Github user srinathshankar commented on a diff in the pull request:
https://github.com/apache/spark/pull/15637#discussion_r85609917
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/aggregate/HistogramEndpoints.scala
---
@@ -0,0 +1,465 @@
+/*
+ * 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 java.nio.ByteBuffer
+
+import scala.collection.immutable.TreeMap
+import scala.collection.mutable
+
+import com.google.common.primitives.{Doubles, Ints, Longs}
+
+import org.apache.spark.sql.catalyst.InternalRow
+import org.apache.spark.sql.catalyst.analysis.TypeCheckResult
+import
org.apache.spark.sql.catalyst.analysis.TypeCheckResult.{TypeCheckFailure,
TypeCheckSuccess}
+import org.apache.spark.sql.catalyst.expressions.{Expression,
ExpressionDescription, Literal}
+import
org.apache.spark.sql.catalyst.expressions.aggregate.ApproximatePercentile.{PercentileDigest,
PercentileDigestSerializer}
+import org.apache.spark.sql.catalyst.util.{ArrayBasedMapData,
QuantileSummaries}
+import org.apache.spark.sql.catalyst.util.QuantileSummaries._
+import org.apache.spark.sql.types.{DataType, _}
+import org.apache.spark.unsafe.types.UTF8String
+
+/**
+ * The HistogramEndpoints function for a column returns bins - (distinct
value, frequency) pairs
+ * of equi-width histogram when the number of distinct values is less than
or equal to the
+ * specified maximum number of bins. Otherwise, for column of string type,
it returns an empty
+ * map; for column of numeric type, it returns endpoints of equi-height
histogram - approximate
+ * percentiles at percentages 0.0, 1/numBins, 2/numBins, ...,
(numBins-1)/numBins, 1.0.
+ *
+ * @param child child expression that can produce column value with
`child.eval(inputRow)`
+ * @param numBinsExpression The maximum number of bins.
+ * @param accuracyExpression Accuracy used in computing approximate
percentiles.
+ */
+@ExpressionDescription(
+ usage =
+ """
+ _FUNC_(col, numBins [, accuracy]) - Returns bins - (distinct value,
frequency) pairs
+ of equi-width histogram when the number of distinct values is less
than or equal to the
+ specified maximum number of bins. Otherwise, for column of string
type, it returns an empty
+ map; for column of numeric type, it returns endpoints of equi-height
histogram - approximate
+ percentiles at percentages 0.0, 1/numBins, 2/numBins, ...,
(numBins-1)/numBins, 1.0. The
+ `accuracy` parameter (default: 10000) is a positive integer literal
which controls percentiles
+ approximation accuracy at the cost of memory. Higher value of
`accuracy` yields better
+ accuracy, `1.0/accuracy` is the relative error of the approximation.
+ """)
+case class HistogramEndpoints(
+ child: Expression,
+ numBinsExpression: Expression,
+ accuracyExpression: Expression,
+ override val mutableAggBufferOffset: Int,
+ override val inputAggBufferOffset: Int) extends
TypedImperativeAggregate[EndpointsDigest] {
+
+ def this(child: Expression, numBinsExpression: Expression,
accuracyExpression: Expression) = {
+ this(child, numBinsExpression, accuracyExpression, 0, 0)
+ }
+
+ def this(child: Expression, numBinsExpression: Expression) = {
+ this(child, numBinsExpression,
Literal(ApproximatePercentile.DEFAULT_PERCENTILE_ACCURACY))
+ }
+
+ // Mark as lazy so that numBinsExpression is not evaluated during tree
transformation.
