use the latest datasketches-java-4.0.0

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/kll/KllDoublesSketchToHistogramPostAggregator.java

@@ -78,7 +78,7 @@ public Object compute(final Map<String, Object> combinedAggregators)
       return histogram;
     }
     final double[] histogram = sketch.getPMF(splitPoints != null ? splitPoints :
-      equallySpacedPoints(numBins, sketch.getMinValue(), sketch.getMaxValue()));
+      equallySpacedPoints(numBins, sketch.getMinItem(), sketch.getMaxItem()));
     for (int i = 0; i < histogram.length; i++) {
       histogram[i] *= sketch.getN(); // scale fractions to counts
     }

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/kll/KllDoublesSketchToRankPostAggregator.java

@@ -83,7 +83,7 @@ public double getValue()
   public Object compute(final Map<String, Object> combinedAggregators)
   {
     final KllDoublesSketch sketch = (KllDoublesSketch) field.compute(combinedAggregators);
-    return sketch.getRank(value);
+    return sketch.isEmpty() ? Double.NaN : sketch.getRank(value);
   }
 
   @Override

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/kll/KllFloatsSketchToHistogramPostAggregator.java

@@ -78,7 +78,7 @@ public Object compute(final Map<String, Object> combinedAggregators)
       return histogram;
     }
     final double[] histogram = sketch.getPMF(splitPoints != null ? splitPoints :
-      equallySpacedPoints(numBins, sketch.getMinValue(), sketch.getMaxValue()));
+      equallySpacedPoints(numBins, sketch.getMinItem(), sketch.getMaxItem()));
     for (int i = 0; i < histogram.length; i++) {
       histogram[i] *= sketch.getN(); // scale fractions to counts
     }

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/kll/KllFloatsSketchToRankPostAggregator.java

@@ -83,7 +83,7 @@ public float getValue()
   public Object compute(final Map<String, Object> combinedAggregators)
   {
     final KllFloatsSketch sketch = (KllFloatsSketch) field.compute(combinedAggregators);
-    return sketch.getRank(value);
+    return sketch.isEmpty() ? Double.NaN : sketch.getRank(value);
   }
 
   @Override

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/quantiles/DoublesSketchAggregatorFactory.java

@@ -23,7 +23,7 @@
 import com.fasterxml.jackson.annotation.JsonInclude;
 import com.fasterxml.jackson.annotation.JsonProperty;
 import com.google.common.annotations.VisibleForTesting;
-import org.apache.datasketches.Util;
+import org.apache.datasketches.common.Util;
 import org.apache.datasketches.quantiles.DoublesSketch;
 import org.apache.datasketches.quantiles.DoublesUnion;
 import org.apache.druid.jackson.DefaultTrueJsonIncludeFilter;
@@ -123,7 +123,7 @@ public DoublesSketchAggregatorFactory(
     }
     this.fieldName = fieldName;
     this.k = k == null ? DEFAULT_K : k;
-    Util.checkIfPowerOf2(this.k, "k");
+    Util.checkIfIntPowerOf2(this.k, "k");
     this.maxStreamLength = maxStreamLength == null ? DEFAULT_MAX_STREAM_LENGTH : maxStreamLength;
     this.shouldFinalize = shouldFinalize == null ? DEFAULT_SHOULD_FINALIZE : shouldFinalize;
     this.cacheTypeId = cacheTypeId;
@@ -240,8 +240,8 @@ public Comparator<DoublesSketch> getComparator()
   public Object combine(final Object lhs, final Object rhs)
   {
     final DoublesUnion union = DoublesUnion.builder().setMaxK(k).build();
-    union.update((DoublesSketch) lhs);
-    union.update((DoublesSketch) rhs);
+    union.union((DoublesSketch) lhs);
+    union.union((DoublesSketch) rhs);
     return union.getResultAndReset();
   }
 
@@ -263,7 +263,7 @@ public void reset(final ColumnValueSelector selector)
       public void fold(final ColumnValueSelector selector)
       {
         final DoublesSketch sketch = (DoublesSketch) selector.getObject();
-        union.update(sketch);
+        union.union(sketch);
       }
 
       @Nullable

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/quantiles/DoublesSketchMergeAggregator.java

