arrow builder: speed up bitsetting for large length #3

atumanov · 2017-03-05T05:24:52Z

No description provided.

wesm · 2017-03-09T13:15:26Z

Can you submit a pull request to Arrow? Thanks

wesm · 2017-03-09T13:16:25Z

https://issues.apache.org/jira/browse/ARROW-607

robertnishihara · 2017-03-09T18:17:47Z

I think we just copied this change from the upstream arrow (from this commit apache@ad01575).

robertnishihara · 2017-03-09T18:18:51Z

We're planning on rebasing on the Arrow master soon.

atumanov · 2017-03-10T01:45:18Z

yes, this was an interim solution until we update arrow. I cherry picked the code and just changed the types to fit our current code base.

…lue data Modified BinaryBuilder::Resize(int64_t) so that when building BinaryArrays with a known size, space is also reserved for value_data_builder_ to prevent internal reallocation. Author: Panchen Xue <pan.panchen.xue@gmail.com> Closes apache#1481 from xuepanchen/master and squashes the following commits: 707b67b [Panchen Xue] ARROW-1712: [C++] Fix lint errors 360e601 [Panchen Xue] Merge branch 'master' of https://github.com/xuepanchen/arrow d4bbd15 [Panchen Xue] ARROW-1712: [C++] Modify test case for BinaryBuilder::ReserveData() and change arguments for offsets_builder_.Resize() 77f8f3c [Panchen Xue] Merge pull request #5 from apache/master bc5db7d [Panchen Xue] ARROW-1712: [C++] Remove unneeded data member in BinaryBuilder and modify test case 5a5b70e [Panchen Xue] Merge pull request #4 from apache/master 8e4c892 [Panchen Xue] Merge pull request #3 from xuepanchen/xuepanchen-arrow-1712 d3c8202 [Panchen Xue] ARROW-1945: [C++] Fix a small typo 0b07895 [Panchen Xue] ARROW-1945: [C++] Add data_capacity_ to track capacity of value data 18f90fb [Panchen Xue] ARROW-1945: [C++] Add data_capacity_ to track capacity of value data bbc6527 [Panchen Xue] ARROW-1945: [C++] Update test case for BinaryBuild data value space reservation 15e045c [Panchen Xue] Add test case for array-test.cc 5a5593e [Panchen Xue] Update again ReserveData(int64_t) method for BinaryBuilder 9b5e805 [Panchen Xue] Update ReserveData(int64_t) method signature for BinaryBuilder 8dd5eaa [Panchen Xue] Update builder.cc b002e0b [Panchen Xue] Remove override keyword from ReserveData(int64_t) method for BinaryBuilder de318f4 [Panchen Xue] Implement ReserveData(int64_t) method for BinaryBuilder e0434e6 [Panchen Xue] Add ReserveData(int64_t) and value_data_capacity() for methods for BinaryBuilder 5ebfb32 [Panchen Xue] Add capacity() method for TypedBufferBuilder 5b73c1c [Panchen Xue] Update again BinaryBuilder::Resize(int64_t capacity) in builder.cc d021c54 [Panchen Xue] Merge pull request #2 from xuepanchen/xuepanchen-arrow-1712 232024e [Panchen Xue] Update BinaryBuilder::Resize(int64_t capacity) in builder.cc c2f8dc4 [Panchen Xue] Merge pull request #1 from apache/master

