This change adds some new utility code to handle shutdown hooks in
Spark. The main goal is to take advantage of Hadoop 2.x's API for
shutdown hooks, which allows Spark to register a hook that will
run before the one that cleans up HDFS clients, and thus avoids
some races that would cause exceptions to show up and other issues
such as failure to properly close event logs.
Unfortunately, Hadoop 1.x does not have such APIs, so in that case
correctness is still left to chance.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#5560 from vanzin/SPARK-6014 and squashes the following commits:
edfafb1 [Marcelo Vanzin] Better scaladoc.
fcaeedd [Marcelo Vanzin] Merge branch 'master' into SPARK-6014
e7039dc [Marcelo Vanzin] [SPARK-6014] [core] Revamp Spark shutdown hooks, fix shutdown races.
Starting with version 0.21.0, Apache Mesos is very noisy if the filter parameter refuse_seconds is set to an invalid value like `-1`.
I have seen systems with millions of log lines like
```
W0420 18:00:48.773059 32352 hierarchical_allocator_process.hpp:589] Using the default value of 'refuse_seconds' to create the refused resources filter because the input value is negative
```
in the Mesos master INFO and WARNING log files.
Therefore the CoarseMesosSchedulerBackend should set the default value for refuse seconds (i.e. 5 seconds) directly.
This is no problem for the fine-grained MesosSchedulerBackend, as it uses the value 1 second for this parameter.
Author: mweindel <m.weindel@usu-software.de>
Closes#5597 from MartinWeindel/master and squashes the following commits:
2f99ffd [mweindel] Avoid warning message about invalid refuse_seconds value in Mesos >=0.21.
add missing comma and space
Author: Alain <aihe@usc.edu>
Closes#5621 from AiHe/tree-node-issue and squashes the following commits:
159a7bb [Alain] [Minor][MLLIB] Fix a minor formatting bug in toString methods in Node.scala
(cherry picked from commit 4508f01890a723f80d631424ff8eda166a13a727)
Signed-off-by: Xiangrui Meng <meng@databricks.com>
SchemaRDD works with ALS.train in 1.2, so we should continue support DataFrames for compatibility. coderxiang
Author: Xiangrui Meng <meng@databricks.com>
Closes#5619 from mengxr/SPARK-7036 and squashes the following commits:
dfcaf5a [Xiangrui Meng] ALS.train should support DataFrames in PySpark
1. Use blas calls to find the dot product between two vectors.
2. Prevent re-computing the L2 norm of the given vector for each word in model.
Author: MechCoder <manojkumarsivaraj334@gmail.com>
Closes#5467 from MechCoder/spark-6065 and squashes the following commits:
dd0b0b2 [MechCoder] Preallocate wordVectors
ffc9240 [MechCoder] Minor
6b74c81 [MechCoder] Switch back to native blas calls
da1642d [MechCoder] Explicit types and indexing
64575b0 [MechCoder] Save indexedmap and a wordvecmat instead of matrix
fbe0108 [MechCoder] Made the following changes 1. Calculate norms during initialization. 2. Use Blas calls from linalg.blas
1350cf3 [MechCoder] [SPARK-6065] Optimize word2vec.findSynonynms using blas calls
The default java options make the call to GenerateMIMAIgnore take
forever to run since it's gc'ing all the time. Improve things by
setting the perm gen size / max heap size to larger values.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#5615 from vanzin/gen-mima-fix and squashes the following commits:
f44e921 [Marcelo Vanzin] [minor] [build] Set java options when generating mima ignores.
This patch modifies several shuffle-related code paths to share and re-use SerializerInstances instead of creating new ones. Some serializers, such as KryoSerializer or SqlSerializer, can be fairly expensive to create or may consume moderate amounts of memory, so it's probably best to avoid unnecessary serializer creation in hot code paths.
The key change in this patch is modifying `getDiskWriter()` / `DiskBlockObjectWriter` to accept `SerializerInstance`s instead of `Serializer`s (which are factories for instances). This allows the disk writer's creator to decide whether the serializer instance can be shared or re-used.
The rest of the patch modifies several write and read paths to use shared serializers. One big win is in `ShuffleBlockFetcherIterator`, where we used to create a new serializer per received block. Similarly, the shuffle write path used to create a new serializer per file even though in many cases only a single thread would be writing to a file at a time.
