[SPARK-15750][MLLib][PYSPARK] Constructing FPGrowth fails when no numPartitions specified in pyspark

mllib/src/main/scala/org/apache/spark/mllib/api/python/PythonMLLibAPI.scala

@@ -572,10 +572,7 @@ private[python] class PythonMLLibAPI extends Serializable {
       data: JavaRDD[java.lang.Iterable[Any]],
       minSupport: Double,
       numPartitions: Int): FPGrowthModel[Any] = {
-    val fpg = new FPGrowth()
-      .setMinSupport(minSupport)
-      .setNumPartitions(numPartitions)
-
+    val fpg = new FPGrowth(minSupport, numPartitions)
     val model = fpg.run(data.rdd.map(_.asScala.toArray))
     new FPGrowthModelWrapper(model)
   }

mllib/src/main/scala/org/apache/spark/mllib/fpm/FPGrowth.scala

@@ -162,7 +162,7 @@ object FPGrowthModel extends Loader[FPGrowthModel[_]] {
  *
  */
 @Since("1.3.0")
-class FPGrowth private (
+class FPGrowth private[spark] (
     private var minSupport: Double,
     private var numPartitions: Int) extends Logging with Serializable {
 

python/pyspark/mllib/tests.py

@@ -57,6 +57,7 @@
     DenseMatrix, SparseMatrix, Vectors, Matrices, MatrixUDT
 from pyspark.mllib.linalg.distributed import RowMatrix
 from pyspark.mllib.classification import StreamingLogisticRegressionWithSGD
+from pyspark.mllib.fpm import FPGrowth
 from pyspark.mllib.recommendation import Rating
 from pyspark.mllib.regression import LabeledPoint, StreamingLinearRegressionWithSGD
 from pyspark.mllib.random import RandomRDDs
@@ -1762,6 +1763,17 @@ def test_pca(self):
                 self.assertEqualUpToSign(pcs.toArray()[:, k - 1], expected_pcs[:, k - 1])
 
 
+class FPGrowthTest(MLlibTestCase):
+
+    def test_fpgrowth(self):
+        data = [["a", "b", "c"], ["a", "b", "d", "e"], ["a", "c", "e"], ["a", "c", "f"]]
+        rdd = self.sc.parallelize(data, 2)
+        model1 = FPGrowth.train(rdd, 0.6, 2)
+        # use default data partition number when numPartitions is not specified
+        model2 = FPGrowth.train(rdd, 0.6)
+        self.assertEqual(sorted(model1.freqItemsets().collect()),
+                         sorted(model2.freqItemsets().collect()))
+
 if __name__ == "__main__":
     from pyspark.mllib.tests import *
     if not _have_scipy: