Implemented python wrappers for Scala functions that don't exist in `ml.features`
Author: Burak Yavuz <brkyvz@gmail.com>
Closes#5991 from brkyvz/ml-feat-PR and squashes the following commits:
adcca55 [Burak Yavuz] add regex tokenizer to __all__
b91cb44 [Burak Yavuz] addressed comments
bd39fd2 [Burak Yavuz] remove addition
b82bd7c [Burak Yavuz] Parity in PySpark for ml.features
Exposes data available in the UI as json over http. Key points:
* new endpoints, handled independently of existing XyzPage classes. Root entrypoint is `JsonRootResource`
* Uses jersey + jackson for routing & converting POJOs into json
* tests against known results in `HistoryServerSuite`
* also fixes some minor issues w/ the UI -- synchronizing on access to `StorageListener` & `StorageStatusListener`, and fixing some inconsistencies w/ the way we handle retained jobs & stages.
Author: Imran Rashid <irashid@cloudera.com>
Closes#5940 from squito/SPARK-3454_better_test_files and squashes the following commits:
1a72ed6 [Imran Rashid] rats
85fdb3e [Imran Rashid] Merge branch 'no_php' into SPARK-3454
1fc65b0 [Imran Rashid] Revert "Revert "[SPARK-3454] separate json endpoints for data in the UI""
1276900 [Imran Rashid] get rid of giant event file, replace w/ smaller one; check both shuffle read & shuffle write
4e12013 [Imran Rashid] just use test case name for expectation file name
863ef64 [Imran Rashid] rename json files to avoid strange file names and not look like php
Based on https://github.com/apache/spark/pull/5478 that provide a PYSPARK_ARCHIVES_PATH env. within this PR, we just should export PYSPARK_ARCHIVES_PATH=/user/spark/pyspark.zip,/user/spark/python/lib/py4j-0.8.2.1-src.zip in conf/spark-env.sh when we don't install PySpark on each node of Yarn. i run python application successfully on yarn-client and yarn-cluster with this PR.
andrewor14 sryza Sephiroth-Lin Can you take a look at this?thanks.
Author: Lianhui Wang <lianhuiwang09@gmail.com>
Closes#5580 from lianhuiwang/SPARK-6869 and squashes the following commits:
66ffa43 [Lianhui Wang] Update Client.scala
c2ad0f9 [Lianhui Wang] Update Client.scala
1c8f664 [Lianhui Wang] Merge remote-tracking branch 'remotes/apache/master' into SPARK-6869
008850a [Lianhui Wang] Merge remote-tracking branch 'remotes/apache/master' into SPARK-6869
f0b4ed8 [Lianhui Wang] Merge remote-tracking branch 'remotes/apache/master' into SPARK-6869
150907b [Lianhui Wang] Merge remote-tracking branch 'remotes/apache/master' into SPARK-6869
20402cd [Lianhui Wang] use ZipEntry
9d87c3f [Lianhui Wang] update scala style
e7bd971 [Lianhui Wang] address vanzin's comments
4b8a3ed [Lianhui Wang] use pyArchivesEnvOpt
e6b573b [Lianhui Wang] address vanzin's comments
f11f84a [Lianhui Wang] zip pyspark archives
5192cca [Lianhui Wang] update import path
3b1e4c8 [Lianhui Wang] address tgravescs's comments
9396346 [Lianhui Wang] put zip to make-distribution.sh
0d2baf7 [Lianhui Wang] update import paths
e0179be [Lianhui Wang] add zip pyspark archives in build or sparksubmit
31e8e06 [Lianhui Wang] update code style
9f31dac [Lianhui Wang] update code and add comments
f72987c [Lianhui Wang] add archives path to PYTHONPATH
Added a new batch named `Substitution` before `Resolution` batch. The motivation for this is there are kind of cases we want to do some substitution on the parsed logical plan before resolve it.
Consider this two cases:
1 CTE, for cte we first build a row logical plan
```
'With Map(q1 -> 'Subquery q1
'Project ['key]
'UnresolvedRelation [src], None)
'Project [*]
'Filter ('key = 5)
'UnresolvedRelation [q1], None
```
In `With` logicalplan here is a map stored the (`q1-> subquery`), we want first take off the with command and substitute the `q1` of `UnresolvedRelation` by the `subquery`
2 Another example is Window function, in window function user may define some windows, we also need substitute the window name of child by the concrete window. this should also done in the Substitution batch.
Author: wangfei <wangfei1@huawei.com>
Closes#5776 from scwf/addbatch and squashes the following commits:
d4b962f [wangfei] added WindowsSubstitution
70f6932 [wangfei] Merge branch 'master' of https://github.com/apache/spark into addbatch
ecaeafb [wangfei] address yhuai's comments
553005a [wangfei] fix test case
0c54798 [wangfei] address comments
29aaaaf [wangfei] fix compile
1c9a092 [wangfei] added Substitution bastch
This only happens if you have `SPARK_PREPEND_CLASSES` set. Then I built it with `build/sbt clean assembly compile` and just ran it with `bin/spark-shell`.
