Our aim is to have pair RDD then by applying reduceByKey() method to aggregate data separately for each key. Why we need to call special version of Spark’s function
reduceByKey()
