Github user BryanCutler commented on a diff in the pull request:
https://github.com/apache/spark/pull/20124#discussion_r159763813
--- Diff:
mllib/src/main/scala/org/apache/spark/ml/tuning/ParamGridBuilder.scala ---
@@ -118,3 +118,63 @@ class ParamGridBuilder @Since("1.2.0") {
paramMaps
}
}
+
+
+object ParamGridBuilder {
+
+ /**
+ * Split the given list of paramMaps in two such that the left list
contains the supplied
+ * params and the right list does not.
+ *
+ * @param paramMaps Array of paramMaps to split
+ * @param params Array of params that will be in the left list
+ * @return Tuple of two Array[ParamMap]
+ */
+ def splitOnParams(paramMaps: Array[ParamMap],
+ params: Array[Param[_]]): (Array[ParamMap],
Array[ParamMap]) = {
+ val leftValues = mutable.Map.empty[Param[_],
mutable.LinkedHashSet[Any]]
--- End diff --
This could be done more efficiently, I just went with a simple routine that
I knew would generate the correct output.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
