logloss -> logLoss

mllib/src/main/scala/org/apache/spark/ml/evaluation/MulticlassClassificationEvaluator.scala

@@ -29,7 +29,7 @@ import org.apache.spark.sql.types.DoubleType
 
 /**
  * Evaluator for multiclass classification, which expects input columns: prediction, label,
- * weight(optional) and probabilityCol(only for log-loss).
+ * weight(optional) and probability(only for logLoss).
  */
 @Since("1.5.0")
 class MulticlassClassificationEvaluator @Since("1.5.0") (@Since("1.5.0") override val uid: String)
@@ -46,7 +46,7 @@ class MulticlassClassificationEvaluator @Since("1.5.0") (@Since("1.5.0") overrid
    * `"weightedPrecision"`, `"weightedRecall"`, `"weightedTruePositiveRate"`,
    * `"weightedFalsePositiveRate"`, `"weightedFMeasure"`, `"truePositiveRateByLabel"`,
    * `"falsePositiveRateByLabel"`, `"precisionByLabel"`, `"recallByLabel"`,
-   * `"fMeasureByLabel"`, `"logloss"`)
+   * `"fMeasureByLabel"`, `"logLoss"`)
    *
    * @group param
    */
@@ -116,7 +116,7 @@ class MulticlassClassificationEvaluator @Since("1.5.0") (@Since("1.5.0") overrid
 
   @Since("3.0.0")
   final val eps: DoubleParam = new DoubleParam(this, "eps",
-    "Log loss is undefined for p=0 or p=1, so probabilities are clipped to " +
+    "LogLoss is undefined for p=0 or p=1, so probabilities are clipped to " +
       "max(eps, min(1 - eps, p)).",
     ParamValidators.inRange(0, 0.5, false, false))
 
@@ -142,7 +142,7 @@ class MulticlassClassificationEvaluator @Since("1.5.0") (@Since("1.5.0") overrid
       lit(1.0)
     }
 
-    val rdd = if ($(metricName) == "logloss") {
+    val rdd = if ($(metricName) == "logLoss") {
       // probabilityCol is only needed to compute logloss
       require(isDefined(probabilityCol) && $(probabilityCol).nonEmpty)
       val p = DatasetUtils.columnToVector(dataset, $(probabilityCol))
@@ -172,7 +172,7 @@ class MulticlassClassificationEvaluator @Since("1.5.0") (@Since("1.5.0") overrid
       case "precisionByLabel" => metrics.precision($(metricLabel))
       case "recallByLabel" => metrics.recall($(metricLabel))
       case "fMeasureByLabel" => metrics.fMeasure($(metricLabel), $(beta))
-      case "logloss" => metrics.logloss($(eps))
+      case "logLoss" => metrics.logLoss($(eps))
     }
   }
 
@@ -193,7 +193,7 @@ object MulticlassClassificationEvaluator
   private val supportedMetricNames = Array("f1", "accuracy", "weightedPrecision", "weightedRecall",
     "weightedTruePositiveRate", "weightedFalsePositiveRate", "weightedFMeasure",
     "truePositiveRateByLabel", "falsePositiveRateByLabel", "precisionByLabel", "recallByLabel",
-    "fMeasureByLabel", "logloss")
+    "fMeasureByLabel", "logLoss")
 
   @Since("1.6.0")
   override def load(path: String): MulticlassClassificationEvaluator = super.load(path)

mllib/src/main/scala/org/apache/spark/mllib/evaluation/MulticlassMetrics.scala

@@ -23,7 +23,7 @@ import scala.collection.mutable
 import org.apache.spark.annotation.Since
 import org.apache.spark.mllib.linalg.{Matrices, Matrix}
 import org.apache.spark.rdd.RDD
-import org.apache.spark.sql.{DataFrame, Row}
+import org.apache.spark.sql.DataFrame
 
 /**
  * Evaluator for multiclass classification.
@@ -37,7 +37,7 @@ class MulticlassMetrics @Since("1.1.0") (predictionAndLabels: RDD[_ <: Product])
   /**
    * An auxiliary constructor taking a DataFrame.
    * @param predictionAndLabels a DataFrame with columns: prediction, label, weight(optional)
-   *                            and probability(only for logloss)
+   *                            and probability(only for logLoss)
    */
   private[mllib] def this(predictionAndLabels: DataFrame) =
     this(predictionAndLabels.rdd.map { r =>
@@ -241,12 +241,12 @@ class MulticlassMetrics @Since("1.1.0") (predictionAndLabels: RDD[_ <: Product])
   lazy val labels: Array[Double] = tpByClass.keys.toArray.sorted
 
   /**
-   * Returns the log-loss, aka logistic loss or cross-entropy loss.
-   * @param eps Log loss is undefined for p=0 or p=1, so probabilities are
+   * Returns the logLoss, aka logistic loss or cross-entropy loss.
+   * @param eps LogLoss is undefined for p=0 or p=1, so probabilities are
    *            clipped to max(eps, min(1 - eps, p)).
    */
   @Since("3.0.0")
-  def logloss(eps: Double = 1e-15): Double = {
+  def logLoss(eps: Double = 1e-15): Double = {
     require(eps > 0 && eps < 0.5, s"eps must be in range (0, 0.5), but got $eps")
     val loss1 = - math.log(eps)
     val loss2 = - math.log(1 - eps)

mllib/src/test/scala/org/apache/spark/ml/evaluation/MulticlassClassificationEvaluatorSuite.scala

@@ -77,7 +77,7 @@ class MulticlassClassificationEvaluatorSuite
     }.toDF("prediction", "label", "probability")
 
     val evaluator = new MulticlassClassificationEvaluator()
-      .setMetricName("logloss")
+      .setMetricName("logLoss")
     assert(evaluator.evaluate(df) ~== 0.9682005730687164 absTol 1e-5)
   }
 }

mllib/src/test/scala/org/apache/spark/mllib/evaluation/MulticlassMetricsSuite.scala