I made a small serializer reuse optimization in CoarseGrainedExecutorBackend as well, since it seemed like a small and obvious improvement.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#5606 from JoshRosen/SPARK-3386 and squashes the following commits:
f661ce7 [Josh Rosen] Remove thread local; add comment instead
64f8398 [Josh Rosen] Use ThreadLocal for serializer instance in CoarseGrainedExecutorBackend
aeb680e [Josh Rosen] [SPARK-3386] Reuse SerializerInstance in shuffle code paths
It's a bug while do query like:
```sql
select d from (select explode(array(1,1)) d from src limit 1) t
```
And it will throws exception like:
```
org.apache.spark.sql.AnalysisException: cannot resolve 'd' given input columns _c0; line 1 pos 7
at org.apache.spark.sql.catalyst.analysis.package$AnalysisErrorAt.failAnalysis(package.scala:42)
at org.apache.spark.sql.catalyst.analysis.CheckAnalysis$$anonfun$apply$3$$anonfun$apply$1.applyOrElse(CheckAnalysis.scala:48)
at org.apache.spark.sql.catalyst.analysis.CheckAnalysis$$anonfun$apply$3$$anonfun$apply$1.applyOrElse(CheckAnalysis.scala:45)
at org.apache.spark.sql.catalyst.trees.TreeNode$$anonfun$transformUp$1.apply(TreeNode.scala:250)
at org.apache.spark.sql.catalyst.trees.TreeNode$$anonfun$transformUp$1.apply(TreeNode.scala:250)
at org.apache.spark.sql.catalyst.trees.CurrentOrigin$.withOrigin(TreeNode.scala:50)
at org.apache.spark.sql.catalyst.trees.TreeNode.transformUp(TreeNode.scala:249)
at org.apache.spark.sql.catalyst.plans.QueryPlan.org$apache$spark$sql$catalyst$plans$QueryPlan$$transformExpressionUp$1(QueryPlan.scala:103)
at org.apache.spark.sql.catalyst.plans.QueryPlan$$anonfun$2$$anonfun$apply$2.apply(QueryPlan.scala:117)
at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244)
at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244)
at scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:47)
at scala.collection.TraversableLike$class.map(TraversableLike.scala:244)
at scala.collection.AbstractTraversable.map(Traversable.scala:105)
at org.apache.spark.sql.catalyst.plans.QueryPlan$$anonfun$2.apply(QueryPlan.scala:116)
at scala.collection.Iterator$$anon$11.next(Iterator.scala:328)
```
To solve the bug, it requires code refactoring for UDTF
The major changes are about:
* Simplifying the UDTF development, UDTF will manage the output attribute names any more, instead, the `logical.Generate` will handle that properly.
* UDTF will be asked for the output schema (data types) during the logical plan analyzing.
Author: Cheng Hao <hao.cheng@intel.com>
Closes#4602 from chenghao-intel/explode_bug and squashes the following commits:
c2a5132 [Cheng Hao] add back resolved for Alias
556e982 [Cheng Hao] revert the unncessary change
002c361 [Cheng Hao] change the rule of resolved for Generate
04ae500 [Cheng Hao] add qualifier only for generator output
5ee5d2c [Cheng Hao] prepend the new qualifier
d2e8b43 [Cheng Hao] Update the code as feedback
ca5e7f4 [Cheng Hao] shrink the commits
liancheng mengxr this is similar to #5146.
Author: Punya Biswal <pbiswal@palantir.com>
Closes#5578 from punya/feature/SPARK-6996 and squashes the following commits:
d56c3e0 [Punya Biswal] Fix imports
c7e308b [Punya Biswal] Support java iterable types in POJOs
5e00685 [Punya Biswal] Support map types in java beans
https://issues.apache.org/jira/browse/SPARK-6969
Author: Yin Huai <yhuai@databricks.com>
Closes#5583 from yhuai/refreshTableRefreshDataCache and squashes the following commits:
1e5142b [Yin Huai] Add todo.
92b2498 [Yin Huai] Minor updates.
367df92 [Yin Huai] Recache data in the command of REFRESH TABLE.
For `GetField` outside `UnresolvedAttribute`, we will throw exception in `Analyzer`.
Author: Wenchen Fan <cloud0fan@outlook.com>
Closes#5588 from cloud-fan/tmp and squashes the following commits:
7ac74d2 [Wenchen Fan] small refactor
It looked weird that up to now there was no way in Spark's Scala API to access fields of `DataFrame/sql.Row` by name, only by their index.
This tries to solve this issue.
Author: vidmantas zemleris <vidmantas@vinted.com>
Closes#5573 from vidma/features/row-with-named-fields and squashes the following commits:
6145ae3 [vidmantas zemleris] [SPARK-6994][SQL] Allow to fetch field values by name on Row
9564ebb [vidmantas zemleris] [SPARK-6994][SQL] Add fieldIndex to schema (StructType)
Since sparse matrices now support a isTransposed flag for row major data, DenseMatrices should do the same.
