add 1 un

mllib/src/main/scala/org/apache/spark/ml/optim/aggregator/HingeAggregator.scala

@@ -22,11 +22,11 @@ import org.apache.spark.ml.feature.Instance
 import org.apache.spark.ml.linalg._
 
 /**
- * LinearSVCAggregator computes the gradient and loss for loss function ("hinge" or
+ * HingeAggregator computes the gradient and loss for loss function ("hinge" or
  * "squared_hinge", as used in binary classification for instances in sparse or dense
  * vector in an online fashion.
  *
- * Two LinearSVCAggregator can be merged together to have a summary of loss and gradient of
+ * Two HingeAggregators can be merged together to have a summary of loss and gradient of
  * the corresponding joint dataset.
  *
  * This class standardizes feature values during computation using bcFeaturesStd.
@@ -50,11 +50,11 @@ private[ml] class HingeAggregator(
   protected override val dim: Int = numFeaturesPlusIntercept
 
   /**
-   * Add a new training instance to this LinearSVCAggregator, and update the loss and gradient
+   * Add a new training instance to this HingeAggregator, and update the loss and gradient
    * of the objective function.
    *
    * @param instance The instance of data point to be added.
-   * @return This LinearSVCAggregator object.
+   * @return This HingeAggregator object.
    */
   def add(instance: Instance): this.type = {
     instance match { case Instance(label, weight, features) =>

mllib/src/test/scala/org/apache/spark/ml/optim/aggregator/HingeAggregatorSuite.scala

@@ -67,7 +67,7 @@ class HingeAggregatorSuite extends SparkFunSuite with MLlibTestSparkContext {
     val interceptArray = Array(2.0)
     val agg = getNewAggregator(instances, Vectors.dense(coefArray ++ interceptArray),
       fitIntercept = true)
-    withClue("LogisticAggregator does not support negative instance weights") {
+    withClue("HingeAggregator does not support negative instance weights") {
       intercept[IllegalArgumentException] {
         agg.add(Instance(1.0, -1.0, Vectors.dense(2.0, 1.0)))
       }
@@ -133,4 +133,18 @@ class HingeAggregatorSuite extends SparkFunSuite with MLlibTestSparkContext {
     assert(gradient ~== agg.gradient relTol 0.01)
   }
 
+  test("check with zero standard deviation") {
+    val instancesConstantFeature = Array(
+      Instance(0.0, 0.1, Vectors.dense(1.0, 2.0)),
+      Instance(1.0, 0.5, Vectors.dense(1.0, 1.0)),
+      Instance(1.0, 0.3, Vectors.dense(1.0, 0.5)))
+    val binaryCoefArray = Array(1.0, 2.0)
+    val intercept = 1.0
+    val aggConstantFeatureBinary = getNewAggregator(instancesConstantFeature,
+      Vectors.dense(binaryCoefArray ++ Array(intercept)), fitIntercept = true)
+    instances.foreach(aggConstantFeatureBinary.add)
+    // constant features should not affect gradient
+    assert(aggConstantFeatureBinary.gradient(0) === 0.0)
+  }
+
 }