Version 0.18.2¶
June 20, 2017
Last release with Python 2.6 support
Scikit-learn 0.18 is the last major release of scikit-learn to support Python 2.6. Later versions of scikit-learn will require Python 2.7 or above.
Changelog¶
Code Contributors¶
Aman Dalmia, Loic Esteve, Nate Guerin, Sergei Lebedev
Version 0.18.1¶
November 11, 2016
Changelog¶
Enhancements¶
Improved
sample_without_replacement
speed by utilizing numpy.random.permutation for most cases. As a result, samples may differ in this release for a fixed random state. Affected estimators:ensemble.BaggingClassifier
ensemble.BaggingRegressor
linear_model.RANSACRegressor
model_selection.RandomizedSearchCV
random_projection.SparseRandomProjection
This also affects the
datasets.make_classification
method.
Bug fixes¶
- Fix issue where
min_grad_norm
andn_iter_without_progress
parameters were not being utilised bymanifold.TSNE
. #6497 by Sebastian Säger - Fix bug for svm’s decision values when
decision_function_shape
isovr
insvm.SVC
.svm.SVC
’s decision_function was incorrect from versions 0.17.0 through 0.18.0. #7724 by Bing Tian Dai - Attribute
explained_variance_ratio
ofdiscriminant_analysis.LinearDiscriminantAnalysis
calculated with SVD and Eigen solver are now of the same length. #7632 by JPFrancoia - Fixes issue in Univariate feature selection where score functions were not accepting multi-label targets. #7676 by Mohammed Affan
- Fixed setting parameters when calling
fit
multiple times onfeature_selection.SelectFromModel
. #7756 by Andreas Müller - Fixes issue in
partial_fit
method ofmulticlass.OneVsRestClassifier
when number of classes used inpartial_fit
was less than the total number of classes in the data. #7786 by Srivatsan Ramesh - Fixes issue in
calibration.CalibratedClassifierCV
where the sum of probabilities of each class for a data was not 1, andCalibratedClassifierCV
now handles the case where the training set has less number of classes than the total data. #7799 by Srivatsan Ramesh - Fix a bug where
sklearn.feature_selection.SelectFdr
did not exactly implement Benjamini-Hochberg procedure. It formerly may have selected fewer features than it should. #7490 by Peng Meng. sklearn.manifold.LocallyLinearEmbedding
now correctly handles integer inputs. #6282 by Jake Vanderplas.- The
min_weight_fraction_leaf
parameter of tree-based classifiers and regressors now assumes uniform sample weights by default if thesample_weight
argument is not passed to thefit
function. Previously, the parameter was silently ignored. #7301 by Nelson Liu. - Numerical issue with
linear_model.RidgeCV
on centered data when n_features > n_samples. #6178 by Bertrand Thirion - Tree splitting criterion classes’ cloning/pickling is now memory safe #7680 by Ibraim Ganiev.
- Fixed a bug where
decomposition.NMF
sets itsn_iters_
attribute in transform(). #7553 by Ekaterina Krivich. sklearn.linear_model.LogisticRegressionCV
now correctly handles string labels. #5874 by Raghav RV.- Fixed a bug where
sklearn.model_selection.train_test_split
raised an error whenstratify
is a list of string labels. #7593 by Raghav RV. - Fixed a bug where
sklearn.model_selection.GridSearchCV
andsklearn.model_selection.RandomizedSearchCV
were not pickleable because of a pickling bug innp.ma.MaskedArray
. #7594 by Raghav RV. - All cross-validation utilities in
sklearn.model_selection
now permit one time cross-validation splitters for thecv
parameter. Also non-deterministic cross-validation splitters (where multiple calls tosplit
produce dissimilar splits) can be used ascv
parameter. Thesklearn.model_selection.GridSearchCV
will cross-validate each parameter setting on the split produced by the firstsplit
call to the cross-validation splitter. #7660 by Raghav RV. - Fix bug where
preprocessing.MultiLabelBinarizer.fit_transform
returned an invalid CSR matrix. #7750 by CJ Carey. - Fixed a bug where
metrics.pairwise.cosine_distances
could return a small negative distance. #7732 by Artsion.
