spark的词频统计之后作升序或者降序排序

scala> b.map((_,1)).reduceByKey(_+_).collect
res26: Array[(String, Int)] = Array((hive,1), (spark,3), (jeff,2), (ruoze,1), (hadoop,1), (hi,1))

 

现有这样的一个已经做好wordcount的RDD，对其进行排序

 

升序：

scala> b.map((_,1)).reduceByKey(_+_).sortBy(_._2).collect
res29: Array[(String, Int)] = Array((hive,1), (ruoze,1), (hadoop,1), (hi,1), (jeff,2), (spark,3))

 

降序：

scala> b.map((_,1)).reduceByKey(_+_).sortBy(_._2,false).collect
res30: Array[(String, Int)] = Array((spark,3), (jeff,2), (hive,1), (ruoze,1), (hadoop,1), (hi,1))