Github user hvanhovell commented on a diff in the pull request:
https://github.com/apache/spark/pull/14136#discussion_r88500845
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/aggregate/Percentile.scala
---
@@ -0,0 +1,201 @@
+/*
+ * 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.spark.sql.AnalysisException
+import org.apache.spark.sql.catalyst.InternalRow
+import org.apache.spark.sql.catalyst.analysis.TypeCheckResult
+import org.apache.spark.sql.catalyst.expressions._
+import org.apache.spark.sql.catalyst.util._
+import org.apache.spark.sql.types._
+import org.apache.spark.util.collection.OpenHashMap
+
+/**
+ * The Percentile aggregate function returns the exact percentile(s) of
numeric column `expr` at
+ * the given percentage(s) with value range in [0.0, 1.0].
+ *
+ * The operator is bound to the slower sort based aggregation path because
the number of elements
+ * and their partial order cannot be determined in advance. Therefore we
have to store all the
+ * elements in memory, and that too many elements can cause GC paused and
eventually OutOfMemory
+ * Errors.
+ *
+ * @param child child expression that produce numeric column value with
`child.eval(inputRow)`
+ * @param percentageExpression Expression that represents a single
percentage value or an array of
+ * percentage values. Each percentage value
must be in the range
+ * [0.0, 1.0].
+ */
+@ExpressionDescription(
+ usage =
+ """
+ _FUNC_(col, percentage) - Returns the exact percentile value of
numeric column `col` at the
+ given percentage. The value of percentage must be between 0.0 and
1.0.
+
+ _FUNC_(col, array(percentage1 [, percentage2]...)) - Returns the
exact percentile value array
+ of numeric column `col` at the given percentage(s). Each value of
the percentage array must
+ be between 0.0 and 1.0.
+ """)
+case class Percentile(
+ child: Expression,
+ percentageExpression: Expression,
+ mutableAggBufferOffset: Int = 0,
+ inputAggBufferOffset: Int = 0) extends ImperativeAggregate {
+
+ def this(child: Expression, percentageExpression: Expression) = {
+ this(child, percentageExpression, 0, 0)
+ }
+
+ override def prettyName: String = "percentile"
+
+ override def withNewMutableAggBufferOffset(newMutableAggBufferOffset:
Int): ImperativeAggregate =
+ copy(mutableAggBufferOffset = newMutableAggBufferOffset)
+
+ override def withNewInputAggBufferOffset(newInputAggBufferOffset: Int):
ImperativeAggregate =
+ copy(inputAggBufferOffset = newInputAggBufferOffset)
+
+ private var counts = new OpenHashMap[Number, Long]
+
+ // Mark as lazy so that percentageExpression is not evaluated during
tree transformation.
+ private lazy val (returnPercentileArray: Boolean, percentages:
Seq[Number]) =
+ evalPercentages(percentageExpression)
+
+ override def children: Seq[Expression] = child :: percentageExpression
:: Nil
+
+ // Returns null for empty inputs
+ override def nullable: Boolean = true
+
+ override def dataType: DataType =
+ if (returnPercentileArray) ArrayType(DoubleType) else DoubleType
+
+ override def inputTypes: Seq[AbstractDataType] =
+ Seq(NumericType, TypeCollection(NumericType, ArrayType))
+
+ override def checkInputDataTypes(): TypeCheckResult =
+ TypeUtils.checkForNumericExpr(child.dataType, "function percentile")
+
+ override def supportsPartial: Boolean = false
+
+ override def aggBufferSchema: StructType =
StructType.fromAttributes(aggBufferAttributes)
+
+ override val aggBufferAttributes: Seq[AttributeReference] = Nil
+
+ override val inputAggBufferAttributes: Seq[AttributeReference] = Nil
+
+ override def initialize(buffer: InternalRow): Unit = {
+ // The counts OpenHashMap will contain values of other groups if we
don't initialize it here.
+ // Since OpenHashMap doesn't support deletions, we have to create a
new instance.
+ counts = new OpenHashMap[Number, Long]
+ }
+
+ private def evalPercentages(expr: Expression): (Boolean, Seq[Number]) = {
+ val (isArrayType, values) = (expr.dataType, expr.eval()) match {
+ case (_, n: Number) => (false, Array(n))
+ case (_, d: Decimal) => (false,
Array(d.toDouble.asInstanceOf[Number]))
+ case (ArrayType(baseType: NumericType, _), arrayData: ArrayData) =>
+ val numericArray = arrayData.toObjectArray(baseType)
+ (true, numericArray.map { x =>
+
baseType.numeric.toDouble(x.asInstanceOf[baseType.InternalType]).asInstanceOf[Number]
+ })
+ case other =>
+ throw new AnalysisException(s"Invalid data type ${other._1} for
parameter percentage")
+ }
+
+ require(values.forall(value => value.doubleValue() >= 0.0 &&
value.doubleValue() <= 1.0),
+ s"Percentage values must be between 0.0 and 1.0, current values =
${values.mkString(", ")}")
+
+ (isArrayType, values)
+ }
+
+ override def update(buffer: InternalRow, input: InternalRow): Unit = {
+ val key = child.eval(input).asInstanceOf[Number]
+
+ // Null values are ignored when computing percentiles.
+ if (key != null) {
+ counts.changeValue(key, 1L, _ + 1L)
+ }
+ }
+
+ override def merge(buffer: InternalRow, inputBuffer: InternalRow): Unit
= {
+ sys.error("Percentile cannot be used in partial aggregations.")
--- End diff --
We might have to reconsider this. If we are reimplementing this as
TypedImperativeAggregate. The reason is that percentile is typically used
without a grouping keys, not having partial aggregation will make a bad
situation even worse by moving all data to a single partition.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