API changes summary¶
Trees and forests
- The
min_weight_fraction_leaf
parameter of tree-based classifiers and regressors now assumes uniform sample weights by default if thesample_weight
argument is not passed to thefit
function. Previously, the parameter was silently ignored. #7301 by Nelson Liu. - Tree splitting criterion classes’ cloning/pickling is now memory safe. #7680 by Ibraim Ganiev.
Linear, kernelized and related models
- Length of
explained_variance_ratio
ofdiscriminant_analysis.LinearDiscriminantAnalysis
changed for both Eigen and SVD solvers. The attribute has now a length of min(n_components, n_classes - 1). #7632 by JPFrancoia - Numerical issue with
linear_model.RidgeCV
on centered data whenn_features > n_samples
. #6178 by Bertrand Thirion
Version 0.18¶
September 28, 2016
Last release with Python 2.6 support
Scikit-learn 0.18 will be the last version of scikit-learn to support Python 2.6. Later versions of scikit-learn will require Python 2.7 or above.
Model Selection Enhancements and API Changes¶
The model_selection module
The new module
sklearn.model_selection
, which groups together the functionalities of formerlysklearn.cross_validation
,sklearn.grid_search
andsklearn.learning_curve
, introduces new possibilities such as nested cross-validation and better manipulation of parameter searches with Pandas.Many things will stay the same but there are some key differences. Read below to know more about the changes.
Data-independent CV splitters enabling nested cross-validation
The new cross-validation splitters, defined in the
sklearn.model_selection
, are no longer initialized with any data-dependent parameters such asy
. Instead they expose asplit
method that takes in the data and yields a generator for the different splits.This change makes it possible to use the cross-validation splitters to perform nested cross-validation, facilitated by
model_selection.GridSearchCV
andmodel_selection.RandomizedSearchCV
utilities.The enhanced cv_results_ attribute
The new
cv_results_
attribute (ofmodel_selection.GridSearchCV
andmodel_selection.RandomizedSearchCV
) introduced in lieu of thegrid_scores_
attribute is a dict of 1D arrays with elements in each array corresponding to the parameter settings (i.e. search candidates).The
cv_results_
dict can be easily imported intopandas
as aDataFrame
for exploring the search results.The
cv_results_
arrays include scores for each cross-validation split (with keys such as'split0_test_score'
), as well as their mean ('mean_test_score'
) and standard deviation ('std_test_score'
).The ranks for the search candidates (based on their mean cross-validation score) is available at
cv_results_['rank_test_score']
.The parameter values for each parameter is stored separately as numpy masked object arrays. The value, for that search candidate, is masked if the corresponding parameter is not applicable. Additionally a list of all the parameter dicts are stored at
cv_results_['params']
.Parameters n_folds and n_iter renamed to n_splits
Some parameter names have changed: The
n_folds
parameter in newmodel_selection.KFold
,model_selection.GroupKFold
(see below for the name change), andmodel_selection.StratifiedKFold
is now renamed ton_splits
. Then_iter
parameter inmodel_selection.ShuffleSplit
, the new classmodel_selection.GroupShuffleSplit
andmodel_selection.StratifiedShuffleSplit
is now renamed ton_splits
.Rename of splitter classes which accepts group labels along with data
The cross-validation splitters
LabelKFold
,LabelShuffleSplit
,LeaveOneLabelOut
andLeavePLabelOut
have been renamed tomodel_selection.GroupKFold
,model_selection.GroupShuffleSplit
,model_selection.LeaveOneGroupOut
andmodel_selection.LeavePGroupsOut
respectively.Note the change from singular to plural form in
model_selection.LeavePGroupsOut
.Fit parameter labels renamed to groups
The
labels
parameter in thesplit
method of the newly renamed splittersmodel_selection.GroupKFold
,model_selection.LeaveOneGroupOut
,model_selection.LeavePGroupsOut
,model_selection.GroupShuffleSplit
is renamed togroups
following the new nomenclature of their class names.Parameter n_labels renamed to n_groups
The parameter
n_labels
in the newly renamedmodel_selection.LeavePGroupsOut
is changed ton_groups
.Training scores and Timing information
cv_results_
also includes the training scores for each cross-validation split (with keys such as'split0_train_score'
), as well as their mean ('mean_train_score'
) and standard deviation ('std_train_score'
). To avoid the cost of evaluating training score, setreturn_train_score=False
.Additionally the mean and standard deviation of the times taken to split, train and score the model across all the cross-validation splits is available at the key
'mean_time'
and'std_time'
respectively.
