Code

[MRG] Openml data loader#11419

Code

Merged

jnothman merged 153 commits intoscikit-learn:masterscikit-learn/scikit-learn:masterfrom

janvanrijn:masterCopy head branch name to clipboard

Aug 15, 2018

Commits

Commits on Oct 10, 2017

• start on openml dataset loader

  amueller

  committed
  268f533

  Copy full SHA for 268f533

Commits on Oct 11, 2017

• working on stuff

  amueller

  committed
  4f3e93e

  Copy full SHA for 4f3e93e

Commits on Oct 12, 2017

• first version working

  amueller

  committed
  fabaa90

  Copy full SHA for fabaa90

• docstrings, use version="active" as default

  amueller

  committed
  1804b14

  Copy full SHA for 1804b14

• add caching to openml loader

  amueller

  committed
  ffd4335

  Copy full SHA for ffd4335

• pep8 annoyance

  amueller

  committed
  fe0904b

  Copy full SHA for fe0904b

• fix download url, allow datasets without target

  amueller

  committed
  eea026e

  Copy full SHA for eea026e

• allow specifying the target column, starting on docs

  amueller

  committed
  bca12e9

  Copy full SHA for bca12e9

• add openml to the narrative docs

  amueller

  committed
  f59ce8b

  Copy full SHA for f59ce8b

• get more people to upload stuff to openml.

  amueller

  committed
  d7dee6d

  Copy full SHA for d7dee6d

• store metadata, convert to dtype object if there is nominal data.

  amueller

  committed
  4c19ad9

  Copy full SHA for 4c19ad9

• fix doctests, add fetch_openml to __init__

  amueller

  committed
  16b7fed

  Copy full SHA for 16b7fed

Commits on Oct 17, 2017

• make arff reading work in python2.7

  amueller

  committed
  b3f6c36

  Copy full SHA for b3f6c36

Commits on Oct 24, 2017

• ignore doctests for now because of unicode issues

  amueller

  committed
  dc401f2

  Copy full SHA for dc401f2

• add version filter.

  amueller

  committed
  d8cfd37

  Copy full SHA for d8cfd37

Commits on Nov 14, 2017

• some typos, addressing joel's comments, working on better errors

  amueller

  committed
  6f6bb57

  Copy full SHA for 6f6bb57

Commits on Nov 15, 2017

• nicer error message on non-existing ID

  amueller

  committed
  b5c72d9

  Copy full SHA for b5c72d9

• minor improvements to data wrangling

  amueller

  committed
  64483f8

  Copy full SHA for 64483f8

• allow downloading inactive datasets if specified by name and version

  amueller

  committed
  26aaff2

  Copy full SHA for 26aaff2

• update mice version 4 dataset id

  amueller

  committed
  b3b9276

  Copy full SHA for b3b9276

• Merge branch 'master' of github.com:scikit-learn/scikit-learn into openml_loader

  amueller

  committed
  b2c283a

  Copy full SHA for b2c283a

• add whatsnew entry

  amueller

  committed
  7e91c71

  Copy full SHA for 7e91c71

• add unicode and normalize whitespace flags to pytest config

  amueller

  committed
  11909d5

  Copy full SHA for 11909d5

Commits on Nov 22, 2017

• Merge branch 'master' into openml_loader

  amueller

  committed
  126f406

  Copy full SHA for 126f406

• add test for fetch_openml

  amueller

  committed
  7e16203

  Copy full SHA for 7e16203

• test error messages

  amueller

  committed
  8dcb26b

  Copy full SHA for 8dcb26b

• fix command for make test-coverage

  amueller

  committed
  0d562b6

  Copy full SHA for 0d562b6

• Merge branch 'fix_test_coverage' into openml_loader

  amueller

  committed
  c2266c5

  Copy full SHA for c2266c5

• make flake8 green

  amueller

  committed
  e274ad3

  Copy full SHA for e274ad3

• py35 compatiility

  amueller

  committed
  eb39a01

  Copy full SHA for eb39a01

Commits on Dec 21, 2017

• Merge branch 'master' into openml_loader

  Show description for c56b549
  amueller

  committed
  c56b549

  Copy full SHA for c56b549

• trying to use CSV interface

  amueller

  committed
  67825e8

  Copy full SHA for 67825e8

Commits on Jul 2, 2018

• Merge branch 'openml_loader' of https://github.com/amueller/scikit-learn into amueller-openml_loader

