There is only a single `stat.py` file for the `mllib.stat` package. We recently added `MultivariateGaussian` under `mllib.stat.distribution` in Scala/Java. It would be nice to refactor `stat.py` and make it easy to expand. Note that `ChiSqTestResult` is moved from `mllib.stat` to `mllib.stat.test`. The latter is used in Scala/Java. It is only used in the return value of `Statistics.chiSqTest`, so this should be an okay change.

davies

Author: Xiangrui Meng <meng@databricks.com>

Closes #4266 from mengxr/py-stat-refactor and squashes the following commits:

1a5e1db [Xiangrui Meng] refactor stat.py

1. Added foreach, foreachPartition, flatMap to DataFrame.
2. Added col() in dsl.
3. Support renaming columns in toDataFrame.
4. Support type inference on arrays (in addition to Seq).
5. Updated mllib to use the new DSL.

Author: Reynold Xin <rxin@databricks.com>

Closes #4260 from rxin/sql-dsl-update and squashes the following commits:

73466c1 [Reynold Xin] Fixed LogisticRegression. Also added better error message for resolve.
fab3ccc [Reynold Xin] Bug fix.
d31fcd2 [Reynold Xin] Style fix.
62608c4 [Reynold Xin] [SQL] Various DataFrame DSL update.

The conversion methods for `BlockMatrix`. Conversions go through `CoordinateMatrix` in order to cause a shuffle so that intermediate operations will be stored on disk and the expensive initial computation will be mitigated.

Author: Burak Yavuz <brkyvz@gmail.com>

Closes #4256 from brkyvz/SPARK-3977PR and squashes the following commits:

4df37fe [Burak Yavuz] moved TODO inside code block
b049c07 [Burak Yavuz] addressed code review feedback v1
66cb755 [Burak Yavuz] added default toBlockMatrix conversion
851f2a2 [Burak Yavuz] added better comments and checks
cdb9895 [Burak Yavuz] [SPARK-3977] Conversion methods for BlockMatrix to other Distributed Matrices

Also removed the literal implicit transformation since it is pretty scary for API design. Instead, created a new lit method for creating literals. This doesn't break anything from a compatibility perspective because Literal was added two days ago.

Author: Reynold Xin <rxin@databricks.com>

Closes #4241 from rxin/df-docupdate and squashes the following commits:

c0f4810 [Reynold Xin] Fix Python merge conflict.
094c7d7 [Reynold Xin] Minor style fix. Reset Python tests.
3c89f4a [Reynold Xin] Package.
dfe6962 [Reynold Xin] Updated Python aggregate.
5dd4265 [Reynold Xin] Made dsl Java callable.
14b3c27 [Reynold Xin] Fix literal expression for symbols.
68b31cb [Reynold Xin] Literal.
4cfeb78 [Reynold Xin] [SPARK-5097][SQL] Address DataFrame code review feedback.

We have seen many use cases of `treeAggregate`/`treeReduce` outside the ML domain. Maybe it is time to move them to Core. pwendell

Author: Xiangrui Meng <meng@databricks.com>

Closes #4228 from mengxr/SPARK-5430 and squashes the following commits:

20ad40d [Xiangrui Meng] exclude tree* from mima
e89a43e [Xiangrui Meng] fix compile and update java doc
3ae1a4b [Xiangrui Meng] add treeReduce/treeAggregate to Python
6f948c5 [Xiangrui Meng] add treeReduce/treeAggregate to JavaRDDLike
d600b6c [Xiangrui Meng] move treeReduce and treeAggregate to core

This PR adds Python API for ML pipeline and parameters. The design doc can be found on the JIRA page. It includes transformers and an estimator to demo the simple text classification example code.

