…n#6365)

…rom classes. (scikit-learn#7606)

The copybutton is the ">>>" on the side of code blocks that enables to
remove the ">>>" at the begining of lines as well as the outputs.

It was broken in the docs of 0.18 due to changes in sphinx. I am
backporting the js used by the Python core docs.

In addition, I am backporting CSS to highlight this button
on hover.

* Change the tests to yield after performing the _named_check.

* Tests affected:

  -test_averaging_multiclass
  -test_averaging_multilabel
  -test_averaging_multilabel_all_zeroes
  -test_averaging_multilabel_all_ones
  -test_sample_weight_invariance

* File updated: sklearn/metrics/tests/test_common.py

* edit releases link from sf to github

* edit link to pypi

…n integer. (scikit-learn#7457)

* Throw an error with explicit message if n_estimators is not an integer.

* Testing for explicit message if n_estimators is not an integer.

* Fixed typo in test for explicit message if n_estimators is an integer.

* Added tests for np.int32 and float input.

* pep8 compliance

* fix function name

* Import numpy to test n_estimators suplied as numpy int32.

…arse inputs (scikit-learn#6116)

* FIX unstable cumsum in utils.random

* equal_nan = true for isclose
since numpy < 1.9 sum is as unstable as cumsum, fallback to np.cumsum

* added axis parameter to stable_cumsum

* FIX unstable sumsum in ensemble.weight_boosting and utils.stats

* FIX axis problem in stable_cumsum

* FIX unstable cumsum in mixture.gmm and mixture.dpgmm

* FIX unstable cumsum in cluster.k_means_, decomposition.pca, and manifold.locally_linear

* FIX unstable sumsum in dataset.samples_generator

* added docstring for parameter axis of stable_cumsum

* added comment for why fall back to np.cumsum when np version < 1.9

* remove unneeded stable_cumsum

* added stable_cumsum's axis testing

* FIX numpy docstring for make_sparse_spd_matrix

* change stable_cumsum from error to warning

* Circle CI: adding 1h timeout for build_doc.sh

* Circle CI: typo + trigger build

* CircleCI: trigger build again

scikit-learn#7691)

…different training sizes (scikit-learn#7506)

* Chooses randomly the indices for different training sizes

* Bring back deleted line

* Rewrote the description of 'shuffle' attribute

* use random.sample instead of np.random.choice

* replace tabs with spaces

* merge to master

* Added shuffle in model-selection's learning_curve method

* Added shuffle for incremental learning + addressed Joel's comment

* Shorten long lines

* Add 2 blank spaces between test cases

* Addressed Joel's review comments

* Added 2 blank lines between methods

* Added non regression test for learning_curve with shuffle

* Fixed indentions

* Fixed space issues

* Modified test cases + small code improvements

* Fix some style issues

* Addressed Joel's comments - removed _shuffle_train_indices, more test cases and added new entry under 0.19/enhancements

* Added some modifications in whats_new.rst

…ion objects (scikit-learn#7680)

* Reorder EllipticEnvelope docstring.

* Fix pep8

)

* print score+fit time instead of just score time when doing cross_validation

* reducing line size

* More clearer log message

…ntSelectorMixin) scikit-learn#2121 (scikit-learn#6181)

* Norm inconsistency between RFE and SelectFromModel (was _LearntSelectorMixin) scikit-learn#2121

* safe_pwr utility

* Norm fix

* Removed safe_pwr

* 1D arrays support for norm fix

* Test case for 2d coef in SelectFromModel

* Fix numpy version requirement for norm fix

* Implement fixes suggested by @jnothman

* Add numpy version requiring the fix.

…ultiTaskLasso (scikit-learn#7741)

Add simple examples showing perfect score, expected value score, and negative score.

…n#7673)

* Clarified error msg in plot_partial_dependence

* Changed err msg for feature[i] out of range. Updated docs.

* Error message shows invalid value.

Debian replaces externals.joblib by an import to joblib, because they
hate duplication. Hence importing from inside joblib doesn't work, and
the line that I removed greats a bug.

See
http://lists.alioth.debian.org/pipermail/neurodebian-users/2016-October/001093.html

…7591)

Raise ValueError if l1_ratio=0 in ElasticNetCV and alphas=None

…ements to the plot. (scikit-learn#7730)

* Small correction of logit->logistic

The function `1 / (1 + np.exp(-x))` is the logistic function. 
The logit function is the inverse of the logistic function : `log(x/(1-x))`

* Add axes ticks, legend and colors to the plot

* Fix flake8 errors

* Rename legend labels as per review

* Fix pep8 error

…cikit-learn#7632)

…st_equal (scikit-learn#7742)

* replaced some assert_true(np.allclose(x, y)) with assert_almost_equal for better error messages.

also some pep8.

* typo fixes

…cikit-learn#7423)

…scikit-learn#7681) (scikit-learn#7683)

* FIX raise an error message when n_groups > actual number of groups (scikit-learn#7681)

This change addresses issue scikit-learn#7681:
- Raise ValueError when n_groups > actual number of unique groups in LeaveOneGroupOut and LeavePGroupsOut.
- Add unit test.

* Make requested changes

- Check error message with `assert_raise_message`
- Pass parameters to `assert_raise_message` instead of defining functions

* Update condition and exception message

* LOF algorithm

add tests and example

fix DepreciationWarning by reshape(1,-1) one-sample data

LOF with inheritance

lof and lof2 return same score

fix bugs

fix bugs

optimized and cosmit

rm lof2

cosmit

rm MixinLOF + fit_predict

fix travis - optimize pairwise_distance like in KNeighborsMixin.kneighbors

add comparison example + doc

LOF -> LocalOutlierFactor
cosmit

change LOF API:
-fit(X).predict() and fit(X).decision_function() do prediction on X without
 considering samples as their own neighbors (ie without considering X as a
 new dataset as does fit(X).predict(X))
-rm fit_predict() method
-add a contamination parameter st predict returns a binary value like other
 anomaly detection algos

cosmit

doc + debug example

correction doc

pass on doc + examples

pep8 + fix warnings

first attempt at fixing API issues

minor changes

takes into account tguillemot advice

-remove pairwise_distance calculation as to heavy in memory
-add benchmarks

cosmit

minor changes + deals with duplicates

fix depreciation warnings

* factorize the two for loops

* take into account @albertthomas88 review and cosmit

* fix doc

* alex review + rebase

* make predict private add outlier_factor_ attribute and update tests

* make fit_predict take y argument

* fix benchmarks file

* update examples

* make decision_function public (rm X=None default)

* fix travis

* take into account tguillemot review + remove useless k_distance function

* fix broken links :meth:`kneighbors`

* cosmit

* whatsnew

* amueller review + remove _local_outlier_factor method

* add n_neighbors_ parameter the effective nb neighbors we use

* make decision_function private and negative_outlier_factor attribute

…t-learn#5876 (scikit-learn#5984)

…arn#7732)

* fix bug with negative values in cosine_distances
clip distances to [0, 2]
set distances between vectors and themselves to 0

* add test

* add test on big random matrix

* use np.diag_indices_from instead of slicing

…lar for LDA).