[SPARK-8744] [ML] Add a public constructor to StringIndexer

It would be helpful to allow users to pass a pre-computed index to create an indexer, rather than always going through StringIndexer to create the model.

Author: Holden Karau <holden@pigscanfly.ca>

Closes #7267 from holdenk/SPARK-8744-StringIndexerModel-should-have-public-constructor.

mllib/src/main/scala/org/apache/spark/ml/feature/StringIndexer.scala

@@ -102,10 +102,12 @@ class StringIndexer(override val uid: String) extends Estimator[StringIndexerMod
  * This is a temporary fix for the case when target labels do not exist during prediction.
  */
 @Experimental
-class StringIndexerModel private[ml] (
+class StringIndexerModel (
     override val uid: String,
     labels: Array[String]) extends Model[StringIndexerModel] with StringIndexerBase {
 
+  def this(labels: Array[String]) = this(Identifiable.randomUID("strIdx"), labels)
+
   private val labelToIndex: OpenHashMap[String, Double] = {
     val n = labels.length
     val map = new OpenHashMap[String, Double](n)

mllib/src/test/scala/org/apache/spark/ml/feature/StringIndexerSuite.scala

@@ -30,7 +30,9 @@ class StringIndexerSuite extends SparkFunSuite with MLlibTestSparkContext {
   test("params") {
     ParamsSuite.checkParams(new StringIndexer)
     val model = new StringIndexerModel("indexer", Array("a", "b"))
+    val modelWithoutUid = new StringIndexerModel(Array("a", "b"))
     ParamsSuite.checkParams(model)
+    ParamsSuite.checkParams(modelWithoutUid)
   }
 
   test("StringIndexer") {