```
...
15/05/07 17:07:30 INFO EventLoggingListener: Logging events to file:/tmp/spark-events/local-1431043649919
15/05/07 17:07:30 INFO SparkILoop: Created spark context..
Spark context available as sc.
java.lang.NoClassDefFoundError: org/apache/hadoop/hive/conf/HiveConf
at java.lang.Class.getDeclaredConstructors0(Native Method)
at java.lang.Class.privateGetDeclaredConstructors(Class.java:2493)
at java.lang.Class.getConstructor0(Class.java:2803)
...
Caused by: java.lang.ClassNotFoundException: org.apache.hadoop.hive.conf.HiveConf
at java.net.URLClassLoader$1.run(URLClassLoader.java:366)
at java.net.URLClassLoader$1.run(URLClassLoader.java:355)
at java.security.AccessController.doPrivileged(Native Method)
at java.net.URLClassLoader.findClass(URLClassLoader.java:354)
at java.lang.ClassLoader.loadClass(ClassLoader.java:425)
at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:308)
at java.lang.ClassLoader.loadClass(ClassLoader.java:358)
... 52 more
<console>:10: error: not found: value sqlContext
import sqlContext.implicits._
^
<console>:10: error: not found: value sqlContext
import sqlContext.sql
^
```
yhuai marmbrus
Author: Andrew Or <andrew@databricks.com>
Closes#5997 from andrewor14/sql-shell-crash and squashes the following commits:
61147e6 [Andrew Or] Also expect NoClassDefFoundError
With 0a2b15ce43, the serialization stream and deserialization stream has enough information to determine it is handling a key-value pari, a key, or a value. It is safe to use `SparkSqlSerializer2` in more cases.
Author: Yin Huai <yhuai@databricks.com>
Closes#5849 from yhuai/serializer2MoreCases and squashes the following commits:
53a5eaa [Yin Huai] Josh's comments.
487f540 [Yin Huai] Use BufferedOutputStream.
8385f95 [Yin Huai] Always create a new row at the deserialization side to work with sort merge join.
c7e2129 [Yin Huai] Update tests.
4513d13 [Yin Huai] Use Serializer2 in more places.
the toArray function of the BoundedPriorityQueue does not necessarily preserve order. Add a counter-example as the test, which would fail the original impl.
Author: Shuo Xiang <shuoxiangpub@gmail.com>
Closes#5990 from coderxiang/topbykey-test and squashes the following commits:
98804c9 [Shuo Xiang] fix bug in topBykey and update test
This PR switches Spark SQL's Hive support to use the isolated hive client interface introduced by #5851, instead of directly interacting with the client. By using this isolated client we can now allow users to dynamically configure the version of Hive that they are connecting to by setting `spark.sql.hive.metastore.version` without the need recompile. This also greatly reduces the surface area for our interaction with the hive libraries, hopefully making it easier to support other versions in the future.
Jars for the desired hive version can be configured using `spark.sql.hive.metastore.jars`, which accepts the following options:
- a colon-separated list of jar files or directories for hive and hadoop.
- `builtin` - attempt to discover the jars that were used to load Spark SQL and use those. This
option is only valid when using the execution version of Hive.
- `maven` - download the correct version of hive on demand from maven.
By default, `builtin` is used for Hive 13.
This PR also removes the test step for building against Hive 12, as this will no longer be required to talk to Hive 12 metastores. However, the full removal of the Shim is deferred until a later PR.
Remaining TODOs:
- Remove the Hive Shims and inline code for Hive 13.
- Several HiveCompatibility tests are not yet passing.
- `nullformatCTAS` - As detailed below, we now are handling CTAS parsing ourselves instead of hacking into the Hive semantic analyzer. However, we currently only handle the common cases and not things like CTAS where the null format is specified.
- `combine1` now leaks state about compression somehow, breaking all subsequent tests. As such we currently add it to the blacklist
- `part_inherit_tbl_props` and `part_inherit_tbl_props_with_star` do not work anymore. We are correctly propagating the information
- "load_dyn_part14.*" - These tests pass when run on their own, but fail when run with all other tests. It seems our `RESET` mechanism may not be as robust as it used to be?
Other required changes:
- `CreateTableAsSelect` no longer carries parts of the HiveQL AST with it through the query execution pipeline. Instead, we parse CTAS during the HiveQL conversion and construct a `HiveTable`. The full parsing here is not yet complete as detailed above in the remaining TODOs. Since the operator is Hive specific, it is moved to the hive package.
- `Command` is simplified to be a trait that simply acts as a marker for a LogicalPlan that should be eagerly evaluated.
