[jira] [Created] (SPARK-22283) withColumn should replaces multiple instances with a single one

["Albert Meltzer (JIRA)" <ji...@apache.org>]

Albert Meltzer created SPARK-22283:
--------------------------------------

             Summary: withColumn should replaces multiple instances with a single one
                 Key: SPARK-22283
                 URL: https://issues.apache.org/jira/browse/SPARK-22283
             Project: Spark
          Issue Type: Bug
          Components: Spark Core
    Affects Versions: 2.2.0
            Reporter: Albert Meltzer


Currently, {{withColumn}} claims to do the following: _"adding a column or replacing the existing column that has the same name."_

Unfortunately, if multiple existing columns have the same name (which is a normal occurrence after a join), this results in multiple replaced -- and retained columns (with the same value) --and messages about an ambiguous column.

The current implementation of {{withColumn}} contains this:

{noformat}
  def withColumn(colName: String, col: Column): DataFrame = {
    val resolver = sparkSession.sessionState.analyzer.resolver
    val output = queryExecution.analyzed.output
    val shouldReplace = output.exists(f => resolver(f.name, colName))
    if (shouldReplace) {
      val columns = output.map { field =>
        if (resolver(field.name, colName)) {
          col.as(colName)
        } else {
          Column(field)
        }
      }
      select(columns : _*)
    } else {
      select(Column("*"), col.as(colName))
    }
  }
{noformat}

Instead, suggest something like this (which replaces all matching fields with a single instance of the new one):

{noformat}
  def withColumn(colName: String, col: Column): DataFrame = {
    val resolver = sparkSession.sessionState.analyzer.resolver
    val output = queryExecution.analyzed.output
    val existing = output.filterNot(f => resolver(f.name, colName)).map(new Column(_))
    select(existing :+ col.as(colName): _*)
  }
{noformat}



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org