Version 0.24#

For a short description of the main highlights of the release, please refer to Release Highlights for scikit-learn 0.24.

Legend for changelogs

Major Feature something big that you couldn’t do before.
Feature something that you couldn’t do before.
Efficiency an existing feature now may not require as much computation or memory.
Enhancement a miscellaneous minor improvement.
Fix something that previously didn’t work as documented – or according to reasonable expectations – should now work.
API Change you will need to change your code to have the same effect in the future; or a feature will be removed in the future.

Version 0.24.2#

April 2021

Changelog#

`sklearn.compose`#

Fix compose.ColumnTransformer.get_feature_names does not call get_feature_names on transformers with an empty column selection. #19579 by Thomas Fan.

`sklearn.cross_decomposition`#

Fix Fixed a regression in cross_decomposition.CCA. #19646 by Thomas Fan.
Fix cross_decomposition.PLSRegression raises warning for constant y residuals instead of a StopIteration error. #19922 by Thomas Fan.

`sklearn.decomposition`#

Fix Fixed a bug in decomposition.KernelPCA’s inverse_transform. #19732 by Kei Ishikawa.

`sklearn.ensemble`#

Fix Fixed a bug in ensemble.HistgradientBoostingRegressor fit with sample_weight parameter and least_absolute_deviation loss function. #19407 by Vadim Ushtanit.

`sklearn.feature_extraction`#

Fix Fixed a bug to support multiple strings for a category when sparse=False in feature_extraction.DictVectorizer. #19982 by guillaume Lemaitre.

`sklearn.gaussian_process`#

Fix Avoid explicitly forming inverse covariance matrix in gaussian_process.gaussianProcessRegressor when set to output standard deviation. With certain covariance matrices this inverse is unstable to compute explicitly. Calling Cholesky solver mitigates this issue in computation. #19939 by Ian Halvic.
Fix Avoid division by zero when scaling constant target in gaussian_process.gaussianProcessRegressor. It was due to a std. dev. equal to 0. Now, such case is detected and the std. dev. is affected to 1 avoiding a division by zero and thus the presence of NaN values in the normalized target. #19703 by @sobkevich, Boris Villazón-Terrazas and Alexandr Fonari.

`sklearn.linear_model`#

Fix : Fixed a bug in linear_model.LogisticRegression: the sample_weight object is not modified anymore. #19182 by Yosuke KOBAYASHI.

`sklearn.metrics`#

Fix metrics.top_k_accuracy_score now supports multiclass problems where only two classes appear in y_true and all the classes are specified in labels. #19721 by Joris Clement.

`sklearn.model_selection`#

Fix model_selection.RandomizedSearchCV and model_selection.gridSearchCV now correctly shows the score for single metrics and verbose > 2. #19659 by Thomas Fan.
Fix Some values in the cv_results_ attribute of model_selection.HalvingRandomSearchCV and model_selection.HalvinggridSearchCV were not properly converted to numpy arrays. #19211 by Nicolas Hug.
Fix The fit method of the successive halving parameter search (model_selection.HalvinggridSearchCV, and model_selection.HalvingRandomSearchCV) now correctly handles the groups parameter. #19847 by Xiaoyu Chai.

`sklearn.multioutput`#

Fix multioutput.MultiOutputRegressor now works with estimators that dynamically define predict during fitting, such as ensemble.StackingRegressor. #19308 by Thomas Fan.

`sklearn.preprocessing`#

Fix Validate the constructor parameter handle_unknown in preprocessing.OrdinalEncoder to only allow for 'error' and 'use_encoded_value' strategies. #19234 by guillaume Lemaitre <glemaitre>.
Fix Fix encoder categories having dtype=’S’ preprocessing.OneHotEncoder and preprocessing.OrdinalEncoder. #19727 by Andrew Delong.
Fix preprocessing.OrdinalEncoder.transform correctly handles unknown values for string dtypes. #19888 by Thomas Fan.
Fix preprocessing.OneHotEncoder.fit no longer alters the drop parameter. #19924 by Thomas Fan.

`sklearn.semi_supervised`#

Fix Avoid NaN during label propagation in LabelPropagation. #19271 by Zhaowei Wang.