TODO:
- [x] handle parameters in LRModel
- [x] unit tests
- [x] missing some docs

CC: davies jkbradley

Author: Xiangrui Meng <meng@databricks.com>
Author: Davies Liu <davies@databricks.com>

Closes #4151 from mengxr/SPARK-4586 and squashes the following commits:

415268e [Xiangrui Meng] remove inherit_doc from __init__
edbd6fe [Xiangrui Meng] move Identifiable to ml.util
44c2405 [Xiangrui Meng] Merge pull request #2 from davies/ml
dd1256b [Xiangrui Meng] Merge remote-tracking branch 'apache/master' into SPARK-4586
14ae7e2 [Davies Liu] fix docs
54ca7df [Davies Liu] fix tests
78638df [Davies Liu] Merge branch 'SPARK-4586' of github.com:mengxr/spark into ml
fc59a02 [Xiangrui Meng] Merge remote-tracking branch 'apache/master' into SPARK-4586
1dca16a [Davies Liu] refactor
090b3a3 [Davies Liu] Merge branch 'master' of github.com:apache/spark into ml
0882513 [Xiangrui Meng] update doc style
a4f4dbf [Xiangrui Meng] add unit test for LR
7521d1c [Xiangrui Meng] add unit tests to HashingTF and Tokenizer
ba0ba1e [Xiangrui Meng] add unit tests for pipeline
0586c7b [Xiangrui Meng] add more comments to the example
5153cff [Xiangrui Meng] simplify java models
036ca04 [Xiangrui Meng] gen numFeatures
46fa147 [Xiangrui Meng] update mllib/pom.xml to include python files in the assembly
1dcc17e [Xiangrui Meng] update code gen and make param appear in the doc
f66ba0c [Xiangrui Meng] make params a property
d5efd34 [Xiangrui Meng] update doc conf and move embedded param map to instance attribute
f4d0fe6 [Xiangrui Meng] use LabeledDocument and Document in example
05e3e40 [Xiangrui Meng] update example
d3e8dbe [Xiangrui Meng] more docs optimize pipeline.fit impl
56de571 [Xiangrui Meng] fix style
d0c5bb8 [Xiangrui Meng] a working copy
bce72f4 [Xiangrui Meng] Merge remote-tracking branch 'apache/master' into SPARK-4586
17ecfb9 [Xiangrui Meng] code gen for shared params
d9ea77c [Xiangrui Meng] update doc
c18dca1 [Xiangrui Meng] make the example working
dadd84e [Xiangrui Meng] add base classes and docs
a3015cf [Xiangrui Meng] add Estimator and Transformer
46eea43 [Xiangrui Meng] a pipeline in python
33b68e0 [Xiangrui Meng] a working LR

SerDeUtil.pairRDDToPython and SerDeUtil.pythonToPairRDD now both support empty RDDs by checking the result of take(1) instead of calling first which throws an exception.

Author: Michael Nazario <mnazario@palantir.com>

Closes #4236 from mnazario/feature/empty-first and squashes the following commits:

a531c0c [Michael Nazario] Added regression tests for SPARK-5441
e3b2fb6 [Michael Nazario] Added acceptance of the empty case

This PR is based on #3255 , fix conflicts and code style.

Closes #3255.

Author: Yandu Oppacher <yandu.oppacher@jadedpixel.com>
Author: Davies Liu <davies@databricks.com>

Closes #3901 from davies/refactor-python-profile-code and squashes the following commits:

b4a9306 [Davies Liu] fix tests
4b79ce8 [Davies Liu] add docstring for profiler_cls
2700e47 [Davies Liu] use BasicProfiler as default
349e341 [Davies Liu] more refactor
6a5d4df [Davies Liu] refactor and fix tests
31bf6b6 [Davies Liu] fix code style
0864b5d [Yandu Oppacher] Remove unused method
76a6c37 [Yandu Oppacher] Added a profile collector to accumulate the profilers per stage
9eefc36 [Yandu Oppacher] Fix doc
9ace076 [Yandu Oppacher] Refactor of profiler, and moved tests around
8739aff [Yandu Oppacher] Code review fixes
9bda3ec [Yandu Oppacher] Refactor profiler code

This happens inside SparkEnv initialization as of #4194

Author: Ryan Williams <ryan.blake.williams@gmail.com>

Closes #4213 from ryan-williams/exec-id-set and squashes the following commits:

b3e4f7b [Ryan Williams] Remove redundant executor-id set() call

Fixes [SPARK-5434](https://issues.apache.org/jira/browse/SPARK-5434).

Simple demonstration of the problem and the fix:

```
$ spacey_path="/path/with some/spaces"
$ dirname $spacey_path
usage: dirname path
$ echo $?
1
$ dirname "$spacey_path"
/path/with some
$ echo $?
0
```

Author: Nicholas Chammas <nicholas.chammas@gmail.com>

Closes #4224 from nchammas/patch-1 and squashes the following commits:

960711a [Nicholas Chammas] [EC2] Preserve spaces in EC2 path

In DriverSuite, we currently set a timeout of 60 seconds. If after this time the process has not terminated, we leak the process because we never destroy it.

In SparkSubmitSuite, we currently do not have a timeout so the test can hang indefinitely.

Author: Andrew Or <andrew@databricks.com>

Closes #4230 from andrewor14/fix-driver-suite and squashes the following commits:

f5c80fd [Andrew Or] Fix timeout behaviors in both suites
8092c36 [Andrew Or] Stop SparkContext after every individual test

[SPARK-4955]With executor dynamic scaling enabled,executor shoude be added or killed in yarn-cluster mode.

With executor dynamic scaling enabled, executor number shoude be added or killed in yarn-cluster mode.so in yarn-cluster mode, ApplicationMaster start a AMActor that add or kill a executor. then YarnSchedulerActor  in YarnSchedulerBackend send message to am's AMActor.
andrewor14 ChengXiangLi tdas

Author: lianhuiwang <lianhuiwang09@gmail.com>

Closes #3962 from lianhuiwang/SPARK-4955 and squashes the following commits:

48d9ebb [lianhuiwang] update with andrewor14's comments
12426af [lianhuiwang] refactor am's code
45da3b0 [lianhuiwang] remove unrelated code
9318fc1 [lianhuiwang] update with andrewor14's comments
08ba473 [lianhuiwang] address andrewor14's comments
265c36d [lianhuiwang] fix small change
f43bda8 [lianhuiwang] fix address andrewor14's comments
7a7767a [lianhuiwang] fix address andrewor14's comments
bbc4d5a [lianhuiwang] address andrewor14's comments
1b029a4 [lianhuiwang] fix bug
7d33791 [lianhuiwang] in AM create a new actorSystem
2164ea8 [lianhuiwang] fix a min bug
6dfeeec [lianhuiwang] in yarn-cluster mode,executor number can be added or killed.

Since Java and Scala both have access to iterate over partitions via the "toLocalIterator" function, python should also have that same ability.

Author: Michael Nazario <mnazario@palantir.com>

Closes #4237 from mnazario/feature/toLocalIterator and squashes the following commits:

1c58526 [Michael Nazario] Fix documentation off by one error
0cdc8f8 [Michael Nazario] Add toLocalIterator to PySpark

SPARK-1934 [CORE] "this" reference escape to "selectorThread" during construction in ConnectionManager

This change reshuffles the order of initialization in `ConnectionManager` so that the last thing that happens is running `selectorThread`, which invokes a method that relies on object state in `ConnectionManager`

zsxwing also reported a similar problem in `BlockManager` in the JIRA, but I can't find a similar pattern there. Maybe it was subsequently fixed?

Author: Sean Owen <sowen@cloudera.com>

Closes #4225 from srowen/SPARK-1934 and squashes the following commits:

c4dec3b [Sean Owen] Init all object state in ConnectionManager constructor before starting thread in constructor that accesses object's state

When we use `make-distribution.sh` with `--with-tachyon` option, Tachyon will be downloaded by `wget` command but some systems don't have `wget` by default (MacOS X doesn't have).
Other scripts like build/mvn, build/sbt support not only `wget` but also `curl` so `make-distribution.sh` should support `curl` too.

Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>

Closes #3988 from sarutak/SPARK-5188 and squashes the following commits:

0f546e0 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-5188
010e884 [Kousuke Saruta] Merge branch 'SPARK-5188' of github.com:sarutak/spark into SPARK-5188
163687e [Kousuke Saruta] Fixed a merge conflict
e24e01b [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-5188
3daf1f1 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-5188
3caa4cb [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-5188
7cc8255 [Kousuke Saruta] Fix to use \$MVN instead of mvn
a3e908b [Kousuke Saruta] Fixed style
2db9fbf [Kousuke Saruta] Removed redirection from the logic which checks the existence of commands
1e4c7e0 [Kousuke Saruta] Used "command" command instead of "type" command
83b49b5 [Kousuke Saruta] Modified make-distribution.sh so that we use curl, not only wget to get tachyon

Author: Sandy Ryza <sandy@cloudera.com>

Closes #4251 from sryza/sandy-spark-5458 and squashes the following commits:

460827a [Sandy Ryza] Python too
d2dc160 [Sandy Ryza] SPARK-5458. Refer to aggregateByKey instead of combineByKey in docs

and

[SPARK-5448][SQL] Make CacheManager a concrete class and field in SQLContext

Author: Reynold Xin <rxin@databricks.com>

Closes #4242 from rxin/sqlCleanup and squashes the following commits:

e351cb2 [Reynold Xin] Fixed toDataFrame.
6545c42 [Reynold Xin] More changes.
728c017 [Reynold Xin] [SPARK-5447][SQL] Replaced reference to SchemaRDD with DataFrame.

This is found through reading RDD from `sc.newAPIHadoopRDD` and writing it back using `rdd.saveAsNewAPIHadoopFile` in pyspark.

It turns out that whenever there are multiple RDD conversions from JavaRDD to PythonRDD then back to JavaRDD, the exception below happens:

```
15/01/16 10:28:31 ERROR Executor: Exception in task 0.0 in stage 3.0 (TID 7)
java.lang.ClassCastException: [Ljava.lang.Object; cannot be cast to java.util.ArrayList
	at org.apache.spark.api.python.SerDeUtil$$anonfun$pythonToJava$1$$anonfun$apply$1.apply(SerDeUtil.scala:157)
	at org.apache.spark.api.python.SerDeUtil$$anonfun$pythonToJava$1$$anonfun$apply$1.apply(SerDeUtil.scala:153)
	at scala.collection.Iterator$$anon$13.hasNext(Iterator.scala:371)
	at scala.collection.Iterator$$anon$10.hasNext(Iterator.scala:308)
```

The test case code below reproduces it:

```
from pyspark.rdd import RDD

dl = [
    (u'2', {u'director': u'David Lean'}),
    (u'7', {u'director': u'Andrew Dominik'})
]

dl_rdd = sc.parallelize(dl)
tmp = dl_rdd._to_java_object_rdd()
tmp2 = sc._jvm.SerDe.javaToPython(tmp)
t = RDD(tmp2, sc)
t.count()

tmp = t._to_java_object_rdd()
tmp2 = sc._jvm.SerDe.javaToPython(tmp)
t = RDD(tmp2, sc)
t.count() # it blows up here during the 2nd time of conversion
```

Author: Winston Chen <wchen@quid.com>

Closes #4146 from wingchen/master and squashes the following commits:

903df7d [Winston Chen] SPARK-5361, update to toSeq based on the PR
5d90a83 [Winston Chen] SPARK-5361, make python pretty, so to pass PEP 8 checks
126be6b [Winston Chen] SPARK-5361, add in test case
4cf1187 [Winston Chen] SPARK-5361, add in test case
9f1a097 [Winston Chen] add in tuple handling while converting form python RDD back to JavaRDD

[SPARK-5291][CORE] Add timestamp and reason why an executor is removed to SparkListenerExecutorAdded and SparkListenerExecutorRemoved

Recently `SparkListenerExecutorAdded` and `SparkListenerExecutorRemoved` are added.
I think it's useful if they have timestamp and the reason why an executor is removed.

Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>

Closes #4082 from sarutak/SPARK-5291 and squashes the following commits:

a026ff2 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-5291
979dfe1 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-5291
cf9f9080 [Kousuke Saruta] Fixed test case
1f2a89b [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-5291
243f2a60 [Kousuke Saruta] Modified MesosSchedulerBackendSuite
a527c35 [Kousuke Saruta] Added timestamp to SparkListenerExecutorAdded

This pull request includes the abstractions for the distributed BlockMatrix representation.
`BlockMatrix` will allow users to store very large matrices in small blocks of local matrices. Specific partitioners, such as `RowBasedPartitioner` and `ColumnBasedPartitioner`, are implemented in order to optimize addition and multiplication operations that will be added in a following PR.

