This fixes a typo that has the effect to not resolve a link

…how to avoid the warning (scikit-learn#13477)

…#13539)

…13497)

* discrete_features str and value check

* Update if logic

* Add discrete_features bad str value test

* Remove unnecessary nested isinstance str check

* Add back nested isinstance str check

* New/updates to tests

* Add v0.21 whats new entry

* Undo v0.21 whats new entry

…alue of epsilon in the code (scikit-learn#13559)

* Update docstring in LinearSVR

Changed the docstring of class LinearSVR to reflect the default value of epsilon in the code.

* Update tree.rst

* Revert "Update tree.rst"

…ativeClustering base class (scikit-learn#13427)

* Changed name n_components to n_connected_components in base class

* Fixed line which exceeded PEP8 max of 79 chars

* Fixed line 818 which exceeded PEP8 max of 79 chars

* Added try and except to provide deprecation warning if  passed

* Updated deprecation and removal version numbers

* Added deprecation of n_components using @Property generator

* Makes FeatureAgglomeration class inherit n_connected_components_ attribute from AgglomerativeClustering class

* Added test for DeprecationWarning when trying to access n_components

* Removed @Property generator causing linting error

* Fixed typo in test

* Fixed flake8 error due to single line between 2 functions

* Test fix attempt

* Edited test function docstring

* Corrected n_components deprecation test docstring

* Fixed line continuation issue in AgglomerativeClustering base class

* Added deprecation message as part of the @deprecated decorator

* Added  attribute deprecation information in the Attributes section of the AgglomerativeClustering base class docstring

* Added test for deprecation warning message

* Added  attribute deprecation information in the Attributes section of the FeatureAgglomeration base class docstring

* Fixed test issue and added longer match string

* Edited n_components_ deprecation message to add double backticks

* Fixed match string to reflect deprecation message change in test

* Added name to list of contributors

* Documented information in v0.21 changelog

* Added cluster parent folder to documentation in v0.21 changelog

* Removed myself from list of core contributors

* Moved |API| subsection to the end of the list, and changed reference to Github username

* Removed n_components deprecation documentation from FeatureAgglomeration and AgglomerativeClustering class dosctrings

* Fix indentation on _fix_connectivity function call

)

…arn#13577)

Minor correction to docstring of the function `load_linnerud()` in `base.py`, which was referring to 'targets' instead of 'target'.

* Add column_or_1d to account for dataframe y

* Add test

* make the diff cleaner

* switched pandas import for 2d array

* change test to a simple comparison between 1d and 2d y

* flake8 errors in tests

* Add warning to acieve consistent behaviour as in other classifiers

* Update whats new file

* remove redundant code, fix regression

* remove unnecessary import

* add comment on test

* address comments

* fix what's new entry

* Update doc/whats_new/v0.21.rst

Co-Authored-By: adrinjalali <adrin.jalali@gmail.com>

…n for gradient boosting (scikit-learn#13193)

* Replace n_node_samples by weighted_n_node_samples in partial dependence computation

* Add tests for both no-op and real sample weights

* Improve naming and remove useless comment

* Fix small test issues

* Fix test for binary classification

* Add test for regressions based on example from initial issue

* Edit whats_new

* 79

* Simplify test code for regression partial dependence

* PEP8

* Facepalm

* Refer to the public function in whats_new

* Make the sample weight test standalone for further reuse

* Fix PR number

* Testing with L1 relative distance computed as averages

* Testing element-wise

* Fix and simplify unit test for binary classification

* Clarify functional test

…learn#8773)

…cikit-learn#13585)

…rn#12583)

…it-learn#13524)

…rn#13529)

* DOC Correct coef_ shape in RidgeClassifier

* address review

…3636)

…cikit-learn#13646)

This change is to work around the hang

 scikit-learn#12263

afflicting Windows on machines with > 62 hyperthreads.

…cikit-learn#13601)

…rn#13620)

1.5+2 = 3.5, not 3.

…ikit-learn#13607)

…ikit-learn#13626)

…t-learn#13543)

…ample (scikit-learn#13383) (scikit-learn#13648)

* Fixing parameter description (for assume_centered)

Fixing parameter description (for assume_centered)

* Update shrunk_covariance_.py

* Update empirical_covariance_.py

)

…scikit-learn#13363)

…3663)

…ikit-learn#13635)

merge fork