Spark (Scala) execute dataframe within for loop

问题

I am using spark 1.6.1 version. I have requirement to execute dataframe in loop.

for ( i <- List ('a','b')){
 val i = sqlContext.sql("SELECT i, col1, col2 FROM DF1")}

I want this dataframe to be executed twice (i = a and i = b).

回答1:

Your code is almost correct. Except two things :

• i is already used in your for loop so don't use it in val i =
• If you want to use the value of i in a string, use String Interpolation

So your code should look like :

for (i <- List ('a','b')) {
  val df = sqlContext.sql(s"SELECT $i, col1, col2 FROM DF1")
  df.show()
}

---

EDIT after author comment :

You can do this with a .map and then a .reduceLeft :

// All your dataframes
val dfs = Seq('a','b').map { i =>
  sqlContext.sql(s"SELECT $i, col1, col2 FROM DF1")
}

// Then you can reduce your dataframes into one
val unionDF = dfs.reduceLeft((dfa, dfb) => 
  dfa.unionAll(dfb)
)

来源：https://stackoverflow.com/questions/42900556/spark-scala-execute-dataframe-within-for-loop