Github user MLnick commented on a diff in the pull request:
https://github.com/apache/spark/pull/10152#discussion_r46806041
--- Diff:
mllib/src/main/scala/org/apache/spark/mllib/feature/Word2Vec.scala ---
@@ -281,16 +280,17 @@ class Word2Vec extends Serializable with Logging {
val expTable = sc.broadcast(createExpTable())
val bcVocab = sc.broadcast(vocab)
val bcVocabHash = sc.broadcast(vocabHash)
-
- val sentences: RDD[Array[Int]] = words.mapPartitions { iter =>
+ //each partition is a collection of sentences, will be translated into
arrays of Index integer
+ val sentences: RDD[Array[Int]] = dataset.mapPartitions { iter =>
--- End diff --
Agree that Word2Vec can be used in settings where the "sentence" is just a
sequence of "things" (e.g. recommendation where the "sentences" could be "page
views" for a user, and could be longer than 1000).
So while I'd tend to agree that the caller should really be responsible for
setting up the "sentence" structure, as @srowen says rather make it
configurable with a default of `1000` to keep backward compatible behavior. You
can create a `set` method for the new parameter and add a `@Since` annotation
(see e.g. the `setMinCount` method added)
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
