MapReduce Streaming Python AWS 30分钟 (5)matrix multiplication

来源:互联网 发布:手机淘宝1元抢购在哪 编辑:程序博客网 时间:2024/06/06 09:18

mapper.py

#!/usr/bin/env python
import sys

def convert_line_to_iter(file_in):
        for line in file_in:
                yield line.strip().split(' ')

def emit_key_value():
        for k, m, n, v in convert_line_to_iter(sys.stdin):
                for i in range(5):
                        if k == 'a':
                                print '%s,%s %s,%s,%s' % (m,i,k,n,v)
                        else:
                                print '%s,%s %s,%s,%s' % (i,n,k,m,v)

emit_key_value()


reducer.py

#!/usr/bin/env python
import sys
from itertools import groupby
from operator import itemgetter


def convert_mapper_o_to_iter(file_in):
        for line in file_in:
                yield line.strip().split(' ')

def emit_final_key_value():
        for key, el in groupby(convert_mapper_o_to_iter(sys.stdin), itemgetter(0)):
                A = {}
                B = {}
                for item in list(el):
                        con = item[1].split(',')
                        if con[0] == 'a':
                                A[int(con[1])] = int(con[2])
                        else:
                                B[int(con[1])] = int(con[2])
                ans = 0
                for i in range(5):
                        if i in A and i in B:
                                ans += A[i]*B[i]
                print key, ans

emit_final_key_value()

原创粉丝点击