This work is based on the ml-matrix repo developed at the AMPLab at UC Berkeley, CA.
https://github.com/amplab/ml-matrix

Additional thanks to rezazadeh, shivaram, and mengxr for guidance on the design.

Author: Burak Yavuz <brkyvz@gmail.com>
Author: Xiangrui Meng <meng@databricks.com>
Author: Burak Yavuz <brkyvz@dn51t42l.sunet>
Author: Burak Yavuz <brkyvz@dn51t4rd.sunet>
Author: Burak Yavuz <brkyvz@dn0a221430.sunet>

Closes #3200 from brkyvz/SPARK-3974 and squashes the following commits:

a8eace2 [Burak Yavuz] Merge pull request #2 from mengxr/brkyvz-SPARK-3974
feb32a7 [Xiangrui Meng] update tests
e1d3ee8 [Xiangrui Meng] minor updates
24ec7b8 [Xiangrui Meng] update grid partitioner
5eecd48 [Burak Yavuz] fixed gridPartitioner and added tests
140f20e [Burak Yavuz] Merge branch 'master' of github.com:apache/spark into SPARK-3974
1694c9e [Burak Yavuz] almost finished addressing comments
f9d664b [Burak Yavuz] updated API and modified partitioning scheme
eebbdf7 [Burak Yavuz] preliminary changes addressing code review
1a63b20 [Burak Yavuz] [SPARK-3974] Remove setPartition method. Isn't required
1e8bb2a [Burak Yavuz] [SPARK-3974] Change return type of cache and persist
239ab4b [Burak Yavuz] [SPARK-3974] Addressed @jkbradley's comments
ba414d2 [Burak Yavuz] [SPARK-3974] fixed frobenius norm
ab6cde0 [Burak Yavuz] [SPARK-3974] Modifications cleaning code up, making size calculation more robust
9ae85aa [Burak Yavuz] [SPARK-3974] Made partitioner a variable inside BlockMatrix instead of a constructor variable
d033861 [Burak Yavuz] [SPARK-3974] Removed SubMatrixInfo and added constructor without partitioner
49b9586 [Burak Yavuz] [SPARK-3974] Updated testing utils from master
645afbe [Burak Yavuz] [SPARK-3974] Pull latest master
b05aabb [Burak Yavuz] [SPARK-3974] Updated tests to reflect changes
19c17e8 [Burak Yavuz] [SPARK-3974] Changed blockIdRow and blockIdCol
589fbb6 [Burak Yavuz] [SPARK-3974] Code review feedback addressed
aa8f086 [Burak Yavuz] [SPARK-3974] Additional comments added
f378e16 [Burak Yavuz] [SPARK-3974] Block Matrix Abstractions ready
b693209 [Burak Yavuz] Ready for Pull request

This commit exists to close the following pull requests on Github:

Closes #1480 (close requested by 'pwendell')
Closes #4205 (close requested by 'kdatta')
Closes #4114 (close requested by 'pwendell')
Closes #3382 (close requested by 'mengxr')
Closes #3933 (close requested by 'mengxr')
Closes #3870 (close requested by 'yhuai')

Author: Ryan Williams <ryan.blake.williams@gmail.com>

Closes #4211 from ryan-williams/sbt0.13.7 and squashes the following commits:

e28476d [Ryan Williams] bump sbt to version to 0.13.7

The current way of shading Guava is a little problematic. Code that
depends on "spark-core" does not see the transitive dependency, yet
classes in "spark-core" actually depend on Guava. So it's a little
tricky to run unit tests that use spark-core classes, since you need
a compatible version of Guava in your dependencies when running the
tests. This can become a little tricky, and is kind of a bad user
experience.

This change modifies the way Guava is shaded so that it's applied
uniformly across the Spark build. This means Guava is shaded inside
spark-core itself, so that the dependency issues above are solved.
Aside from that, all Spark sub-modules have their Guava references
relocated, so that they refer to the relocated classes now packaged
inside spark-core. Before, this was only done by the time the assembly
was built, so projects that did not end up inside the assembly (such
as streaming backends) could still reference the original location
of Guava classes.

