Commit Graph

95 Commits

Author SHA1 Message Date
Raghav R V e3afc0e8c9 MAINT move custom error/warning classes into sklearn.exceptions
ENH NonBLASDotWarning -> EfficiencyWarning; Improve error message
DOC Add exceptions module to modules/classes.rst
MAINT Move ConvergenceWarning, UndefinedMetricWarning et al into exceptions
MAINT Remove ChangedBehaviorWarning from base
DOC/FIX Improve DataConversionWarning's docstring
2015-10-19 22:35:35 +02:00
TomDLT 628ed2c867 ENH remove some warnings in test suit 2015-09-16 18:27:58 +02:00
Raghav R V 664d78eb7c MAINT Remove support for the deprecated sequence of sequences
MAINT Remove sequence of sequence support from datasets
MAINT Remove return_indicator param
MAINT Remove multilabel-seq test in OVR
MAINT Remove multilable-seq test in check_cv
MAINT Remove multilabel seq test in label_binarizer
TST type_of_target returns "unknown" for multilabel-sequence types
TST _check_targets should raise a ValueError
DOC show multilabel indicator as an example; remove return_indicator param
DOC use consistent lower case y for target
2015-08-11 22:26:44 +05:30
Joel Nothman febefb0894 FIX avoid memory cost when sampling from large parameter grids 2015-06-25 08:10:31 +10:00
Raghav R V cd2ee7e454 MAINT docstring --> comments to prevent nose from using doc in verbose mode 2015-03-21 11:16:49 +05:30
Andreas Mueller 2e2eebf4ab Make ParameterSampler sample without replacement if all parameters are given as lists. 2015-01-15 10:55:31 -05:00
Andreas Mueller e2d163a148 address @jnothman's comments 2014-12-30 10:28:07 -05:00
Andreas Mueller ed26caeb40 make gridsearch_no_predict test stricter. 2014-12-29 20:51:30 -05:00
Andreas Mueller 6dce07733d remove bad estimator test as it was enforcing too many restrictions, instead test that we don't enforce these restrictions. 2014-12-29 20:40:34 -05:00
Andreas Mueller 3d1ede915a TST add tests for nd grid-search can train_test_split 2014-12-25 23:08:44 +11:00
Lars Buitinck 40c03d4880 COSMIT pep8 fixes to sklearn/tests 2014-12-18 16:40:01 +01:00
Joel Nothman e00fd83a83 DOC narrative docs for grid search's robustness to failure 2014-11-27 10:53:29 +11:00
Michal Romaniuk 58be184f5f Enable grid search with classifiers that may throw an error on individual fits. 2014-10-02 10:03:17 +10:00
Joel Nothman 00c75859be FIX more intuitive behavior for *SearchCV.score 2014-08-20 23:09:22 +10:00
Andreas Mueller b080d93b5e ENH don't convert dataframes in grid search and cross-validation 2014-07-18 10:02:14 +02:00
Arnaud Joly 8de00d857f Merge pull request #3411 from lesteve/scheduled-removal-from-0.15
[MRG] Remove deprecated parameters whose removal was scheduled for 0.15
2014-07-17 16:57:30 +02:00
lesteve b12d1ac487 Remove deprecated 'score_func' and 'loss_func' parameters from sklearn.metrics.scorer.check_scoring. Amend the code in all the other places they were used. 2014-07-17 13:53:06 +01:00
Andreas Mueller e242ebc5c7 ENH rename parameters in MockListClassifiers. 2014-07-16 16:20:10 +02:00
Andreas Mueller 3c7818e8cf ENH allow y to be a list in GridSearchCV, cross_val_score and train_test_split. 2014-07-16 10:59:57 +02:00
Andreas Mueller c7db70ac9f removed unused variables and imports, add used imports where missing. 2014-06-08 21:44:01 +02:00
Yung Siang Liau 3f9dff942d FIX: Add allow_nans option to check_arrays
Grid search and cross validation should not panic when seeing NaNs
in the input arrays, because that breaks Imputer.
Fixes #2774 and #3044.
2014-04-13 16:59:52 +02:00
Andrew Tulloch 8eeda198d2 [tests] Use Python 3 zip/Python 2 zip consistently in tests 2014-03-03 13:02:17 +00:00
Lars Buitinck 0f5685b8ee FIX predict_proba status on SGD and SVC when disabled
Using a property trick suggested by @mblondel. The following now
works:

    >>> clf = SVC(probability=False)
    >>> hasattr(clf, "predict_proba")
    False

This simplifies the smoke tests.

