如何将mllib als结果导出到文本文件Spark中

Question

我是Spark的新手。我正在尝试运行协同过滤算法，我做到了。现在我想知道如何将用户推荐放入文本文件或mysql数据集中？

例如在我的火花壳中：

scala> topKRecs.mkString("\n")
res3: String = 
Rating(101168482,976126,17.762857135595247)
Rating(101168482,298077,15.618423165870723)
Rating(101168482,269474,15.20514216060495)
Rating(101168482,992979001,14.874869183497482)
Rating(101168482,302470001,14.794531500898763)
Rating(101168482,303592,14.779064871786169)
Rating(101168482,1538670001,13.96013790139799)
Rating(101168482,458919,13.905207127505795)
Rating(101168482,1141423,13.882807601308972)
Rating(101168482,992981003,13.844550596026654)

Answer 1

你可以通过两种方式做到这一点：   1存储评级对象，即：

topKRecs.saveAsObjectFile("path")

2将值存储为输出文件

topKRecs.map(rating => (rating.user, rating.product, rating.rating)).saveAsTextFile("path")