@@ -77,7 +77,7 @@ static void updateUnion(ColumnValueSelector selector, DoublesUnion union)
       return;
     }
     if (object instanceof DoublesSketch) {
-      union.update((DoublesSketch) object);
+      union.union((DoublesSketch) object);
     } else {
       union.update(selector.getDouble());
     }

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/quantiles/DoublesSketchMergeVectorAggregator.java

@@ -58,7 +58,7 @@ public void aggregate(final ByteBuffer buf, final int position, final int startR
     for (int i = startRow; i < endRow; i++) {
       final DoublesSketch sketch = (DoublesSketch) vector[i];
       if (sketch != null) {
-        union.update(sketch);
+        union.union(sketch);
       }
     }
   }
@@ -80,7 +80,7 @@ public void aggregate(
       if (sketch != null) {
         final int position = positions[i] + positionOffset;
         final DoublesUnion union = helper.getSketchAtPosition(buf, position);
-        union.update(sketch);
+        union.union(sketch);
       }
     }
   }

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/quantiles/DoublesSketchToHistogramPostAggregator.java

@@ -78,7 +78,7 @@ public Object compute(final Map<String, Object> combinedAggregators)
       return histogram;
     }
     final double[] histogram = sketch.getPMF(splitPoints != null ? splitPoints :
-      equallySpacedPoints(numBins, sketch.getMinValue(), sketch.getMaxValue()));
+      equallySpacedPoints(numBins, sketch.getMinItem(), sketch.getMaxItem()));
     for (int i = 0; i < histogram.length; i++) {
       histogram[i] *= sketch.getN(); // scale fractions to counts
     }

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/quantiles/DoublesSketchToQuantilePostAggregator.java

@@ -83,7 +83,7 @@ public double getFraction()
   public Object compute(final Map<String, Object> combinedAggregators)
   {
     final DoublesSketch sketch = (DoublesSketch) field.compute(combinedAggregators);
-    return sketch.getQuantile(fraction);
+    return sketch.isEmpty() ? Double.NaN : sketch.getQuantile(fraction);
   }
 
   @Override

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/quantiles/DoublesSketchToRankPostAggregator.java

@@ -83,7 +83,7 @@ public double getValue()
   public Object compute(final Map<String, Object> combinedAggregators)
   {
     final DoublesSketch sketch = (DoublesSketch) field.compute(combinedAggregators);
-    return sketch.getRank(value);
+    return sketch.isEmpty() ? Double.NaN : sketch.getRank(value);
   }
 
   @Override

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/theta/SketchAggregator.java

@@ -19,7 +19,7 @@
 
 package org.apache.druid.query.aggregation.datasketches.theta;
 
-import org.apache.datasketches.Family;
+import org.apache.datasketches.common.Family;
 import org.apache.datasketches.theta.SetOperation;
 import org.apache.datasketches.theta.Union;
 import org.apache.druid.common.config.NullHandling;

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/theta/SketchAggregatorFactory.java

@@ -22,10 +22,11 @@
 import com.fasterxml.jackson.annotation.JsonProperty;
 import com.google.common.base.Preconditions;
 import com.google.common.primitives.Ints;
-import org.apache.datasketches.Family;
-import org.apache.datasketches.Util;
+import org.apache.datasketches.common.Family;
+import org.apache.datasketches.common.Util;
 import org.apache.datasketches.theta.SetOperation;
 import org.apache.datasketches.theta.Union;
+import org.apache.datasketches.thetacommon.ThetaUtil;
 import org.apache.druid.java.util.common.StringUtils;
 import org.apache.druid.query.aggregation.AggregateCombiner;
 import org.apache.druid.query.aggregation.Aggregator;
@@ -52,7 +53,7 @@ public abstract class SketchAggregatorFactory extends AggregatorFactory
 
   // Smallest number of entries in an Aggregator. Each entry is a long. Based on the constructor of
   // HeapQuickSelectSketch and used by guessAggregatorHeapFootprint.
-  private static final int MIN_ENTRIES_PER_AGGREGATOR = 1 << Util.MIN_LG_ARR_LONGS;
+  private static final int MIN_ENTRIES_PER_AGGREGATOR = 1 << ThetaUtil.MIN_LG_ARR_LONGS;
 