@trxcllnt

This PR moves the `Table` class out of the Vector hierarchy and adds optimized dataframe operations to it. Currently implements an optimized `scan()` method, `filter(predicate)`, `count()`, and `countBy(column_name)` (only works on dictionary-encoded columns). Some usage examples, based on the file generated by `js/test/data/tables/generate.py`: ``` js > let table = Table.from(...); > table.count() 1000000 > table.filter(col('lat').gteq(0)).count() 499718 > table.countBy('origin').toJSON() { Charlottesville: 166839, 'New York': 166251, 'San Francisco': 166642, Seattle: 166659, 'Terre Haute': 166756, 'Washington, DC': 166853 } > table.filter(col('lng').gteq(0)).countBy('origin').toJSON() { Charlottesville: 83109, 'New York': 83221, 'San Francisco': 83515, Seattle: 83362, 'Terre Haute': 83314, 'Washington, DC': 83479 } ``` There are performance tests for the dataframe operations, to run them you must first generate the test data by running `npm run create:perfdata`. The PR also includes @trxcllnt's refactor of the JS implementation to make it more closely resemble the C++ implementation. This refactor resolves multiple JIRAs: ARROW-1903, ARROW-1898, ARROW-1502, ARROW-1952 (partially), and ARROW-1985 Author: Paul Taylor <paul.e.taylor@me.com> Author: Brian Hulette <brian.hulette@ccri.com> Author: Brian Hulette <hulettbh@gmail.com> Closes apache#1482 from TheNeuralBit/table-scan-perf and squashes the following commits: 52f1e0e [Brian Hulette] <, > are not commutative, misc cleanup 04b1838 [Brian Hulette] even more table tests 16b9ccb [Brian Hulette] Merge pull request #4 from trxcllnt/js-cpp-refactor fe300df [Paul Taylor] fix closure es5/umd toString() iterator 3d5240a [Paul Taylor] fix more externs 10c48ad [Paul Taylor] Merge branch 'table-scan-perf' of github.com:ccri/arrow into js-cpp-refactor dbe7f81 [Brian Hulette] Add more Table unit tests 1910962 [Brian Hulette] Add optional bind callback to scan 5bdf17f [Brian Hulette] Fix perf 8cf2473 [Brian Hulette] Merge remote-tracking branch 'origin/master' into table-scan-perf 4a41b18 [Paul Taylor] add src/predicate to the list of exports we should save from uglify 5a91fab [Paul Taylor] add more view, predicate externs f6adfb3 [Brian Hulette] Create predicate namespace f7bb0ed [Paul Taylor] Merge branch 'table-scan-perf' of github.com:ccri/arrow into js-cpp-refactor e148ee4 [Paul Taylor] Merge branch 'extern-woes' into js-cpp-refactor 25cdc4a [Paul Taylor] add src/predicate to the list of exports we should save from uglify dc7c728 [Paul Taylor] add more view, predicate externs 25e6af7 [Brian Hulette] Create predicate namespace 579ab1f [Brian Hulette] Merge pull request #2 from trxcllnt/js-cpp-refactor f3cde1a [Paul Taylor] fix lint 9769773 [Paul Taylor] fix vector perf tests 016ba78 [Brian Hulette] Merge pull request #1 from trxcllnt/js-cpp-refactor 272d293 [Paul Taylor] Merge pull request #4 from ccri/empty-table 7bc7363 [Brian Hulette] Fix exception for empty Table 8ddce0a [Paul Taylor] check bounds in getChildAt(i) to avoid NPEs f1dead0 [Paul Taylor] compute chunked nested childData list correctly 18807c6 [Paul Taylor] rename ChunkData's fields so it's more clear they're not semantically similar to other similarly named fields 7e43b78 [Paul Taylor] add test:integration npm script a5f200f [Paul Taylor] Merge pull request #3 from ccri/table-from-struct c8cd286 [Brian Hulette] Add Table.fromStruct a00415e [Brian Hulette] Fix perf 54d4f5b [Paul Taylor] lazily allocate table and recordbatch columns, support NestedView's getChildAt(i) method in ChunkedView 40b3638 [Paul Taylor] run integration tests with local data for coverage stats fe31ee0 [Paul Taylor] slice the flat data values before returning an iterator of them e537789 [Paul Taylor] make it easier to run all integration tests from local data c0fd2f9 [Paul Taylor] use the dictionary of the last chunked vector list for chunked dictionary vectors e33c068 [Paul Taylor] Merge pull request #2 from ccri/fixed-size-list 5bb63af [Brian Hulette] Don't read OFFSET vector for FixedSizeList 614b688 [Paul Taylor] add asEpochMs to date and timestamp vectors 87334a5 [Paul Taylor] Merge branch 'table-scan-perf' of github.com:ccri/arrow into js-cpp-refactor b7f5bfb [Paul Taylor] rename numRows to length, add table.getColumn() e81082f [Paul Taylor] export vector views, allow cloning data as another type 700a47c [Paul Taylor] export visitors e859e13 [Paul Taylor] fix package.json bin entry 0620cfd [Brian Hulette] use Math.fround 0126dc4 [Brian Hulette] Don't recompute total length e761eee [Brian Hulette] Rename asJSON to toJSON 6c91ed4 [Paul Taylor] Merge branch 'master' of github.com:apache/arrow into js-cpp-refactor-merge_with-table-scan-perf d2b18d5 [Paul Taylor] Merge remote-tracking branch 'ccri/table-scan-perf' into js-cpp-refactor-merge_with-table-scan-perf f3f3b86 [Paul Taylor] rename table.ts to recordbatch.ts in preparation for merging latest e3f629d [Paul Taylor] fix rest of the mangling issues fa7c17a [Paul Taylor] passing all tests except es5 umd mangler ones e20decd [Brian Hulette] Add license headers edcbdbe [Brian Hulette] cleanup 20717d5 [Brian Hulette] Fixed countBy(string) 7244887 [Brian Hulette] Add table unit tests... 6719147 [Brian Hulette] Add DataFrame.countBy operation 2f4a349 [Brian Hulette] Minor tweaks 2e118ab [Brian Hulette] linter a788db3 [Brian Hulette] Cleanup a9fff89 [Brian Hulette] Move Table out of the Vector hierarchy 1d60aa1 [Brian Hulette] Moved DataFrame ops to Table. DataFrame is now an interface e8979ba [Brian Hulette] Refactor DataFrame to extend Vector<StructRow> 6a41d68 [Brian Hulette] clean up table benchmarks 2744c63 [Brian Hulette] Remove Chunked/Simple DataFrame distinction aa999f8 [Brian Hulette] Add DictionaryVector optimization for equals predicate 4d9e8c0 [Brian Hulette] Add concept of predicates for filtering dataframes 796f45d [Brian Hulette] add DataFrame filter and count ops 30f0330 [Brian Hulette] Add basic DataFrame impl ... a1edac2 [Brian Hulette] Add perf tests for table scans d18d915 [Paul Taylor] fix struct and map rows 61dc699 [Paul Taylor] WIP -- refactor types to closer match arrow-cpp 62db338 [Paul Taylor] update dependencies and add es6+ umd targets to jest transform ignore patterns to fix ci 6ff18e9 [Paul Taylor] ship es2015 commonJS in main package to avoid confusion 74e828a [Paul Taylor] fix typings issues (ARROW-1903)