Changelog¶
New features¶
Classifiers and Regressors
- The Gaussian Process module has been reimplemented and now offers classification
and regression estimators through
gaussian_process.GaussianProcessClassifier
andgaussian_process.GaussianProcessRegressor
. Among other things, the new implementation supports kernel engineering, gradient-based hyperparameter optimization or sampling of functions from GP prior and GP posterior. Extensive documentation and examples are provided. By Jan Hendrik Metzen. - Added new supervised learning algorithm: Multi-layer Perceptron #3204 by Issam H. Laradji
- Added
linear_model.HuberRegressor
, a linear model robust to outliers. #5291 by Manoj Kumar. - Added the
multioutput.MultiOutputRegressor
meta-estimator. It converts single output regressors to multi-output regressors by fitting one regressor per output. By Tim Head.
Other estimators
- New
mixture.GaussianMixture
andmixture.BayesianGaussianMixture
replace former mixture models, employing faster inference for sounder results. #7295 by Wei Xue and Thierry Guillemot. - Class
decomposition.RandomizedPCA
is now factored intodecomposition.PCA
and it is available calling with parametersvd_solver='randomized'
. The default number ofn_iter
for'randomized'
has changed to 4. The old behavior of PCA is recovered bysvd_solver='full'
. An additional solver callsarpack
and performs truncated (non-randomized) SVD. By default, the best solver is selected depending on the size of the input and the number of components requested. #5299 by Giorgio Patrini. - Added two functions for mutual information estimation:
feature_selection.mutual_info_classif
andfeature_selection.mutual_info_regression
. These functions can be used infeature_selection.SelectKBest
andfeature_selection.SelectPercentile
as score functions. By Andrea Bravi and Nikolay Mayorov. - Added the
ensemble.IsolationForest
class for anomaly detection based on random forests. By Nicolas Goix. - Added
algorithm="elkan"
tocluster.KMeans
implementing Elkan’s fast K-Means algorithm. By Andreas Müller.
Model selection and evaluation
- Added
metrics.cluster.fowlkes_mallows_score
, the Fowlkes Mallows Index which measures the similarity of two clusterings of a set of points By Arnaud Fouchet and Thierry Guillemot. - Added
metrics.calinski_harabaz_score
, which computes the Calinski and Harabaz score to evaluate the resulting clustering of a set of points. By Arnaud Fouchet and Thierry Guillemot. - Added new cross-validation splitter
model_selection.TimeSeriesSplit
to handle time series data. #6586 by YenChen Lin - The cross-validation iterators are replaced by cross-validation splitters
available from
sklearn.model_selection
, allowing for nested cross-validation. See Model Selection Enhancements and API Changes for more information. #4294 by Raghav RV.
Enhancements¶
Trees and ensembles
- Added a new splitting criterion for
tree.DecisionTreeRegressor
, the mean absolute error. This criterion can also be used inensemble.ExtraTreesRegressor
,ensemble.RandomForestRegressor
, and the gradient boosting estimators. #6667 by Nelson Liu. - Added weighted impurity-based early stopping criterion for decision tree growth. #6954 by Nelson Liu
- The random forest, extra tree and decision tree estimators now has a
method
decision_path
which returns the decision path of samples in the tree. By Arnaud Joly. - A new example has been added unveiling the decision tree structure. By Arnaud Joly.
- Random forest, extra trees, decision trees and gradient boosting estimator
accept the parameter
min_samples_split
andmin_samples_leaf
provided as a percentage of the training samples. By yelite and Arnaud Joly. - Gradient boosting estimators accept the parameter
criterion
to specify to splitting criterion used in built decision trees. #6667 by Nelson Liu. - The memory footprint is reduced (sometimes greatly) for
ensemble.bagging.BaseBagging
and classes that inherit from it, i.e,ensemble.BaggingClassifier
,ensemble.BaggingRegressor
, andensemble.IsolationForest
, by dynamically generating attributeestimators_samples_
only when it is needed. By David Staub. - Added
n_jobs
andsample_weight
parameters forensemble.VotingClassifier
to fit underlying estimators in parallel. #5805 by Ibraim Ganiev.
Linear, kernelized and related models
- In
linear_model.LogisticRegression
, the SAG solver is now available in the multinomial case. #5251 by Tom Dupre la Tour. linear_model.RANSACRegressor
,svm.LinearSVC
andsvm.LinearSVR
now supportsample_weight
. By Imaculate.- Add parameter
loss
tolinear_model.RANSACRegressor
to measure the error on the samples for every trial. By Manoj Kumar. - Prediction of out-of-sample events with Isotonic Regression
(
isotonic.IsotonicRegression
) is now much faster (over 1000x in tests with synthetic data). By Jonathan Arfa. - Isotonic regression (
isotonic.IsotonicRegression
) now uses a better algorithm to avoid O(n^2) behavior in pathological cases, and is also generally faster (##6691). By Antony Lee. naive_bayes.GaussianNB
now accepts data-independent class-priors through the parameterpriors
. By Guillaume Lemaitre.linear_model.ElasticNet
andlinear_model.Lasso
now works withnp.float32
input data without converting it intonp.float64
. This allows to reduce the memory consumption. #6913 by YenChen Lin.semi_supervised.LabelPropagation
andsemi_supervised.LabelSpreading
now accept arbitrary kernel functions in addition to stringsknn
andrbf
. #5762 by Utkarsh Upadhyay.
Decomposition, manifold learning and clustering
- Added
inverse_transform
function todecomposition.NMF
to compute data matrix of original shape. By Anish Shah. cluster.KMeans
andcluster.MiniBatchKMeans
now works withnp.float32
andnp.float64
input data without converting it. This allows to reduce the memory consumption by usingnp.float32
. #6846 by Sebastian Säger and YenChen Lin.
Preprocessing and feature selection
preprocessing.RobustScaler
now acceptsquantile_range
parameter. #5929 by Konstantin Podshumok.feature_extraction.FeatureHasher
now accepts string values. #6173 by Ryad Zenine and Devashish Deshpande.- Keyword arguments can now be supplied to
func
inpreprocessing.FunctionTransformer
by means of thekw_args
parameter. By Brian McFee. feature_selection.SelectKBest
andfeature_selection.SelectPercentile
now accept score functions that take X, y as input and return only the scores. By Nikolay Mayorov.
Model evaluation and meta-estimators
multiclass.OneVsOneClassifier
andmulticlass.OneVsRestClassifier
now supportpartial_fit
. By Asish Panda and Philipp Dowling.- Added support for substituting or disabling
pipeline.Pipeline
andpipeline.FeatureUnion
components using theset_params
interface that powerssklearn.grid_search
. See Selecting dimensionality reduction with Pipeline and GridSearchCV By Joel Nothman and Robert McGibbon. - The new
cv_results_
attribute ofmodel_selection.GridSearchCV
(andmodel_selection.RandomizedSearchCV
) can be easily imported into pandas as aDataFrame
. Ref Model Selection Enhancements and API Changes for more information. #6697 by Raghav RV. - Generalization of
model_selection.cross_val_predict
. One can pass method names such as predict_proba to be used in the cross validation framework instead of the default predict. By Ori Ziv and Sears Merritt. - The training scores and time taken for training followed by scoring for
each search candidate are now available at the
cv_results_
dict. See Model Selection Enhancements and API Changes for more information. #7325 by Eugene Chen and Raghav RV.
Metrics
- Added
labels
flag tometrics.log_loss
to explicitly provide the labels when the number of classes iny_true
andy_pred
differ. #7239 by Hong Guangguo with help from Mads Jensen and Nelson Liu. - Support sparse contingency matrices in cluster evaluation
(
metrics.cluster.supervised
) to scale to a large number of clusters. #7419 by Gregory Stupp and Joel Nothman. - Add
sample_weight
parameter tometrics.matthews_corrcoef
. By Jatin Shah and Raghav RV. - Speed up
metrics.silhouette_score
by using vectorized operations. By Manoj Kumar. - Add
sample_weight
parameter tometrics.confusion_matrix
. By Bernardo Stein.
Miscellaneous
- Added
n_jobs
parameter tofeature_selection.RFECV
to compute the score on the test folds in parallel. By Manoj Kumar - Codebase does not contain C/C++ cython generated files: they are generated during build. Distribution packages will still contain generated C/C++ files. By Arthur Mensch.
- Reduce the memory usage for 32-bit float input arrays of
utils.sparse_func.mean_variance_axis
andutils.sparse_func.incr_mean_variance_axis
by supporting cython fused types. By YenChen Lin. - The
ignore_warnings
now accept a category argument to ignore only the warnings of a specified type. By Thierry Guillemot. - Added parameter
return_X_y
and return type(data, target) : tuple
option toload_iris
dataset #7049,load_breast_cancer
dataset #7152,load_digits
dataset,load_diabetes
dataset,load_linnerud
dataset,load_boston
dataset #7154 by Manvendra Singh. - Simplification of the
clone
function, deprecate support for estimators that modify parameters in__init__
. #5540 by Andreas Müller. - When unpickling a scikit-learn estimator in a different version than the one
the estimator was trained with, a
UserWarning
is raised, see the documentation on model persistence for more details. (#7248) By Andreas Müller.
Bug fixes¶
Trees and ensembles
- Random forest, extra trees, decision trees and gradient boosting
won’t accept anymore
min_samples_split=1
as at least 2 samples are required to split a decision tree node. By Arnaud Joly ensemble.VotingClassifier
now raisesNotFittedError
ifpredict
,transform
orpredict_proba
are called on the non-fitted estimator. by Sebastian Raschka.- Fix bug where
ensemble.AdaBoostClassifier
andensemble.AdaBoostRegressor
would perform poorly if therandom_state
was fixed (#7411). By Joel Nothman. - Fix bug in ensembles with randomization where the ensemble would not
set
random_state
on base estimators in a pipeline or similar nesting. (#7411). Note, results forensemble.BaggingClassifier
ensemble.BaggingRegressor
,ensemble.AdaBoostClassifier
andensemble.AdaBoostRegressor
will now differ from previous versions. By Joel Nothman.
Linear, kernelized and related models
- Fixed incorrect gradient computation for
loss='squared_epsilon_insensitive'
inlinear_model.SGDClassifier
andlinear_model.SGDRegressor
(#6764). By Wenhua Yang. - Fix bug in
linear_model.LogisticRegressionCV
wheresolver='liblinear'
did not acceptclass_weights='balanced
. (#6817). By Tom Dupre la Tour. - Fix bug in
neighbors.RadiusNeighborsClassifier
where an error occurred when there were outliers being labelled and a weight function specified (#6902). By LeonieBorne. - Fix
linear_model.ElasticNet
sparse decision function to match output with dense in the multioutput case.
