cboumalh commented on code in PR #51298: URL: https://github.com/apache/spark/pull/51298#discussion_r2331149232
########## sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/expressions/aggregate/ThetasketchesAggSuite.scala: ########## @@ -0,0 +1,175 @@ +/* + * 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 scala.collection.immutable.NumericRange +import scala.util.Random + +import org.apache.spark.SparkFunSuite +import org.apache.spark.sql.catalyst.InternalRow +import org.apache.spark.sql.catalyst.expressions.{BoundReference, ThetaSketchEstimate} +import org.apache.spark.sql.catalyst.util.ArrayData +import org.apache.spark.sql.types.{ArrayType, BinaryType, DataType, DoubleType, FloatType, IntegerType, LongType, StringType} +import org.apache.spark.unsafe.types.UTF8String + +class ThetasketchesAggSuite extends SparkFunSuite { + + def simulateUpdateMerge( + dataType: DataType, + input: Seq[Any], + numSketches: Integer = 5): (Long, NumericRange[Long]) = { + + // Create a map of the agg function instances. + val aggFunctionMap = Seq + .tabulate(numSketches)(index => { + val sketch = new ThetaSketchAgg(BoundReference(0, dataType, nullable = true)) + index -> (sketch, sketch.createAggregationBuffer()) + }) + .toMap + + // Randomly update agg function instances. + input.map(value => { + val (aggFunction, aggBuffer) = aggFunctionMap(Random.nextInt(numSketches)) + aggFunction.update(aggBuffer, InternalRow(value)) + }) + + def serializeDeserialize( + tuple: (ThetaSketchAgg, ThetaSketchState)): (ThetaSketchAgg, ThetaSketchState) = { + val (agg, buf) = tuple + val serialized = agg.serialize(buf) + (agg, agg.deserialize(serialized)) + } + + // Simulate serialization -> deserialization -> merge. + val mapValues = aggFunctionMap.values + val (mergedAgg, UnionAggregationBuffer(mergedBuf)) = + mapValues.tail.foldLeft(mapValues.head)((prev, cur) => { + val (prevAgg, prevBuf) = serializeDeserialize(prev) + val (_, curBuf) = serializeDeserialize(cur) + + (prevAgg, prevAgg.merge(prevBuf, curBuf)) + }) + + val estimator = ThetaSketchEstimate(BoundReference(0, BinaryType, nullable = true)) + val estimate = + estimator.eval(InternalRow(mergedBuf.getResult.toByteArrayCompressed)).asInstanceOf[Long] + ( + estimate, + mergedBuf.getResult.getLowerBound(3).toLong to mergedBuf.getResult.getUpperBound(3).toLong) + } + + test("SPARK-52407: Test min/max values of supported datatypes") { + val intRange = Integer.MIN_VALUE to Integer.MAX_VALUE by 10000000 + val (intEstimate, intEstimateRange) = simulateUpdateMerge(IntegerType, intRange) + assert(intEstimate == intRange.size || intEstimateRange.contains(intRange.size.toLong)) + + val longRange = Long.MinValue to Long.MaxValue by 1000000000000000L + val (longEstimate, longEstimateRange) = simulateUpdateMerge(LongType, longRange) + assert(longEstimate == longRange.size || longEstimateRange.contains(longRange.size.toLong)) + + val stringRange = Seq.tabulate(1000)(i => UTF8String.fromString(Random.nextString(i + 1))) + val (stringEstimate, stringEstimateRange) = simulateUpdateMerge(StringType, stringRange) + assert( + stringEstimate == stringRange.size || + stringEstimateRange.contains(stringRange.size.toLong)) + + val binaryRange = + Seq.tabulate(1000)(i => UTF8String.fromString(Random.nextString(i + 1)).getBytes) + val (binaryEstimate, binaryEstimateRange) = simulateUpdateMerge(BinaryType, binaryRange) + assert( + binaryEstimate == binaryRange.size || + binaryEstimateRange.contains(binaryRange.size.toLong)) + + val floatRange = (1 to 1000).map(_.toFloat) + val (floatEstimate, floatRangeEst) = simulateUpdateMerge(FloatType, floatRange) + assert(floatEstimate == floatRange.size || floatRangeEst.contains(floatRange.size.toLong)) + + val doubleRange = (1 to 1000).map(_.toDouble) + val (doubleEstimate, doubleRangeEst) = simulateUpdateMerge(DoubleType, doubleRange) + assert(doubleEstimate == doubleRange.size || doubleRangeEst.contains(doubleRange.size.toLong)) + + val arrayIntRange = (1 to 500).map(i => ArrayData.toArrayData(Array(i, i + 1))) + val (arrayIntEstimate, arrayIntRangeEst) = + simulateUpdateMerge(ArrayType(IntegerType), arrayIntRange) + assert( + arrayIntEstimate == arrayIntRange.size || + arrayIntRangeEst.contains(arrayIntRange.size.toLong)) + + val arrayLongRange = + (1 to 500).map(i => ArrayData.toArrayData(Array(i.toLong, (i + 1).toLong))) + val (arrayLongEstimate, arrayLongRangeEst) = + simulateUpdateMerge(ArrayType(LongType), arrayLongRange) + assert( + arrayLongEstimate == arrayLongRange.size || + arrayLongRangeEst.contains(arrayLongRange.size.toLong)) + } + + test("SPARK-52407: Test lgNomEntries results in downsampling sketches during Union") { + // Create a sketch with larger configuration (more precise). + val aggFunc1 = new ThetaSketchAgg(BoundReference(0, IntegerType, nullable = true), 12) + val sketch1 = aggFunc1.createAggregationBuffer() + (0 to 100).map(i => aggFunc1.update(sketch1, InternalRow(i))) + val binary1 = aggFunc1.eval(sketch1) + + // Create a sketch with smaller configuration (less precise). + val aggFunc2 = new ThetaSketchAgg(BoundReference(0, IntegerType, nullable = true), 10) + val sketch2 = aggFunc2.createAggregationBuffer() + (0 to 100).map(i => aggFunc2.update(sketch2, InternalRow(i))) + val binary2 = aggFunc2.eval(sketch2) + + // Union the sketches. + val unionAgg = new ThetaUnionAgg(BoundReference(0, BinaryType, nullable = true), 12) Review Comment: The issue is that Theta sketches does not allow any access to the nominal value past the initial aggregation (creation of the sketch). The nominal entries value does not carry any actual value/meaning in theory past this point. Because of this, the next best thing in my opinion is to give the power to the developers to assign the value for each operation accordingly. If they want it constant across all methods they can either create wrappers for each of the methods with that specific value, or just assign it to every call through a CONST defined somewhere in their package. -- 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: reviews-unsubscr...@spark.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
