Spark：如何将RDD [（Long，Iterable [String]）]转换为RDD [（Long，String）]？

Question

如何将此RDD [（Long，Iterable [String]）]转换为...

(852403,Set(PT0000094043, PT0000097083, PT0000036162))
(357331,Set(PT0000068829, PT0000094042, PT0000066859))

RDD [（Long，String）]是这样的吗？

(852403, PT0000094043)
(852403, PT0000097083)
(852403, PT0000036162)
(357331, PT0000068829)
(357331, PT0000094042)
(357331, PT0000066859)

Answer 1

尝试flatMapValues：

rdd.flatMapValues(identity)

或flatMap：

rdd.flatMap{ case (k, vs) => vs.map(v => (k, v)) }