+ private lazy val numBins: Int =
numBinsExpression.eval().asInstanceOf[Int]
+
+ private lazy val percentages: Array[Double] = {
+ val array = new Array[Double](numBins + 1)
+ for (i <- 0 to numBins) {
+ array(i) = i / numBins.toDouble
+ }
+ array
+ }
+
+ private lazy val accuracy: Int =
accuracyExpression.eval().asInstanceOf[Int]
+
+ override def inputTypes: Seq[AbstractDataType] = {
+ Seq(TypeCollection(NumericType, TimestampType, DateType, StringType),
IntegerType, IntegerType)
+ }
+
+ override def checkInputDataTypes(): TypeCheckResult = {
+ val defaultCheck = super.checkInputDataTypes()
+ if (defaultCheck.isFailure) {
+ defaultCheck
+ } else if (!numBinsExpression.foldable ||
!accuracyExpression.foldable) {
+ TypeCheckFailure("The maximum number of bins or accuracy provided
must be a constant literal")
+ } else if (numBins < 2) {
+ TypeCheckFailure(
+ "The maximum number of bins provided must be a positive integer
literal >= 2 " +
+ s"(current value = $numBins)")
+ } else if (accuracy <= 0) {
+ TypeCheckFailure(
+ s"The accuracy provided must be a positive integer literal
(current value = $accuracy)")
+ } else {
+ TypeCheckSuccess
+ }
+ }
+
+ override def update(buffer: EndpointsDigest, input: InternalRow): Unit =
{
+ if (buffer.invalid) {
+ return
+ }
+ val evaluated = child.eval(input)
+ if (evaluated != null) {
+ buffer.update(child.dataType, evaluated, numBins)
+ }
+ }
+
+ override def merge(buffer: EndpointsDigest, other: EndpointsDigest):
Unit = {
+ if (buffer.invalid) return
+ if (other.invalid) {
+ buffer.invalid = true
+ buffer.clear()
+ return
+ }
+ buffer.merge(other, numBins)
+ }
+
+ override def eval(buffer: EndpointsDigest): Any = {
+ if (buffer.invalid) {
+ // return empty map
+ ArrayBasedMapData(Map.empty)
+ } else {
+ buffer match {
+ case stringDigest: StringEndpointsDigest =>
+ // sort the result to make it more readable
+ val sorted = TreeMap[UTF8String, Long](stringDigest.bins.toSeq:
_*)
+ ArrayBasedMapData(sorted.keys.toArray, sorted.values.toArray)
+ case numericDigest: NumericEndpointsDigest =>
+ if (!numericDigest.mapInvalid) {
+ val sorted = TreeMap[Double, Long](numericDigest.bins.toSeq:
_*)
+ ArrayBasedMapData(sorted.keys.toArray, sorted.values.toArray)
+ } else {
+ val percentiles =
numericDigest.percentileDigest.getPercentiles(percentages)
+ // we only need percentiles, this is for constructing MapData
+ val padding = new Array[Long](percentiles.length)
+ ArrayBasedMapData(percentiles, padding)
+ }
+ }
+ }
+ }
+
+ override def serialize(buffer: EndpointsDigest): Array[Byte] = {
+ buffer match {
+ case stringDigest: StringEndpointsDigest =>
StringEndpointsDigest.serialize(stringDigest)
+ case numericDigest: NumericEndpointsDigest =>
NumericEndpointsDigest.serialize(numericDigest)
+ }
+ }
+
+ override def deserialize(bytes: Array[Byte]): EndpointsDigest = {
+ child.dataType match {
+ case StringType => StringEndpointsDigest.deserialize(bytes)
+ case _ => NumericEndpointsDigest.deserialize(bytes)
+ }
+ }
+
+ override def createAggregationBuffer(): EndpointsDigest = {
+ child.dataType match {
+ case StringType =>
+ StringEndpointsDigest()
+ case _ =>
+ NumericEndpointsDigest(new PercentileDigest(1.0D / accuracy))
+ }
+ }
+
+ override def withNewMutableAggBufferOffset(newMutableAggBufferOffset:
Int): HistogramEndpoints = {
+ copy(mutableAggBufferOffset = newMutableAggBufferOffset)
+ }
+
+ override def withNewInputAggBufferOffset(newInputAggBufferOffset: Int):
HistogramEndpoints = {
+ copy(inputAggBufferOffset = newInputAggBufferOffset)
+ }
+
+ override def nullable: Boolean = false
+
+ override def dataType: DataType = {
+ child.dataType match {
+ case StringType => MapType(StringType, LongType)
+ case _ => MapType(DoubleType, LongType)
+ }
+ }
+
+ override def children: Seq[Expression] = Seq(child, numBinsExpression,
accuracyExpression)
+
+ override def prettyName: String = "histogram_endpoints"
+}
+
+trait EndpointsDigest {
+
+ // Mark this EndpointsDigest as invalid when:
+ // 1. for string type - the size of the hashmap (ndv of the column)
exceeds numBins;
+ // 2. for numeric type - Some Decimal value out of the range of Double
occurs.
+ var invalid: Boolean = false
+
+ def update(dataType: DataType, value: Any, numBins: Int): Unit
+ def merge(otherDigest: EndpointsDigest, numBins: Int): Unit
+ def clear(): Unit
+
+ // Updates baseMap, and returns success or not.
+ def updateMap[T](baseMap: mutable.HashMap[T, Long], value: T, numBins:
Int): Boolean = {
+ if (baseMap.contains(value)) {
--- End diff --
Isn't this just
`mergeMap(baseMap, (value -> 1), numBins)`
---
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]
