python 推荐系统 hadoop_hadoop1-构建电影推荐系统

计算结果：

[root@hadoop ~]# hadoop dfs -cat /output/*

1       107,5.0

1       106,18.0

1       105,15.5

1       104,33.5

1       103,39.0

1       102,31.5

1       101,44.0

2       107,4.0

2       106,20.5

2       105,15.5

2       104,36.0

2       103,41.5

2       102,32.5

2       101,45.5

3       107,15.5

3       106,16.5

3       105,26.0

3       104,38.0

3       103,24.5

3       102,18.5

3       101,40.0

4       107,9.5

4       106,33.0

4       105,26.0

4       104,55.0

4       103,53.5

4       102,37.0

4       101,63.0

5       107,11.5

5       106,34.5

5       105,32.0

5       104,59.0

5       103,56.5

5       102,42.5

5       101,68.0

--------------------------------------------------------------------------------

参考：http://www.aboutyun.com/thread-8155-1-1.html

这样我们就自己编程实现了MapReduce化基于物品的协同过滤算法。