@@ -206,14 +206,14 @@ class MulticlassMetricsSuite extends SparkFunSuite with MLlibTestSparkContext {
         (prediction, label, weight, probability)
     }
     val metrics = new MulticlassMetrics(rdd)
-    assert(metrics.logloss() ~== 0.16145936283256573 relTol delta)
+    assert(metrics.logLoss() ~== 0.16145936283256573 relTol delta)
 
     val rdd2 = rdd.map {
       case (prediction: Double, label: Double, weight: Double, probability: Array[Double]) =>
         (prediction, label, 1.0, probability)
     }
     val metrics2 = new MulticlassMetrics(rdd2)
-    assert(metrics2.logloss() ~== 0.21616187468057912 relTol delta)
+    assert(metrics2.logLoss() ~== 0.21616187468057912 relTol delta)
   }
 
   test("MulticlassMetrics supports multi-class log-loss") {
@@ -245,13 +245,13 @@ class MulticlassMetricsSuite extends SparkFunSuite with MLlibTestSparkContext {
         (prediction, label, weight, probability)
     }
     val metrics = new MulticlassMetrics(rdd)
-    assert(metrics.logloss() ~== 1.3529429766879466 relTol delta)
+    assert(metrics.logLoss() ~== 1.3529429766879466 relTol delta)
 
     val rdd2 = rdd.map {
       case (prediction: Double, label: Double, weight: Double, probability: Array[Double]) =>
         (prediction, label, 1.0, probability)
     }
     val metrics2 = new MulticlassMetrics(rdd2)
-    assert(metrics2.logloss() ~== 0.9682005730687164 relTol delta)
+    assert(metrics2.logLoss() ~== 0.9682005730687164 relTol delta)
   }
 }

python/pyspark/ml/evaluation.py

@@ -317,7 +317,7 @@ class MulticlassClassificationEvaluator(JavaEvaluator, HasLabelCol, HasPredictio
                                         HasProbabilityCol, JavaMLReadable, JavaMLWritable):
     """
     Evaluator for Multiclass Classification, which expects input
-    columns: prediction, label, weight(optional) and probabilityCol(only for log-loss).
+    columns: prediction, label, weight(optional) and probabilityCol(only for logLoss).
 
     >>> scoreAndLabels = [(0.0, 0.0), (0.0, 1.0), (0.0, 0.0),
     ...     (1.0, 0.0), (1.0, 1.0), (1.0, 1.0), (1.0, 1.0), (2.0, 2.0), (2.0, 0.0)]
@@ -352,7 +352,7 @@ class MulticlassClassificationEvaluator(JavaEvaluator, HasLabelCol, HasPredictio
     ...     "label", "weight", "probability"])
     >>> evaluator = MulticlassClassificationEvaluator(predictionCol="prediction",
     ...     probabilityCol="probability")
-    >>> evaluator.setMetricName("logloss")
+    >>> evaluator.setMetricName("logLoss")
     MulticlassClassificationEvaluator...
     >>> evaluator.evaluate(dataset)
     0.9682...
@@ -364,7 +364,7 @@ class MulticlassClassificationEvaluator(JavaEvaluator, HasLabelCol, HasPredictio
                        "(f1|accuracy|weightedPrecision|weightedRecall|weightedTruePositiveRate|"
                        "weightedFalsePositiveRate|weightedFMeasure|truePositiveRateByLabel|"
                        "falsePositiveRateByLabel|precisionByLabel|recallByLabel|fMeasureByLabel|"
-                       "logloss)",
+                       "logLoss)",
                        typeConverter=TypeConverters.toString)
     metricLabel = Param(Params._dummy(), "metricLabel",
                         "The class whose metric will be computed in truePositiveRateByLabel|"
@@ -376,7 +376,7 @@ class MulticlassClassificationEvaluator(JavaEvaluator, HasLabelCol, HasPredictio
                  " Must be > 0. The default value is 1.",
                  typeConverter=TypeConverters.toFloat)
     eps = Param(Params._dummy(), "eps",
-                "Log loss is undefined for p=0 or p=1, so probabilities are clipped to "
+                "LogLoss is undefined for p=0 or p=1, so probabilities are clipped to "
                 "max(eps, min(1 - eps, p)). "
                 "Must be in range (0, 0.5). The default value is 1e-15.",
                 typeConverter=TypeConverters.toFloat)

python/pyspark/mllib/evaluation.py

@@ -244,7 +244,7 @@ class MulticlassMetrics(JavaModelWrapper):
     ...      (1.0, 1.0, 1.0, [0.1, 0.8, 0.1]), (0.0, 2.0, 1.0, [0.9, 0.05, 0.05]),
     ...      (0.0, 0.0, 1.0, [0.8, 0.2, 0.0]), (1.0, 1.0, 1.0, [0.3, 0.65, 0.05])])
     >>> metrics = MulticlassMetrics(predictionAndLabelsWithProbabilities)
-    >>> metrics.logloss()
+    >>> metrics.logLoss()
     0.9682...
 
     .. versionadded:: 1.4.0
@@ -366,11 +366,11 @@ def weightedFMeasure(self, beta=None):
             return self.call("weightedFMeasure", beta)
 
     @since('3.0.0')
-    def logloss(self, eps=1e-15):
+    def logLoss(self, eps=1e-15):
         """
-        Returns weighted log-loss.
+        Returns weighted logLoss.
         """
-        return self.call("logloss", eps)
+        return self.call("logLoss", eps)
 
 
 class RankingMetrics(JavaModelWrapper):