I am contributing to [Arrow 3731](https://issues.apache.org/jira/browse/ARROW-3731). This PR has the minimum functionality to read parquet files into an arrow::Table, which can then be converted to a tibble. Multiple parquet files can be read inside `lapply`, and then concatenated at the end. Steps to compile 1) Build arrow and parquet c++ projects 2) In R run `devtools::load_all()` What I could use help with: The biggest challenge for me is my lack of experience with pkg-config. The R library has a `configure` file which uses pkg-config to figure out what c++ libraries to link to. Currently, `configure` looks up the Arrow project and links to -larrow only. We need it to also link to -lparquet. I do not know how to modify pkg-config's metadata to let it know to link to both -larrow and -lparquet Author: Jeffrey Wong <jeffreyw@netflix.com> Author: Romain Francois <romain@purrple.cat> Author: jeffwong-nflx <jeffreyw@netflix.com> Closes apache#3230 from jeffwong-nflx/master and squashes the following commits: c67fa3d <jeffwong-nflx> Merge pull request #3 from jeffwong-nflx/cleanup 1df3026 <Jeffrey Wong> don't hard code -larrow and -lparquet 8ccaa51 <Jeffrey Wong> cleanup 75ba5c9 <Jeffrey Wong> add contributor 56adad2 <jeffwong-nflx> Merge pull request #2 from romainfrancois/3731/parquet-2 7d6e64d <Romain Francois> read_parquet() only reading one parquet file, and gains a `as_tibble` argument e936b44 <Romain Francois> need parquet on travis too ff260c5 <Romain Francois> header was too commented, renamed to parquet.cpp 9e1897f <Romain Francois> styling etc ... 456c5d2 <Jeffrey Wong> read parquet files 22d89dd <Jeffrey Wong> hardcode -larrow and -lparquet

