Skip to content

[SPARK-15764][SQL] Replace N^2 loop in BindReferences #13505

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
JoshRosen wants to merge 14 commits into from
Closed

[SPARK-15764][SQL] Replace N^2 loop in BindReferences #13505

Show file tree
Hide file tree
Changes from 4 commits
Commits
Show all changes
14 commits
Select commit Hold shift + click to select a range
6216e94
Replace N^2 loop in BindReferences.
JoshRosen Jun 4, 2016
b1a7646
Fix Java 7 compilation issue.
JoshRosen Jun 4, 2016
38e8a99
Whoops, getOrDefault is also new in Java 8.
JoshRosen Jun 4, 2016
0b412b0
Amortize map construction.
JoshRosen Jun 4, 2016
bc17587
Merge remote-tracking branch 'origin/master' into bind-references-imp…
JoshRosen Jun 4, 2016
e7c4150
Fix similar loop in InMemoryTableScanExec.
JoshRosen Jun 4, 2016
210dbd3
Rename getOrdinal to getOrdinalWithExprId
JoshRosen Jun 4, 2016
dd94e29
Delete some related yet dead code in AttributeMap.
JoshRosen Jun 4, 2016
b933fe0
Code style and readability improvements.
JoshRosen Jun 4, 2016
4efd3ee
Comments.
JoshRosen Jun 4, 2016
5504b6c
Address Eric's comments.
JoshRosen Jun 5, 2016
bdb68ad
Merge remote-tracking branch 'origin/master' into bind-references-imp…
JoshRosen Jun 6, 2016
99197b7
Add missing `@transient` annotations.
JoshRosen Jun 6, 2016
5e9c258
Also extend Serializable.
JoshRosen Jun 6, 2016
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/BoundAttribute.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -82,16 +82,16 @@ object BindReferences extends Logging {

def bindReference[A <: Expression](
expression: A,
input: Seq[Attribute],
input: AttributeSeq,
allowFailures: Boolean = false): A = {
expression.transform { case a: AttributeReference =>
attachTree(a, "Binding attribute") {
val ordinal = input.indexWhere(_.exprId == a.exprId)
val ordinal = input.getOrdinal(a.exprId)
if (ordinal == -1) {
if (allowFailures) {
a
} else {
sys.error(s"Couldn't find $a in ${input.mkString("[", ",", "]")}")
sys.error(s"Couldn't find $a in ${input.attrs.mkString("[", ",", "]")}")
}
} else {
BoundReference(ordinal, a.dataType, input(ordinal).nullable)
Expand Down
22 changes: 21 additions & 1 deletion sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/package.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -86,11 +86,31 @@ package object expressions {
/**
* Helper functions for working with `Seq[Attribute]`.
*/
implicit class AttributeSeq(attrs: Seq[Attribute]) {
implicit class AttributeSeq(val attrs: Seq[Attribute]) {
/** Creates a StructType with a schema matching this `Seq[Attribute]`. */
def toStructType: StructType = {
StructType(attrs.map(a => StructField(a.name, a.dataType, a.nullable)))
}

private lazy val inputArr = attrs.toArray

private lazy val inputToOrdinal = {
val map = new java.util.HashMap[ExprId, Int](inputArr.length * 2)
Copy link
Member

@kiszk kiszk Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why *2 is necessary?
I think that the size of map's entry is up to attrs.size since the max number of calling map.put() is equal to `attrs.size. Isattrs.size``equal to``inputArr.legnth``?

Copy link
Contributor Author

@JoshRosen JoshRosen Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The goal was to avoid having to rehash the elements of the hash map once the number of inserted keys exceeded the default 0.75 load factor.

Copy link
Contributor Author

@JoshRosen JoshRosen Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's probably clearer to use Guava's newHashMapWithExpectedSize instead: https://google.github.io/guava/releases/snapshot/api/docs/com/google/common/collect/Maps.html#newHashMapWithExpectedSize(int)

Copy link
Contributor

@ericl ericl Jun 5, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

+1 on withExpectedSize

var index = 0
attrs.foreach { attr =>
if (!map.containsKey(attr.exprId)) {
map.put(attr.exprId, index)
}
index += 1
Copy link
Member

@kiszk kiszk Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Which style is better, this style or a style to use zipWithIndex?

Copy link
Contributor Author

@JoshRosen JoshRosen Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This was just a minor perf. optimization since this method was in a hot path.

Copy link
Contributor Author

@JoshRosen JoshRosen Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

On the other hand, accessing a variable that's outside of the inner closure is likely to be expensive, too. It's probably fastest to just iterate over the elements of inputArr.

Copy link
Member

@kiszk kiszk Jun 5, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Got it. I did not have no preference. Good to hear a reason for this decision.

}
map
}

def apply(ordinal: Int): Attribute = inputArr(ordinal)

def getOrdinal(exprId: ExprId): Int = {
Copy link
Contributor Author

@JoshRosen JoshRosen Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I suppose this needs documentation.

Copy link
Contributor

@rxin rxin Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yup ...

Option(inputToOrdinal.get(exprId)).getOrElse(-1)
}
}

/**
Expand Down
2 changes: 1 addition & 1 deletion sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/QueryPlan.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -296,7 +296,7 @@ abstract class QueryPlan[PlanType <: QueryPlan[PlanType]] extends TreeNode[PlanT
/**
* All the attributes that are used for this plan.
*/
lazy val allAttributes: Seq[Attribute] = children.flatMap(_.output)
lazy val allAttributes: AttributeSeq = children.flatMap(_.output)
Copy link
Contributor Author

@JoshRosen JoshRosen Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ericl and I found another layer of polynomial looping: in QueryPlan.cleanArgs we take every expression in the query plan and bind its references against allAttributes, which can be huge. If we turn this into an AttributeSeq once and build the map inside of that wrapper then we amortize that cost and remove this expensive loop.

Copy link
Contributor Author

@JoshRosen JoshRosen Jun 4, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should probably construct the AttributeSeq outside of the loop in the various projection operators, too, although that doesn't appear to be as serious a bottleneck yet.


private def cleanExpression(e: Expression): Expression = e match {
case a: Alias =>
Expand Down
2 changes: 1 addition & 1 deletion sql/core/src/main/scala/org/apache/spark/sql/execution/aggregate/HashAggregateExec.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,7 +49,7 @@ case class HashAggregateExec(

require(HashAggregateExec.supportsAggregate(aggregateBufferAttributes))

override lazy val allAttributes: Seq[Attribute] =
override lazy val allAttributes: AttributeSeq =
child.output ++ aggregateBufferAttributes ++ aggregateAttributes ++
aggregateExpressions.flatMap(_.aggregateFunction.inputAggBufferAttributes)

Expand Down