gistfile1.scala

import org.apache.spark.sql.types.{DoubleType,LongType,ShortType, IntegerType, StructField,TimestampType, StructType,StringType,NumericType,BooleanType}
import  org.apache.hadoop.fs.{FileSystem,Path}

val sqlContext = new org.apache.spark.sql.SQLContext(sc)
import sqlContext.implicits._ 

def csvToDF(file: Path, delimiter : String,charset: String = "UTF8", useHeader: Boolean = true, schema: Option[StructType] = None) = { 

  val df = schema match {
          case Some(schema) => sqlContext.read
            .format("com.databricks.spark.csv")
            .option("header",useHeader.toString()) // Use first line of all files as header
            .option("inferSchema", "true") // Automatical¤ly infer data types
            .option("delimiter",delimiter)
            .option("charset",charset)
            .option("mode","DROPMALFORMED")
            .option("treatEmptyValuesAsNulls","true")
            .option("nullValue","NULL")
            .schema(schema)
            .load(file.toString() )

          case None => sqlContext.read
            .format("com.databricks.spark.csv")
            .option("header",useHeader.toString()) // Use first line of all files as header
            .option("inferSchema", "true") // Automatical¤ly infer data types
            .option("delimiter",delimiter)
            .option("charset",charset)
            .option("mode","DROPMALFORMED")
            .option("treatEmptyValuesAsNulls","true")
            .option("nullValue","NULL")
            .load(file.toString() )
        }
       
 
   df
}


val dfmetoder = csvToDF(new Path("/data/thepath/metoder.csv"),",",useHeader = false,schema=None)

import org.apache.spark.sql.SaveMode
dfmetoder.write.mode(SaveMode.Append).parquet("/data/thepath/p/metoder" )