Decomposition, manifold learning and clustering
decomposition.RandomizedPCA
default number of iterated_power is 4 instead of 3. #5141 by Giorgio Patrini.utils.extmath.randomized_svd
performs 4 power iterations by default, instead or 0. In practice this is enough for obtaining a good approximation of the true eigenvalues/vectors in the presence of noise. When n_components is small (< .1 * min(X.shape)
) n_iter is set to 7, unless the user specifies a higher number. This improves precision with few components. #5299 by Giorgio Patrini.- Whiten/non-whiten inconsistency between components of
decomposition.PCA
anddecomposition.RandomizedPCA
(now factored into PCA, see the New features) is fixed. components_ are stored with no whitening. #5299 by Giorgio Patrini. - Fixed bug in
manifold.spectral_embedding
where diagonal of unnormalized Laplacian matrix was incorrectly set to 1. #4995 by Peter Fischer. - Fixed incorrect initialization of
utils.arpack.eigsh
on all occurrences. Affectscluster.bicluster.SpectralBiclustering
,decomposition.KernelPCA
,manifold.LocallyLinearEmbedding
, andmanifold.SpectralEmbedding
(#5012). By Peter Fischer. - Attribute
explained_variance_ratio_
calculated with the SVD solver ofdiscriminant_analysis.LinearDiscriminantAnalysis
now returns correct results. By JPFrancoia
Preprocessing and feature selection
preprocessing.data._transform_selected
now always passes a copy ofX
to transform function whencopy=True
(#7194). By Caio Oliveira.
Model evaluation and meta-estimators
model_selection.StratifiedKFold
now raises error if all n_labels for individual classes is less than n_folds. #6182 by Devashish Deshpande.- Fixed bug in
model_selection.StratifiedShuffleSplit
where train and test sample could overlap in some edge cases, see #6121 for more details. By Loic Esteve. - Fix in
sklearn.model_selection.StratifiedShuffleSplit
to return splits of sizetrain_size
andtest_size
in all cases (#6472). By Andreas Müller. - Cross-validation of
OneVsOneClassifier
andOneVsRestClassifier
now works with precomputed kernels. #7350 by Russell Smith. - Fix incomplete
predict_proba
method delegation frommodel_selection.GridSearchCV
tolinear_model.SGDClassifier
(#7159) by Yichuan Liu.
Metrics
- Fix bug in
metrics.silhouette_score
in which clusters of size 1 were incorrectly scored. They should get a score of 0. By Joel Nothman. - Fix bug in
metrics.silhouette_samples
so that it now works with arbitrary labels, not just those ranging from 0 to n_clusters - 1. - Fix bug where expected and adjusted mutual information were incorrect if
cluster contingency cells exceeded
2**16
. By Joel Nothman. metrics.pairwise.pairwise_distances
now converts arrays to boolean arrays when required inscipy.spatial.distance
. #5460 by Tom Dupre la Tour.- Fix sparse input support in
metrics.silhouette_score
as well as example examples/text/document_clustering.py. By YenChen Lin. metrics.roc_curve
andmetrics.precision_recall_curve
no longer roundy_score
values when creating ROC curves; this was causing problems for users with very small differences in scores (#7353).
Miscellaneous
model_selection.tests._search._check_param_grid
now works correctly with all types that extends/implements Sequence (except string), including range (Python 3.x) and xrange (Python 2.x). #7323 by Viacheslav Kovalevskyi.utils.extmath.randomized_range_finder
is more numerically stable when many power iterations are requested, since it applies LU normalization by default. Ifn_iter<2
numerical issues are unlikely, thus no normalization is applied. Other normalization options are available:'none', 'LU'
and'QR'
. #5141 by Giorgio Patrini.- Fix a bug where some formats of
scipy.sparse
matrix, and estimators with them as parameters, could not be passed tobase.clone
. By Loic Esteve. datasets.load_svmlight_file
now is able to read long int QID values. #7101 by Ibraim Ganiev.
API changes summary¶
Linear, kernelized and related models
residual_metric
has been deprecated inlinear_model.RANSACRegressor
. Useloss
instead. By Manoj Kumar.- Access to public attributes
.X_
and.y_
has been deprecated inisotonic.IsotonicRegression
. By Jonathan Arfa.