  Show description for e4ab363
  janvanrijn

  committed
  e4ab363

  Copy full SHA for e4ab363

• Merge branch 'amueller-openml_loader'

  janvanrijn

  committed
  7894622

  Copy full SHA for 7894622

• packed liac-arff and added appropriate utility functions

  janvanrijn

  committed
  f6e6c8e

  Copy full SHA for f6e6c8e

Commits on Jul 3, 2018

• extended unit tests

  janvanrijn

  committed
  64724d1

  Copy full SHA for 64724d1

• added comment

  janvanrijn

  committed
  4ac219c

  Copy full SHA for 4ac219c

Commits on Jul 6, 2018

• added sparse arff

  janvanrijn

  committed
  73ff417

  Copy full SHA for 73ff417

• added mock files

  janvanrijn

  committed
  a18dcff

  Copy full SHA for a18dcff

• mocked data description and data features

  janvanrijn

  committed
  f8ad349

  Copy full SHA for f8ad349

Commits on Jul 9, 2018

• changed mocking structure to not depend on unittest.mock (for Python 2 compatibility)

  janvanrijn

  committed
  720d4f6

  Copy full SHA for 720d4f6

Commits on Jul 12, 2018

• removed panda dependency

  janvanrijn

  committed
  8a4732b

  Copy full SHA for 8a4732b

Commits on Jul 14, 2018

• small updated incorporating several of the comments on the pr by @rth and @jnothman

  janvanrijn

  committed
  593ed17

  Copy full SHA for 593ed17

• improved parameterization of fetch openml (split id and name into two arguments)

  Show description for 60ee9a2
  janvanrijn

  committed
  60ee9a2

  Copy full SHA for 60ee9a2

• extended functionality and unit tests for sparse matrices

  janvanrijn

  committed
  fe7cfc9

  Copy full SHA for fe7cfc9

• added __init__.py for externals/liac-arff (neccessary for Python 2)

  janvanrijn

  committed
  be9332b

  Copy full SHA for be9332b

• removed monkeypatch context manager from with block

  janvanrijn

  committed
  3a537b3

  Copy full SHA for 3a537b3

• adapted test paths

  janvanrijn

  committed
  757552a

  Copy full SHA for 757552a

• monkeypatched anneal test

  janvanrijn

  committed
  266e722

  Copy full SHA for 266e722

Commits on Jul 15, 2018

• fixed travis doc test and line size

  janvanrijn

  committed
  c7c83a6

  Copy full SHA for c7c83a6

• fix indent for unit test

  janvanrijn

  committed
  280292c

  Copy full SHA for 280292c

• doctest

  janvanrijn

  committed
  63a6167

  Copy full SHA for 63a6167

• mocked more unit tests

  janvanrijn

  committed
  8856a91

  Copy full SHA for 8856a91

• mocked ALL unit tests

  janvanrijn

  committed
  6a834e8

  Copy full SHA for 6a834e8

• fixed doc test, added unit test with ignore attributes and row id

  janvanrijn

  committed
  d5f0441

  Copy full SHA for d5f0441

• added openml reference paper to documentation

  janvanrijn

  committed
  f13ed63

  Copy full SHA for f13ed63

• suggestions by @amueller

  janvanrijn

  committed
  579d10e

  Copy full SHA for 579d10e

Commits on Jul 16, 2018

• moved liac-arff to general externals directory (for appveyor)

  janvanrijn

  committed
  2dc3e2c

  Copy full SHA for 2dc3e2c

• improved docstring based on Joris' comments

  janvanrijn

  committed
  2659dd9

  Copy full SHA for 2659dd9

• changed order of arguments

  janvanrijn

  committed
  403576b

  Copy full SHA for 403576b

• openml doc improvements

  janvanrijn

  committed
  5c65cb7

  Copy full SHA for 5c65cb7

• improved doc+tests

  janvanrijn

  committed
  f8cf13f

  Copy full SHA for f8cf13f

• code improvements requested by Joris

  janvanrijn

  committed
  bf10a86

  Copy full SHA for bf10a86

• doc fixes addressing review comments

  janvanrijn

  committed
  6d4bd90

  Copy full SHA for 6d4bd90

Commits on Jul 17, 2018

• incorporated many suggestions from the code review:

  Show description for 3f8128a
  janvanrijn

  committed
  3f8128a

  Copy full SHA for 3f8128a

• final touch on json load factorization

  Show description for 7ee5329
  janvanrijn

  committed
  7ee5329

  Copy full SHA for 7ee5329

• more replies to reviews

  janvanrijn

  committed
  aedfefd

  Copy full SHA for aedfefd

• changed arff import

  janvanrijn

  committed
  defb1a6

  Copy full SHA for defb1a6

• fixed line lengths

  janvanrijn

  committed
  b731b0d

  Copy full SHA for b731b0d

• extended unit test by checking the output data arrays

  janvanrijn

  committed
  c7faee2

  Copy full SHA for c7faee2

• removed unused variable

  janvanrijn

  committed
  a8dec37

  Copy full SHA for a8dec37

• aligned whatsnew

  janvanrijn

  committed
  26b5c28

  Copy full SHA for 26b5c28

• changed line size

  janvanrijn

  committed
  440eb86

  Copy full SHA for 440eb86

• name lower case

  Show description for 68ffece
  janvanrijn

  committed
  68ffece

  Copy full SHA for 68ffece

• renamed mock files (to lowercase)

  Show description for 1cec31b
  janvanrijn

  committed
  1cec31b

  Copy full SHA for 1cec31b

• added support for multi-target classification

  janvanrijn

  committed
  b5cce8c

  Copy full SHA for b5cce8c

• fix visual indent error

  janvanrijn

  committed
  02108c4

  Copy full SHA for 02108c4

• gzipped test data

  janvanrijn

  committed
  a01e8ba

  Copy full SHA for a01e8ba

Commits on Jul 18, 2018

• Merge pull request #2 from scikit-learn/master

  Show description for b45fea6
  janvanrijn

  authored
  b45fea6

  Copy full SHA for b45fea6

Commits on Jul 19, 2018

• url open fix

  janvanrijn

  committed
  6ea8cfb

  Copy full SHA for 6ea8cfb

• added more tests

  janvanrijn

  committed
  6e68177

  Copy full SHA for 6e68177

Commits on Jul 20, 2018

• fix flake8 stuff

  janvanrijn

  committed
  29046e1

  Copy full SHA for 29046e1

• added sanity check unit test

  janvanrijn

  committed
  40b5bb1

  Copy full SHA for 40b5bb1

• small fixes for travis

  janvanrijn

  committed
  7c921c0

  Copy full SHA for 7c921c0

• see if loading a non-binary file works in gzip

  janvanrijn

  committed
  004b831

  Copy full SHA for 004b831

• fix unit test?

  janvanrijn

  committed
  02ae09f

  Copy full SHA for 02ae09f

• added more explicit error when unicode error arises

  janvanrijn

  committed
  09b0fd8

  Copy full SHA for 09b0fd8

• changed back to 'rb'