Added a note to SVC stating that this should be done in new code.
2014-02-27 18:30:35 +01:00
Lars Buitinck c997cdee9f COSMIT pep8 + full stop police
metrics/tests/test_score_objects.py made stricter by replacing . with \.
in regular expressions.
2014-02-01 15:28:53 +01:00
Alexander Fabisch 4e968a2c1e Check error message 2014-01-16 12:43:16 +09:00
Alexander Fabisch 526d1dd945 check_scorable returns scorer 2014-01-16 12:43:16 +09:00
balu ca6b42cffd fix visual indent 2013-08-31 12:21:30 +05:30
Mathieu Blondel 057db3b0fa COSMIT: change variable name. 2013-08-15 22:45:17 +09:00
Mathieu Blondel 1ac7c58b82 Fix grid search test. 2013-08-15 22:45:17 +09:00
Arnaud Joly cf73a9ed85 ENH more explicit name for auc + consistency for scorer, fix #2096 2013-08-07 10:23:37 +02:00
Olivier Grisel b0035bf9e8 Rename cv_scores(_) back to grid_scores(_) to keep the name free for a future refactoring 2013-07-28 17:10:14 +02:00
Arnaud Joly 387951da0f DOC credit + fix typo + wording + use mathplotlib.pyplot
DOC fix typo

DOC wording

DOC typo and use mathplotlib.pyplot
2013-07-26 16:01:17 +02:00
Arnaud Joly f4eded634d DOC overall improvements
Typo

COSMIT remove debug print

DOC better docstring

Typo

Remove aborted examples

DOC fix numpy doc rendering

DOC fix the figure grid has been transposed

DOC precise which estimators

DOC polish example

DOC wording
2013-07-26 16:01:04 +02:00
Arnaud Joly 7e3f7ba376 ENH TST random search with multioutput data 2013-07-26 15:57:35 +02:00
Arnaud Joly 351784a843 ENH TST grid search with multioutput 2013-07-26 15:57:35 +02:00
Arnaud Joly 2109b1be6d ENH consistency mse=> mean_squared_error ari => adjusted_rand_score 2013-07-26 11:00:22 +02:00
Arnaud Joly 8367727d6a COSMIT 2013-07-26 10:24:34 +02:00
Arnaud Joly 0bce8a1698 TST grid_search_cv works with multioutput data 2013-07-25 18:33:16 +02:00
Gael Varoquaux 0c461e4bf3 ENH: better error message on scoring 2013-07-25 11:35:23 +02:00
Lars Buitinck 393b99636b ENH simplify the Scorer API
A Scorer is now a function that returns a score that should be maximized.
2013-07-24 14:47:47 +02:00
Justin Vincent d420aafb0f TST + PY3 various fixes
* Little hack in fixes.py that can be removed when NumPy's
  https://github.com/numpy/numpy/issues/3484 is fixed.
* Little cleanup to the fixes implementation.
* Change the numpy workaround and added a test. Fixed some xrange usage.
* Fix the weight boosting test.
* None is not comparable in python3
* Fix a test that depended on dictionary iteration order.
* Python3 removed execfile
* More test fixes.

Fixes #2122.
2013-07-05 20:17:23 +02:00
Lars Buitinck 0c3ae9cce0 ENH allow empty grid in ParameterGrid
Fixes #2048.
2013-06-22 12:06:41 +02:00
Joel Nothman c012d4fe4b FIX use Python 2.6-compatible str.format 2013-06-10 14:31:13 +10:00
Olivier Grisel 5d7aaaec50 FIX inconsistent cv_scores_ generation for randomized search and re-add example 2013-06-10 13:45:52 +10:00
Andreas Mueller fb890311e9 FIX broken scorer, add non-regression test. 2013-05-04 13:52:57 +02:00
Andreas Mueller a1d4d184fe COSMIT pep8 2013-04-02 13:30:55 +02:00
Joel Nothman 364db7a68e FIX make *SearchCV picklable
Define CVScoreTuple in the module namespace.
2013-04-02 10:33:51 +02:00
Joel Nothman 83f67ee57e ENH/FIX make best_estimator_'s predict functions available in parameter search
Avoids copying an unpicklable method to parameter search's __dict__.

Also adds decision_function and transform where only predict and predict_proba were available before.
2013-04-02 10:33:50 +02:00
Raul Garreta 095085a986 PY3: used six.moves.xrange to fix xrange 2013-03-19 18:02:34 +01:00
Olivier Grisel 5a5268b89f PY3: fix remaining cStringIO imports 2013-03-19 17:57:12 +01:00