From 4895f65447aa2338729fccb5200efa29a9d62163 Mon Sep 17 00:00:00 2001 From: Davies Liu Date: Fri, 7 Nov 2014 20:53:03 -0800 Subject: [SPARK-4304] [PySpark] Fix sort on empty RDD This PR fix sortBy()/sortByKey() on empty RDD. This should be back ported into 1.1/1.2 Author: Davies Liu Closes #3162 from davies/fix_sort and squashes the following commits: 84f64b7 [Davies Liu] add tests 52995b5 [Davies Liu] fix sortByKey() on empty RDD (cherry picked from commit 7779109796c90d789464ab0be35917f963bbe867) Signed-off-by: Josh Rosen Conflicts: python/pyspark/tests.py --- python/pyspark/rdd.py | 2 ++ python/pyspark/tests.py | 3 +++ 2 files changed, 5 insertions(+) (limited to 'python') diff --git a/python/pyspark/rdd.py b/python/pyspark/rdd.py index 3f81550bbb..ac8ceff055 100644 --- a/python/pyspark/rdd.py +++ b/python/pyspark/rdd.py @@ -598,6 +598,8 @@ class RDD(object): # the key-space into bins such that the bins have roughly the same # number of (key, value) pairs falling into them rddSize = self.count() + if not rddSize: + return self # empty RDD maxSampleSize = numPartitions * 20.0 # constant from Spark's RangePartitioner fraction = min(maxSampleSize / max(rddSize, 1), 1.0) samples = self.sample(False, fraction, 1).map(lambda (k, v): k).collect() diff --git a/python/pyspark/tests.py b/python/pyspark/tests.py index 5cea1b03ea..b4a9c59790 100644 --- a/python/pyspark/tests.py +++ b/python/pyspark/tests.py @@ -470,6 +470,9 @@ class TestRDDFunctions(PySparkTestCase): self.assertEquals(([1, "b"], [5]), rdd.histogram(1)) self.assertRaises(TypeError, lambda: rdd.histogram(2)) + def test_sort_on_empty_rdd(self): + self.assertEqual([], self.sc.parallelize(zip([], [])).sortByKey().collect()) + def test_sample(self): rdd = self.sc.parallelize(range(0, 100), 4) wo = rdd.sample(False, 0.1, 2).collect() -- cgit v1.2.3