Decomposition, manifold learning and clustering
- The old
mixture.DPGMM
is deprecated in favor of the newmixture.BayesianGaussianMixture
(with the parameterweight_concentration_prior_type='dirichlet_process'
). The new class solves the computational problems of the old class and computes the Gaussian mixture with a Dirichlet process prior faster than before. #7295 by Wei Xue and Thierry Guillemot. - The old
mixture.VBGMM
is deprecated in favor of the newmixture.BayesianGaussianMixture
(with the parameterweight_concentration_prior_type='dirichlet_distribution'
). The new class solves the computational problems of the old class and computes the Variational Bayesian Gaussian mixture faster than before. #6651 by Wei Xue and Thierry Guillemot. - The old
mixture.GMM
is deprecated in favor of the newmixture.GaussianMixture
. The new class computes the Gaussian mixture faster than before and some of computational problems have been solved. #6666 by Wei Xue and Thierry Guillemot.
Model evaluation and meta-estimators
- The
sklearn.cross_validation
,sklearn.grid_search
andsklearn.learning_curve
have been deprecated and the classes and functions have been reorganized into thesklearn.model_selection
module. Ref Model Selection Enhancements and API Changes for more information. #4294 by Raghav RV. - The
grid_scores_
attribute ofmodel_selection.GridSearchCV
andmodel_selection.RandomizedSearchCV
is deprecated in favor of the attributecv_results_
. Ref Model Selection Enhancements and API Changes for more information. #6697 by Raghav RV. - The parameters
n_iter
orn_folds
in old CV splitters are replaced by the new parametern_splits
since it can provide a consistent and unambiguous interface to represent the number of train-test splits. #7187 by YenChen Lin. classes
parameter was renamed tolabels
inmetrics.hamming_loss
. #7260 by Sebastián Vanrell.- The splitter classes
LabelKFold
,LabelShuffleSplit
,LeaveOneLabelOut
andLeavePLabelsOut
are renamed tomodel_selection.GroupKFold
,model_selection.GroupShuffleSplit
,model_selection.LeaveOneGroupOut
andmodel_selection.LeavePGroupsOut
respectively. Also the parameterlabels
in thesplit
method of the newly renamed splittersmodel_selection.LeaveOneGroupOut
andmodel_selection.LeavePGroupsOut
is renamed togroups
. Additionally inmodel_selection.LeavePGroupsOut
, the parametern_labels
is renamed ton_groups
. #6660 by Raghav RV. - Error and loss names for
scoring
parameters are now prefixed by'neg_'
, such asneg_mean_squared_error
. The unprefixed versions are deprecated and will be removed in version 0.20. #7261 by Tim Head.
Code Contributors¶
Aditya Joshi, Alejandro, Alexander Fabisch, Alexander Loginov, Alexander Minyushkin, Alexander Rudy, Alexandre Abadie, Alexandre Abraham, Alexandre Gramfort, Alexandre Saint, alexfields, Alvaro Ulloa, alyssaq, Amlan Kar, Andreas Mueller, andrew giessel, Andrew Jackson, Andrew McCulloh, Andrew Murray, Anish Shah, Arafat, Archit Sharma, Ariel Rokem, Arnaud Joly, Arnaud Rachez, Arthur Mensch, Ash Hoover, asnt, b0noI, Behzad Tabibian, Bernardo, Bernhard Kratzwald, Bhargav Mangipudi, blakeflei, Boyuan Deng, Brandon Carter, Brett Naul, Brian McFee, Caio Oliveira, Camilo Lamus, Carol Willing, Cass, CeShine Lee, Charles Truong, Chyi-Kwei Yau, CJ Carey, codevig, Colin Ni, Dan Shiebler, Daniel, Daniel Hnyk, David Ellis, David Nicholson, David Staub, David Thaler, David Warshaw, Davide Lasagna, Deborah, definitelyuncertain, Didi Bar-Zev, djipey, dsquareindia, edwinENSAE, Elias Kuthe, Elvis DOHMATOB, Ethan White, Fabian Pedregosa, Fabio Ticconi, fisache, Florian Wilhelm, Francis, Francis O’Donovan, Gael Varoquaux, Ganiev Ibraim, ghg, Gilles Louppe, Giorgio Patrini, Giovanni Cherubin, Giovanni Lanzani, Glenn Qian, Gordon Mohr, govin-vatsan, Graham Clenaghan, Greg Reda, Greg Stupp, Guillaume Lemaitre, Gustav Mörtberg, halwai, Harizo Rajaona, Harry Mavroforakis, hashcode55, hdmetor, Henry Lin, Hobson Lane, Hugo Bowne-Anderson, Igor Andriushchenko, Imaculate, Inki Hwang, Isaac Sijaranamual, Ishank Gulati, Issam Laradji, Iver Jordal, jackmartin, Jacob Schreiber, Jake Vanderplas, James Fiedler, James Routley, Jan Zikes, Janna Brettingen, jarfa, Jason Laska, jblackburne, jeff levesque, Jeffrey Blackburne, Jeffrey04, Jeremy Hintz, jeremynixon, Jeroen, Jessica Yung, Jill-Jênn Vie, Jimmy Jia, Jiyuan Qian, Joel Nothman, johannah, John, John Boersma, John Kirkham, John Moeller, jonathan.striebel, joncrall, Jordi, Joseph Munoz, Joshua Cook, JPFrancoia, jrfiedler, JulianKahnert, juliathebrave, kaichogami, KamalakerDadi, Kenneth Lyons, Kevin Wang, kingjr, kjell, Konstantin Podshumok, Kornel Kielczewski, Krishna Kalyan, krishnakalyan3, Kvle Putnam, Kyle Jackson, Lars Buitinck, ldavid, LeiG, LeightonZhang, Leland McInnes, Liang-Chi Hsieh, Lilian Besson, lizsz, Loic Esteve, Louis Tiao, Léonie Borne, Mads Jensen, Maniteja Nandana, Manoj Kumar, Manvendra Singh, Marco, Mario Krell, Mark Bao, Mark Szepieniec, Martin Madsen, MartinBpr, MaryanMorel, Massil, Matheus, Mathieu Blondel, Mathieu Dubois, Matteo, Matthias Ekman, Max Moroz, Michael Scherer, michiaki ariga, Mikhail Korobov, Moussa Taifi, mrandrewandrade, Mridul Seth, nadya-p, Naoya Kanai, Nate George, Nelle Varoquaux, Nelson Liu, Nick James, NickleDave, Nico, Nicolas Goix, Nikolay Mayorov, ningchi, nlathia, okbalefthanded, Okhlopkov, Olivier Grisel, Panos Louridas, Paul Strickland, Perrine Letellier, pestrickland, Peter Fischer, Pieter, Ping-Yao, Chang, practicalswift, Preston Parry, Qimu Zheng, Rachit Kansal, Raghav RV, Ralf Gommers, Ramana.S, Rammig, Randy Olson, Rob Alexander, Robert Lutz, Robin Schucker, Rohan Jain, Ruifeng Zheng, Ryan Yu, Rémy Léone, saihttam, Saiwing Yeung, Sam Shleifer, Samuel St-Jean, Sartaj Singh, Sasank Chilamkurthy, saurabh.bansod, Scott Andrews, Scott Lowe, seales, Sebastian Raschka, Sebastian Saeger, Sebastián Vanrell, Sergei Lebedev, shagun Sodhani, shanmuga cv, Shashank Shekhar, shawpan, shengxiduan, Shota, shuckle16, Skipper Seabold, sklearn-ci, SmedbergM, srvanrell, Sébastien Lerique, Taranjeet, themrmax, Thierry, Thierry Guillemot, Thomas, Thomas Hallock, Thomas Moreau, Tim Head, tKammy, toastedcornflakes, Tom, TomDLT, Toshihiro Kamishima, tracer0tong, Trent Hauck, trevorstephens, Tue Vo, Varun, Varun Jewalikar, Viacheslav, Vighnesh Birodkar, Vikram, Villu Ruusmann, Vinayak Mehta, walter, waterponey, Wenhua Yang, Wenjian Huang, Will Welch, wyseguy7, xyguo, yanlend, Yaroslav Halchenko, yelite, Yen, YenChenLin, Yichuan Liu, Yoav Ram, Yoshiki, Zheng RuiFeng, zivori, Óscar Nájera