https://issues.apache.org/jira/browse/ARROW-3965 This creates an object which configures the BaseAllocator and Calendar used during to configure the translation from a JDBC ResultSet to an Arrow vector. Author: Mike Pigott <mpigott@gmail.com> Author: Michael Pigott <mikepigott@users.noreply.github.com> Closes apache#3133 from mikepigott/jdbc-to-arrow-config and squashes the following commits: be95426 <Mike Pigott> ARROW-3965: JDBC-To-Arrow Config Builder javadocs. d6c64a7 <Mike Pigott> ARROW-3965: JdbcToArrowConfigBuilder d7ca982 <Mike Pigott> Merge branch 'master' into jdbc-to-arrow-config 789c8c8 <Michael Pigott> Merge pull request #4 from apache/master e5b19ee <Michael Pigott> Merge pull request #3 from apache/master 3b17c29 <Michael Pigott> Merge pull request #2 from apache/master 5b1b364 <Mike Pigott> Merge branch 'master' into jdbc-to-arrow-config 881c6c8 <Michael Pigott> Merge pull request #1 from apache/master bb3165b <Mike Pigott> Updating the function calls to use the JdbcToArrowConfig versions. 68c91e7 <Mike Pigott> Modifying the jdbcToArrowSchema and jdbcToArrowVectors methods to receive JdbcToArrowConfig objects. 8d6cf00 <Mike Pigott> Documentation for public static VectorSchemaRoot sqlToArrow(Connection connection, String query, JdbcToArrowConfig config) 4f1260c <Mike Pigott> Adding documentation for public static VectorSchemaRoot sqlToArrow(ResultSet resultSet, JdbcToArrowConfig config) df632e3 <Mike Pigott> Updating the SQL tests to include JdbcToArrowConfig versions. b270044 <Mike Pigott> Updated validaton & documentation, and unit tests for the new JdbcToArrowConfig. da77cbe <Mike Pigott> Creating a configuration class for the JDBC-to-Arrow converter.

https://issues.apache.org/jira/browse/ARROW-3923 Hello! I was reading through the JDBC source code and I noticed that a java.util.Calendar was required for creating an Arrow Schema and Arrow Vectors from a JDBC ResultSet, when none is required. This change makes the Calendar optional. Unit Tests: The existing SureFire plugin configuration uses a UTC calendar for the database, which is the default Calendar in the existing code. Likewise, no changes to the unit tests are required to provide adequate coverage for the change. Author: Michael Pigott <mikepigott@users.noreply.github.com> Author: Mike Pigott <mpigott@gmail.com> Closes apache#3066 from mikepigott/jdbc-timestamp-no-calendar and squashes the following commits: 4d95da0 <Mike Pigott> ARROW-3923: Supporting a null Calendar in the config, and reverting the breaking change. cd9a230 <Mike Pigott> Merge branch 'master' into jdbc-timestamp-no-calendar 509a1cc <Michael Pigott> Merge pull request #5 from apache/master 789c8c8 <Michael Pigott> Merge pull request #4 from apache/master e5b19ee <Michael Pigott> Merge pull request #3 from apache/master 3b17c29 <Michael Pigott> Merge pull request #2 from apache/master 881c6c8 <Michael Pigott> Merge pull request #1 from apache/master 089cff4 <Mike Pigott> Format fixes a58a4a5 <Mike Pigott> Fixing calendar usage. e12832a <Mike Pigott> Allowing for timestamps without a time zone.

