scala> df.select(colString, colShort, colShort.equalTo(colString)).explain == Physical Plan == LocalTableScan [tier_id#3, tier_id#56, (CAST(tier_id AS SMALLINT) = tier_id)#50]
On Wed, Feb 6, 2019 at 6:19 PM Russell Spitzer <russell.spit...@gmail.com> wrote: > Run an "explain" instead of show, i'm betting it's casting tier_id to a > small_int to do the comparison > > On Wed, Feb 6, 2019 at 9:31 AM Artur Sukhenko <artur.sukhe...@gmail.com> > wrote: > >> Hello guys, >> I am migrating from Spark 1.6 to 2.2 and have this issue: >> I am casting string to short and comparing them with equal . >> Original code is: >> ... when(col(fieldName).equalTo(castedValueCol), castedValueCol). >> >> otherwise(defaultErrorValueCol) >> >> Reproduce (version 2.3.0.cloudera4): >> scala> val df = Seq("3.15").toDF("tier_id") >> df: org.apache.spark.sql.DataFrame = [tier_id: string] >> >> scala> val colShort = col("tier_id").cast(ShortType) >> colShort: org.apache.spark.sql.Column = CAST(tier_id AS SMALLINT) >> >> scala> val colString = col("tier_id") >> colString: org.apache.spark.sql.Column = tier_id >> >> scala> res4.select(colString, colShort, colShort.equalTo(colString)).show >> +-------+-------+-------------------------------------+ >> |tier_id|tier_id|(CAST(tier_id AS SMALLINT) = tier_id)| >> +-------+-------+-------------------------------------+ >> | 3.15| 3| true| >> +-------+-------+-------------------------------------+ >> scala> >> >> Why is this? >> -- >> -- >> Artur Sukhenko >> > -- -- Artur Sukhenko