Author: MechCoder <manojkumarsivaraj334@gmail.com>
Closes#5455 from MechCoder/spark-6845 and squashes the following commits:
525c370 [MechCoder] minor
004a37f [MechCoder] Cast boolean to int
151f3b6 [MechCoder] [WIP] Add isTransposed to pickle DenseMatrix
cc0b90a [MechCoder] [SPARK-6845] Add isTranposed flag to DenseMatrix
SPARK-3276 Added a new configuration parameter ``spark.streaming.minRememberDuration``, with a default value of 1 minute.
So that when a Spark Streaming application is started, an arbitrary number of minutes can be taken as threshold for remembering.
Author: emres <emre.sevinc@gmail.com>
Closes#5438 from emres/SPARK-3276 and squashes the following commits:
766f938 [emres] SPARK-3276 Switched to using newly added getTimeAsSeconds method.
affee1d [emres] SPARK-3276 Changed the property name and variable name for minRememberDuration
c9d58ca [emres] SPARK-3276 Minor code re-formatting.
1c53ba9 [emres] SPARK-3276 Started to use ssc.conf rather than ssc.sparkContext.getConf, and also getLong method directly.
bfe0acb [emres] SPARK-3276 Moved the minRememberDurationMin to the class
daccc82 [emres] SPARK-3276 Changed the property name to reflect the unit of value and reduced number of fields.
43cc1ce [emres] SPARK-3276 Added a new configuration parameter spark.streaming.minRemember duration, with a default value of 1 minute.
CoGroupPartition, part of CoGroupedRDD, includes references to each RDD that the CoGroupedRDD narrowly depends on, and a reference to the ShuffleHandle. The partition is serialized separately from the RDD, so when the RDD and partition arrive on the worker, the references in the partition and in the RDD no longer point to the same object.
This is a relatively minor performance issue (the closure can be 2x larger than it needs to be because the rdds and partitions are serialized twice; see numbers below) but is more annoying as a developer issue (this is where I ran into): if any state is stored in the RDD or ShuffleHandle on the worker side, subtle bugs can appear due to the fact that the references to the RDD / ShuffleHandle in the RDD and in the partition point to separate objects. I'm not sure if this is enough of a potential future problem to fix this old and central part of the code, so hoping to get input from others here.
I did some simple experiments to see how much this effects closure size. For this example:
$ val a = sc.parallelize(1 to 10).map((_, 1))
$ val b = sc.parallelize(1 to 2).map(x => (x, 2*x))
$ a.cogroup(b).collect()
the closure was 1902 bytes with current Spark, and 1129 bytes after my change. The difference comes from eliminating duplicate serialization of the shuffle handle.
For this example:
$ val sortedA = a.sortByKey()
$ val sortedB = b.sortByKey()
$ sortedA.cogroup(sortedB).collect()
the closure was 3491 bytes with current Spark, and 1333 bytes after my change. Here, the difference comes from eliminating duplicate serialization of the two RDDs for the narrow dependencies.
The ShuffleHandle includes the ShuffleDependency, so this difference will get larger if a ShuffleDependency includes a serializer, a key ordering, or an aggregator (all set to None by default). It would also get bigger for a big RDD -- although I can't think of any examples where the RDD object gets large. The difference is not affected by the size of the function the user specifies, which (based on my understanding) is typically the source of large task closures.
Author: Kay Ousterhout <kayousterhout@gmail.com>
Closes#4145 from kayousterhout/SPARK-5360 and squashes the following commits:
85156c3 [Kay Ousterhout] Better comment the narrowDeps parameter
cff0209 [Kay Ousterhout] Fixed spelling issue
658e1af [Kay Ousterhout] [SPARK-5360] Eliminate duplicate objects in serialized CoGroupedRDD
A simple truncation in integer division (on rates over 1000 messages / second) causes the existing implementation to sleep for 0 milliseconds, then call itself recursively; this causes what is essentially an infinite recursion, since the base case of the calculated amount of time having elapsed can't be reached before available stack space is exhausted. A fix to this truncation error is included in this patch.
However, even with the defect patched, the accuracy of the existing implementation is abysmal (the error bounds of the original test were effectively [-30%, +10%], although this fact was obscured by hard-coded error margins); as such, when the error bounds were tightened down to [-5%, +5%], the existing implementation failed to meet the new, tightened, requirements. Therefore, an industry-vetted solution (from Guava) was used to get the adapted tests to pass.
Author: David McGuire <david.mcguire2@nike.com>
Closes#5559 from dmcguire81/master and squashes the following commits:
d29d2e0 [David McGuire] Back out to +/-5% error margins, for flexibility in timing
8be6934 [David McGuire] Fix spacing per code review
90e98b9 [David McGuire] Address scalastyle errors
29011bd [David McGuire] Further ratchet down the error margins
b33b796 [David McGuire] Eliminate dependency on even distribution by BlockGenerator
8f2934b [David McGuire] Remove arbitrary thread timing / cooperation code
70ee310 [David McGuire] Use Thread.yield(), since Thread.sleep(0) is system-dependent
82ee46d [David McGuire] Replace guard clause with nested conditional
2794717 [David McGuire] Replace the RateLimiter with the Guava implementation
38f3ca8 [David McGuire] Ratchet down the error rate to +/- 5%; tests fail
24b1bc0 [David McGuire] Fix truncation in integer division causing infinite recursion
d6e1079 [David McGuire] Stack overflow error in RateLimiter on rates over 1000/s
Model import/export for IsotonicRegression
Author: Yanbo Liang <ybliang8@gmail.com>
Closes#5270 from yanboliang/spark-5990 and squashes the following commits:
872028d [Yanbo Liang] fix code style
f80ec1b [Yanbo Liang] address comments
49600cc [Yanbo Liang] address comments
429ff7d [Yanbo Liang] store each interval as a record
2b2f5a1 [Yanbo Liang] Model import/export for IsotonicRegression
This PR enable auto_convert in JavaGateway, then we could register a converter for a given types, for example, date and datetime.
There are two bugs related to auto_convert, see [1] and [2], we workaround it in this PR.
[1] https://github.com/bartdag/py4j/issues/160
[2] https://github.com/bartdag/py4j/issues/161
cc rxin JoshRosen
Author: Davies Liu <davies@databricks.com>
Closes#5570 from davies/py4j_date and squashes the following commits:
eb4fa53 [Davies Liu] fix tests in python 3
d17d634 [Davies Liu] rollback changes in mllib
2e7566d [Davies Liu] convert tuple into ArrayList
ceb3779 [Davies Liu] Update rdd.py
3c373f3 [Davies Liu] support date and datetime by auto_convert
cb094ff [Davies Liu] enable auto convert
Deprecated `spark.akka.num.retries`, `spark.akka.retry.wait`, `spark.akka.askTimeout`, `spark.akka.lookupTimeout`, and added `spark.rpc.num.retries`, `spark.rpc.retry.wait`, `spark.rpc.askTimeout`, `spark.rpc.lookupTimeout`.
Author: zsxwing <zsxwing@gmail.com>
Closes#5595 from zsxwing/SPARK-6490 and squashes the following commits:
e0d80a9 [zsxwing] Use getTimeAsMs and getTimeAsSeconds and other minor fixes
31dbe69 [zsxwing] Add spark.rpc.* and deprecate spark.akka.*
JIRA https://issues.apache.org/jira/browse/SPARK-6635
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#5541 from viirya/replace_with_column and squashes the following commits:
b539c7b [Liang-Chi Hsieh] For comment.
72f35b1 [Liang-Chi Hsieh] DataFrame.withColumn can replace original column with identical column name.
The commit message is pretty self-explanatory.
Author: BenFradet <benjamin.fradet@gmail.com>
Closes#5600 from BenFradet/master and squashes the following commits:
108492d [BenFradet] [doc][streaming] Fixed broken link in mllib section
The contribution is my original work. I license the work to the project under the project's open source license.
Small typo in the programming guide.
Author: Eric Chiang <eric.chiang.m@gmail.com>
Closes#5599 from ericchiang/docs-typo and squashes the following commits:
1177942 [Eric Chiang] fixed doc
It should load from `testInput` instead of `input` for test data.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#5594 from viirya/use_testinput and squashes the following commits:
5e8b174 [Liang-Chi Hsieh] Fix style.
b60b475 [Liang-Chi Hsieh] Use testInput.
Author: Elisey Zanko <elisey.zanko@gmail.com>
Closes#5361 from 31z4/spark-6661 and squashes the following commits:
73c5d79 [Elisey Zanko] Python type errors should print type, not object
Currently we rely on the assumption that an exception will be raised and the channel closed if two endpoints cannot communicate over a Netty TCP channel. However, this guarantee does not hold in all network environments, and [SPARK-6962](https://issues.apache.org/jira/browse/SPARK-6962) seems to point to a case where only the server side of the connection detected a fault.
This patch improves robustness of fetch/rpc requests by having an explicit timeout in the transport layer which closes the connection if there is a period of inactivity while there are outstanding requests.
NB: This patch is actually only around 50 lines added if you exclude the testing-related code.
Author: Aaron Davidson <aaron@databricks.com>
Closes#5584 from aarondav/timeout and squashes the following commits:
8699680 [Aaron Davidson] Address Reynold's comments
37ce656 [Aaron Davidson] [SPARK-7003] Improve reliability of connection failure detection between Netty block transfer service endpoints
The current implementation call the default constructor of mllib.feature.StandarScaler without the possibility to specify withMean or withStd options.
Author: jrabary <Jaonary@gmail.com>
Closes#4704 from jrabary/master and squashes the following commits:
fae8568 [jrabary] style fix
8896b0e [jrabary] Comments fix
ef96d73 [jrabary] style fix
8e52607 [jrabary] style fix
edd9d48 [jrabary] Fix default param initialization
17e1a76 [jrabary] Fix default param initialization
298f405 [jrabary] Typo fix
45ed914 [jrabary] Add withMean and withStd params to StandarScaler
* Fix the page title in Isotonic regression documents (Naive Bayes -> Isotonic regression)
* Add a newline character at the end of the file
Author: dobashim <dobashim@oss.nttdata.co.jp>
Closes#5581 from dobashim/master and squashes the following commits:
d54a041 [dobashim] Fix typo of the page title in Isotonic regression documents
Title says it all.
cc rxin tdas
Author: zsxwing <zsxwing@gmail.com>
Closes#5554 from zsxwing/SPARK-6979 and squashes the following commits:
5304350 [zsxwing] Fix NotSerializableException
e9d3479 [zsxwing] Add blank lines
633e279 [zsxwing] Fix NotSerializableException
e496ace [zsxwing] Replace JobGenerator.eventActor with EventLoop
ec6ec58 [zsxwing] Fix the import order
ce0fa73 [zsxwing] Replace JobScheduler.eventActor with EventLoop
A subtask of [SPARK-5293](https://issues.apache.org/jira/browse/SPARK-5293)
Author: zsxwing <zsxwing@gmail.com>
Closes#5557 from zsxwing/SPARK-6983 and squashes the following commits:
e777e9f [zsxwing] Update ReceiverTrackerActor to use the new Rpc interface
If `StreamingKMeans` is not `Serializable`, we cannot do checkpoint for applications that using `StreamingKMeans`. So we should make it `Serializable`.
Author: zsxwing <zsxwing@gmail.com>
Closes#5582 from zsxwing/SPARK-6998 and squashes the following commits:
67c2a14 [zsxwing] Make StreamingKMeans 'Serializable'
cc andrewor14
Author: GuoQiang Li <witgo@qq.com>
Closes#5548 from witgo/SPARK-6963 and squashes the following commits:
964aea7 [GuoQiang Li] review commits
b08b3c9 [GuoQiang Li] Flaky test: o.a.s.ContextCleanerSuite automatically cleanup checkpoint
The default method will use Guava's Ordering instead of
java.util.Comparator.naturalOrder() because it's not available
in Java 7, only in Java 8.
Author: Olivier Girardot <o.girardot@lateral-thoughts.com>
Closes#5571 from ogirardot/master and squashes the following commits:
7fe2e9e [Olivier Girardot] SPARK-6993 : Add default min, max methods for JavaDoubleRDD
Just fixed a doc.
Author: Gaurav Nanda <gaurav324@gmail.com>
Closes#5576 from gaurav324/master and squashes the following commits:
8a7323f [Gaurav Nanda] Fixed doc
Per the discussion in the comments on [this commit](f17d43b033 (commitcomment-10780649)), this PR allows the Python lint script to reuse `pep8.py` when possible.
Author: Nicholas Chammas <nicholas.chammas@gmail.com>
Closes#5561 from nchammas/save-dem-pep8-bytes and squashes the following commits:
b7c91e6 [Nicholas Chammas] reuse pep8.py
My previous fix (force a selector wakeup) didn't seem to work since
I ran into the hang again. So change the code a bit to be more
explicit about the condition when the selector thread should exit.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#5566 from vanzin/conn-mgr-hang and squashes the following commits:
ddb2c03 [Marcelo Vanzin] [core] [minor] Make sure ConnectionManager stops.
This patch is fixing the Java examples for Spark SQL when defining
programmatically a Schema and mapping Rows.
Author: Olivier Girardot <o.girardot@lateral-thoughts.com>
Closes#5569 from ogirardot/branch-1.3 and squashes the following commits:
c29e58d [Olivier Girardot] SPARK-6992 : Fix documentation example for Spark SQL on StructType
(cherry picked from commit c9b1ba4b16a7afe93d45bf75b128cc0dd287ded0)
Signed-off-by: Reynold Xin <rxin@databricks.com>
`numExecutors` checking is failed when dynamic allocation is enabled with default configuration. Details can be seen is [SPARK-6975](https://issues.apache.org/jira/browse/SPARK-6975). sryza, please help me to review this, not sure is this the correct way, I think previous you change this part :)
Author: jerryshao <saisai.shao@intel.com>
Closes#5551 from jerryshao/SPARK-6975 and squashes the following commits:
4335da1 [jerryshao] Change according to the comments
77bdcbd [jerryshao] Fix argument validation error
I didn't find many deprecated configs after a grep-based search,
but the ones I could find were moved to the centralized location
in SparkConf.
While there, I deprecated a couple more HS configs that mentioned
time units.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#5562 from vanzin/SPARK-5933 and squashes the following commits:
dcb617e7 [Marcelo Vanzin] [SPARK-5933] [core] Move config deprecation warnings to SparkConf.
- Defined executorCores from "spark.mesos.executor.cores"
- Changed the amount of mesosExecutor's cores to executorCores.
- Added new configuration option on running-on-mesos.md
Author: Jongyoul Lee <jongyoul@gmail.com>
Closes#5063 from jongyoul/SPARK-6350 and squashes the following commits:
9238d6e [Jongyoul Lee] [SPARK-6350][Mesos] Make mesosExecutorCores configurable in mesos "fine-grained" mode - Fixed docs - Changed configuration name - Made mesosExecutorCores private
2d41241 [Jongyoul Lee] [SPARK-6350][Mesos] Make mesosExecutorCores configurable in mesos "fine-grained" mode - Fixed docs
89edb4f [Jongyoul Lee] [SPARK-6350][Mesos] Make mesosExecutorCores configurable in mesos "fine-grained" mode - Fixed docs
8ba7694 [Jongyoul Lee] [SPARK-6350][Mesos] Make mesosExecutorCores configurable in mesos "fine-grained" mode - Fixed docs
7549314 [Jongyoul Lee] [SPARK-6453][Mesos] Some Mesos*Suite have a different package with their classes - Fixed docs
4ae7b0c [Jongyoul Lee] [SPARK-6453][Mesos] Some Mesos*Suite have a different package with their classes - Removed TODO
c27efce [Jongyoul Lee] [SPARK-6453][Mesos] Some Mesos*Suite have a different package with their classes - Fixed Mesos*Suite for supporting integer WorkerOffers - Fixed Documentation
1fe4c03 [Jongyoul Lee] [SPARK-6453][Mesos] Some Mesos*Suite have a different package with their classes - Change available resources of cpus to integer value beacuse WorkerOffer support the amount cpus as integer value
5f3767e [Jongyoul Lee] Revert "[SPARK-6350][Mesos] Make mesosExecutorCores configurable in mesos "fine-grained" mode"
4b7c69e [Jongyoul Lee] [SPARK-6350][Mesos] Make mesosExecutorCores configurable in mesos "fine-grained" mode - Changed configruation name and description from "spark.mesos.executor.cores" to "spark.executor.frameworkCores"
0556792 [Jongyoul Lee] [SPARK-6350][Mesos] Make mesosExecutorCores configurable in mesos "fine-grained" mode - Defined executorCores from "spark.mesos.executor.cores" - Changed the amount of mesosExecutor's cores to executorCores. - Added new configuration option on running-on-mesos.md
I've added a static getOrCreate method to the static SparkContext object that allows one to either retrieve a previously created SparkContext or to instantiate a new one with the provided config. The method accepts an optional SparkConf to make usage intuitive.
Still working on a test for this, basically want to create a new context from scratch, then ensure that subsequent calls don't overwrite that.
Author: Ilya Ganelin <ilya.ganelin@capitalone.com>
Closes#5501 from ilganeli/SPARK-6703 and squashes the following commits:
db9a963 [Ilya Ganelin] Closing second spark context
1dc0444 [Ilya Ganelin] Added ref equality check
8c884fa [Ilya Ganelin] Made getOrCreate synchronized
cb0c6b7 [Ilya Ganelin] Doc updates and code cleanup
270cfe3 [Ilya Ganelin] [SPARK-6703] Documentation fixes
15e8dea [Ilya Ganelin] Updated comments and added MiMa Exclude
0e1567c [Ilya Ganelin] Got rid of unecessary option for AtomicReference
dfec4da [Ilya Ganelin] Changed activeContext to AtomicReference
733ec9f [Ilya Ganelin] Fixed some bugs in test code
8be2f83 [Ilya Ganelin] Replaced match with if
e92caf7 [Ilya Ganelin] [SPARK-6703] Added test to ensure that getOrCreate both allows creation, retrieval, and a second context if desired
a99032f [Ilya Ganelin] Spacing fix
d7a06b8 [Ilya Ganelin] Updated SparkConf class to add getOrCreate method. Started test suite implementation
As a follow up PR to #5544.
cc davies
Author: Reynold Xin <rxin@databricks.com>
Closes#5558 from rxin/sort-doc-improvement and squashes the following commits:
f4c276f [Reynold Xin] Review feedback.
d2dcf24 [Reynold Xin] Minor fix to SPARK-6958: Improve Python docstring for DataFrame.sort.
This patch includes :
* adding how to use map after an sql query using javaRDD
* fixing the first few java examples that were written in Scala
Thank you for your time,
Olivier.
Author: Olivier Girardot <o.girardot@lateral-thoughts.com>
Closes#5564 from ogirardot/branch-1.3 and squashes the following commits:
9f8d60e [Olivier Girardot] SPARK-6988 : Fix documentation regarding DataFrames using the Java API
(cherry picked from commit 6b528dc139da594ef2e651d84bd91fe0f738a39d)
Signed-off-by: Reynold Xin <rxin@databricks.com>
This is a PR for cleaning up and finalizing the DecisionTree API. PRs for ensembles will follow once this is merged.
### Goal
Here is the description copied from the JIRA (for both trees and ensembles):
> **Issue**: The APIs for DecisionTree and ensembles (RandomForests and GradientBoostedTrees) have been experimental for a long time. The API has become very convoluted because trees and ensembles have many, many variants, some of which we have added incrementally without a long-term design.
> **Proposal**: This JIRA is for discussing changes required to finalize the APIs. After we discuss, I will make a PR to update the APIs and make them non-Experimental. This will require making many breaking changes; see the design doc for details.
> **[Design doc](https://docs.google.com/document/d/1rJ_DZinyDG3PkYkAKSsQlY0QgCeefn4hUv7GsPkzBP4)** : This outlines current issues and the proposed API.
Overall code layout:
* The old API in mllib.tree.* will remain the same.
* The new API will reside in ml.classification.* and ml.regression.*
### Summary of changes
Old API
* Exactly the same, except I made 1 method in Loss private (but that is not a breaking change since that method was introduced after the Spark 1.3 release).
New APIs
* Under Pipeline API
* The new API preserves functionality, except:
* New API does NOT store prob (probability of label in classification). I want to have it store the full vector of probabilities but feel that should be in a later PR.
* Use abstractions for parameters, estimators, and models to avoid code duplication
* Limit parameters to relevant algorithms
* For enum-like types, only expose Strings
* We can make these pluggable later on by adding new parameters. That is a far-future item.
Test suites
* I organized DecisionTreeSuite, but I made absolutely no changes to the tests themselves.
* The test suites for the new API only test (a) similarity with the results of the old API and (b) elements of the new API.
* After code is moved to this new API, we should move the tests from the old suites which test the internals.
### Details
#### Changed names
Parameters
* useNodeIdCache -> cacheNodeIds
#### Other changes
* Split: Changed categories to set instead of list
#### Non-decision tree changes
* AttributeGroup
* Added parentheses to toMetadata, toStructField methods (These were removed in a previous PR, but I ran into 1 issue with the Scala compiler not being able to disambiguate between a toMetadata method with no parentheses and a toMetadata method which takes 1 argument.)
* Attributes
* Renamed: toMetadata -> toMetadataImpl
* Added toMetadata methods which return ML metadata (keyed with “ML_ATTR”)
* NominalAttribute: Added getNumValues method which examines both numValues and values.
* Params.inheritValues: Checks whether the parent param really belongs to the child (to allow Estimator-Model pairs with different sets of parameters)
### Questions for reviewers
* Is "DecisionTreeClassificationModel" too long a name?
* Is this OK in the docs?
```
class DecisionTreeRegressor extends TreeRegressor[DecisionTreeRegressionModel] with DecisionTreeParams[DecisionTreeRegressor] with TreeRegressorParams[DecisionTreeRegressor]
```
### Future
We should open up the abstractions at some point. E.g., it would be useful to be able to set tree-related parameters in 1 place and then pass those to multiple tree-based algorithms.
Follow-up JIRAs will be (in this order):
* Tree ensembles
* Deprecate old tree code
* Move DecisionTree implementation code to new API.
* Move tests from the old suites which test the internals.
* Update programming guide
* Python API
* Change RandomForest* to always use bootstrapping, even when numTrees = 1
* Provide the probability of the predicted label for classification. After we move code to the new API and update it to maintain probabilities for all labels, then we can add the probabilities to the new API.