Author: Michael Armbrust <michael@databricks.com>
Closes#5876 from marmbrus/useIsolatedClient and squashes the following commits:
258d000 [Michael Armbrust] really really correct path handling
e56fd4a [Michael Armbrust] getAbsolutePath
5a259f5 [Michael Armbrust] fix typos
81bb366 [Michael Armbrust] comments from vanzin
5f3945e [Michael Armbrust] Merge remote-tracking branch 'origin/master' into useIsolatedClient
4b5cd41 [Michael Armbrust] yin's comments
f5de7de [Michael Armbrust] cleanup
11e9c72 [Michael Armbrust] better coverage in versions suite
7e8f010 [Michael Armbrust] better error messages and jar handling
e7b3941 [Michael Armbrust] more permisive checking for function registration
da91ba7 [Michael Armbrust] Merge remote-tracking branch 'origin/master' into useIsolatedClient
5fe5894 [Michael Armbrust] fix serialization suite
81711c4 [Michael Armbrust] Initial support for running without maven
1d8ae44 [Michael Armbrust] fix final tests?
1c50813 [Michael Armbrust] more comments
a3bee70 [Michael Armbrust] Merge remote-tracking branch 'origin/master' into useIsolatedClient
a6f5df1 [Michael Armbrust] style
ab07f7e [Michael Armbrust] WIP
4d8bf02 [Michael Armbrust] Remove hive 12 compilation
8843a25 [Michael Armbrust] [SPARK-6908] [SQL] Use isolated Hive client
If jobs are dropped by SparkListener, at least we can show the job ids in BatchPage. Screenshot:
Author: zsxwing <zsxwing@gmail.com>
Closes#5840 from zsxwing/SPARK-7305 and squashes the following commits:
aca0ba6 [zsxwing] Fix the code style
718765e [zsxwing] Make generateNormalJobRow private
8073b03 [zsxwing] Merge branch 'master' into SPARK-7305
83dec11 [zsxwing] Make BatchPage show friendly information when jobs are dropped by SparkListener
Author: tedyu <yuzhihong@gmail.com>
Closes#5897 from tedyu/master and squashes the following commits:
473bf9d [tedyu] Address Josh's review comments
1719c5b [tedyu] Correct upper bound in for loop
b51dcaf [tedyu] Add unit test in BitSetSuite for BitSet#anySet()
83f9f87 [tedyu] Merge branch 'master' of github.com:apache/spark
817e3f9 [tedyu] Replace constant 8 with SIZE_OF_LONG
75a467b [tedyu] Correct offset for UNSAFE.getLong()
855374b [tedyu] Remove second loop since bitSetWidthInBytes is WORD aligned
093b7a4 [tedyu] Use UNSAFE.getLong() to speed up BitSetMethods#anySet()
63ee050 [tedyu] Use UNSAFE.getLong() to speed up BitSetMethods#anySet()
4ca0ef6 [tedyu] Use UNSAFE.getLong() to speed up BitSetMethods#anySet()
3e9b6919 [tedyu] Use UNSAFE.getLong() to speed up BitSetMethods#anySet()
Avoid translating to CaseWhen and evaluate the key expression many times.
Author: Wenchen Fan <cloud0fan@outlook.com>
Closes#5979 from cloud-fan/condition and squashes the following commits:
3ce54e1 [Wenchen Fan] add CaseKeyWhen
Go through the context classloader when reflecting on user types in ScalaReflection.
Replaced calls to `typeOf` with `typeTag[T].in(mirror)`. The convenience method assumes
all types can be found in the classloader that loaded scala-reflect (the primordial
classloader). This assumption is not valid in all contexts (sbt console, Eclipse launchers).
Fixed SPARK-5281
Author: Iulian Dragos <jaguarul@gmail.com>
Closes#5981 from dragos/issue/mirrors-missing-requirement-error and squashes the following commits:
d103e70 [Iulian Dragos] Go through the context classloader when reflecting on user types in ScalaReflection
JIRA: https://issues.apache.org/jira/browse/SPARK-7277
As automatically determining the number of reducers is not supported (`mapred.reduce.tasks` is set to `-1`), we should throw exception to users.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#5811 from viirya/no_neg_reduce_tasks and squashes the following commits:
e518f96 [Liang-Chi Hsieh] Consider other wrong setting values.
fd9c817 [Liang-Chi Hsieh] Merge remote-tracking branch 'upstream/master' into no_neg_reduce_tasks
4ede705 [Liang-Chi Hsieh] Throw exception instead of warning message.
68a1c70 [Liang-Chi Hsieh] Show warning message if mapred.reduce.tasks is set to -1.
`Star` and `MultiAlias` just used in `analyzer` and them will be substituted after analyze, So just like `Alias` they do not need extend `Attribute`
Author: scwf <wangfei1@huawei.com>
Closes#5928 from scwf/attribute and squashes the following commits:
73a0560 [scwf] star and multialias do not need extend attribute
The compression is based on storage. brkyvz
Author: Xiangrui Meng <meng@databricks.com>
Closes#5985 from mengxr/SPARK-6948 and squashes the following commits:
df56a00 [Xiangrui Meng] update python tests
6d90d45 [Xiangrui Meng] compress vectors in VectorAssembler
Add
1. Class methods squared_dist
3. parse
4. norm
5. numNonzeros
6. copy
I made a few vectorizations wrt squared_dist and dot as well. I have added support for SparseMatrix serialization in a separate PR (https://github.com/apache/spark/pull/5775) and plan to complete support for Matrices in another PR.
