Last active
August 20, 2018 22:00
-
-
Save HeartSaVioR/6606bceceaceaff439db0853fa591501 to your computer and use it in GitHub Desktop.
A bit tricky result of Spark SQL query result (Tested with 2.3.0)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
///////////////////////////////////////////////////////////////////////////////////////////// | |
// 1. select with swapping columns, and apply where | |
///////////////////////////////////////////////////////////////////////////////////////////// | |
import spark.implicits._ | |
import org.apache.spark.sql.{DataFrame, Dataset} | |
case class Hello(id: Int, name1: String, name2: String) | |
val ds = List(Hello(1, "Alice", "Bob"), Hello(2, "Bob", "Alice")).toDS | |
val swapRenamedDF: DataFrame = ds.select($"id", $"name1".as("name2"), $"name2".as("name1")) | |
swapRenamedDF.explain | |
swapRenamedDF.where($"name1" === "Alice").show | |
swapRenamedDF.where($"name1" === "Alice").explain | |
swapRenamedDF.where($"name2" === "Alice").show | |
swapRenamedDF.where($"name2" === "Alice").explain | |
/* | |
> swapRenamedDF.explain | |
== Physical Plan == | |
LocalTableScan [id#3, name2#7, name1#8] | |
> swapRenamedDF.where($"name1" === "Alice").show | |
+---+-----+-----+ | |
| id|name2|name1| | |
+---+-----+-----+ | |
| 2| Bob|Alice| | |
+---+-----+-----+ | |
> swapRenamedDF.where($"name1" === "Alice").explain | |
== Physical Plan == | |
*(1) Project [id#3, name1#4 AS name2#7, name2#5 AS name1#8] | |
+- *(1) Filter (isnotnull(name2#5) && (name2#5 = Alice)) | |
+- LocalTableScan [id#3, name1#4, name2#5] | |
> swapRenamedDF.where($"name2" === "Alice").show | |
+---+-----+-----+ | |
| id|name2|name1| | |
+---+-----+-----+ | |
| 1|Alice| Bob| | |
+---+-----+-----+ | |
> swapRenamedDF.where($"name2" === "Alice").explain | |
== Physical Plan == | |
*(1) Project [id#3, name1#4 AS name2#7, name2#5 AS name1#8] | |
+- *(1) Filter (isnotnull(name1#4) && (name1#4 = Alice)) | |
+- LocalTableScan [id#3, name1#4, name2#5] | |
*/ | |
///////////////////////////////////////////////////////////////////////////////////////////// | |
// 2. apply where, and select with swapping columns | |
///////////////////////////////////////////////////////////////////////////////////////////// | |
import spark.implicits._ | |
import org.apache.spark.sql.{DataFrame, Dataset} | |
case class Hello(id: Int, name1: String, name2: String) | |
val ds = List(Hello(1, "Alice", "Bob"), Hello(2, "Bob", "Alice")).toDS | |
val applyWhereToName1ThenSelectDF = ds.where($"name1" === "Alice").select($"id", $"name1".as("name2"), $"name2".as("name1")) | |
applyWhereToName1ThenSelectDF.show | |
applyWhereToName1ThenSelectDF.explain | |
val applyWhereToName2ThenSelectDF = ds.where($"name2" === "Alice").select($"id", $"name1".as("name2"), $"name2".as("name1")) | |
applyWhereToName2ThenSelectDF.show | |
applyWhereToName2ThenSelectDF.explain | |
/* | |
> applyWhereToName1ThenSelectDF.show | |
+---+-----+-----+ | |
| id|name2|name1| | |
+---+-----+-----+ | |
| 1|Alice| Bob| | |
+---+-----+-----+ | |
> applyWhereToName1ThenSelectDF.explain | |
== Physical Plan == | |
*(1) Project [id#3, name1#4 AS name2#8, name2#5 AS name1#9] | |
+- *(1) Filter (isnotnull(name1#4) && (name1#4 = Alice)) | |
+- LocalTableScan [id#3, name1#4, name2#5] | |
> applyWhereToName2ThenSelectDF.show | |
+---+-----+-----+ | |
| id|name2|name1| | |
+---+-----+-----+ | |
| 2| Bob|Alice| | |
+---+-----+-----+ | |
> applyWhereToName2ThenSelectDF.explain | |
== Physical Plan == | |
*(1) Project [id#3, name1#4 AS name2#26, name2#5 AS name1#27] | |
+- *(1) Filter (isnotnull(name2#5) && (name2#5 = Alice)) | |
+- LocalTableScan [id#3, name1#4, name2#5] | |
*/ |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment