pyspark --Action算子

pyspark --Action算子

countByKey


reduce


fold

first


take 取前n个(位置上)


top

7.takeSample

7. takeOrdered

    rdd = sc.parallelize([1,3,2,4,7,9,6],1)
    print(rdd.takeOrdered(3))  # 1,2,3
    print(rdd.takeOrdered(3,lambda x:-x)) # 9,7,6

8.foreach
8. saveAsTextFile

9. foreachPartition

    rdd = sc.parallelize([1,3,2,4,7,9,6],3)
    def rid10(data):
        print("-------------------")
        result = list()
        for i in data:
            result.append(i*10)
        print(result)
    
    rdd.foreachPartition(rid10)

groupByKey与reduceByKey

总结：

partitionBy