https://issues.apache.org/jira/browse/ARROW-3966 This change includes apache#3133, and supports a new configuration item called "Include Metadata." If true, metadata from the JDBC ResultSetMetaData object is pulled along to the Schema Field Metadata. For now, this includes: * Catalog Name * Table Name * Column Name * Column Type Name Author: Mike Pigott <mpigott@gmail.com> Author: Michael Pigott <mikepigott@users.noreply.github.com> Closes apache#3134 from mikepigott/jdbc-column-metadata and squashes the following commits: 02f2f34 <Mike Pigott> ARROW-3966: Picking up lost change to support null calendars. 7049c36 <Mike Pigott> Merge branch 'master' into jdbc-column-metadata e9a9b2b <Michael Pigott> Merge pull request apache#6 from apache/master 65741a9 <Mike Pigott> ARROW-3966: Code review feedback cc6cc88 <Mike Pigott> ARROW-3966: Using a 1:N loop instead of a 0:N-1 loop for fewer index offsets in code. cfb2ba6 <Mike Pigott> ARROW-3966: Using a helper method for building a UTC calendar with root locale. 2928513 <Mike Pigott> ARROW-3966: Moving the metadata flag assignment into the builder. 69022c2 <Mike Pigott> ARROW-3966: Fixing merge. 4a6de86 <Mike Pigott> Merge branch 'master' into jdbc-column-metadata 509a1cc <Michael Pigott> Merge pull request #5 from apache/master 789c8c8 <Michael Pigott> Merge pull request #4 from apache/master e5b19ee <Michael Pigott> Merge pull request #3 from apache/master 3b17c29 <Michael Pigott> Merge pull request #2 from apache/master d847ebc <Mike Pigott> Fixing file location 1ceac9e <Mike Pigott> Merge branch 'master' into jdbc-column-metadata 881c6c8 <Michael Pigott> Merge pull request #1 from apache/master 03091a8 <Mike Pigott> Unit tests for including result set metadata. 72d64cc <Mike Pigott> Affirming the field metadata is empty when the configuration excludes field metadata. 7b4527c <Mike Pigott> Test for the include-metadata flag in the configuration. 7e9ce37 <Mike Pigott> Merge branch 'jdbc-to-arrow-config' into jdbc-column-metadata bb3165b <Mike Pigott> Updating the function calls to use the JdbcToArrowConfig versions. a6fb1be <Mike Pigott> Fixing function call 5bfd6a2 <Mike Pigott> Merge branch 'jdbc-to-arrow-config' into jdbc-column-metadata 68c91e7 <Mike Pigott> Modifying the jdbcToArrowSchema and jdbcToArrowVectors methods to receive JdbcToArrowConfig objects. b5b0cb1 <Mike Pigott> Merge branch 'jdbc-to-arrow-config' into jdbc-column-metadata 8d6cf00 <Mike Pigott> Documentation for public static VectorSchemaRoot sqlToArrow(Connection connection, String query, JdbcToArrowConfig config) 4f1260c <Mike Pigott> Adding documentation for public static VectorSchemaRoot sqlToArrow(ResultSet resultSet, JdbcToArrowConfig config) e34a9e7 <Mike Pigott> Fixing formatting. fe097c8 <Mike Pigott> Merge branch 'jdbc-to-arrow-config' into jdbc-column-metadata df632e3 <Mike Pigott> Updating the SQL tests to include JdbcToArrowConfig versions. b270044 <Mike Pigott> Updated validaton & documentation, and unit tests for the new JdbcToArrowConfig. da77cbe <Mike Pigott> Creating a configuration class for the JDBC-to-Arrow converter. a78c770 <Mike Pigott> Updating Javadocs. 523387f <Mike Pigott> Updating the API to support an optional 'includeMetadata' field. 5af1b5b <Mike Pigott> Separating out the field-type creation from the field creation.

