Add filtering for kNN vector indexer test scenarios (#130751)

* Add filtering for kNN vector indexer test scenarios

* [CI] Auto commit changes from spotless

---------

Co-authored-by: elasticsearchmachine <[email protected]>

Author: benwtrent

qa/vector/src/main/java/org/elasticsearch/test/knn/CmdLineArgs.java

@@ -46,6 +46,8 @@ record CmdLineArgs(
     int indexThreads,
     boolean reindex,
     boolean forceMerge,
+    float filterSelectivity,
+    long seed,
     VectorSimilarityFunction vectorSpace,
     int quantizeBits,
     VectorEncoding vectorEncoding,
@@ -75,6 +77,8 @@ record CmdLineArgs(
     static final ParseField VECTOR_ENCODING_FIELD = new ParseField("vector_encoding");
     static final ParseField DIMENSIONS_FIELD = new ParseField("dimensions");
     static final ParseField EARLY_TERMINATION_FIELD = new ParseField("early_termination");
+    static final ParseField FILTER_SELECTIVITY_FIELD = new ParseField("filter_selectivity");
+    static final ParseField SEED_FIELD = new ParseField("seed");
 
     static CmdLineArgs fromXContent(XContentParser parser) throws IOException {
         Builder builder = PARSER.apply(parser, null);
@@ -106,6 +110,8 @@ static CmdLineArgs fromXContent(XContentParser parser) throws IOException {
         PARSER.declareString(Builder::setVectorEncoding, VECTOR_ENCODING_FIELD);
         PARSER.declareInt(Builder::setDimensions, DIMENSIONS_FIELD);
         PARSER.declareBoolean(Builder::setEarlyTermination, EARLY_TERMINATION_FIELD);
+        PARSER.declareFloat(Builder::setFilterSelectivity, FILTER_SELECTIVITY_FIELD);
+        PARSER.declareLong(Builder::setSeed, SEED_FIELD);
     }
 
     @Override
@@ -136,6 +142,9 @@ public XContentBuilder toXContent(XContentBuilder builder, Params params) throws
         builder.field(QUANTIZE_BITS_FIELD.getPreferredName(), quantizeBits);
         builder.field(VECTOR_ENCODING_FIELD.getPreferredName(), vectorEncoding.name().toLowerCase(Locale.ROOT));
         builder.field(DIMENSIONS_FIELD.getPreferredName(), dimensions);
+        builder.field(EARLY_TERMINATION_FIELD.getPreferredName(), earlyTermination);
+        builder.field(FILTER_SELECTIVITY_FIELD.getPreferredName(), filterSelectivity);
+        builder.field(SEED_FIELD.getPreferredName(), seed);
         return builder.endObject();
     }
 
@@ -167,6 +176,8 @@ static class Builder {
         private VectorEncoding vectorEncoding = VectorEncoding.FLOAT32;
         private int dimensions;
         private boolean earlyTermination;
+        private float filterSelectivity = 1f;
+        private long seed = 1751900822751L;
 
         public Builder setDocVectors(String docVectors) {
             this.docVectors = PathUtils.get(docVectors);
@@ -278,6 +289,16 @@ public Builder setEarlyTermination(Boolean patience) {
             return this;
         }
 
+        public Builder setFilterSelectivity(float filterSelectivity) {
+            this.filterSelectivity = filterSelectivity;
+            return this;
+        }
+
+        public Builder setSeed(long seed) {
+            this.seed = seed;
+            return this;
+        }
+
         public CmdLineArgs build() {
             if (docVectors == null) {
                 throw new IllegalArgumentException("Document vectors path must be provided");
@@ -305,6 +326,8 @@ public CmdLineArgs build() {
                 indexThreads,
                 reindex,
                 forceMerge,
+                filterSelectivity,
+                seed,
                 vectorSpace,
                 quantizeBits,
                 vectorEncoding,

qa/vector/src/main/java/org/elasticsearch/test/knn/KnnIndexTester.java

@@ -178,10 +178,20 @@ public static void main(String[] args) throws Exception {
                 ? cmdLineArgs.nProbes()
                 : new int[] { 0 };
             String indexType = cmdLineArgs.indexType().name().toLowerCase(Locale.ROOT);
-            Results indexResults = new Results(cmdLineArgs.docVectors().getFileName().toString(), indexType, cmdLineArgs.numDocs());
+            Results indexResults = new Results(
+                cmdLineArgs.docVectors().getFileName().toString(),
+                indexType,
+                cmdLineArgs.numDocs(),
+                cmdLineArgs.filterSelectivity()
+            );
             Results[] results = new Results[nProbes.length];
             for (int i = 0; i < nProbes.length; i++) {
-                results[i] = new Results(cmdLineArgs.docVectors().getFileName().toString(), indexType, cmdLineArgs.numDocs());
+                results[i] = new Results(
+                    cmdLineArgs.docVectors().getFileName().toString(),
+                    indexType,
+                    cmdLineArgs.numDocs(),
+                    cmdLineArgs.filterSelectivity()
+                );
             }
             logger.info("Running KNN index tester with arguments: " + cmdLineArgs);
             Codec codec = createCodec(cmdLineArgs);
@@ -244,7 +254,8 @@ public String toString() {
                 "avg_cpu_count",
                 "QPS",
                 "recall",
-                "visited" };
+                "visited",
+                "filter_selectivity" };
 
             // Calculate appropriate column widths based on headers and data
 
@@ -274,7 +285,8 @@ public String toString() {
                     String.format(Locale.ROOT, "%.2f", queryResult.avgCpuCount),
                     String.format(Locale.ROOT, "%.2f", queryResult.qps),
                     String.format(Locale.ROOT, "%.2f", queryResult.avgRecall),
-                    String.format(Locale.ROOT, "%.2f", queryResult.averageVisited) };
+                    String.format(Locale.ROOT, "%.2f", queryResult.averageVisited),
+                    String.format(Locale.ROOT, "%.2f", queryResult.filterSelectivity), };
             }
 
             printBlock(sb, searchHeaders, queryResultsArray);
@@ -339,6 +351,7 @@ private int[] calculateColumnWidths(String[] headers, String[]... data) {
     static class Results {
         final String indexType, indexName;
         final int numDocs;
+        final float filterSelectivity;
         long indexTimeMS;
         long forceMergeTimeMS;
         int numSegments;
@@ -350,10 +363,11 @@ static class Results {
         double netCpuTimeMS;
         double avgCpuCount;
 
-        Results(String indexName, String indexType, int numDocs) {
+        Results(String indexName, String indexType, int numDocs, float filterSelectivity) {
             this.indexName = indexName;
             this.indexType = indexType;
             this.numDocs = numDocs;
+            this.filterSelectivity = filterSelectivity;
         }
     }