From d83995271bbf319735efccfd0a36c99b10243d7a Mon Sep 17 00:00:00 2001 From: Holden Karau Date: Mon, 22 Jun 2015 22:40:19 -0700 Subject: [PATCH] [SPARK-7781] [MLLIB] gradient boosted trees.train regressor missing max bins Author: Holden Karau Closes #6331 from holdenk/SPARK-7781-GradientBoostedTrees.trainRegressor-missing-max-bins and squashes the following commits: 2894695 [Holden Karau] remove extra blank line 2573e8d [Holden Karau] Update the scala side of the pythonmllibapi and make the test a bit nicer too 3a09170 [Holden Karau] add maxBins to to the train method as well af7f274 [Holden Karau] Add maxBins to GradientBoostedTrees.trainRegressor and correctly mention the default of 32 in other places where it mentioned 100 (cherry picked from commit 164fe2aa44993da6c77af6de5efdae47a8b3958c) Signed-off-by: Joseph K. Bradley --- .../mllib/api/python/PythonMLLibAPI.scala | 4 +++- python/pyspark/mllib/tests.py | 7 ++++++ python/pyspark/mllib/tree.py | 22 ++++++++++++------- 3 files changed, 24 insertions(+), 9 deletions(-) diff --git a/mllib/src/main/scala/org/apache/spark/mllib/api/python/PythonMLLibAPI.scala b/mllib/src/main/scala/org/apache/spark/mllib/api/python/PythonMLLibAPI.scala index 16f3131796709..d1b2c98a547ed 100644 --- a/mllib/src/main/scala/org/apache/spark/mllib/api/python/PythonMLLibAPI.scala +++ b/mllib/src/main/scala/org/apache/spark/mllib/api/python/PythonMLLibAPI.scala @@ -685,12 +685,14 @@ private[python] class PythonMLLibAPI extends Serializable { lossStr: String, numIterations: Int, learningRate: Double, - maxDepth: Int): GradientBoostedTreesModel = { + maxDepth: Int, + maxBins: Int): GradientBoostedTreesModel = { val boostingStrategy = BoostingStrategy.defaultParams(algoStr) boostingStrategy.setLoss(Losses.fromString(lossStr)) boostingStrategy.setNumIterations(numIterations) boostingStrategy.setLearningRate(learningRate) boostingStrategy.treeStrategy.setMaxDepth(maxDepth) + boostingStrategy.treeStrategy.setMaxBins(maxBins) boostingStrategy.treeStrategy.categoricalFeaturesInfo = categoricalFeaturesInfo.asScala.toMap val cached = data.rdd.persist(StorageLevel.MEMORY_AND_DISK) diff --git a/python/pyspark/mllib/tests.py b/python/pyspark/mllib/tests.py index 7a113f8751ff8..4335143a8dd59 100644 --- a/python/pyspark/mllib/tests.py +++ b/python/pyspark/mllib/tests.py @@ -444,6 +444,13 @@ def test_regression(self): except ValueError: self.fail() + # Verify that maxBins is being passed through + GradientBoostedTrees.trainRegressor( + rdd, categoricalFeaturesInfo=categoricalFeaturesInfo, numIterations=4, maxBins=32) + with self.assertRaises(Exception) as cm: + GradientBoostedTrees.trainRegressor( + rdd, categoricalFeaturesInfo=categoricalFeaturesInfo, numIterations=4, maxBins=1) + class StatTests(MLlibTestCase): # SPARK-4023 diff --git a/python/pyspark/mllib/tree.py b/python/pyspark/mllib/tree.py index cfcbea573fd22..372b86a7c95d9 100644 --- a/python/pyspark/mllib/tree.py +++ b/python/pyspark/mllib/tree.py @@ -299,7 +299,7 @@ def trainClassifier(cls, data, numClasses, categoricalFeaturesInfo, numTrees, 1 internal node + 2 leaf nodes. (default: 4) :param maxBins: maximum number of bins used for splitting features - (default: 100) + (default: 32) :param seed: Random seed for bootstrapping and choosing feature subsets. :return: RandomForestModel that can be used for prediction @@ -377,7 +377,7 @@ def trainRegressor(cls, data, categoricalFeaturesInfo, numTrees, featureSubsetSt 1 leaf node; depth 1 means 1 internal node + 2 leaf nodes. (default: 4) :param maxBins: maximum number of bins used for splitting - features (default: 100) + features (default: 32) :param seed: Random seed for bootstrapping and choosing feature subsets. :return: RandomForestModel that can be used for prediction @@ -435,16 +435,17 @@ class GradientBoostedTrees(object): @classmethod def _train(cls, data, algo, categoricalFeaturesInfo, - loss, numIterations, learningRate, maxDepth): + loss, numIterations, learningRate, maxDepth, maxBins): first = data.first() assert isinstance(first, LabeledPoint), "the data should be RDD of LabeledPoint" model = callMLlibFunc("trainGradientBoostedTreesModel", data, algo, categoricalFeaturesInfo, - loss, numIterations, learningRate, maxDepth) + loss, numIterations, learningRate, maxDepth, maxBins) return GradientBoostedTreesModel(model) @classmethod def trainClassifier(cls, data, categoricalFeaturesInfo, - loss="logLoss", numIterations=100, learningRate=0.1, maxDepth=3): + loss="logLoss", numIterations=100, learningRate=0.1, maxDepth=3, + maxBins=32): """ Method to train a gradient-boosted trees model for classification. @@ -467,6 +468,8 @@ def trainClassifier(cls, data, categoricalFeaturesInfo, :param maxDepth: Maximum depth of the tree. E.g., depth 0 means 1 leaf node; depth 1 means 1 internal node + 2 leaf nodes. (default: 3) + :param maxBins: maximum number of bins used for splitting + features (default: 32) DecisionTree requires maxBins >= max categories :return: GradientBoostedTreesModel that can be used for prediction @@ -499,11 +502,12 @@ def trainClassifier(cls, data, categoricalFeaturesInfo, [1.0, 0.0] """ return cls._train(data, "classification", categoricalFeaturesInfo, - loss, numIterations, learningRate, maxDepth) + loss, numIterations, learningRate, maxDepth, maxBins) @classmethod def trainRegressor(cls, data, categoricalFeaturesInfo, - loss="leastSquaresError", numIterations=100, learningRate=0.1, maxDepth=3): + loss="leastSquaresError", numIterations=100, learningRate=0.1, maxDepth=3, + maxBins=32): """ Method to train a gradient-boosted trees model for regression. @@ -522,6 +526,8 @@ def trainRegressor(cls, data, categoricalFeaturesInfo, contribution of each estimator. The learning rate should be between in the interval (0, 1]. (default: 0.1) + :param maxBins: maximum number of bins used for splitting + features (default: 32) DecisionTree requires maxBins >= max categories :param maxDepth: Maximum depth of the tree. E.g., depth 0 means 1 leaf node; depth 1 means 1 internal node + 2 leaf nodes. (default: 3) @@ -556,7 +562,7 @@ def trainRegressor(cls, data, categoricalFeaturesInfo, [1.0, 0.0] """ return cls._train(data, "regression", categoricalFeaturesInfo, - loss, numIterations, learningRate, maxDepth) + loss, numIterations, learningRate, maxDepth, maxBins) def _test():