Load and save a text file as follows:

      val input = 
      sc.textFile("hdfs://namenodeHostName:8020/repos/spark/README.md") 
      val wholeInput = 
      sc.wholeTextFiles("file://home/padma/salesFiles") 
      val result = wholeInput.mapValues{value => val nums = value.split
      (" ").map(x => x.toDouble) 
      nums.sum/nums.size.toDouble} 
      result.saveAsTextFile("/home/Padma/outputFile.txt") 

For loading a JSON file, the people.json input file is taken from the SPARK_HOME folder whose location is /spark-1.6.0/examples/src/main/resource/people.json. Now, loading and saving a JSON file looks like this:

        // Loading JSON file 
        import com.fasterxml.jackson.module.scala.DefaultScalaModule 
        import com.fasterxml.jackson.module.scala.
        experimental.ScalaObjectMapper
        import com.fasterxml.jackson.databind.ObjectMapper 
        import com.fasterxml.jackson.module.databind.
            DeserializatiuonFeature 
        ... 
        case class Person(name:String, age:Int) 
        ... 
        val jsonInput =
        sc.textFile(""hdfs://namenode:9000/data/people.json") 
        val result = jsonInput.flatMap(record => {
        try{Some(mapper.readValue(record, classOf[Person])) 
        } 
        catch{ 
        case e:Exception => None 
        }} ) 
        result.filter(person =>
        person.age>15).map(mapper.writeValueAsString(_)).
        saveAsTextFile(output File) 

To load and save a CSV file, let's take the stocks data:

       IBM,20160113,133.5,134.279999,131.100006,131.169998,4672300 
       GOOG,20160113,730.849976,734.73999,698.609985,700.559998,2468300 
       MSFT,20160113,53.799999,54.07,51.299999,51.639999,66119000 
       MSFT,20160112,52.759998,53.099998,52.060001,52.779999,35650700 
       YHOO,20160113,30.889999,31.17,29.33,29.440001,16593700 
       . 
       . 
       import java.io.StringReader 
       import au.com.bytecode.opencsv.CSVReader 
       ... 
       case class Stocks(name:String, totalPrice:Long) 
       ... 
       val input = sc.textFile("hdfs://namenodeHostName:8020 
       /data/stocks.txt") 
       val result = input.map{line => val reader = new CSVReader(new 
       StringReader(line)) 
       reader.readAll().map(x => Stocks(x(0), x(6))) 
       } 
       result.map(stock => Array(stock.name, stock.
       totalPrice)).mapPartitions {stock =>  
       val stringWriter = new StringWriter 
       val csvWriter = new CSVWriter(stringWriter) 
       csvWriter.writeAll(people.toList) 
       Iterator(stringWriter.toString) 
       }.saveAsTextFilehdfs://namenode:9000/CSVOutputFile") 

Now, let's see the way sequenceFile is loaded and saved:

       val data = sc.sequenceFile(inputFile, classOf[Text],  
       classOf[IntWritable]).map{case(x,y) => (x.toString, y.get())} 
       val input = sc.parallelize(List(("Panda",3),("Kay",6),
       ("Snail",2))) 
       input.saveAsSequenceFilehdfs://namenode:9000/
       sequenceOutputFile")