Skip to content
Snippets Groups Projects
Commit 78bb7f80 authored by Sandeep Singh's avatar Sandeep Singh Committed by Joseph K. Bradley
Browse files

[SPARK-18274][ML][PYSPARK] Memory leak in PySpark JavaWrapper

## What changes were proposed in this pull request?
In`JavaWrapper `'s destructor make Java Gateway dereference object in destructor, using `SparkContext._active_spark_context._gateway.detach`
Fixing the copying parameter bug, by moving the `copy` method from `JavaModel` to `JavaParams`

## How was this patch tested?
```scala
import random, string
from pyspark.ml.feature import StringIndexer

l = [(''.join(random.choice(string.ascii_uppercase) for _ in range(10)), ) for _ in range(int(7e5))]  # 700000 random strings of 10 characters
df = spark.createDataFrame(l, ['string'])

for i in range(50):
    indexer = StringIndexer(inputCol='string', outputCol='index')
    indexer.fit(df)
```
* Before: would keep StringIndexer strong reference, causing GC issues and is halted midway
After: garbage collection works as the object is dereferenced, and computation completes
* Mem footprint tested using profiler
* Added a parameter copy related test which was failing before.

Author: Sandeep Singh <sandeep@techaddict.me>
Author: jkbradley <joseph.kurata.bradley@gmail.com>

Closes #15843 from techaddict/SPARK-18274.
parent e6534847
No related branches found
No related tags found
No related merge requests found
......@@ -390,6 +390,24 @@ class ParamTests(PySparkTestCase):
self.assertEqual(model.getWindowSize(), 6)
class EvaluatorTests(SparkSessionTestCase):
def test_java_params(self):
"""
This tests a bug fixed by SPARK-18274 which causes multiple copies
of a Params instance in Python to be linked to the same Java instance.
"""
evaluator = RegressionEvaluator(metricName="r2")
df = self.spark.createDataFrame([Row(label=1.0, prediction=1.1)])
evaluator.evaluate(df)
self.assertEqual(evaluator._java_obj.getMetricName(), "r2")
evaluatorCopy = evaluator.copy({evaluator.metricName: "mae"})
evaluator.evaluate(df)
evaluatorCopy.evaluate(df)
self.assertEqual(evaluator._java_obj.getMetricName(), "r2")
self.assertEqual(evaluatorCopy._java_obj.getMetricName(), "mae")
class FeatureTests(SparkSessionTestCase):
def test_binarizer(self):
......
......@@ -71,6 +71,10 @@ class JavaParams(JavaWrapper, Params):
__metaclass__ = ABCMeta
def __del__(self):
if SparkContext._active_spark_context:
SparkContext._active_spark_context._gateway.detach(self._java_obj)
def _make_java_param_pair(self, param, value):
"""
Makes a Java parm pair.
......@@ -180,6 +184,25 @@ class JavaParams(JavaWrapper, Params):
% stage_name)
return py_stage
def copy(self, extra=None):
"""
Creates a copy of this instance with the same uid and some
extra params. This implementation first calls Params.copy and
then make a copy of the companion Java pipeline component with
extra params. So both the Python wrapper and the Java pipeline
component get copied.
:param extra: Extra parameters to copy to the new instance
:return: Copy of this instance
"""
if extra is None:
extra = dict()
that = super(JavaParams, self).copy(extra)
if self._java_obj is not None:
that._java_obj = self._java_obj.copy(self._empty_java_param_map())
that._transfer_params_to_java()
return that
@inherit_doc
class JavaEstimator(JavaParams, Estimator):
......@@ -256,21 +279,3 @@ class JavaModel(JavaTransformer, Model):
super(JavaModel, self).__init__(java_model)
if java_model is not None:
self._resetUid(java_model.uid())
def copy(self, extra=None):
"""
Creates a copy of this instance with the same uid and some
extra params. This implementation first calls Params.copy and
then make a copy of the companion Java model with extra params.
So both the Python wrapper and the Java model get copied.
:param extra: Extra parameters to copy to the new instance
:return: Copy of this instance
"""
if extra is None:
extra = dict()
that = super(JavaModel, self).copy(extra)
if self._java_obj is not None:
that._java_obj = self._java_obj.copy(self._empty_java_param_map())
that._transfer_params_to_java()
return that
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment