用Python统计单词的个数写wordcount

'''
Created on 2018年9月22日

@author: Administrator
'''
from pyspark.conf import SparkConf
from pyspark.context import SparkContext
from pyspark.streaming.tests import result
from test.test_importlib.namespace_pkgs.both_portions.foo import one
from com.bjsxt.spark.wc import getCurrentSiteTop3User, getSiteIfo

def shows(s):
    print(s)

def wordcount(line):
    word=line.flatMap(lambda lines:lines.split(" "),True)
    pairword=word.map(lambda words:(words,1),True)
    result=pairword.reduceByKey(lambda v1,v2:(v1+v2),True)
    result.foreach(lambda t:shows(t))
    


if __name__ == '__main__':
    conf=SparkConf().setAppName("wordcount").setMaster("local")
    sc=SparkContext(conf=conf)
    line=sc.textFile("./wprds")
#         统计单词数量
    wordcount(line)
    

版本二：

@author: Administrator
'''
from pyspark.conf import SparkConf
from pyspark.context import SparkContext

def show(s):
    print(s)

def Wordcount(line):
    rdd1=line.flatMap(lambda s:s.split(" ")).map(lambda s:(_,1)).reduceByKey(lambda v1,v2:(v1+v2)).foreach(lambda ss:show(ss))

if __name__ == '__main__':
    conf=SparkConf().setMaster("local").setamAppName("test");
    sc=SparkContext(conf=conf)
    line=sc.textFile("./words")
    
    #统计单词个数
    Wordcount(line)