[SPARK-33959] Improve the statistics estimation of the Tail

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 3.2.0
• Fix Version/s: 3.2.0
• Component/s: SQL
• Assignee: Yuming Wang
• Reporter: Yuming Wang
• Created: 02/Jan/21 09:29
• Updated: 12/Dec/22 18:10
• Resolved: 03/Jan/21 01:59

Description

spark.sql("set spark.sql.cbo.enabled=true")
spark.range(100).selectExpr("id as a", "id as b", "id as c", "id as e").write.saveAsTable("t1")
println(Tail(Literal(5), spark.sql("SELECT * FROM t1").queryExecution.logical).queryExecution.explainString(org.apache.spark.sql.execution.CostMode))

Current:

== Optimized Logical Plan ==
Tail 5, Statistics(sizeInBytes=3.8 KiB)
+- Relation[a#24L,b#25L,c#26L,e#27L] parquet, Statistics(sizeInBytes=3.8 KiB)

Expected:

== Optimized Logical Plan ==
Tail 5, Statistics(sizeInBytes=200.0 B, rowCount=5)
+- Relation[a#24L,b#25L,c#26L,e#27L] parquet, Statistics(sizeInBytes=3.8 KiB)

Issue Links

• links to: [Github] Pull Request #30991 (wangyum)
• links to: [Github] Pull Request #30991 (wangyum)