[
https://issues.apache.org/jira/browse/SPARK-54841?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
loong updated SPARK-54841:
--------------------------
Description: FilterEstimation.scala use
EstimationUtils#numBinsHoldingRange() to calculate how many bins in the
range.While there is a small logical issue, when calculate lowerPart, if
lowerBin.hi equals to lowerBound and lowerBoundInclusive is false, then
lowerPart should be 0, but now it is 1/lowerBin.ndv . Similarly higherPart.
(was: FilterEstimation.scala use EstimationUtils#numBinsHoldingRange() to
calculate how many bins in the range.But there is a little )
> Make FilterEstimation more accurate when using histogram
> --------------------------------------------------------
>
> Key: SPARK-54841
> URL: https://issues.apache.org/jira/browse/SPARK-54841
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 4.0.1
> Reporter: loong
> Priority: Major
>
> FilterEstimation.scala use EstimationUtils#numBinsHoldingRange() to calculate
> how many bins in the range.While there is a small logical issue, when
> calculate lowerPart, if lowerBin.hi equals to lowerBound and
> lowerBoundInclusive is false, then lowerPart should be 0, but now it is
> 1/lowerBin.ndv . Similarly higherPart.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