   // Largest preamble size for the sketch stored in an Aggregator, in bytes. Based on Util.getMaxUnionBytes.
   private static final int LONGEST_POSSIBLE_PREAMBLE_BYTES = Family.UNION.getMaxPreLongs() << 3;
@@ -68,7 +69,7 @@ public SketchAggregatorFactory(String name, String fieldName, Integer size, byte
     this.fieldName = Preconditions.checkNotNull(fieldName, "Must have a valid, non-null fieldName");
 
     this.size = size == null ? DEFAULT_MAX_SKETCH_SIZE : size;
-    Util.checkIfPowerOf2(this.size, "size");
+    Util.checkIfIntPowerOf2(this.size, "size");
 
     this.cacheId = cacheId;
   }
@@ -197,7 +198,7 @@ public int guessAggregatorHeapFootprint(long rows)
       expectedEntries = maxEntries;
     } else {
       // rows is within int range since it's <= maxEntries, so casting is OK.
-      expectedEntries = Math.max(MIN_ENTRIES_PER_AGGREGATOR, Util.ceilingPowerOf2(Ints.checkedCast(rows)));
+      expectedEntries = Math.max(MIN_ENTRIES_PER_AGGREGATOR, Util.ceilingIntPowerOf2(Ints.checkedCast(rows)));
     }
 
     // 8 bytes per entry + largest possible preamble.

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/theta/SketchBufferAggregatorHelper.java

@@ -21,7 +21,7 @@
 
 import it.unimi.dsi.fastutil.ints.Int2ObjectMap;
 import it.unimi.dsi.fastutil.ints.Int2ObjectOpenHashMap;
-import org.apache.datasketches.Family;
+import org.apache.datasketches.common.Family;
 import org.apache.datasketches.memory.WritableMemory;
 import org.apache.datasketches.theta.SetOperation;
 import org.apache.datasketches.theta.Union;

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/theta/SketchHolder.java

@@ -23,7 +23,7 @@
 import com.google.common.collect.Ordering;
 import com.google.common.primitives.Doubles;
 import com.google.common.primitives.Longs;
-import org.apache.datasketches.Family;
+import org.apache.datasketches.common.Family;
 import org.apache.datasketches.memory.Memory;
 import org.apache.datasketches.theta.AnotB;
 import org.apache.datasketches.theta.Intersection;

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/theta/SketchSetPostAggregator.java

@@ -21,7 +21,7 @@
 
 import com.fasterxml.jackson.annotation.JsonCreator;
 import com.fasterxml.jackson.annotation.JsonProperty;
-import org.apache.datasketches.Util;
+import org.apache.datasketches.common.Util;
 import org.apache.druid.java.util.common.IAE;
 import org.apache.druid.query.aggregation.AggregatorFactory;
 import org.apache.druid.query.aggregation.PostAggregator;
@@ -55,7 +55,7 @@ public SketchSetPostAggregator(
     this.fields = fields;
     this.func = SketchHolder.Func.valueOf(func);
     this.maxSketchSize = maxSize == null ? SketchAggregatorFactory.DEFAULT_MAX_SKETCH_SIZE : maxSize;
-    Util.checkIfPowerOf2(this.maxSketchSize, "size");
+    Util.checkIfIntPowerOf2(this.maxSketchSize, "size");
 
     if (fields.size() <= 1) {
       throw new IAE("Illegal number of fields[%s], must be > 1", fields.size());

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/tuple/ArrayOfDoublesSketchAggregatorFactory.java

@@ -22,7 +22,8 @@
 import com.fasterxml.jackson.annotation.JsonCreator;
 import com.fasterxml.jackson.annotation.JsonProperty;
 import com.google.common.base.Preconditions;
-import org.apache.datasketches.Util;
+import org.apache.datasketches.common.Util;
+import org.apache.datasketches.thetacommon.ThetaUtil;
 import org.apache.datasketches.tuple.arrayofdoubles.ArrayOfDoublesSetOperationBuilder;
 import org.apache.datasketches.tuple.arrayofdoubles.ArrayOfDoublesSketch;
 import org.apache.datasketches.tuple.arrayofdoubles.ArrayOfDoublesUnion;
@@ -74,8 +75,8 @@ public ArrayOfDoublesSketchAggregatorFactory(
   {
     this.name = Preconditions.checkNotNull(name, "Must have a valid, non-null aggregator name");
     this.fieldName = Preconditions.checkNotNull(fieldName, "Must have a valid, non-null fieldName");
-    this.nominalEntries = nominalEntries == null ? Util.DEFAULT_NOMINAL_ENTRIES : nominalEntries;
-    Util.checkIfPowerOf2(this.nominalEntries, "nominalEntries");
+    this.nominalEntries = nominalEntries == null ? ThetaUtil.DEFAULT_NOMINAL_ENTRIES : nominalEntries;
+    Util.checkIfIntPowerOf2(this.nominalEntries, "nominalEntries");
     this.metricColumns = metricColumns;
     this.numberOfValues = numberOfValues == null ? (metricColumns == null ? 1 : metricColumns.size()) : numberOfValues;
     if (metricColumns != null && metricColumns.size() != this.numberOfValues) {