Author: MechCoder <manojkumarsivaraj334@gmail.com>
Closes#5872 from MechCoder/local_linalg_api and squashes the following commits:
a8ff1e0 [MechCoder] minor
ce3e53e [MechCoder] Add error message for parser
1bd3c04 [MechCoder] Robust parser and removed unnecessary methods
f779561 [MechCoder] [SPARK-7328] Pyspark.mllib.linalg.Vectors: Missing items
This is an addition to #5729.
Here's an example with ALS.
<img src="https://issues.apache.org/jira/secure/attachment/12731039/tooltip.png" width="400px"></img>
Author: Andrew Or <andrew@databricks.com>
Closes#5957 from andrewor14/viz-hover2 and squashes the following commits:
60e3758 [Andrew Or] Add tooltips for RDDs on job page
This is an addition to #5729.
If you click into a stage from the DAG viz on the job page, you might expect to expand on the stage. However, once you get to the stage page, you actually have to expand the DAG viz there yourself.
This patch makes this happen automatically. It's a small UX improvement.
Author: Andrew Or <andrew@databricks.com>
Closes#5958 from andrewor14/viz-auto-expand and squashes the following commits:
03cd157 [Andrew Or] Automatically expand DAG viz if from job page
Using the existing docker support for mesos, also enabling the mesos cluster mode scheduler to launch Spark drivers in docker images as well.
This also allows the executors launched by the drivers to be also in the same Docker image by passing the docker settings.
Author: Timothy Chen <tnachen@gmail.com>
Closes#5917 from tnachen/spark_cluster_docker and squashes the following commits:
1e842f5 [Timothy Chen] Add docker support for launching drivers in mesos cluster mode.
scala has deterministic naming-scheme for the generated methods which return default arguments . here one of the default argument of overloaded method has to be removed
Author: Tijo Thomas <tijoparacka@gmail.com>
Closes#5966 from tijoparacka/fix_compilation_error_in_scala2.11 and squashes the following commits:
c90bba8 [Tijo Thomas] Fixed compilation error in scala 2.11
This is a follow up of #5827 to remove the additional `SparkSQLParser`
Author: Cheng Hao <hao.cheng@intel.com>
Closes#5965 from chenghao-intel/remove_sparksqlparser and squashes the following commits:
509a233 [Cheng Hao] Remove the HiveQlQueryExecution
a5f9e3b [Cheng Hao] Remove the duplicated SparkSQLParser
This patch simply removes a `cache()` on an intermediate RDD when evaluating Python UDFs.
Author: ksonj <kson@siberie.de>
Closes#5973 from ksonj/udf and squashes the following commits:
db5b564 [ksonj] removed TODO about cleaning up
fe70c54 [ksonj] Remove cache() causing memory leak
Address marmbrus and scwf's comments in #5604.
Author: Yin Huai <yhuai@databricks.com>
Closes#5945 from yhuai/windowFollowup and squashes the following commits:
0ef879d [Yin Huai] Add collectFirst to TreeNode.
2373968 [Yin Huai] wip
4a16df9 [Yin Huai] Address minor comments for [SPARK-1442].
https://issues.apache.org/jira/browse/SPARK-6093
Author: Yanbo Liang <ybliang8@gmail.com>
Closes#5941 from yanboliang/spark-6093 and squashes the following commits:
6934af3 [Yanbo Liang] change to @property
aac3bc5 [Yanbo Liang] Add RegressionMetrics in PySpark/MLlib
This patch adds a proxy call from PySpark to the Spark SQL coalesce function and this patch comes out of a discussion on devspark with rxin
This contribution is my original work and i license the work to the project under the project's open source license.
Olivier.
Author: Olivier Girardot <o.girardot@lateral-thoughts.com>
Closes#5698 from ogirardot/master and squashes the following commits:
d9a4439 [Olivier Girardot] SPARK-7118 Add the coalesce Spark SQL function available in PySpark
The wrapper required the implementation of the `ArrayParam`, because `Array[T]` is hard to obtain from Python. `ArrayParam` has an extra function called `wCast` which is an internal function to obtain `Array[T]` from `Seq[T]`
Author: Burak Yavuz <brkyvz@gmail.com>
Author: Xiangrui Meng <meng@databricks.com>
Closes#5930 from brkyvz/ml-feat and squashes the following commits:
73e745f [Burak Yavuz] Merge pull request #3 from mengxr/SPARK-7388
c221db9 [Xiangrui Meng] overload StringArrayParam.w
c81072d [Burak Yavuz] addressed comments
99c2ebf [Burak Yavuz] add to python_shared_params
39ecb07 [Burak Yavuz] fix scalastyle
7f7ea2a [Burak Yavuz] [SPARK-7388][SPARK-7383] wrapper for VectorAssembler in Python
Thank nadavoosh point this out in #5590
Author: Daoyuan Wang <daoyuan.wang@intel.com>
Closes#5877 from adrian-wang/jdbcrdd and squashes the following commits:
cc11900 [Daoyuan Wang] avoid NPE in jdbcrdd
Params.setDefault taking a set of ParamPairs should be annotated with varargs. I thought it would not work before, but it apparently does.
