[SPARK-10736] [ML] Use 1 for all ratings if $(ratingCol) = ""

For some implicit dataset, ratings may not exist in the training data. In this case, we can assume all observed pairs to be positive and treat their ratings as 1. This should happen when users set ```ratingCol``` to an empty string.

Author: Yanbo Liang <ybliang8@gmail.com>

Closes #8937 from yanboliang/spark-10736.

mllib/src/main/scala/org/apache/spark/ml/recommendation/ALS.scala

@@ -315,9 +315,9 @@ class ALS(override val uid: String) extends Estimator[ALSModel] with ALSParams {
 
   override def fit(dataset: DataFrame): ALSModel = {
     import dataset.sqlContext.implicits._
+    val r = if ($(ratingCol) != "") col($(ratingCol)).cast(FloatType) else lit(1.0f)
     val ratings = dataset
-      .select(col($(userCol)).cast(IntegerType), col($(itemCol)).cast(IntegerType),
-        col($(ratingCol)).cast(FloatType))
+      .select(col($(userCol)).cast(IntegerType), col($(itemCol)).cast(IntegerType), r)
       .map { row =>
         Rating(row.getInt(0), row.getInt(1), row.getFloat(2))
       }