`sklearn.tree`#

Fix Fix a bug in fit of tree.BaseDecisionTree that caused segmentation faults under certain conditions. fit now deep copies the Criterion object to prevent shared concurrent accesses. #19580 by Samuel Brice and Alex Adamson and Wil Yegelwel.

`sklearn.utils`#

Fix Better contains the CSS provided by utils.estimator_html_repr by giving CSS ids to the html representation. #19417 by Thomas Fan.

Version 0.24.1#

January 2021

Packaging#

The 0.24.0 scikit-learn wheels were not working with MacOS <1.15 due to libomp. The version of libomp used to build the wheels was too recent for older macOS versions. This issue has been fixed for 0.24.1 scikit-learn wheels. Scikit-learn wheels published on PyPI.org now officially support macOS 10.13 and later.

Changelog#

`sklearn.metrics`#

Fix Fix numerical stability bug that could happen in metrics.adjusted_mutual_info_score and metrics.mutual_info_score with NumPy 1.20+. #19179 by Thomas Fan.

`sklearn.semi_supervised`#

Fix semi_supervised.SelfTrainingClassifier is now accepting meta-estimator (e.g. ensemble.StackingClassifier). The validation of this estimator is done on the fitted estimator, once we know the existence of the method predict_proba. #19126 by guillaume Lemaitre.

Version 0.24.0#

December 2020

Changed models#

The following estimators and functions, when fit with the same data and parameters, may produce different models from the previous version. This often occurs due to changes in the modelling logic (bug fixes or enhancements), or in random sampling procedures.

Fix decomposition.KernelPCA behaviour is now more consistent between 32-bits and 64-bits data when the kernel has small positive eigenvalues.
Fix decomposition.TruncatedSVD becomes deterministic by exposing a random_state parameter.
Fix linear_model.Perceptron when penalty='elasticnet'.
Fix Change in the random sampling procedures for the center initialization of cluster.KMeans.

Details are listed in the changelog below.

(While we are trying to better inform users by providing this information, we cannot assure that this list is complete.)

Changelog#

`sklearn.base`#

Fix base.BaseEstimator.get_params now will raise an AttributeError if a parameter cannot be retrieved as an instance attribute. Previously it would return None. #17448 by Juan Carlos Alfaro Jiménez.

`sklearn.calibration`#

Efficiency calibration.CalibratedClassifierCV.fit now supports parallelization via joblib.Parallel using argument n_jobs. #17107 by Julien Jerphanion.
Enhancement Allow calibration.CalibratedClassifierCV use with prefit pipeline.Pipeline where data is not X is not array-like, sparse matrix or dataframe at the start. #17546 by Lucy Liu.
Enhancement Add ensemble parameter to calibration.CalibratedClassifierCV, which enables implementation of calibration via an ensemble of calibrators (current method) or just one calibrator using all the data (similar to the built-in feature of sklearn.svm estimators with the probabilities=True parameter). #17856 by Lucy Liu and Andrea Esuli.

`sklearn.cluster`#

Enhancement cluster.AgglomerativeClustering has a new parameter compute_distances. When set to True, distances between clusters are computed and stored in the distances_ attribute even when the parameter distance_threshold is not used. This new parameter is useful to produce dendrogram visualizations, but introduces a computational and memory overhead. #17984 by Michael Riedmann, Emilie Delattre, and Francesco Casalegno.
Enhancement cluster.SpectralClustering and cluster.spectral_clustering have a new keyword argument verbose. When set to True, additional messages will be displayed which can aid with debugging. #18052 by Sean O. Stalley.
Enhancement Added cluster.kmeans_plusplus as public function. Initialization by KMeans++ can now be called separately to generate initial cluster centroids. #17937 by @g-walsh
API Change cluster.MiniBatchKMeans attributes, counts_ and init_size_, are deprecated and will be removed in 1.1 (renaming of 0.26). #17864 by Jérémie du Boisberranger.

`sklearn.compose`#

Fix compose.ColumnTransformer will skip transformers the column selector is a list of bools that are False. #17616 by Thomas Fan.
Fix compose.ColumnTransformer now displays the remainder in the diagram display. #18167 by Thomas Fan.
Fix compose.ColumnTransformer enforces strict count and order of column names between fit and transform by raising an error instead of a warning, following the deprecation cycle. #18256 by Madhura Jayratne.

