cboumalh commented on code in PR #52883: URL: https://github.com/apache/spark/pull/52883#discussion_r2550268830
########## sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/aggregate/tuplesketchesAggregates.scala: ########## @@ -0,0 +1,843 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package org.apache.spark.sql.catalyst.expressions.aggregate + +import org.apache.datasketches.tuple.{Intersection, Sketch, Summary, Union, UpdatableSketch, UpdatableSketchBuilder, UpdatableSummary} + +import org.apache.spark.SparkUnsupportedOperationException +import org.apache.spark.sql.catalyst.InternalRow +import org.apache.spark.sql.catalyst.expressions.{ExpectsInputTypes, Expression, ExpressionDescription, Literal} +import org.apache.spark.sql.catalyst.expressions.aggregate.TypedImperativeAggregate +import org.apache.spark.sql.catalyst.trees.{QuaternaryLike, TernaryLike} +import org.apache.spark.sql.catalyst.util.{ArrayData, CollationFactory, ThetaSketchUtils} +import org.apache.spark.sql.errors.QueryExecutionErrors +import org.apache.spark.sql.internal.types.StringTypeWithCollation +import org.apache.spark.sql.types.{AbstractDataType, ArrayType, BinaryType, DataType, DoubleType, FloatType, IntegerType, LongType, StringType, StructType} +import org.apache.spark.unsafe.types.UTF8String + +sealed trait TupleSketchState { + def serialize(): Array[Byte] + def eval(): Array[Byte] +} +case class UpdatableTupleSketchBuffer[U, S <: UpdatableSummary[U]](sketch: UpdatableSketch[U, S]) + extends TupleSketchState { + override def serialize(): Array[Byte] = sketch.compact.toByteArray + override def eval(): Array[Byte] = sketch.compact.toByteArray +} +case class UnionTupleAggregationBuffer[S <: Summary](union: Union[S]) extends TupleSketchState { + override def serialize(): Array[Byte] = union.getResult.toByteArray + override def eval(): Array[Byte] = union.getResult.toByteArray +} +case class IntersectionTupleAggregationBuffer[S <: Summary](intersection: Intersection[S]) + extends TupleSketchState { + override def serialize(): Array[Byte] = intersection.getResult.toByteArray + override def eval(): Array[Byte] = intersection.getResult.toByteArray +} +case class FinalizedTupleSketch[S <: Summary](sketch: Sketch[S]) extends TupleSketchState { + override def serialize(): Array[Byte] = sketch.toByteArray + override def eval(): Array[Byte] = sketch.toByteArray +} + +/** + * The TupleSketchAgg function utilizes a Datasketches TupleSketch instance to count a + * probabilistic approximation of the number of unique values in a given column with associated + * summary values, and outputs the binary representation of the TupleSketch. + * + * See [[https://datasketches.apache.org/docs/Tuple/TupleSketches.html]] for more information. + * + * @param child + * child expression (struct with key and summary value) against which unique counting will occur + * @param lgNomEntriesExpr + * the log-base-2 of nomEntries decides the number of buckets for the sketch + * @param summaryType + * the type of summary (double, integer, string) Review Comment: Yes that's fair. I thought a bit about this and the reason I went with the single function was to limit the number of SQL functions created which will cause a good amount of redundancy. If we break it down per summary type, this will will result in ~33 functions instead of 11 (tuple has set operations with other tuple and theta sketches leads to an explosion). That said, I agree it would improve type safety and align better with Spark SQL conventions. I’m open to refactoring toward typed variants if we think that’s the right direction! One challenge would be functions like `tuple_sketch_estimate`, breaking that function down by summary type would be a bit awkward. We’d need to decide whether to encode the summary type in the function name or infer it from the sketch input. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
