pyspark的transformation和action练习002

import findspark
findspark.init()
import pyspark 
sc = pyspark.SparkContext(appName='rdd_excercise_2')

from decimal import Decimal
acTransList =sc.parallelize([("SB10001",1000),("SB10002",1200),("SB10003",8000),("SB10004",4000),("SB10004",300),("SB10003",10000),("SB10002",500),("SB10002",56),("SB10001",1000),("SB10003",1000),("SB10004",-1000)])
acTransList.collect()
# 计算每个账号的交易汇总，reduceByKey
# 根据汇总的Key排序 sortByKey
# 查看api文档里的groupByKey，了解与reduceByKey的区别

[('SB10001', 1000),
 ('SB10002', 1200),
 ('SB10003', 8000),
 ('SB10004', 4000),
 ('SB10004', 300),
 ('SB10003', 10000),
 ('SB10002', 500),
 ('SB10002', 56),
 ('SB10001', 1000),
 ('SB10003', 1000),
 ('SB10004', -1000)]

# 计算每个账号的交易汇总，reduceByKey
acTransList.reduceByKey(lambda x,y : (x+y)).collect()

[('SB10003', 19000), ('SB10001', 2000), ('SB10004', 3300), ('SB10002', 1756)]

# 根据汇总的Key排序 sortByKey
acTransList_2 = acTransList.reduceByKey(lambda x,y : (x+y))
acTransList_2.sortByKey(ascending = False).collect()

[('SB10004', 3300), ('SB10003', 19000), ('SB10002', 1756), ('SB10001', 2000)]

#查看api文档里的groupByKey，了解与reduceByKey的区别
acTransList_2.groupByKey().collect()

[('SB10003', <pyspark.resultiterable.ResultIterable at 0x28094351a58>),
 ('SB10001', <pyspark.resultiterable.ResultIterable at 0x28094351940>),
 ('SB10004', <pyspark.resultiterable.ResultIterable at 0x280943573c8>),
 ('SB10002', <pyspark.resultiterable.ResultIterable at 0x28094357438>)]

sorted(acTransList_2.groupByKey().collect())

[('SB10001', <pyspark.resultiterable.ResultIterable at 0x28094357be0>),
 ('SB10002', <pyspark.resultiterable.ResultIterable at 0x2809436c860>),
 ('SB10003', <pyspark.resultiterable.ResultIterable at 0x280943579e8>),
 ('SB10004', <pyspark.resultiterable.ResultIterable at 0x2809436ca90>)]