`sklearn.covariance`#

API Change Deprecates cv_alphas_ in favor of cv_results_['alphas'] and grid_scores_ in favor of split scores in cv_results_ in covariance.graphicalLassoCV. cv_alphas_ and grid_scores_ will be removed in version 1.1 (renaming of 0.26). #16392 by Thomas Fan.

`sklearn.cross_decomposition`#

Fix Fixed a bug in cross_decomposition.PLSSVD which would sometimes return components in the reversed order of importance. #17095 by Nicolas Hug.
Fix Fixed a bug in cross_decomposition.PLSSVD, cross_decomposition.CCA, and cross_decomposition.PLSCanonical, which would lead to incorrect predictions for est.transform(Y) when the training data is single-target. #17095 by Nicolas Hug.
Fix Increases the stability of cross_decomposition.CCA #18746 by Thomas Fan.
API Change The bounds of the n_components parameter is now restricted:
- into [1, min(n_samples, n_features, n_targets)], for cross_decomposition.PLSSVD, cross_decomposition.CCA, and cross_decomposition.PLSCanonical.
- into [1, n_features] or cross_decomposition.PLSRegression.
An error will be raised in 1.1 (renaming of 0.26). #17095 by Nicolas Hug.
API Change For cross_decomposition.PLSSVD, cross_decomposition.CCA, and cross_decomposition.PLSCanonical, the x_scores_ and y_scores_ attributes were deprecated and will be removed in 1.1 (renaming of 0.26). They can be retrieved by calling transform on the training data. The norm_y_weights attribute will also be removed. #17095 by Nicolas Hug.
API Change For cross_decomposition.PLSRegression, cross_decomposition.PLSCanonical, cross_decomposition.CCA, and cross_decomposition.PLSSVD, the x_mean_, y_mean_, x_std_, and y_std_ attributes were deprecated and will be removed in 1.1 (renaming of 0.26). #18768 by Maren Westermann.
Fix decomposition.TruncatedSVD becomes deterministic by using the random_state. It controls the weights’ initialization of the underlying ARPACK solver. :pr:` #18302` by gaurav Desai and Ivan Panico.

`sklearn.datasets`#

Feature datasets.fetch_openml now validates md5 checksum of arff files downloaded or cached to ensure data integrity. #14800 by Shashank Singh and Joel Nothman.
Enhancement datasets.fetch_openml now allows argument as_frame to be ‘auto’, which tries to convert returned data to pandas DataFrame unless data is sparse. #17396 by Jiaxiang.
Enhancement datasets.fetch_covtype now supports the optional argument as_frame; when it is set to True, the returned Bunch object’s data and frame members are pandas DataFrames, and the target member is a pandas Series. #17491 by Alex Liang.
Enhancement datasets.fetch_kddcup99 now supports the optional argument as_frame; when it is set to True, the returned Bunch object’s data and frame members are pandas DataFrames, and the target member is a pandas Series. #18280 by Alex Liang and guillaume Lemaitre.
Enhancement datasets.fetch_20newsgroups_vectorized now supports loading as a pandas DataFrame by setting as_frame=True. #17499 by Brigitta Sipőcz and guillaume Lemaitre.
API Change The default value of as_frame in datasets.fetch_openml is changed from False to ‘auto’. #17610 by Jiaxiang.

`sklearn.decomposition`#

API Change For decomposition.NMF, the init value, when ‘init=None’ and n_components <= min(n_samples, n_features) will be changed from 'nndsvd' to 'nndsvda' in 1.1 (renaming of 0.26). #18525 by Chiara Marmo.
Enhancement decomposition.FactorAnalysis now supports the optional argument rotation, which can take the value None, 'varimax' or 'quartimax'. #11064 by Jona Sassenhagen.
Enhancement decomposition.NMF now supports the optional parameter regularization, which can take the values None, ‘components’, ‘transformation’ or ‘both’, in accordance with decomposition.NMF.non_negative_factorization. #17414 by Bharat Raghunathan.
Fix decomposition.KernelPCA behaviour is now more consistent between 32-bits and 64-bits data input when the kernel has small positive eigenvalues. Small positive eigenvalues were not correctly discarded for 32-bits data. #18149 by Sylvain Marié.
Fix Fix decomposition.SparseCoder such that it follows scikit-learn API and support cloning. The attribute components_ is deprecated in 0.24 and will be removed in 1.1 (renaming of 0.26). This attribute was redundant with the dictionary attribute and constructor parameter. #17679 by Xavier Dupré.
Fix decomposition.TruncatedSVD.fit_transform consistently returns the same as decomposition.TruncatedSVD.fit followed by decomposition.TruncatedSVD.transform. #18528 by Albert Villanova del Moral and Ruifeng Zheng.

`sklearn.discriminant_analysis`#

Enhancement discriminant_analysis.LinearDiscriminantAnalysis can now use custom covariance estimate by setting the covariance_estimator parameter. #14446 by Hugo Richard.

`sklearn.ensemble`#

Major Feature ensemble.HistgradientBoostingRegressor and ensemble.HistgradientBoostingClassifier now have native support for categorical features with the categorical_features parameter. #18394 by Nicolas Hug and Thomas Fan.
Feature ensemble.HistgradientBoostingRegressor and ensemble.HistgradientBoostingClassifier now support the method staged_predict, which allows monitoring of each stage. #16985 by Hao Chun Chang.
Efficiency break cyclic references in the tree nodes used internally in ensemble.HistgradientBoostingRegressor and ensemble.HistgradientBoostingClassifier to allow for the timely garbage collection of large intermediate datastructures and to improve memory usage in fit. #18334 by Olivier grisel Nicolas Hug, Thomas Fan and Andreas Müller.
Efficiency Histogram initialization is now done in parallel in ensemble.HistgradientBoostingRegressor and ensemble.HistgradientBoostingClassifier which results in speed improvement for problems that build a lot of nodes on multicore machines. #18341 by Olivier grisel, Nicolas Hug, Thomas Fan, and Egor Smirnov.
Fix Fixed a bug in ensemble.HistgradientBoostingRegressor and ensemble.HistgradientBoostingClassifier which can now accept data with uint8 dtype in predict. #18410 by Nicolas Hug.
API Change The parameter n_classes_ is now deprecated in ensemble.gradientBoostingRegressor and returns 1. #17702 by Simona Maggio.
API Change Mean absolute error (‘mae’) is now deprecated for the parameter criterion in ensemble.gradientBoostingRegressor and ensemble.gradientBoostingClassifier. #18326 by Madhura Jayaratne.

`sklearn.exceptions`#

API Change exceptions.ChangedBehaviorWarning and exceptions.NonBLASDotWarning are deprecated and will be removed in 1.1 (renaming of 0.26). #17804 by Adrin Jalali.

`sklearn.feature_extraction`#

Enhancement feature_extraction.DictVectorizer accepts multiple values for one categorical feature. #17367 by Peng Yu and Chiara Marmo.
Fix feature_extraction.text.CountVectorizer raises an issue if a custom token pattern which capture more than one group is provided. #15427 by gangesh gudmalwar and Erin R Hoffman.

`sklearn.feature_selection`#

Feature Added feature_selection.SequentialFeatureSelector which implements forward and backward sequential feature selection. #6545 by Sebastian Raschka and #17159 by Nicolas Hug.
Feature A new parameter importance_getter was added to feature_selection.RFE, feature_selection.RFECV and feature_selection.SelectFromModel, allowing the user to specify an attribute name/path or a callable for extracting feature importance from the estimator. #15361 by Venkatachalam N.
Efficiency Reduce memory footprint in feature_selection.mutual_info_classif and feature_selection.mutual_info_regression by calling neighbors.KDTree for counting nearest neighbors. #17878 by Noel Rogers.
Enhancement feature_selection.RFE supports the option for the number of n_features_to_select to be given as a float representing the percentage of features to select. #17090 by Lisa Schwetlick and Marija Vlajic Wheeler.

`sklearn.gaussian_process`#

Enhancement A new method gaussian_process.kernel._check_bounds_params is called after fitting a gaussian Process and raises a ConvergenceWarning if the bounds of the hyperparameters are too tight. #12638 by Sylvain Lannuzel.