CrossValidator.transform should call transformSchema since the underlying Model might be a PipelineModel
CC: mengxr
Author: Joseph K. Bradley <joseph@databricks.com>
Closes#5960 from jkbradley/params-cleanups and squashes the following commits:
118b158 [Joseph K. Bradley] Params.setDefault taking a set of ParamPairs should be annotated with varargs. I thought it would not work before, but it apparently does. CrossValidator.transform should call transformSchema since the underlying Model might be a PipelineModel
Small changes, primarily to allow us more flexibility in the future:
* Rename "tau_0" to "tau0"
* Mark LDAOptimizer trait sealed and DeveloperApi.
* Mark LDAOptimizer subclasses as final.
* Mark setOptimizer (the one taking an LDAOptimizer) and getOptimizer as DeveloperApi since we may need to change them in the future
CC: hhbyyh
Author: Joseph K. Bradley <joseph@databricks.com>
Closes#5956 from jkbradley/onlinelda-cleanups and squashes the following commits:
f4be508 [Joseph K. Bradley] added newline
f4003e4 [Joseph K. Bradley] Changes: * Rename "tau_0" to "tau0" * Mark LDAOptimizer trait sealed and DeveloperApi. * Mark LDAOptimizer subclasses as final. * Mark setOptimizer (the one taking an LDAOptimizer) and getOptimizer as DeveloperApi since we may need to change them in the future
Author: Shiti <ssaxena.ece@gmail.com>
Closes#5867 from Shiti/spark-7295 and squashes the following commits:
71a9913 [Shiti] implementation for bitwise and,or, not and xor on Column with tests and docs
In environments like notebooks, the SparkContext is managed by the underlying infrastructure and it is expected that the SparkContext will not be stopped. However, StreamingContext.stop() calls SparkContext.stop() as a non-intuitive side-effect. This PR adds a configuration in SparkConf that sets the default StreamingContext stop behavior. It should be such that the existing behavior does not change for existing users.
Author: Tathagata Das <tathagata.das1565@gmail.com>
Closes#5929 from tdas/SPARK-7217 and squashes the following commits:
869a763 [Tathagata Das] Changed implementation.
685fe00 [Tathagata Das] Added configuration
Author: Tathagata Das <tathagata.das1565@gmail.com>
Closes#5961 from tdas/SPARK-7430 and squashes the following commits:
d654978 [Tathagata Das] Fix scala style
fbf7174 [Tathagata Das] Added more verbose assert failure messages.
6aea07a [Tathagata Das] Ensure SynchronizedBuffer is used in every TestSuiteBase
This patch comprises of a few related pieces of work:
* Schema inference is performed directly on the JSON token stream
* `String => Row` conversion populate Spark SQL structures without intermediate types
* Projection pushdown is implemented via CatalystScan for DataFrame queries
* Support for the legacy parser by setting `spark.sql.json.useJacksonStreamingAPI` to `false`
Performance improvements depend on the schema and queries being executed, but it should be faster across the board. Below are benchmarks using the last.fm Million Song dataset:
```
Command | Baseline | Patched
---------------------------------------------------|----------|--------
import sqlContext.implicits._ | |
val df = sqlContext.jsonFile("/tmp/lastfm.json") | 70.0s | 14.6s
df.count() | 28.8s | 6.2s
df.rdd.count() | 35.3s | 21.5s
df.where($"artist" === "Robert Hood").collect() | 28.3s | 16.9s
```
To prepare this dataset for benchmarking, follow these steps:
```
# Fetch the datasets from http://labrosa.ee.columbia.edu/millionsong/lastfm
wget http://labrosa.ee.columbia.edu/millionsong/sites/default/files/lastfm/lastfm_test.zip \
http://labrosa.ee.columbia.edu/millionsong/sites/default/files/lastfm/lastfm_train.zip
# Decompress and combine, pipe through `jq -c` to ensure there is one record per line
unzip -p lastfm_test.zip lastfm_train.zip | jq -c . > lastfm.json
```
Author: Nathan Howell <nhowell@godaddy.com>
Closes#5801 from NathanHowell/json-performance and squashes the following commits:
26fea31 [Nathan Howell] Recreate the baseRDD each for each scan operation
a7ebeb2 [Nathan Howell] Increase coverage of inserts into a JSONRelation
e06a1dd [Nathan Howell] Add comments to the `useJacksonStreamingAPI` config flag
6822712 [Nathan Howell] Split up JsonRDD2 into multiple objects
fa8234f [Nathan Howell] Wrap long lines
b31917b [Nathan Howell] Rename `useJsonRDD2` to `useJacksonStreamingAPI`
15c5d1b [Nathan Howell] JSONRelation's baseRDD need not be lazy
f8add6e [Nathan Howell] Add comments on lack of support for precision and scale DecimalTypes
fa0be47 [Nathan Howell] Remove unused default case in the field parser
80dba17 [Nathan Howell] Add comments regarding null handling and empty strings
842846d [Nathan Howell] Point the empty schema inference test at JsonRDD2
ab6ee87 [Nathan Howell] Add projection pushdown support to JsonRDD/JsonRDD2
f636c14 [Nathan Howell] Enable JsonRDD2 by default, add a flag to switch back to JsonRDD
0bbc445 [Nathan Howell] Improve JSON parsing and type inference performance
7ca70c1 [Nathan Howell] Eliminate arrow pattern, replace with pattern matches
According to the R manual: https://stat.ethz.ch/R-manual/R-devel/library/base/html/Startup.html,
" if a function .First is found on the search path, it is executed as .First(). Finally, function .First.sys() in the base package is run. This calls require to attach the default packages specified by options("defaultPackages")."
