Github user akopich commented on a diff in the pull request:
https://github.com/apache/spark/pull/18924#discussion_r140199136
--- Diff:
mllib/src/main/scala/org/apache/spark/mllib/clustering/LDAOptimizer.scala ---
@@ -503,17 +518,15 @@ final class OnlineLDAOptimizer extends LDAOptimizer {
}
/**
- * Update alpha based on `gammat`, the inferred topic distributions for
documents in the
- * current mini-batch. Uses Newton-Rhapson method.
+ * Update alpha based on `logphat`.
--- End diff --
Please, check out the updated PR.
I have added `val stats: RDD[(BDM[Double], Option[BDV[Double]], Int)] =
batch.mapPartitions {...}`.
Unfortunately, we cannot have the aggregation operation in a purely
in-place manner now since `Int` is immutable. Shouldn't be a big deal since
matrices and vectors are still updated in place.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
