Ulrich zink created SPARK-18163:
-----------------------------------
Summary: Union unexpected behaviour when generating data frames
programatically
Key: SPARK-18163
URL: https://issues.apache.org/jira/browse/SPARK-18163
Project: Spark
Issue Type: Bug
Components: SQL
Affects Versions: 2.0.1
Reporter: Ulrich zink
//expected behaviour
val df1 = Seq((1,2),(3,4)).toDF("a","b")
val df2 = Seq((5,6)).toDF("a","b")
df1.union(df2).show()
+---+---+
| a| b|
+---+---+
| 1| 2|
| 3| 4|
| 5| 6|
+---+---+
// When generating the data frames programmatically
val nInst = 2
val fltr = 1
case class Instrument(id: Long, value: Double)
def dataset (nst:Int,fltrVal:Int) = sqlContext.range(0, nst).select(($"id"),
round(abs(randn)).alias("value")).as[Instrument].filter('value
> fltrVal)
val df3 = dataset(nInst,fltr)
val df4 = dataset(nInst,fltr)
df3.show()
df4.show()
df3.union(df4).show()
df3: org.apache.spark.sql.Dataset[Instrument] = [id: bigint, value: double]
+---+-----+
| id|value|
+---+-----+
| 0| 1.0|
| 1| 1.0|
+---+-----+
df4: org.apache.spark.sql.Dataset[Instrument] = [id: bigint, value: double]
+---+-----+
| id|value|
+---+-----+
| 0| 1.0|
| 1| 0.0|
+---+-----+
+---+-----+
| id|value|
+---+-----+
| 0| 1.0|
| 1| 1.0|
| 0| 1.0|
| 1| 2.0|
+---+-----+
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
