[SPARK-18481][ML] ML 2.1 QA: Remove deprecated methods for ML

mllib/src/main/scala/org/apache/spark/ml/classification/GBTClassifier.scala

@@ -201,6 +201,11 @@ class GBTClassificationModel private[ml](
   @Since("1.4.0")
   override def trees: Array[DecisionTreeRegressionModel] = _trees
 
+  /**
+   * Number of trees in ensemble
+   */
+  val getNumTrees: Int = trees.length
+
   @Since("1.4.0")
   override def treeWeights: Array[Double] = _treeWeights
 

mllib/src/main/scala/org/apache/spark/ml/classification/LogisticRegression.scala

@@ -40,7 +40,7 @@ import org.apache.spark.mllib.util.MLUtils
 import org.apache.spark.rdd.RDD
 import org.apache.spark.sql.{DataFrame, Dataset, Row}
 import org.apache.spark.sql.functions.{col, lit}
-import org.apache.spark.sql.types.DoubleType
+import org.apache.spark.sql.types.{DataType, DoubleType, StructType}
 import org.apache.spark.storage.StorageLevel
 import org.apache.spark.util.VersionUtils
 
@@ -176,8 +176,12 @@ private[classification] trait LogisticRegressionParams extends ProbabilisticClas
     }
   }
 
-  override def validateParams(): Unit = {
+  override protected def validateAndTransformSchema(
+      schema: StructType,
+      fitting: Boolean,
+      featuresDataType: DataType): StructType = {
     checkThresholdConsistency()
+    super.validateAndTransformSchema(schema, fitting, featuresDataType)
   }
 }
 

mllib/src/main/scala/org/apache/spark/ml/classification/RandomForestClassifier.scala

@@ -158,7 +158,7 @@ class RandomForestClassificationModel private[ml] (
     @Since("1.6.0") override val numFeatures: Int,
     @Since("1.5.0") override val numClasses: Int)
   extends ProbabilisticClassificationModel[Vector, RandomForestClassificationModel]
-  with RandomForestClassificationModelParams with TreeEnsembleModel[DecisionTreeClassificationModel]
+  with RandomForestClassifierParams with TreeEnsembleModel[DecisionTreeClassificationModel]
   with MLWritable with Serializable {
 
   require(_trees.nonEmpty, "RandomForestClassificationModel requires at least 1 tree.")
@@ -221,15 +221,6 @@ class RandomForestClassificationModel private[ml] (
     }
   }
 
-  /**
-   * Number of trees in ensemble
-   *
-   * @deprecated  Use [[getNumTrees]] instead.  This method will be removed in 2.1.0
-   */
-  // TODO: Once this is removed, then this class can inherit from RandomForestClassifierParams
-  @deprecated("Use getNumTrees instead.  This method will be removed in 2.1.0.", "2.0.0")
-  val numTrees: Int = trees.length
-
   @Since("1.4.0")
   override def copy(extra: ParamMap): RandomForestClassificationModel = {
     copyValues(new RandomForestClassificationModel(uid, _trees, numFeatures, numClasses), extra)

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

@@ -216,13 +216,6 @@ final class ChiSqSelectorModel private[ml] (
   @Since("1.6.0")
   def setOutputCol(value: String): this.type = set(outputCol, value)
 
-  /**
-   * @group setParam
-   */
-  @Since("1.6.0")
-  @deprecated("labelCol is not used by ChiSqSelectorModel.", "2.0.0")
-  def setLabelCol(value: String): this.type = set(labelCol, value)
-
   @Since("2.0.0")
   override def transform(dataset: Dataset[_]): DataFrame = {
     val transformedSchema = transformSchema(dataset.schema, logging = true)

mllib/src/main/scala/org/apache/spark/ml/param/params.scala

@@ -546,21 +546,6 @@ trait Params extends Identifiable with Serializable {
       .map(m => m.invoke(this).asInstanceOf[Param[_]])
   }
 
-  /**
-   * Validates parameter values stored internally.
-   * Raise an exception if any parameter value is invalid.
-   *
-   * This only needs to check for interactions between parameters.
-   * Parameter value checks which do not depend on other parameters are handled by
-   * `Param.validate()`. This method does not handle input/output column parameters;
-   * those are checked during schema validation.
-   * @deprecated Will be removed in 2.1.0. All the checks should be merged into transformSchema
-   */
-  @deprecated("Will be removed in 2.1.0. Checks should be merged into transformSchema.", "2.0.0")
-  def validateParams(): Unit = {
-    // Do nothing by default.  Override to handle Param interactions.
-  }
-
   /**
    * Explains a param.
    * @param param input param, must belong to this instance.

