", e);
                 }
-                queryValues[i] = dimValue.getValue().doubleValue();
+                queryValues[i] = dimValue.getValue().floatValue();
                 i++;
             }
             // Sort dimensions in the ascending order and sort values in the same order as their corresponding dimensions
-            sortSparseDimsDoubleValues(queryDims, queryValues, n);
+            sortSparseDimsFloatValues(queryDims, queryValues, n);
         }
     }
 
@@ -317,7 +317,7 @@ public double cosineSimilaritySparse(VectorScriptDocValues.SparseVectorScriptDoc
         }
     }
 
-    private static double intDotProductSparse(double[] v1Values, int[] v1Dims, float[] v2Values, int[] v2Dims) {
+    private static double intDotProductSparse(float[] v1Values, int[] v1Dims, float[] v2Values, int[] v2Dims) {
         double v1v2DotProduct = 0;
         int v1Index = 0;
         int v2Index = 0;

x-pack/plugin/vectors/src/test/java/org/elasticsearch/xpack/vectors/query/ScoreScriptUtilsTests.java

@@ -36,11 +36,11 @@ public void testDenseVectorFunctions() {
         BytesRef encodedDocVector =  mockEncodeDenseVector(docVector);
         VectorScriptDocValues.DenseVectorScriptDocValues dvs = mock(VectorScriptDocValues.DenseVectorScriptDocValues.class);
         when(dvs.getEncodedValue()).thenReturn(encodedDocVector);
-        List<Number> queryVector = Arrays.asList(0.5, 111.3, -13.0, 14.8, -156.0);
+        List<Number> queryVector = Arrays.asList(0.5f, 111.3f, -13.0f, 14.8f, -156.0f);
 
         // test dotProduct
         double result = dotProduct(queryVector, dvs);
-        assertEquals("dotProduct result is not equal to the expected value!", 65425.626, result, 0.001);
+        assertEquals("dotProduct result is not equal to the expected value!", 65425.624, result, 0.001);
 
         // test cosineSimilarity
         CosineSimilarity cosineSimilarity = new CosineSimilarity(queryVector);
@@ -91,7 +91,7 @@ public void testSparseVectorFunctions() {
         // test dotProduct
         DotProductSparse docProductSparse = new DotProductSparse(queryVector);
         double result = docProductSparse.dotProductSparse(dvs);
-        assertEquals("dotProductSparse result is not equal to the expected value!", 65425.626, result, 0.001);
+        assertEquals("dotProductSparse result is not equal to the expected value!", 65425.624, result, 0.001);
 
         // test cosineSimilarity
         CosineSimilaritySparse cosineSimilaritySparse = new CosineSimilaritySparse(queryVector);
@@ -128,7 +128,7 @@ public void testSparseVectorMissingDimensions1() {
         // test dotProduct
         DotProductSparse docProductSparse = new DotProductSparse(queryVector);
         double result = docProductSparse.dotProductSparse(dvs);
-        assertEquals("dotProductSparse result is not equal to the expected value!", 65425.626, result, 0.001);
+        assertEquals("dotProductSparse result is not equal to the expected value!", 65425.624, result, 0.001);
 
         // test cosineSimilarity
         CosineSimilaritySparse cosineSimilaritySparse = new CosineSimilaritySparse(queryVector);
@@ -165,7 +165,7 @@ public void testSparseVectorMissingDimensions2() {
         // test dotProduct
         DotProductSparse docProductSparse = new DotProductSparse(queryVector);
         double result = docProductSparse.dotProductSparse(dvs);
-        assertEquals("dotProductSparse result is not equal to the expected value!", 65425.626, result, 0.001);
+        assertEquals("dotProductSparse result is not equal to the expected value!", 65425.624, result, 0.001);
 
         // test cosineSimilarity
         CosineSimilaritySparse cosineSimilaritySparse = new CosineSimilaritySparse(queryVector);