  janvanrijn

  committed
  bc7fd91

  Copy full SHA for bc7fd91

• solving all my unicode problems

  janvanrijn

  committed
  26a5f51

  Copy full SHA for 26a5f51

• removed str hack, proper way to handle unicode across Python versions

  janvanrijn

  committed
  e267f16

  Copy full SHA for e267f16

• improved path.join handling, so Windows also likes the OpenML extension

  janvanrijn

  committed
  931c93f

  Copy full SHA for 931c93f

• removed strict string check

  janvanrijn

  committed
  cdca931

  Copy full SHA for cdca931

• changed assertion in sanity check test

  janvanrijn

  committed
  be25e36

  Copy full SHA for be25e36

• replaced last occurance of str into string_types

  janvanrijn

  committed
  1582da5

  Copy full SHA for 1582da5

• test to see if appveyor handles ungzipped files

  janvanrijn

  committed
  0f4d22c

  Copy full SHA for 0f4d22c

• changed abspath to realpath

  janvanrijn

  committed
  817a3dc

  Copy full SHA for 817a3dc

• appveyor: added debug info

  janvanrijn

  committed
  25f9fce

  Copy full SHA for 25f9fce

Commits on Jul 21, 2018

• added arff.gz and json.gz to manifest.in

  janvanrijn

  committed
  3f32a93

  Copy full SHA for 3f32a93

• moved openml test files to data directory

  janvanrijn

  committed
  381fd45

  Copy full SHA for 381fd45

• added additional unit test for multi-target

  janvanrijn

  committed
  73c7fdb

  Copy full SHA for 73c7fdb

• changed joblib import

  janvanrijn

  committed
  07b2a7e

  Copy full SHA for 07b2a7e

• emotions unit test to gzip format

  janvanrijn

  committed
  6393f0a

  Copy full SHA for 6393f0a

• review by Roman

  janvanrijn

  committed
  12a538d

  Copy full SHA for 12a538d

• extended default feature check

  janvanrijn

  committed
  f71548e

  Copy full SHA for f71548e

• explicit naming for feature dicts

  janvanrijn

  committed
  2cf14ea

  Copy full SHA for 2cf14ea

• typo fix

  janvanrijn

  committed
  73544da

  Copy full SHA for 73544da

• removed sys import (for flake test)