mllib/src/main/scala/org/apache/spark/ml/regression/GBTRegressor.scala

@@ -183,6 +183,11 @@ class GBTRegressionModel private[ml](
   @Since("1.4.0")
   override def trees: Array[DecisionTreeRegressionModel] = _trees
 
+  /**
+   * Number of trees in ensemble
+   */
+  val getNumTrees: Int = trees.length
+
   @Since("1.4.0")
   override def treeWeights: Array[Double] = _treeWeights
 

mllib/src/main/scala/org/apache/spark/ml/regression/LinearRegression.scala

@@ -605,9 +605,6 @@ class LinearRegressionSummary private[regression] (
     private val privateModel: LinearRegressionModel,
     private val diagInvAtWA: Array[Double]) extends Serializable {
 
-  @deprecated("The model field is deprecated and will be removed in 2.1.0.", "2.0.0")
-  val model: LinearRegressionModel = privateModel
-
   @transient private val metrics = new RegressionMetrics(
     predictions
       .select(col(predictionCol), col(labelCol).cast(DoubleType))

mllib/src/main/scala/org/apache/spark/ml/regression/RandomForestRegressor.scala

@@ -144,7 +144,7 @@ class RandomForestRegressionModel private[ml] (
     private val _trees: Array[DecisionTreeRegressionModel],
     override val numFeatures: Int)
   extends PredictionModel[Vector, RandomForestRegressionModel]
-  with RandomForestRegressionModelParams with TreeEnsembleModel[DecisionTreeRegressionModel]
+  with RandomForestRegressorParams with TreeEnsembleModel[DecisionTreeRegressionModel]
   with MLWritable with Serializable {
 
   require(_trees.nonEmpty, "RandomForestRegressionModel requires at least 1 tree.")
@@ -181,14 +181,6 @@ class RandomForestRegressionModel private[ml] (
     _trees.map(_.rootNode.predictImpl(features).prediction).sum / getNumTrees
   }
 
-  /**
-   * Number of trees in ensemble
-   * @deprecated  Use [[getNumTrees]] instead.  This method will be removed in 2.1.0
-   */
-  // TODO: Once this is removed, then this class can inherit from RandomForestRegressorParams
-  @deprecated("Use getNumTrees instead.  This method will be removed in 2.1.0.", "2.0.0")
-  val numTrees: Int = trees.length
-
   @Since("1.4.0")
   override def copy(extra: ParamMap): RandomForestRegressionModel = {
     copyValues(new RandomForestRegressionModel(uid, _trees, numFeatures), extra).setParent(parent)

mllib/src/main/scala/org/apache/spark/ml/tree/treeModels.scala

@@ -95,11 +95,6 @@ private[ml] trait TreeEnsembleModel[M <: DecisionTreeModel] {
   /** Trees in this ensemble. Warning: These have null parent Estimators. */
   def trees: Array[M]
 
-  /**
-   * Number of trees in ensemble
-   */
-  val getNumTrees: Int = trees.length
-
   /** Weights for each tree, zippable with [[trees]] */
   def treeWeights: Array[Double]
 

mllib/src/main/scala/org/apache/spark/ml/tree/treeParams.scala

@@ -317,8 +317,28 @@ private[ml] trait TreeEnsembleParams extends DecisionTreeParams {
   }
 }
 
-/** Used for [[RandomForestParams]] */
-private[ml] trait HasFeatureSubsetStrategy extends Params {
+/**
+ * Parameters for Random Forest algorithms.
+ */
+private[ml] trait RandomForestParams extends TreeEnsembleParams {
+
+  /**
+   * Number of trees to train (>= 1).
+   * If 1, then no bootstrapping is used.  If > 1, then bootstrapping is done.
+   * TODO: Change to always do bootstrapping (simpler).  SPARK-7130
+   * (default = 20)
+   * @group param
+   */
+  final val numTrees: IntParam = new IntParam(this, "numTrees", "Number of trees to train (>= 1)",
+    ParamValidators.gtEq(1))
+
+  setDefault(numTrees -> 20)
+
+  /** @group setParam */
+  def setNumTrees(value: Int): this.type = set(numTrees, value)
+
+  /** @group getParam */
+  final def getNumTrees: Int = $(numTrees)
 
