spark UnionEstimation 源码

spark UnionEstimation 代码

文件路径：/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/statsEstimation/UnionEstimation.scala

/*
 * 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.plans.logical.statsEstimation

import org.apache.spark.sql.catalyst.expressions.{Attribute, AttributeMap}
import org.apache.spark.sql.catalyst.plans.logical.{ColumnStat, Statistics, Union}
import org.apache.spark.sql.types._

/**
 * Estimate the number of output rows by doing the sum of output rows for each child of union,
 * and estimate min and max stats for each column by finding the overall min and max of that
 * column coming from its children.
 */
object UnionEstimation {
  import EstimationUtils._

  private def createStatComparator(dt: DataType): (Any, Any) => Boolean = dt match {
    case ByteType => (a: Any, b: Any) =>
      ByteType.ordering.lt(a.asInstanceOf[Byte], b.asInstanceOf[Byte])
    case ShortType => (a: Any, b: Any) =>
      ShortType.ordering.lt(a.asInstanceOf[Short], b.asInstanceOf[Short])
    case IntegerType => (a: Any, b: Any) =>
      IntegerType.ordering.lt(a.asInstanceOf[Int], b.asInstanceOf[Int])
    case LongType => (a: Any, b: Any) =>
      LongType.ordering.lt(a.asInstanceOf[Long], b.asInstanceOf[Long])
    case FloatType => (a: Any, b: Any) =>
      FloatType.ordering.lt(a.asInstanceOf[Float], b.asInstanceOf[Float])
    case DoubleType => (a: Any, b: Any) =>
      DoubleType.ordering.lt(a.asInstanceOf[Double], b.asInstanceOf[Double])
    case _: DecimalType => (a: Any, b: Any) =>
      dt.asInstanceOf[DecimalType].ordering.lt(a.asInstanceOf[Decimal], b.asInstanceOf[Decimal])
    case DateType => (a: Any, b: Any) =>
      DateType.ordering.lt(a.asInstanceOf[DateType.InternalType],
        b.asInstanceOf[DateType.InternalType])
    case TimestampType => (a: Any, b: Any) =>
      TimestampType.ordering.lt(a.asInstanceOf[TimestampType.InternalType],
        b.asInstanceOf[TimestampType.InternalType])
    case TimestampNTZType => (a: Any, b: Any) =>
      TimestampNTZType.ordering.lt(a.asInstanceOf[TimestampNTZType.InternalType],
        b.asInstanceOf[TimestampNTZType.InternalType])
    case i: AnsiIntervalType => (a: Any, b: Any) =>
      i.ordering.lt(a.asInstanceOf[i.InternalType], b.asInstanceOf[i.InternalType])
    case _ =>
      throw new IllegalStateException(s"Unsupported data type: ${dt.catalogString}")
  }

  private def isTypeSupported(dt: DataType): Boolean = dt match {
    case ByteType | IntegerType | ShortType | FloatType | LongType |
         DoubleType | DateType | _: DecimalType | TimestampType | TimestampNTZType |
         _: AnsiIntervalType => true
    case _ => false
  }

  def estimate(union: Union): Option[Statistics] = {
    val sizeInBytes = union.children.map(_.stats.sizeInBytes).sum
    val outputRows = if (rowCountsExist(union.children: _*)) {
      Some(union.children.map(_.stats.rowCount.get).sum)
    } else {
      None
    }

    val newMinMaxStats = computeMinMaxStats(union)
    val newNullCountStats = computeNullCountStats(union)
    val newAttrStats = {
      val baseStats = AttributeMap(newMinMaxStats)
      val overwriteStats = newNullCountStats.map { case attrStat@(attr, stat) =>
        baseStats.get(attr).map { baseStat =>
          attr -> baseStat.copy(nullCount = stat.nullCount)
        }.getOrElse(attrStat)
      }
      AttributeMap(newMinMaxStats ++ overwriteStats)
    }

    Some(
      Statistics(
        sizeInBytes = sizeInBytes,
        rowCount = outputRows,
        attributeStats = newAttrStats))
  }

  private def computeMinMaxStats(union: Union): Seq[(Attribute, ColumnStat)] = {
    val unionOutput = union.output
    val attrToComputeMinMaxStats = union.children.map(_.output).transpose.zipWithIndex.filter {
      case (attrs, outputIndex) => isTypeSupported(unionOutput(outputIndex).dataType) &&
        // checks if all the children has min/max stats for an attribute
        attrs.zipWithIndex.forall {
          case (attr, childIndex) =>
            val attrStats = union.children(childIndex).stats.attributeStats
            attrStats.get(attr).isDefined && attrStats(attr).hasMinMaxStats
        }
    }
    attrToComputeMinMaxStats.map {
      case (attrs, outputIndex) =>
        val dataType = unionOutput(outputIndex).dataType
        val statComparator = createStatComparator(dataType)
        val minMaxValue = attrs.zipWithIndex.foldLeft[(Option[Any], Option[Any])]((None, None)) {
          case ((minVal, maxVal), (attr, childIndex)) =>
            val colStat = union.children(childIndex).stats.attributeStats(attr)
            val min = if (minVal.isEmpty || statComparator(colStat.min.get, minVal.get)) {
              colStat.min
            } else {
              minVal
            }
            val max = if (maxVal.isEmpty || statComparator(maxVal.get, colStat.max.get)) {
              colStat.max
            } else {
              maxVal
            }
            (min, max)
        }
        val newStat = ColumnStat(min = minMaxValue._1, max = minMaxValue._2)
        unionOutput(outputIndex) -> newStat
    }
  }

  private def computeNullCountStats(union: Union): Seq[(Attribute, ColumnStat)] = {
    val unionOutput = union.output
    val attrToComputeNullCount = union.children.map(_.output).transpose.zipWithIndex.filter {
      case (attrs, _) => attrs.zipWithIndex.forall {
        case (attr, childIndex) =>
          val attrStats = union.children(childIndex).stats.attributeStats
          attrStats.get(attr).isDefined && attrStats(attr).nullCount.isDefined
      }
    }
    attrToComputeNullCount.map {
      case (attrs, outputIndex) =>
        val firstStat = union.children.head.stats.attributeStats(attrs.head)
        val firstNullCount = firstStat.nullCount.get
        val colWithNullStatValues = attrs.zipWithIndex.tail.foldLeft[BigInt](firstNullCount) {
          case (totalNullCount, (attr, childIndex)) =>
            val colStat = union.children(childIndex).stats.attributeStats(attr)
            totalNullCount + colStat.nullCount.get
        }
        val newStat = ColumnStat(nullCount = Some(colWithNullStatValues))
        unionOutput(outputIndex) -> newStat
    }
  }
}

相关信息

spark 源码目录

相关文章

spark AggregateEstimation 源码

spark BasicStatsPlanVisitor 源码

spark EstimationUtils 源码

spark FilterEstimation 源码

spark JoinEstimation 源码

spark LogicalPlanStats 源码

spark ProjectEstimation 源码

spark SizeInBytesOnlyStatsPlanVisitor 源码

spark ValueInterval 源码