Github user MLnick commented on a diff in the pull request:
https://github.com/apache/spark/pull/18305#discussion_r124238050
--- Diff:
mllib/src/main/scala/org/apache/spark/ml/optim/loss/RDDLossFunction.scala ---
@@ -62,8 +62,8 @@ private[ml] class RDDLossFunction[
val newAgg = instances.treeAggregate(thisAgg)(seqOp, combOp,
aggregationDepth)
val gradient = newAgg.gradient
val regLoss = regularization.map { regFun =>
- val (regLoss, regGradient) = regFun.calculate(coefficients.data)
- BLAS.axpy(1.0, Vectors.dense(regGradient), gradient)
+ val (regLoss, regGradient) =
regFun.calculate(Vectors.fromBreeze(coefficients))
--- End diff --
So here, for example, we need to convert from Breeze. We could just make
the L2 reg also for Breeze vector? It feels like we should just be able to
compose the loss function and reg loss function together.
---
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]