   /**
    * The number of features to consider for splits at each tree node.
@@ -364,38 +384,6 @@ private[ml] trait HasFeatureSubsetStrategy extends Params {
   final def getFeatureSubsetStrategy: String = $(featureSubsetStrategy).toLowerCase
 }
 
-/**
- * Used for [[RandomForestParams]].
- * This is separated out from [[RandomForestParams]] because of an issue with the
- * `numTrees` method conflicting with this Param in the Estimator.
- */
-private[ml] trait HasNumTrees extends Params {
-
-  /**
-   * Number of trees to train (>= 1).
-   * If 1, then no bootstrapping is used.  If > 1, then bootstrapping is done.
-   * TODO: Change to always do bootstrapping (simpler).  SPARK-7130
-   * (default = 20)
-   * @group param
-   */
-  final val numTrees: IntParam = new IntParam(this, "numTrees", "Number of trees to train (>= 1)",
-    ParamValidators.gtEq(1))
-
-  setDefault(numTrees -> 20)
-
-  /** @group setParam */
-  def setNumTrees(value: Int): this.type = set(numTrees, value)
-
-  /** @group getParam */
-  final def getNumTrees: Int = $(numTrees)
-}
-
-/**
- * Parameters for Random Forest algorithms.
- */
-private[ml] trait RandomForestParams extends TreeEnsembleParams
-  with HasFeatureSubsetStrategy with HasNumTrees
-
 private[spark] object RandomForestParams {
   // These options should be lowercase.
   final val supportedFeatureSubsetStrategies: Array[String] =
@@ -405,15 +393,9 @@ private[spark] object RandomForestParams {
 private[ml] trait RandomForestClassifierParams
   extends RandomForestParams with TreeClassifierParams
 
-private[ml] trait RandomForestClassificationModelParams extends TreeEnsembleParams
-  with HasFeatureSubsetStrategy with TreeClassifierParams
-
 private[ml] trait RandomForestRegressorParams
   extends RandomForestParams with TreeRegressorParams
 
-private[ml] trait RandomForestRegressionModelParams extends TreeEnsembleParams
-  with HasFeatureSubsetStrategy with TreeRegressorParams
-
 /**
  * Parameters for Gradient-Boosted Tree algorithms.
  *
@@ -443,12 +425,11 @@ private[ml] trait GBTParams extends TreeEnsembleParams with HasMaxIter with HasS
    * (default = 0.1)
    * @group setParam
    */
-  def setStepSize(value: Double): this.type = set(stepSize, value)
-
-  override def validateParams(): Unit = {
+  def setStepSize(value: Double): this.type = {
     require(ParamValidators.inRange(0, 1, lowerInclusive = false, upperInclusive = true)(
-      getStepSize), "GBT parameter stepSize should be in interval (0, 1], " +
-      s"but it given invalid value $getStepSize.")
+      value), "GBT parameter stepSize should be in interval (0, 1], " +
+      s"but it given invalid value $value.")
+    set(stepSize, value)
   }
 
   /** (private[ml]) Create a BoostingStrategy instance to use with the old API. */

mllib/src/main/scala/org/apache/spark/ml/util/ReadWrite.scala

@@ -46,7 +46,7 @@ private[util] sealed trait BaseReadWrite {
    * Sets the Spark SQLContext to use for saving/loading.
    */
   @Since("1.6.0")
-  @deprecated("Use session instead", "2.0.0")
+  @deprecated("Use session instead, This method will be removed in 2.2.0.", "2.0.0")
   def context(sqlContext: SQLContext): this.type = {
     optionSparkSession = Option(sqlContext.sparkSession)
     this