CC: mengxr manishamde codedeft chouqin MechCoder
Author: Joseph K. Bradley <joseph@databricks.com>
Closes#5530 from jkbradley/dt-api-dt and squashes the following commits:
6aae255 [Joseph K. Bradley] Changed tree abstractions not to take type parameters, and for setters to return this.type instead
ec17947 [Joseph K. Bradley] Updates based on code review. Main changes were: moving public types from ml.impl.tree to ml.tree, modifying CategoricalSplit to take an Array of categories but store a Set internally, making more types sealed or final
5626c81 [Joseph K. Bradley] style fixes
f8fbd24 [Joseph K. Bradley] imported reorg of DecisionTreeSuite from old PR. small cleanups
7ef63ed [Joseph K. Bradley] Added DecisionTreeRegressor, test suites, and example (for real this time)
e11673f [Joseph K. Bradley] Added DecisionTreeRegressor, test suites, and example
119f407 [Joseph K. Bradley] added DecisionTreeClassifier example
0bdc486 [Joseph K. Bradley] fixed issues after param PR was merged
f9fbb60 [Joseph K. Bradley] Done with DecisionTreeClassifier, but no save/load yet. Need to add example as well
2532c9a [Joseph K. Bradley] partial move to spark.ml API, not done yet
c72c1a0 [Joseph K. Bradley] Copied changes for common items, plus DecisionTreeClassifier from original PR
Currently, when Spark launches the Yarn AM, the process will use
the local Hadoop configuration on the node where the AM launches,
if one is present. A more correct approach is to use the same
configuration used to launch the Spark job, since the user may
have made modifications (such as adding app-specific configs).
The approach taken here is to use the distributed cache to make
all files in the Hadoop configuration directory available to the
AM. This is a little overkill since only the AM needs them (the
executors use the broadcast Hadoop configuration from the driver),
but is the easier approach.
Even though only a few files in that directory may end up being
used, all of them are uploaded. This allows supporting use cases
such as when auxiliary configuration files are used for SSL
configuration, or when uploading a Hive configuration directory.
Not all of these may be reflected in a o.a.h.conf.Configuration object,
but may be needed when a driver in cluster mode instantiates, for
example, a HiveConf object instead.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#4142 from vanzin/SPARK-2669 and squashes the following commits:
f5434b9 [Marcelo Vanzin] Merge branch 'master' into SPARK-2669
013f0fb [Marcelo Vanzin] Review feedback.
f693152 [Marcelo Vanzin] Le sigh.
ed45b7d [Marcelo Vanzin] Zip all config files and upload them as an archive.
5927b6b [Marcelo Vanzin] Merge branch 'master' into SPARK-2669
cbb9fb3 [Marcelo Vanzin] Remove stale test.
e3e58d0 [Marcelo Vanzin] Merge branch 'master' into SPARK-2669
e3d0613 [Marcelo Vanzin] Review feedback.
34bdbd8 [Marcelo Vanzin] Fix test.
022a688 [Marcelo Vanzin] Merge branch 'master' into SPARK-2669
a77ddd5 [Marcelo Vanzin] Merge branch 'master' into SPARK-2669
79221c7 [Marcelo Vanzin] [SPARK-2669] [yarn] Distribute client configuration to AM.
```
select(['cola', 'colb'])
groupby(['colA', 'colB'])
groupby([df.colA, df.colB])
df.sort('A', ascending=True)
df.sort(['A', 'B'], ascending=True)
df.sort(['A', 'B'], ascending=[1, 0])
```
cc rxin
Author: Davies Liu <davies@databricks.com>
Closes#5544 from davies/compatibility and squashes the following commits:
4944058 [Davies Liu] add docstrings
adb2816 [Davies Liu] Merge branch 'master' of github.com:apache/spark into compatibility
bcbbcab [Davies Liu] support ascending as list
8dabdf0 [Davies Liu] improve API compatibility to pandas
In driver now will start a server socket and use a wildcard ip, use 127.0.0.0 is more reasonable, as we only use it by local Python process.
/cc davies
Author: linweizhong <linweizhong@huawei.com>
Closes#5256 from Sephiroth-Lin/SPARK-6604 and squashes the following commits:
7b3c633 [linweizhong] rephrase
sbin/spark-daemon.sh used
ps -p "$TARGET_PID" -o args=
to figure out whether the process running with the expected PID is actually a Spark
daemon. When running with a large classpath, the output of ps gets
truncated and the check fails spuriously.
This weakens the check to see if it's a java command (which is something
we do in other parts of the script) rather than looking for the specific
main class name. This means that SPARK-4832 might happen under a
slightly broader range of circumstances (a java program happened to
reuse the same PID), but it seems worthwhile compared to failing
consistently with a large classpath.
Author: Punya Biswal <pbiswal@palantir.com>
Closes#5535 from punya/feature/SPARK-6952 and squashes the following commits:
7ea12d1 [Punya Biswal] Handle long args when detecting PID reuse