  janvanrijn

  committed
  887d328

  Copy full SHA for 887d328

• fix Travix fail

  janvanrijn

  committed
  56f478e

  Copy full SHA for 56f478e

• make travis run again

  janvanrijn

  committed
  1df22eb

  Copy full SHA for 1df22eb

• removed unused variable

  janvanrijn

  committed
  8e42e09

  Copy full SHA for 8e42e09

Commits on Jul 23, 2018

• incorporated comments by Joel

  janvanrijn

  committed
  96c627b

  Copy full SHA for 96c627b

• incorporated last comment

  janvanrijn

  committed
  e6fe403

  Copy full SHA for e6fe403

• sparse data support

  janvanrijn

  committed
  c08065b

  Copy full SHA for c08065b

Commits on Jul 24, 2018

• comments by Joel

  janvanrijn

  committed
  94f86c1

  Copy full SHA for 94f86c1

• Merge branch 'sparse_data' into master

  janvanrijn

  authored
  312ecd9

  Copy full SHA for 312ecd9

• resolved merge errors

  janvanrijn

  committed
  546c4b0

  Copy full SHA for 546c4b0

• fixes flake8 error

  janvanrijn

  committed
  30662b4

  Copy full SHA for 30662b4

Commits on Jul 25, 2018

• comments by Joel

  janvanrijn

  committed
  8d02af2

  Copy full SHA for 8d02af2

• changed arff.loads to arff.load

  janvanrijn

  committed
  197685b

  Copy full SHA for 197685b

• reverted loads for PY3

  janvanrijn

  committed
  8d9d7dc

  Copy full SHA for 8d9d7dc

• implemented caching as recommended by Joel

  janvanrijn

  committed
  5218c67

  Copy full SHA for 5218c67

• changed function signature, better testing for cache and small bugfixes

  janvanrijn

  committed
  857042f

  Copy full SHA for 857042f

• Merge pull request #5 from janvanrijn/new_caching

  Show description for 82eb0fd
  janvanrijn

  authored
  82eb0fd

  Copy full SHA for 82eb0fd

• changed hashes into file structure

  janvanrijn

  committed
  bff90e5

  Copy full SHA for bff90e5

• Merge pull request #6 from janvanrijn/new_caching

  Show description for 4f655d3
  janvanrijn

  authored
  4f655d3

  Copy full SHA for 4f655d3

Commits on Jul 26, 2018

• comments by Joel

  janvanrijn

  committed
  5bec402

  Copy full SHA for 5bec402

Commits on Jul 27, 2018

• comments by Joel

  janvanrijn

  committed
  6c4c920

  Copy full SHA for 6c4c920

• flake8 correction

  janvanrijn

  committed
  eae5f62

  Copy full SHA for eae5f62

Commits on Jul 30, 2018

• extended test files with expected missing values

  janvanrijn

  committed
  7a01418

  Copy full SHA for 7a01418

Commits on Jul 31, 2018

• encode categoricals

  janvanrijn

  committed
  426bea4

  Copy full SHA for 426bea4

Commits on Aug 2, 2018

• renamed variable names

  janvanrijn

  committed
  15c2b48

  Copy full SHA for 15c2b48

• Merge pull request #8 from janvanrijn/encode

  Show description for 3f62bef
  janvanrijn

  authored
  3f62bef

  Copy full SHA for 3f62bef

Commits on Aug 6, 2018

• Targets not encoded; Towards metadata about nominals

  Show description for cea4f64
  jnothman

  committed
  cea4f64

  Copy full SHA for cea4f64

• handle no target

  jnothman

  committed
  cea54a4

  Copy full SHA for cea54a4

• Merge pull request #9 from jnothman/fetch_openml

  Show description for 350a64f
  janvanrijn

  authored
  350a64f

  Copy full SHA for 350a64f

Commits on Aug 7, 2018

• Make target dtype an object in classification

  Show description for 9a4d1d8
  jnothman

  committed
  9a4d1d8

  Copy full SHA for 9a4d1d8

• Test categories attribute, and some fixes

  jnothman

  committed
  2892be6

  Copy full SHA for 2892be6

• Merge pull request #10 from jnothman/fetch_openml

  Show description for 778fbd4
  janvanrijn

  authored
  778fbd4

  Copy full SHA for 778fbd4

• added nominal values to openml mock files, renamed data list mock files with data_list__ prefix

  janvanrijn

  committed
  25bae94

  Copy full SHA for 25bae94

• Merge branch 'master' into master

  janvanrijn

  authored
  87b8bd7

  Copy full SHA for 87b8bd7

Commits on Aug 8, 2018

• upgraded liac arff,

  Show description for 2881936
  janvanrijn

  committed
  2881936

  Copy full SHA for 2881936

• tiny typo fix

  janvanrijn

  committed
  d7ce597

  Copy full SHA for d7ce597

• flake8 fix

  janvanrijn

  committed
  261448d

  Copy full SHA for 261448d

• incorporated additional raise from https://github.com/renatopp/liac-arff/pull/80

  janvanrijn

  committed
  7624af7

  Copy full SHA for 7624af7

Commits on Aug 13, 2018

• DOC note return value is experimental

  jnothman

  authored
  40c11dd

  Copy full SHA for 40c11dd

• Merge branch 'master' into master

  jnothman

  authored
  c91eef4

  Copy full SHA for c91eef4

• Fix flake8

  lesteve

  committed
  4f7003b

  Copy full SHA for 4f7003b

Commits on Aug 15, 2018

• Remove obsolete FIXME

  jnothman

  committed
  a9f3a2c

  Copy full SHA for a9f3a2c

• Better error message with string attributes

  jnothman

  committed
  bd1a189

  Copy full SHA for bd1a189

• cosmit

  jnothman

  committed
  143c19b

  Copy full SHA for 143c19b

• DOC clean what's new merge mess

  jnothman

  committed
  1789421

  Copy full SHA for 1789421

• DOC clean what's new merge mess more

  jnothman

  committed
  088ddeb

  Copy full SHA for 088ddeb

• Fix urlopen has no __exit__ in Python 2

  jnothman

  committed
  fd9fba0

  Copy full SHA for fd9fba0