In .First() in profile/shell.R, we load SparkR package. This means SparkR package is loaded before default packages. If there are same names in default packages, they will overwrite those in SparkR. This is why filter() in SparkR is masked by filter() in stats, which is usually in the default package list.
We need to make sure SparkR is loaded after default packages. The solution is to append SparkR to default packages, instead of loading SparkR in .First().
BTW, I'd like to discuss our policy on how to solve name conflict. Previously, we rename API names from Scala API if there is name conflict with base or other commonly-used packages. However, from long term perspective, this is not good for API stability, because we can't predict name conflicts, for example, if in the future a name added in base package conflicts with an API in SparkR? So the better policy is to keep API name same as Scala's without worrying about name conflicts. When users use SparkR, they should load SparkR as last package, so that all API names are effective. Use can explicitly use :: to refer to hidden names from other packages. If we agree on this, I can submit a JIRA issue to change back some rename API methods, for example, DataFrame.sortDF().
Author: Sun Rui <rui.sun@intel.com>
Closes#5938 from sun-rui/SPARK-6812 and squashes the following commits:
b569145 [Sun Rui] [SPARK-6812][SparkR] filter() on DataFrame does not work as expected.
Temporarily disable flaky doctest for CrossValidator. jkbradley
Author: Xiangrui Meng <meng@databricks.com>
Closes#5962 from mengxr/disable-pyspark-cv-test and squashes the following commits:
5db7e5b [Xiangrui Meng] disable cv doctest
The bug is because SPARK-7139 removed some codes from SPARK-7112 unintentionally here: 1854ac326a (diff-5c8651dd78abd20439b8eb938175075dL72)
This PR just added them back and added some assertions in the tests to verify it.
Author: zsxwing <zsxwing@gmail.com>
Closes#5950 from zsxwing/SPARK-7405 and squashes the following commits:
675f5d9 [zsxwing] Fix the bug that ReceiverInputDStream doesn't report InputInfo
This is a follow-up patch for #5729.
**[SPARK-7408]** Move as much style code from JS to CSS as possible
**[SPARK-7377]** Fix JS error if a job / stage contains only one RDD
**[SPARK-7371]** Decrease emphasis on RDD on stage page as requested by mateiz pwendell
This patch also includes general code clean up.
<img src="https://issues.apache.org/jira/secure/attachment/12730992/before-after.png" width="500px"></img>
Author: Andrew Or <andrew@databricks.com>
Closes#5954 from andrewor14/viz-emphasize-rdd and squashes the following commits:
3c0d4f0 [Andrew Or] Guard against JS error by rendering arrows only if needed
f23e15b [Andrew Or] Merge branch 'master' of github.com:apache/spark into viz-emphasize-rdd
565801f [Andrew Or] Clean up code
9dab5f0 [Andrew Or] Move styling from JS to CSS + clean up code
107c0b6 [Andrew Or] Tweak background color, stroke width, font size etc.
