Github user srowen commented on a diff in the pull request:
https://github.com/apache/spark/pull/19988#discussion_r157342991
--- Diff:
mllib/src/main/scala/org/apache/spark/ml/optim/NormalEquationSolver.scala ---
@@ -109,6 +109,11 @@ private[optim] class QuasiNewtonSolver(
state = states.next()
arrayBuilder += state.adjustedValue
}
+ if (state == null || state.searchFailed) {
+ val msg = s"${optimizer.getClass.getName} failed."
+ logError(msg)
--- End diff --
Yeah, I'm thinking logWarning is more appropriate. I know the existing code
had logError, and could change too.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