…igNum - Use stride in `DataBufferBuilder.set` to ensure that Int64Builder is consistent with and without BigNum. - Use `WideBufferBuilder` for `Int64` and `Uint64` builders, even when `BigInt` is not available. Moves check for `BigInt` availability into `WideBufferBuilder`. (Thanks to Paul Taylor) Author: Brian Hulette <hulettbh@gmail.com> Author: ptaylor <paul.e.taylor@me.com> Closes apache#4691 from TheNeuralBit/js-data-buffer-builder-stride and squashes the following commits: 7862612 <Brian Hulette> Add clarifying comment for int64 generation in builder tests b08ac1b <Brian Hulette> Merge pull request #3 from trxcllnt/js/data-buffer-builder-64bit-stride 2e4ce7a <Brian Hulette> Use stride in DataBufferBuilder.set 4567d07 <ptaylor> use WideBufferBuilder for Int64 and Uint64 builders dc9ed3a <ptaylor> update package-lock.json

This updates the language in `install_arrow()` to follow the README revision that will land in https://github.com/apache/arrow/pull/4948/files#diff-563b2cb2c8c2d51b2ff6b177e2d84286R33. The [Jira ticket](https://issues.apache.org/jira/browse/ARROW-6142) requested three things; this is `#2` in the list. On `#1`, I defer to the C++ installation docs, which are already included in the install_arrow message, rather than duplicating content here. `#3` is out of scope. Closes apache#5027 from nealrichardson/no-ppa and squashes the following commits: 80b142e <Neal Richardson> s/arrow/Arrow/ 44c9659 <Neal Richardson> Tweak language again 36cfe28 <Neal Richardson> Further linux install revisions 79bd7e0 <Neal Richardson> One more PPurge 63f75bd <Neal Richardson> Revise install_arrow instructions for Linux Authored-by: Neal Richardson <neal.p.richardson@gmail.com> Signed-off-by: Sutou Kouhei <kou@clear-code.com>

builder: speed up bitsetting for large length

bdae7a2

pcmoritz merged commit ae9c5d9 into pcmoritz:master Mar 5, 2017

atumanov deleted the fast-bitsetting branch March 10, 2017 01:46

pcmoritz added a commit that referenced this pull request Dec 23, 2018

Replace std::future with boost::future (#3)

8ea756d

pcmoritz added a commit that referenced this pull request Feb 5, 2019

Replace std::future with boost::future (#3)

bd26d55

toddfarmer mentioned this pull request Mar 13, 2017

[C++] Speed up bitsetting in ArrayBuilder::UnsafeSetNotNull toddfarmer/arrow-migration#493

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

arrow builder: speed up bitsetting for large length #3

arrow builder: speed up bitsetting for large length #3

Uh oh!

atumanov commented Mar 5, 2017

Uh oh!

wesm commented Mar 9, 2017

Uh oh!

wesm commented Mar 9, 2017

Uh oh!

robertnishihara commented Mar 9, 2017

Uh oh!

robertnishihara commented Mar 9, 2017

Uh oh!

atumanov commented Mar 10, 2017

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

arrow builder: speed up bitsetting for large length #3

arrow builder: speed up bitsetting for large length #3

Uh oh!

Conversation

atumanov commented Mar 5, 2017

Uh oh!

wesm commented Mar 9, 2017

Uh oh!

wesm commented Mar 9, 2017

Uh oh!

robertnishihara commented Mar 9, 2017

Uh oh!

robertnishihara commented Mar 9, 2017

Uh oh!

atumanov commented Mar 10, 2017

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants