export numFeatures in ML PredictionModel

python/pyspark/ml/base.py

@@ -116,16 +116,3 @@ class Model(Transformer):
     """
 
     __metaclass__ = ABCMeta
-
-
-class HasNumFeaturesModel:
-    """
-    Provides getter of the number of features especially for model class
-    It should be mixin with JavaModel.
-    """
-    @property
-    def numFeatures(self):
-        """
-        The number of features used to train the model.
-        """
-        return self._call_java("numFeatures")

python/pyspark/ml/classification.py

@@ -20,7 +20,6 @@
 
 from pyspark import since, keyword_only
 from pyspark.ml import Estimator, Model
-from pyspark.ml.base import HasNumFeaturesModel
 from pyspark.ml.param.shared import *
 from pyspark.ml.regression import DecisionTreeModel, DecisionTreeRegressionModel, \
     RandomForestParams, TreeEnsembleModels, TreeEnsembleParams
@@ -66,6 +65,8 @@ class LogisticRegression(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredicti
     DenseVector([5.5...])
     >>> model.intercept
     -2.68...
+    >>> model.numFeatures
+    1
     >>> test0 = sc.parallelize([Row(features=Vectors.dense(-1.0))]).toDF()
     >>> result = model.transform(test0).head()
     >>> result.prediction
@@ -93,6 +94,8 @@ class LogisticRegression(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredicti
     True
     >>> model.intercept == model2.intercept
     True
+    >>> model.numFeatures == model2.numFeatures
+    True
 
     .. versionadded:: 1.3.0
     """
@@ -215,7 +218,7 @@ def _checkThresholdConsistency(self):
                                  " threshold (%g) and thresholds (equivalent to %g)" % (t2, t))
 
 
-class LogisticRegressionModel(HasNumFeaturesModel, JavaModel, JavaMLWritable, JavaMLReadable):
+class LogisticRegressionModel(JavaModel, JavaMLWritable, JavaMLReadable):
     """
     .. note:: Experimental
 
@@ -240,6 +243,14 @@ def intercept(self):
         """
         return self._call_java("intercept")
 
+    @property
+    @since("2.0.0")
+    def numFeatures(self):
+        """
+        Number of features the model was trained on.
+        """
+        return self._call_java("numFeatures")
+
     @property
     @since("2.0.0")
     def summary(self):
@@ -525,6 +536,8 @@ class DecisionTreeClassifier(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPred
     1
     >>> model.featureImportances
     SparseVector(1, {0: 1.0})
+    >>> model.numFeatures
+    1
     >>> print(model.toDebugString)
     DecisionTreeClassificationModel (uid=...) of depth 1 with 3 nodes...
     >>> test0 = spark.createDataFrame([(Vectors.dense(-1.0),)], ["features"])
@@ -549,8 +562,8 @@ class DecisionTreeClassifier(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPred
     >>> model2 = DecisionTreeClassificationModel.load(model_path)
     >>> model.featureImportances == model2.featureImportances
     True
-    >>> model.numFeatures
-    1
+    >>> model.numFeatures == model2.numFeatures
+    True
 
     .. versionadded:: 1.4.0
     """
@@ -600,8 +613,7 @@ def _create_model(self, java_model):
 
 
 @inherit_doc
-class DecisionTreeClassificationModel(HasNumFeaturesModel, DecisionTreeModel, JavaMLWritable,
-                                      JavaMLReadable):
+class DecisionTreeClassificationModel(DecisionTreeModel, JavaMLWritable, JavaMLReadable):
     """
     .. note:: Experimental
 
@@ -631,6 +643,14 @@ def featureImportances(self):
         """
         return self._call_java("featureImportances")
 
+    @property
+    @since("2.0.0")
+    def numFeatures(self):
+        """
+        Number of features the model was trained on.
+        """
+        return self._call_java("numFeatures")
+
 
 @inherit_doc
 class RandomForestClassifier(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredictionCol, HasSeed,
@@ -672,6 +692,8 @@ class RandomForestClassifier(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPred
     >>> test1 = spark.createDataFrame([(Vectors.sparse(1, [0], [1.0]),)], ["features"])
     >>> model.transform(test1).head().prediction
     1.0
+    >>> model.numFeatures
+    1
     >>> model.trees
     [DecisionTreeClassificationModel (uid=...) of depth..., DecisionTreeClassificationModel...]
     >>> rfc_path = temp_path + "/rfc"
@@ -734,8 +756,7 @@ def _create_model(self, java_model):
         return RandomForestClassificationModel(java_model)
 
 
-class RandomForestClassificationModel(TreeEnsembleModels, JavaMLWritable, JavaMLReadable,
-                                      HasNumFeaturesModel):
+class RandomForestClassificationModel(TreeEnsembleModels, JavaMLWritable, JavaMLReadable):
     """
     .. note:: Experimental
 
@@ -759,6 +780,14 @@ def featureImportances(self):
         """
         return self._call_java("featureImportances")
 
+    @property
+    @since("2.0.0")
+    def numFeatures(self):
+        """
+        Number of features the model was trained on.
+        """
+        return self._call_java("numFeatures")
+
     @property
     @since("2.0.0")
     def trees(self):
@@ -811,6 +840,8 @@ class GBTClassifier(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredictionCol
     1.0
     >>> model.totalNumNodes
     15
+    >>> model.numFeatures
+    1
     >>> print(model.toDebugString)
     GBTClassificationModel (uid=...)...with 5 trees...
     >>> gbtc_path = temp_path + "gbtc"
@@ -892,7 +923,7 @@ def getLossType(self):
         return self.getOrDefault(self.lossType)
 
 
-class GBTClassificationModel(TreeEnsembleModels, JavaMLWritable, JavaMLReadable, HasNumFeaturesModel):
+class GBTClassificationModel(TreeEnsembleModels, JavaMLWritable, JavaMLReadable):
     """
     .. note:: Experimental
 
@@ -916,6 +947,14 @@ def featureImportances(self):
         """
         return self._call_java("featureImportances")
 
+    @property
+    @since("2.0.0")
+    def numFeatures(self):
+        """
+        Number of features the model was trained on.
+        """
+        return self._call_java("numFeatures")
+
     @property
     @since("2.0.0")
     def trees(self):
@@ -961,6 +1000,8 @@ class NaiveBayes(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredictionCol, H
     >>> test1 = sc.parallelize([Row(features=Vectors.sparse(2, [0], [1.0]))]).toDF()
     >>> model.transform(test1).head().prediction
     1.0
+    >>> model.numFeatures
+    2
     >>> nb_path = temp_path + "/nb"
     >>> nb.save(nb_path)
     >>> nb2 = NaiveBayes.load(nb_path)
@@ -979,7 +1020,7 @@ class NaiveBayes(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredictionCol, H
     >>> result.prediction
     0.0
     >>> model.numFeatures == model2.numFeatures
-    2
+    True
 
     .. versionadded:: 1.5.0
     """
@@ -1052,7 +1093,7 @@ def getModelType(self):
         return self.getOrDefault(self.modelType)
 
 
-class NaiveBayesModel(JavaModel, JavaMLWritable, JavaMLReadable, HasNumFeaturesModel):
+class NaiveBayesModel(JavaModel, JavaMLWritable, JavaMLReadable):
     """
     .. note:: Experimental
 
@@ -1077,6 +1118,14 @@ def theta(self):
         """
         return self._call_java("theta")
 
+    @property
+    @since("2.0.0")
+    def numFeatures(self):
+        """
+        Number of features the model was trained on.
+        """
+        return self._call_java("numFeatures")
+
 
 @inherit_doc
 class MultilayerPerceptronClassifier(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredictionCol,
@@ -1102,6 +1151,8 @@ class MultilayerPerceptronClassifier(JavaEstimator, HasFeaturesCol, HasLabelCol,
     [2, 2, 2]
     >>> model.weights.size
     12
+    >>> model.numFeatures
+    2
     >>> testDF = spark.createDataFrame([
     ...     (Vectors.dense([1.0, 0.0]),),
     ...     (Vectors.dense([0.0, 0.0]),)], ["features"])
@@ -1255,8 +1306,7 @@ def getInitialWeights(self):
         return self.getOrDefault(self.initialWeights)
 
 
-class MultilayerPerceptronClassificationModel(JavaModel, JavaMLWritable, JavaMLReadable,
-                                              HasNumFeaturesModel):
+class MultilayerPerceptronClassificationModel(JavaModel, JavaMLWritable, JavaMLReadable):
     """
     .. note:: Experimental
 
@@ -1281,6 +1331,14 @@ def weights(self):
         """
         return self._call_java("weights")
 
+    @property
+    @since("2.0.0")
+    def numFeatures(self):
+        """
+        Number of features the model was trained on.
+        """
+        return self._call_java("numFeatures")
+
 
 class OneVsRestParams(HasFeaturesCol, HasLabelCol, HasPredictionCol):
     """