The Guava classes are added to the "first" artifact Spark generates
(network-common), so that all downstream modules have the needed
classes available. Since "network-common" is a dependency of spark-core,
all Spark apps should get the relocated classes automatically.

Author: Marcelo Vanzin <vanzin@cloudera.com>

Closes #3658 from vanzin/SPARK-4809 and squashes the following commits:

3c93e42 [Marcelo Vanzin] Shade Guava in the network-common artifact.
5d69ec9 [Marcelo Vanzin] Merge branch 'master' into SPARK-4809
b3104fc [Marcelo Vanzin] Add comment.
941848f [Marcelo Vanzin] Merge branch 'master' into SPARK-4809
f78c48a [Marcelo Vanzin] Merge branch 'master' into SPARK-4809
8053dd4 [Marcelo Vanzin] Merge branch 'master' into SPARK-4809
107d7da [Marcelo Vanzin] Add fix for SPARK-5052 (PR #3874).
40b8723 [Marcelo Vanzin] Merge branch 'master' into SPARK-4809
4a4ed42 [Marcelo Vanzin] [SPARK-4809] Rework Guava library shading.

Author: Reynold Xin <rxin@databricks.com>

Closes #4235 from rxin/df-tests1 and squashes the following commits:

f341db6 [Reynold Xin] [SPARK-5097][SQL] Test cases for DataFrame expressions.

This pull request redesigns the existing Spark SQL dsl, which already provides data frame like functionalities.

TODOs:
With the exception of Python support, other tasks can be done in separate, follow-up PRs.
- [ ] Audit of the API
- [ ] Documentation
- [ ] More test cases to cover the new API
- [x] Python support
- [ ] Type alias SchemaRDD

Author: Reynold Xin <rxin@databricks.com>
Author: Davies Liu <davies@databricks.com>

Closes #4173 from rxin/df1 and squashes the following commits:

0a1a73b [Reynold Xin] Merge branch 'df1' of github.com:rxin/spark into df1
23b4427 [Reynold Xin] Mima.
828f70d [Reynold Xin] Merge pull request #7 from davies/df
257b9e6 [Davies Liu] add repartition
6bf2b73 [Davies Liu] fix collect with UDT and tests
e971078 [Reynold Xin] Missing quotes.
b9306b4 [Reynold Xin] Remove removeColumn/updateColumn for now.
a728bf2 [Reynold Xin] Example rename.
e8aa3d3 [Reynold Xin] groupby -> groupBy.
9662c9e [Davies Liu] improve DataFrame Python API
4ae51ea [Davies Liu] python API for dataframe
1e5e454 [Reynold Xin] Fixed a bug with symbol conversion.
2ca74db [Reynold Xin] Couple minor fixes.
ea98ea1 [Reynold Xin] Documentation & literal expressions.
2b22684 [Reynold Xin] Got rid of IntelliJ problems.
02bbfbc [Reynold Xin] Tightening imports.
ffbce66 [Reynold Xin] Fixed compilation error.
59b6d8b [Reynold Xin] Style violation.
b85edfb [Reynold Xin] ALS.
8c37f0a [Reynold Xin] Made MLlib and examples compile
6d53134 [Reynold Xin] Hive module.
d35efd5 [Reynold Xin] Fixed compilation error.
ce4a5d2 [Reynold Xin] Fixed test cases in SQL except ParquetIOSuite.
66d5ef1 [Reynold Xin] SQLContext minor patch.
c9bcdc0 [Reynold Xin] Checkpoint: SQL module compiles!

...mbineFileSplits

Author: Sandy Ryza <sandy@cloudera.com>

Closes #4050 from sryza/sandy-spark-5199 and squashes the following commits:

864514b [Sandy Ryza] Add tests and fix bug
0d504f1 [Sandy Ryza] Prettify
915c7e6 [Sandy Ryza] Get metrics from all filesystems
cdbc3e8 [Sandy Ryza] SPARK-5199. Input metrics should show up for InputFormats that return CombineFileSplits

fix python example of ALS in guide, use Rating instead of np.array.

