Github user sethah commented on a diff in the pull request:
https://github.com/apache/spark/pull/16377#discussion_r94288799
--- Diff:
mllib/src/main/scala/org/apache/spark/ml/tree/impl/RandomForest.scala ---
@@ -713,6 +713,15 @@ private[spark] object RandomForest extends Logging {
binAggregates.metadata.numSplits(featureIndex) != 0
}
+ // If no valid splits for features, then this split is invalid,
+ // return invalid information gain stats. Take any split and continue.
+ if (validFeatureSplits.isEmpty) {
+ val parentImpurityCalculator =
binAggregates.getImpurityCalculator(0, 0)
+ // Splits is empty, so arbitrarily choose to split on first feature
with any threshold
+ return (new ContinuousSplit(0, 0),
+ ImpurityStats.getInvalidImpurityStats(parentImpurityCalculator))
+ }
+
// For each (feature, split), calculate the gain, and select the best
(feature, split).
val (bestSplit, bestSplitStats) =
--- End diff --
I slightly prefer to avoid using `if (...) return` and instead do:
````scala
val splitsAndImpurityInfo = validFeatureSplits.map { ... }
val (bestSplit, bestSplitStats) = if (splitsAndImpurityInfo.isEmpty) { ...
} else { ... }
(bestSplit, bestSplitStats)
````
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