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/tuple/ArrayOfDoublesSketchSetOpPostAggregator.java

@@ -21,7 +21,8 @@
 
 import com.fasterxml.jackson.annotation.JsonCreator;
 import com.fasterxml.jackson.annotation.JsonProperty;
-import org.apache.datasketches.Util;
+import org.apache.datasketches.common.Util;
+import org.apache.datasketches.thetacommon.ThetaUtil;
 import org.apache.datasketches.tuple.arrayofdoubles.ArrayOfDoublesSketch;
 import org.apache.druid.java.util.common.IAE;
 import org.apache.druid.query.aggregation.AggregatorUtil;
@@ -58,9 +59,9 @@ public ArrayOfDoublesSketchSetOpPostAggregator(
   {
     super(name, fields);
     this.operation = ArrayOfDoublesSketchOperations.Operation.valueOf(operation);
-    this.nominalEntries = nominalEntries == null ? Util.DEFAULT_NOMINAL_ENTRIES : nominalEntries;
+    this.nominalEntries = nominalEntries == null ? ThetaUtil.DEFAULT_NOMINAL_ENTRIES : nominalEntries;
     this.numberOfValues = numberOfValues == null ? 1 : numberOfValues;
-    Util.checkIfPowerOf2(this.nominalEntries, "size");
+    Util.checkIfIntPowerOf2(this.nominalEntries, "size");
 
     if (fields.size() <= 1) {
       throw new IAE("Illegal number of fields[%d], must be > 1", fields.size());

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/tuple/sql/ArrayOfDoublesSketchSetBaseOperatorConversion.java

@@ -28,7 +28,7 @@
 import org.apache.calcite.sql.SqlOperator;
 import org.apache.calcite.sql.type.OperandTypes;
 import org.apache.calcite.sql.type.SqlOperandCountRanges;
-import org.apache.datasketches.Util;
+import org.apache.datasketches.thetacommon.ThetaUtil;
 import org.apache.druid.java.util.common.StringUtils;
 import org.apache.druid.query.aggregation.PostAggregator;
 import org.apache.druid.query.aggregation.datasketches.tuple.ArrayOfDoublesSketchSetOpPostAggregator;
@@ -92,7 +92,7 @@ public PostAggregator toPostAggregator(
       nominalEntries = ((Number) RexLiteral.value(potentialNominalEntriesArg)).intValue();
       metricExpressionEndIndex = lastArgIndex - 1;
     } else {
-      nominalEntries = Util.DEFAULT_NOMINAL_ENTRIES;
+      nominalEntries = ThetaUtil.DEFAULT_NOMINAL_ENTRIES;
       metricExpressionEndIndex = lastArgIndex;
     }
 

extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/tuple/sql/ArrayOfDoublesSketchSqlAggregator.java

@@ -31,7 +31,7 @@
 import org.apache.calcite.sql.type.InferTypes;
 import org.apache.calcite.sql.type.OperandTypes;
 import org.apache.calcite.util.Optionality;
-import org.apache.datasketches.Util;
+import org.apache.datasketches.thetacommon.ThetaUtil;
 import org.apache.druid.java.util.common.ISE;
 import org.apache.druid.query.aggregation.AggregatorFactory;
 import org.apache.druid.query.aggregation.datasketches.tuple.ArrayOfDoublesSketchAggregatorFactory;
@@ -99,7 +99,7 @@ public Aggregation toDruidAggregation(
       nominalEntries = ((Number) RexLiteral.value(potentialNominalEntriesArg)).intValue();
       metricExpressionEndIndex = lastArgIndex - 1;
     } else {
-      nominalEntries = Util.DEFAULT_NOMINAL_ENTRIES;
+      nominalEntries = ThetaUtil.DEFAULT_NOMINAL_ENTRIES;
       metricExpressionEndIndex = lastArgIndex;
     }
 

extensions-core/datasketches/src/test/java/org/apache/druid/query/aggregation/datasketches/hll/HllSketchObjectStrategyTest.java