Author: Davies Liu <davies@databricks.com>

Closes #4226 from davies/fix_als_guide and squashes the following commits:

1433d76 [Davies Liu] fix python example of als in guide

Here's one way to make the hashes match what Maven's plugins would create. It takes a little extra footwork since OS X doesn't have the same command line tools. An alternative is just to make Maven output these of course - would that be better? I ask in case there is a reason I'm missing, like, we need to hash files that Maven doesn't build.

Author: Sean Owen <sowen@cloudera.com>

Closes #4161 from srowen/SPARK-5308 and squashes the following commits:

70d09d0 [Sean Owen] Use $(...) syntax
e25eff8 [Sean Owen] Generate MD5, SHA1 hashes in a format like Maven's plugin

Support for transposing local matrices added. The `.transpose` function creates a new object re-using the backing array(s) but switches `numRows` and `numCols`. Operations check the flag `.isTransposed` to see whether the indexing in `values` should be modified.

This PR will pave the way for transposing `BlockMatrix`.

Author: Burak Yavuz <brkyvz@gmail.com>

Closes #4109 from brkyvz/SPARK-5321 and squashes the following commits:

87ab83c [Burak Yavuz] fixed scalastyle
caf4438 [Burak Yavuz] addressed code review v3
c524770 [Burak Yavuz] address code review comments 2
77481e8 [Burak Yavuz] fixed MiMa
f1c1742 [Burak Yavuz] small refactoring
ccccdec [Burak Yavuz] fixed failed test
dd45c88 [Burak Yavuz] addressed code review
a01bd5f [Burak Yavuz] [SPARK-5321] Fixed MiMa issues
2a63593 [Burak Yavuz] [SPARK-5321] fixed bug causing failed gemm test
c55f29a [Burak Yavuz] [SPARK-5321] Support for transposing local matrices cleaned up
c408c05 [Burak Yavuz] [SPARK-5321] Support for transposing local matrices added

The current implementation in Vectors.sqdist is not efficient because of allocating temp arrays. There is also a bug in the code `v1.indices.length / v1.size < 0.5`. This pr fixes the bug and refactors sqdist without allocating new arrays.

Author: Liang-Chi Hsieh <viirya@gmail.com>

Closes #4217 from viirya/fix_sqdist and squashes the following commits:

e8b0b3d [Liang-Chi Hsieh] For review comments.
314c424 [Liang-Chi Hsieh] Fix sqdist bug.

I've added support for sampling_rate not equal to 1.0 . I have two major questions.

1. A Scala style test is failing, since the number of parameters now exceed 10.
2. I would like suggestions to understand how to test this.

Author: MechCoder <manojkumarsivaraj334@gmail.com>

Closes #4073 from MechCoder/spark-3726 and squashes the following commits:

8012fb2 [MechCoder] Add test in Strategy
e0e0d9c [MechCoder] TST: Add better test
d1df1b2 [MechCoder] Add test to verify subsampling behavior
a7bfc70 [MechCoder] [SPARK-3726] Allow sampling_rate not equal to 1.0

... decision tree model

Labels loaded from libsvm files are mapped to 0.0 if they are negative labels because they should be nonnegative value.

Author: lewuathe <lewuathe@me.com>

Closes #3975 from Lewuathe/map-negative-label-to-positive and squashes the following commits:

12d1d59 [lewuathe] [SPARK-5119] Fix code styles
6d9a18a [lewuathe] [SPARK-5119] Organize test codes
62a150c [lewuathe] [SPARK-5119] Modify Impurities throw exceptions with negatie labels
3336c21 [lewuathe] [SPARK-5119] java.lang.ArrayIndexOutOfBoundsException on trying to train decision tree model

Fixes problems with incorrect method signatures related to shaded classes. For discussion see the jira issue.

Author: Elmer Garduno <elmerg@google.com>

Closes #3874 from elmer-garduno/fix_guava_signatures and squashes the following commits:

aa5d8e0 [Elmer Garduno] Unshade common/base[Function|Supplier] classes to fix guava methods signatures.

