Export - CSV (All fields)
Export - CSV (Current fields)
SPARK-8771
Actor system deprecation tag uses deprecated deprecation tag
SPARK-8769
toLocalIterator should mention it results in many jobs
Update comments that contain R statements in ml.logisticRegressionSuiteSPARK-8120
Typos in warning message in sql/types.py
Need to update download link for Hive 0.13.1 jars (HiveComparisonTest)
StringIndexerModel should have public constructor- SPARK-9490
MLlib evaluation metrics guide example python code uses deprecated print statement
LocalLDAModel logLikelihood
Add checkInputColumn back to Params and print more infoSPARK-9212
update Netty version to "4.0.29.Final" for Netty Metrics
Spelling error in Strategy.defaultStrategy
Join example fix in streaming-programming-guide.md
Methods in Unevaluable should be final- SPARK-9797
Document StreamingLinearRegressionWithSGD default parameter values
Require Maven 3.3.3+ in the buildSPARK-9657
PrefixSpan getMaxPatternLength should return an Int
Update BinaryClassificationEvaluator to use setRawPredictionCol
Make log messages in ExecutorRunnable more readable
Improve error message for decision trees when numExamples < maxCategoriesPerFeature
Fix definition of residual in LinearRegressionSummary
Tiny cleanups to Params, Pipeline
Remove highly unnecessary accumulators stringify methods
Rename configuration `spark.streaming.minRememberDuration` to `spark.streaming.fileStream.minRememberDuration`
Percent of pruned partitions is shown wrong
Use @deprecated instead of @Deprecated in Scala code
MinMaxScaler missing getters and docs
StopWordsRemover.stopWords should use StringArrayParam
RFormula are missing setters for featuresCol and labelCol
LDAUtils should be private
QRDecomposition should use QType and RType for type names instead of UType and VType
Spark UI Shows incorrect memory under Yarn
Update crosstab doc for pairs that have no occurrences
Improve the Spark Configuration document about `spark.kryoserializer.buffer`
SPARK_EXECUTOR_CORES is not read in SparkSubmit- SPARK-8457
Documentation for N-Gram feature transformer 
SPARK-5562
LDA should handle empty documents
Java unit test for PCA transformer
Potential issues with the constant hashCode
Spark Standalone master json API's worker number is not match web UI number
Add support for cutoff to RandomForestClassifier
Reduce latency between executor requests and RM heartbeat
Received block event count through the StreamingListener can be negative
Add a setting to disable SparkSQL parquet schema merge by using datasource API
Add spark-ec2 options to assign launched instances into IAM roles and to set instance-initiated shutdown behavior
Deprecate callUDF in favor of udf
Improve MLlib Local Matrix Documentation.
Simplify and clean up the checkpointing code
string function: translate
Make AsynchronousListenerBus.waitUntilEmpty throw TimeoutException if timeout
Shuffle code cleanup: remove BlockStoreShuffleFetcher class