Skip to content

fix: stddev_pop should not directly return 0.0 when count is 1.0 #1184

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

Merged
merged 4 commits into from
Dec 19, 2024
Merged

fix: stddev_pop should not directly return 0.0 when count is 1.0 #1184

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
824e190
add test
viirya Dec 15, 2024
f14dd90
fix
viirya Dec 19, 2024
6dc46fd
fix
viirya Dec 19, 2024
611eca6
fix
viirya Dec 19, 2024
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: 2 additions & 4 deletions native/spark-expr/src/variance.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -245,10 +245,8 @@ impl Accumulator for VarianceAccumulator {

Ok(ScalarValue::Float64(match self.count {
count if count == 0.0 => None,
count if count == 1.0 => {
if let StatsType::Population = self.stats_type {
Some(0.0)
} else if self.null_on_divide_by_zero {
count if count == 1.0 && StatsType::Sample == self.stats_type => {
if self.null_on_divide_by_zero {
Copy link
Member Author

Choose a reason for hiding this comment

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

I checked Spark StddevPop expression. Its evaluation expression is:

override val evaluateExpression: Expression = {
    If(n === 0.0, Literal.create(null, DoubleType), sqrt(m2 / n))
  }

The special handling is only for Spark StddevSamp (i.e., StatsType::Population in Comet).

None
} else {
Some(f64::NAN)
Expand Down
17 changes: 17 additions & 0 deletions spark/src/test/scala/org/apache/comet/exec/CometAggregateSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,6 +38,23 @@ import org.apache.comet.CometSparkSessionExtensions.isSpark34Plus
class CometAggregateSuite extends CometTestBase with AdaptiveSparkPlanHelper {
import testImplicits._

test("stddev_pop should return NaN for some cases") {
withSQLConf(
CometConf.COMET_EXEC_SHUFFLE_ENABLED.key -> "true",
CometConf.COMET_EXPR_STDDEV_ENABLED.key -> "true") {
Seq(true, false).foreach { nullOnDivideByZero =>
withSQLConf("spark.sql.legacy.statisticalAggregate" -> nullOnDivideByZero.toString) {

val data: Seq[(Float, Int)] = Seq((Float.PositiveInfinity, 1))
withParquetTable(data, "tbl", false) {
val df = sql("SELECT stddev_pop(_1), stddev_pop(_2) FROM tbl")
checkSparkAnswer(df)
}
}
}
}
}

test("count with aggregation filter") {
withSQLConf(
CometConf.COMET_ENABLED.key -> "true",
Expand Down
Loading