@@ -19,7 +19,7 @@
 
 package org.apache.druid.query.aggregation.datasketches.hll;
 
-import org.apache.datasketches.SketchesArgumentException;
+import org.apache.datasketches.common.SketchesArgumentException;
 import org.apache.datasketches.hll.HllSketch;
 import org.apache.druid.java.util.common.StringUtils;
 import org.junit.Assert;
@@ -53,7 +53,7 @@ public void testSafeRead()
 
       final ByteBuffer buf2 = ByteBuffer.wrap(garbage2).order(ByteOrder.LITTLE_ENDIAN);
       Assert.assertThrows(
-          IndexOutOfBoundsException.class,
+          Exception.class, // can throw either SketchesArgumentException or IndexOutOfBoundsException
           () -> objectStrategy.fromByteBufferSafe(buf2, garbage2.length).getSketch().copy()
       );
     }
@@ -62,7 +62,7 @@ public void testSafeRead()
     final byte[] garbage = new byte[]{0x01, 0x02};
     final ByteBuffer buf3 = ByteBuffer.wrap(garbage).order(ByteOrder.LITTLE_ENDIAN);
     Assert.assertThrows(
-        IndexOutOfBoundsException.class,
+        SketchesArgumentException.class,
         () -> objectStrategy.fromByteBufferSafe(buf3, garbage.length).getSketch().copy()
     );
 

extensions-core/datasketches/src/test/java/org/apache/druid/query/aggregation/datasketches/kll/KllDoublesSketchComplexMetricSerdeTest.java

@@ -55,7 +55,7 @@ public void testExtractorOnPositiveNumber()
         "foo"
     );
     Assert.assertEquals(1, sketch.getNumRetained());
-    Assert.assertEquals(777d, sketch.getMaxValue(), 0.01d);
+    Assert.assertEquals(777d, sketch.getMaxItem(), 0.01d);
   }
 
   @Test
@@ -68,7 +68,7 @@ public void testExtractorOnNegativeNumber()
         "foo"
     );
     Assert.assertEquals(1, sketch.getNumRetained());
-    Assert.assertEquals(-133d, sketch.getMaxValue(), 0.01d);
+    Assert.assertEquals(-133d, sketch.getMaxItem(), 0.01d);
   }
 
   @Test
@@ -81,7 +81,7 @@ public void testExtractorOnDecimalNumber()
         "foo"
     );
     Assert.assertEquals(1, sketch.getNumRetained());
-    Assert.assertEquals(3.1d, sketch.getMaxValue(), 0.01d);
+    Assert.assertEquals(3.1d, sketch.getMaxItem(), 0.01d);
   }
 
   @Test
@@ -94,7 +94,7 @@ public void testExtractorOnLeadingDecimalNumber()
         "foo"
     );
     Assert.assertEquals(1, sketch.getNumRetained());
-    Assert.assertEquals(0.1d, sketch.getMaxValue(), 0.01d);
+    Assert.assertEquals(0.1d, sketch.getMaxItem(), 0.01d);
   }
 
   @Test

extensions-core/datasketches/src/test/java/org/apache/druid/query/aggregation/datasketches/kll/KllDoublesSketchToCDFPostAggregatorTest.java

@@ -142,7 +142,7 @@ public void normalCase()
     final PostAggregator postAgg = new KllDoublesSketchToCDFPostAggregator(
         "cdf",
         new FieldAccessPostAggregator("field", "sketch"),
-        new double[] {4} // half of the distribution is below 4
+        new double[] {3} // half of the distribution is less or equals 3
     );
 
     final double[] cdf = (double[]) postAgg.compute(fields);

extensions-core/datasketches/src/test/java/org/apache/druid/query/aggregation/datasketches/kll/KllDoublesSketchToRankPostAggregatorTest.java

@@ -120,7 +120,7 @@ public void normalCase()
     final PostAggregator postAgg = new KllDoublesSketchToRankPostAggregator(
         "rank",
         new FieldAccessPostAggregator("field", "sketch"),
-        4
+        3
     );
 
     final double rank = (double) postAgg.compute(fields);