project/MimaExcludes.scala

@@ -867,6 +867,32 @@ object MimaExcludes {
       // [SPARK-12221] Add CPU time to metrics
       ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.status.api.v1.TaskMetrics.this"),
       ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.status.api.v1.TaskMetricDistributions.this")
+    ) ++ Seq(
+      // [SPARK-18481] ML 2.1 QA: Remove deprecated methods for ML
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.PipelineStage.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.param.JavaParams.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.param.Params.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.classification.GBTClassificationModel.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.classification.LogisticRegression.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.classification.GBTClassifier.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.classification.LogisticRegressionModel.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.classification.RandomForestClassificationModel.numTrees"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.feature.ChiSqSelectorModel.setLabelCol"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.evaluation.Evaluator.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.regression.GBTRegressor.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.regression.GBTRegressionModel.validateParams"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.regression.LinearRegressionSummary.model"),
+      ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.ml.regression.RandomForestRegressionModel.numTrees"),
+      ProblemFilters.exclude[MissingTypesProblem]("org.apache.spark.ml.classification.RandomForestClassifier"),
+      ProblemFilters.exclude[MissingTypesProblem]("org.apache.spark.ml.classification.RandomForestClassificationModel"),
+      ProblemFilters.exclude[MissingTypesProblem]("org.apache.spark.ml.regression.RandomForestRegressor"),
+      ProblemFilters.exclude[MissingTypesProblem]("org.apache.spark.ml.regression.RandomForestRegressionModel"),
+      ProblemFilters.exclude[FinalMethodProblem]("org.apache.spark.ml.classification.RandomForestClassificationModel.getNumTrees"),
+      ProblemFilters.exclude[FinalMethodProblem]("org.apache.spark.ml.regression.RandomForestRegressionModel.getNumTrees"),
+      ProblemFilters.exclude[IncompatibleResultTypeProblem]("org.apache.spark.ml.classification.RandomForestClassificationModel.numTrees"),
+      ProblemFilters.exclude[IncompatibleResultTypeProblem]("org.apache.spark.ml.classification.RandomForestClassificationModel.setFeatureSubsetStrategy"),
+      ProblemFilters.exclude[IncompatibleResultTypeProblem]("org.apache.spark.ml.regression.RandomForestRegressionModel.numTrees"),
+      ProblemFilters.exclude[IncompatibleResultTypeProblem]("org.apache.spark.ml.regression.RandomForestRegressionModel.setFeatureSubsetStrategy")
     )
   }
 

python/pyspark/ml/util.py

@@ -78,7 +78,14 @@ def overwrite(self):
         raise NotImplementedError("MLWriter is not yet implemented for type: %s" % type(self))
 
     def context(self, sqlContext):
-        """Sets the SQL context to use for saving."""
+        """
+        Sets the SQL context to use for saving.
+        .. note:: Deprecated in 2.1, use session instead.
+        """
+        raise NotImplementedError("MLWriter is not yet implemented for type: %s" % type(self))
+
+    def session(self, sparkSession):
+        """Sets the Spark Session to use for saving."""
         raise NotImplementedError("MLWriter is not yet implemented for type: %s" % type(self))
 
 
@@ -105,10 +112,19 @@ def overwrite(self):
         return self
 
     def context(self, sqlContext):
-        """Sets the SQL context to use for saving."""
+        """
+        Sets the SQL context to use for saving.
+        .. note:: Deprecated in 2.1, use session instead.
+        """
+        warnings.warn("Deprecated in 2.1, use session instead.")
         self._jwrite.context(sqlContext._ssql_ctx)
         return self
 
+    def session(self, sparkSession):
+        """Sets the Spark Session to use for saving."""
+        self._jwrite.session(sparkSession._jsparkSession)
+        return self
+
 
 @inherit_doc
 class MLWritable(object):
@@ -155,7 +171,14 @@ def load(self, path):
         raise NotImplementedError("MLReader is not yet implemented for type: %s" % type(self))
 
     def context(self, sqlContext):
-        """Sets the SQL context to use for loading."""
+        """
+        Sets the SQL context to use for loading.
+        .. note:: Deprecated in 2.1, use session instead.
+        """
+        raise NotImplementedError("MLReader is not yet implemented for type: %s" % type(self))
+
+    def session(self, sparkSession):
+        """Sets the Spark Session to use for loading."""
         raise NotImplementedError("MLReader is not yet implemented for type: %s" % type(self))
 
 
@@ -180,10 +203,19 @@ def load(self, path):
         return self._clazz._from_java(java_obj)
 
     def context(self, sqlContext):
-        """Sets the SQL context to use for loading."""
+        """
+        Sets the SQL context to use for loading.
+        .. note:: Deprecated in 2.1, use session instead.
+        """
+        warnings.warn("Deprecated in 2.1, use session instead.")
         self._jread.context(sqlContext._ssql_ctx)
         return self
 
+    def session(self, sparkSession):
+        """Sets the Spark Session to use for loading."""
+        self._jread.session(sparkSession._jsparkSession)
+        return self
+
     @classmethod
     def _java_loader_class(cls, clazz):
         """