This reenables and fixes this test, after addressing two issues:

- The Semaphore that was intended to be shared locally was being serialized and copied; it's now a static member in the companion object as in other tests
- Later changes to Spark means that cancelling the first task will not cancel the shared stage and therefore the second task should succeed

Author: Sean Owen <sowen@cloudera.com>

Closes #4180 from srowen/SPARK-960 and squashes the following commits:

43da66f [Sean Owen] Fix 'two jobs sharing the same stage' test and reenable it: truly share a Semaphore locally as intended, and update expectation of failure in non-cancelled task

Storing command in variables is tricky in bash, use an array
to handle all issues with spaces, quoting, etc.
See: http://mywiki.wooledge.org/BashFAQ/050

Author: David Y. Ross <dyross@gmail.com>

Closes #4126 from dyross/dyr-fix-make-distribution and squashes the following commits:

4ce522b [David Y. Ross] Fix command spaces issue in make-distribution.sh

Defer use of log4j class until it's known that log4j 1.2 is being used. This may avoid dealing with log4j dependencies for callers that reroute slf4j to another logging framework. The only change is to push one half of the check in the original `if` condition inside. This is a trivial change, may or may not actually solve a problem, but I think it's all that makes sense to do for SPARK-4147.

Author: Sean Owen <sowen@cloudera.com>

Closes #4190 from srowen/SPARK-4147 and squashes the following commits:

4e99942 [Sean Owen] Defer use of log4j class until it's known that log4j 1.2 is being used. This may avoid dealing with log4j dependencies for callers that reroute slf4j to another logging framework.

build/mvn will automatically download tarball of maven. But currently, the URL is invalid.

Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>

Closes #4124 from sarutak/SPARK-5339 and squashes the following commits:

6e96121 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-5339
0e012d1 [Kousuke Saruta] Updated Maven version to 3.2.5
ca26499 [Kousuke Saruta] Fixed URL of the tarball of Maven

j.u.c.ConcurrentHashMap is more battle tested.

cc rxin JoshRosen pwendell

Author: Davies Liu <davies@databricks.com>

Closes #4208 from davies/safe-conf and squashes the following commits:

c2182dc [Davies Liu] address comments, fix tests
3a1d821 [Davies Liu] fix test
da14ced [Davies Liu] Merge branch 'master' of github.com:apache/spark into safe-conf
ae4d305 [Davies Liu] change to j.u.c.ConcurrentMap
f8fa1cf [Davies Liu] change to TrieMap
a1d769a [Davies Liu] make SparkConf thread-safe

[SPARK-5384][mllib] Vectors.sqdist returns inconsistent results for sparse/dense vectors when the vectors have different lengths

JIRA issue: https://issues.apache.org/jira/browse/SPARK-5384
Currently `Vectors.sqdist` return inconsistent result for sparse/dense vectors when the vectors have different lengths, please refer to JIRA for sample

PR scope:
Unify the sqdist logic for dense/sparse vectors and fix the inconsistency, also remove the possible sparse to dense conversion in the original code.

For reviewers:
Maybe we should first discuss what's the correct behavior.
1. Vectors for sqdist must have the same length, like in breeze?
2. If they can have different lengths, what's the correct result for sqdist? (should the extra part get into calculation?)

I'll update PR with more optimization and additional ut afterwards. Thanks.

Author: Yuhao Yang <hhbyyh@gmail.com>

Closes #4183 from hhbyyh/fixDouble and squashes the following commits:

1f17328 [Yuhao Yang] limit PR scope to size constraints only
54cbf97 [Yuhao Yang] fix Vectors.sqdist inconsistence

https://issues.apache.org/jira/browse/SPARK-5268

In CoarseGrainedExecutorBackend, we subscribe DisassociatedEvent in executor backend actor and exit the program upon receive such event...

let's consider the following case

The user may develop an Akka-based program which starts the actor with Spark's actor system and communicate with an external actor system (e.g. an Akka-based receiver in spark streaming which communicates with an external system) If the external actor system fails or disassociates with the actor within spark's system with purpose, we may receive DisassociatedEvent and the executor is restarted.

This is not the expected behavior.....

----

This is a simple fix to check the event before making the quit decision

Author: CodingCat <zhunansjtu@gmail.com>

Closes #4063 from CodingCat/SPARK-5268 and squashes the following commits:

4d7d48e [CodingCat] simplify the log
18c36f4 [CodingCat] more descriptive log
f299e0b [CodingCat] clean log
1632e79 [CodingCat] check whether DisassociatedEvent is relevant before quit