1610c62 [Andrew Or] Implement cluster padding for stage page
Otherwise it will throw exception:
```
Exception in thread "main" kafka.common.FailedToSendMessageException: Failed to send messages after 3 tries.
at kafka.producer.async.DefaultEventHandler.handle(DefaultEventHandler.scala:90)
at kafka.producer.Producer.send(Producer.scala:77)
at org.apache.spark.examples.streaming.KafkaWordCountProducer$.main(KafkaWordCount.scala:96)
at org.apache.spark.examples.streaming.KafkaWordCountProducer.main(KafkaWordCount.scala)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:606)
at org.apache.spark.deploy.SparkSubmit$.org$apache$spark$deploy$SparkSubmit$$runMain(SparkSubmit.scala:623)
at org.apache.spark.deploy.SparkSubmit$.doRunMain$1(SparkSubmit.scala:169)
at org.apache.spark.deploy.SparkSubmit$.submit(SparkSubmit.scala:192)
at org.apache.spark.deploy.SparkSubmit$.main(SparkSubmit.scala:111)
at org.apache.spark.deploy.SparkSubmit.main(SparkSubmit.scala)
```
Author: jerryshao <saisai.shao@intel.com>
Closes#5936 from jerryshao/SPARK-7396 and squashes the following commits:
270bbe2 [jerryshao] Fix Kafka Produce throw Exception issue
This PR also makes some of the DataFrame to RDD methods private as the RDD class is private in 1.4
cc rxin pwendell
Author: Shivaram Venkataraman <shivaram@cs.berkeley.edu>
Closes#5949 from shivaram/sparkr-examples and squashes the following commits:
6c42fdc [Shivaram Venkataraman] Remove SparkR RDD examples, add dataframe examples
Changes:
* Update protected prediction methods, following design doc. **<--most interesting change**
* Changed abstract classes for Estimator and Model to be public. Added DeveloperApi tag. (I kept the traits for Estimator/Model Params private.)
* Changed ProbabilisticClassificationModel method names to use probability instead of probabilities.
CC: mengxr shivaram etrain
Author: Joseph K. Bradley <joseph@databricks.com>
Closes#5913 from jkbradley/public-dev-api and squashes the following commits:
e9aa0ea [Joseph K. Bradley] moved findMax to DenseVector and renamed to argmax. fixed bug for vector of length 0
15b9957 [Joseph K. Bradley] renamed probabilities to probability in method names
5cda84d [Joseph K. Bradley] regenerated sharedParams
7d1877a [Joseph K. Bradley] Made spark.ml prediction abstractions public. Organized their prediction methods for efficient computation of multiple output columns.
Author: Yin Huai <yhuai@databricks.com>
Closes#5951 from yhuai/fixBuildMaven and squashes the following commits:
fdde183 [Yin Huai] Move HiveWindowFunctionQuerySuite.scala to hive compatibility dir.
This patch extends the `Serializer` interface with a new `Private` API which allows serializers to indicate whether they support relocation of serialized objects in serializer stream output.
This relocatibilty property is described in more detail in `Serializer.scala`, but in a nutshell a serializer supports relocation if reordering the bytes of serialized objects in serialization stream output is equivalent to having re-ordered those elements prior to serializing them. The optimized shuffle path introduced in #4450 and #5868 both rely on serializers having this property; this patch just centralizes the logic for determining whether a serializer has this property. I also added tests and comments clarifying when this works for KryoSerializer.
This change allows the optimizations in #4450 to be applied for shuffles that use `SqlSerializer2`.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#5924 from JoshRosen/SPARK-7311 and squashes the following commits:
50a68ca [Josh Rosen] Address minor nits
0a7ebd7 [Josh Rosen] Clarify reason why SqlSerializer2 supports this serializer
123b992 [Josh Rosen] Cleanup for submitting as standalone patch.
4aa61b2 [Josh Rosen] Add missing newline
2c1233a [Josh Rosen] Small refactoring of SerializerPropertiesSuite to enable test re-use:
0ba75e6 [Josh Rosen] Add tests for serializer relocation property.
450fa21 [Josh Rosen] Back out accidental log4j.properties change
86d4dcd [Josh Rosen] Flag that SparkSqlSerializer2 supports relocation
b9624ee [Josh Rosen] Expand serializer API and use new function to help control when new UnsafeShuffle path is used.
Adding more information about the implementation...
This PR is adding the support of window functions to Spark SQL (specifically OVER and WINDOW clause). For every expression having a OVER clause, we use a WindowExpression as the container of a WindowFunction and the corresponding WindowSpecDefinition (the definition of a window frame, i.e. partition specification, order specification, and frame specification appearing in a OVER clause).
# Implementation #
The high level work flow of the implementation is described as follows.
* Query parsing: In the query parse process, all WindowExpressions are originally placed in the projectList of a Project operator or the aggregateExpressions of an Aggregate operator. It makes our changes to simple and keep all of parsing rules for window functions at a single place (nodesToWindowSpecification). For the WINDOWclause in a query, we use a WithWindowDefinition as the container as the mapping from the name of a window specification to a WindowSpecDefinition. This changes is similar with our common table expression support.
* Analysis: The query analysis process has three steps for window functions.
* Resolve all WindowSpecReferences by replacing them with WindowSpecReferences according to the mapping table stored in the node of WithWindowDefinition.
* Resolve WindowFunctions in the projectList of a Project operator or the aggregateExpressions of an Aggregate operator. For this PR, we use Hive's functions for window functions because we will have a major refactoring of our internal UDAFs and it is better to switch our UDAFs after that refactoring work.
* Once we have resolved all WindowFunctions, we will use ResolveWindowFunction to extract WindowExpressions from projectList and aggregateExpressions and then create a Window operator for every distinct WindowSpecDefinition. With this choice, at the execution time, we can rely on the Exchange operator to do all of work on reorganizing the table and we do not need to worry about it in the physical Window operator. An example analyzed plan is shown as follows
```
sql("""
SELECT
year, country, product, sales,
avg(sales) over(partition by product) avg_product,
sum(sales) over(partition by country) sum_country
FROM sales
ORDER BY year, country, product
""").explain(true)
== Analyzed Logical Plan ==
Sort [year#34 ASC,country#35 ASC,product#36 ASC], true
Project [year#34,country#35,product#36,sales#37,avg_product#27,sum_country#28]
Window [year#34,country#35,product#36,sales#37,avg_product#27], [HiveWindowFunction#org.apache.hadoop.hive.ql.udf.generic.GenericUDAFSum(sales#37) WindowSpecDefinition [country#35], [], ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING AS sum_country#28], WindowSpecDefinition [country#35], [], ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING
Window [year#34,country#35,product#36,sales#37], [HiveWindowFunction#org.apache.hadoop.hive.ql.udf.generic.GenericUDAFAverage(sales#37) WindowSpecDefinition [product#36], [], ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING AS avg_product#27], WindowSpecDefinition [product#36], [], ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING
Project [year#34,country#35,product#36,sales#37]
MetastoreRelation default, sales, None
```
* Query planning: In the process of query planning, we simple generate the physical Window operator based on the logical Window operator. Then, to prepare the executedPlan, the EnsureRequirements rule will add Exchange and Sort operators if necessary. The EnsureRequirements rule will analyze the data properties and try to not add unnecessary shuffle and sort. The physical plan for the above example query is shown below.
```
== Physical Plan ==
Sort [year#34 ASC,country#35 ASC,product#36 ASC], true
Exchange (RangePartitioning [year#34 ASC,country#35 ASC,product#36 ASC], 200), []
Window [year#34,country#35,product#36,sales#37,avg_product#27], [HiveWindowFunction#org.apache.hadoop.hive.ql.udf.generic.GenericUDAFSum(sales#37) WindowSpecDefinition [country#35], [], ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING AS sum_country#28], WindowSpecDefinition [country#35], [], ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING
Exchange (HashPartitioning [country#35], 200), [country#35 ASC]
Window [year#34,country#35,product#36,sales#37], [HiveWindowFunction#org.apache.hadoop.hive.ql.udf.generic.GenericUDAFAverage(sales#37) WindowSpecDefinition [product#36], [], ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING AS avg_product#27], WindowSpecDefinition [product#36], [], ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING
Exchange (HashPartitioning [product#36], 200), [product#36 ASC]
HiveTableScan [year#34,country#35,product#36,sales#37], (MetastoreRelation default, sales, None), None
```
* Execution time: At execution time, a physical Window operator buffers all rows in a partition specified in the partition spec of a OVER clause. If necessary, it also maintains a sliding window frame. The current implementation tries to buffer the input parameters of a window function according to the window frame to avoid evaluating a row multiple times.
# Future work #
Here are three improvements that are not hard to add:
* Taking advantage of the window frame specification to reduce the number of rows buffered in the physical Window operator. For some cases, we only need to buffer the rows appearing in the sliding window. But for other cases, we will not be able to reduce the number of rows buffered (e.g. ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING).
* When aRAGEN frame is used, for <value> PRECEDING and <value> FOLLOWING, it will be great if the <value> part is an expression (we can start with Literal). So, when the data type of ORDER BY expression is a FractionalType, we can support FractionalType as the type <value> (<value> still needs to be evaluated as a positive value).
* When aRAGEN frame is used, we need to support DateType and TimestampType as the data type of the expression appearing in the order specification. Then, the <value> part of <value> PRECEDING and <value> FOLLOWING can support interval types (once we support them).
This is a joint work with guowei2 and yhuai
Thanks hbutani hvanhovell for his comments
Thanks scwf for his comments and unit tests
Author: Yin Huai <yhuai@databricks.com>
Closes#5604 from guowei2/windowImplement and squashes the following commits:
76fe1c8 [Yin Huai] Implementation.
aa2b0ae [Yin Huai] Tests.
huangjs
Acutally spark sql will first go through analysis period, in which we do widen types and promote strings, and then optimization, where constant IN will be converted into INSET.
So it turn out that we only need to fix this for IN.
Author: Daoyuan Wang <daoyuan.wang@intel.com>
Closes#4945 from adrian-wang/inset and squashes the following commits:
71e05cc [Daoyuan Wang] minor fix
581fa1c [Daoyuan Wang] mysql way
f3f7baf [Daoyuan Wang] address comments
5eed4bc [Daoyuan Wang] promote string and do widen types for IN