`sklearn.impute`#

Feature impute.SimpleImputer now supports a list of strings when strategy='most_frequent' or strategy='constant'. #17526 by Ayako YAgI and Juan Carlos Alfaro Jiménez.
Feature Added method impute.SimpleImputer.inverse_transform to revert imputed data to original when instantiated with add_indicator=True. #17612 by Srimukh Sripada.
Fix replace the default values in impute.IterativeImputer of min_value and max_value parameters to -np.inf and np.inf, respectively instead of None. However, the behaviour of the class does not change since None was defaulting to these values already. #16493 by Darshan N.
Fix impute.IterativeImputer will not attempt to set the estimator’s random_state attribute, allowing to use it with more external classes. #15636 by David Cortes.
Efficiency impute.SimpleImputer is now faster with object dtype array. when strategy='most_frequent' in SimpleImputer. #18987 by David Katz.

`sklearn.inspection`#

Feature inspection.partial_dependence and inspection.plot_partial_dependence now support calculating and plotting Individual Conditional Expectation (ICE) curves controlled by the kind parameter. #16619 by Madhura Jayratne.
Feature Add sample_weight parameter to inspection.permutation_importance. #16906 by Roei Kahny.
API Change Positional arguments are deprecated in inspection.PartialDependenceDisplay.plot and will error in 1.1 (renaming of 0.26). #18293 by Thomas Fan.

`sklearn.isotonic`#

Feature Expose fitted attributes X_thresholds_ and y_thresholds_ that hold the de-duplicated interpolation thresholds of an isotonic.IsotonicRegression instance for model inspection purpose. #16289 by Masashi Kishimoto and Olivier grisel.
Enhancement isotonic.IsotonicRegression now accepts 2d array with 1 feature as input array. #17379 by Jiaxiang.
Fix Add tolerance when determining duplicate X values to prevent inf values from being predicted by isotonic.IsotonicRegression. #18639 by Lucy Liu.

`sklearn.kernel_approximation`#

Feature Added class kernel_approximation.PolynomialCountSketch which implements the Tensor Sketch algorithm for polynomial kernel feature map approximation. #13003 by Daniel López Sánchez.
Efficiency kernel_approximation.Nystroem now supports parallelization via joblib.Parallel using argument n_jobs. #18545 by Laurenz Reitsam.

`sklearn.linear_model`#

Feature linear_model.LinearRegression now forces coefficients to be all positive when positive is set to True. #17578 by Joseph Knox, Nelle Varoquaux and Chiara Marmo.
Enhancement linear_model.RidgeCV now supports finding an optimal regularization value alpha for each target separately by setting alpha_per_target=True. This is only supported when using the default efficient leave-one-out cross-validation scheme cv=None. #6624 by Marijn van Vliet.
Fix Fixes bug in linear_model.TheilSenRegressor where predict and score would fail when fit_intercept=False and there was one feature during fitting. #18121 by Thomas Fan.
Fix Fixes bug in linear_model.ARDRegression where predict was raising an error when normalize=True and return_std=True because X_offset_ and X_scale_ were undefined. #18607 by fhaselbeck.
Fix Added the missing l1_ratio parameter in linear_model.Perceptron, to be used when penalty='elasticnet'. This changes the default from 0 to 0.15. #18622 by Haesun Park.

`sklearn.manifold`#

Efficiency Fixed #10493. Improve Local Linear Embedding (LLE) that raised MemoryError exception when used with large inputs. #17997 by Bertrand Maisonneuve.
Enhancement Add square_distances parameter to manifold.TSNE, which provides backward compatibility during deprecation of legacy squaring behavior. Distances will be squared by default in 1.1 (renaming of 0.26), and this parameter will be removed in 1.3. #17662 by Joshua Newton.
Fix manifold.MDS now correctly sets its _pairwise attribute. #18278 by Thomas Fan.

`sklearn.metrics`#

Feature Added metrics.cluster.pair_confusion_matrix implementing the confusion matrix arising from pairs of elements from two clusterings. #17412 by Uwe F Mayer.
Feature new metric metrics.top_k_accuracy_score. It’s a generalization of metrics.top_k_accuracy_score, the difference is that a prediction is considered correct as long as the true label is associated with one of the k highest predicted scores. metrics.accuracy_score is the special case of k = 1. #16625 by geoffrey Bolmier.
Feature Added metrics.det_curve to compute Detection Error Tradeoff curve classification metric. #10591 by Jeremy Karnowski and Daniel Mohns.
Feature Added metrics.plot_det_curve and metrics.DetCurveDisplay to ease the plot of DET curves. #18176 by guillaume Lemaitre.
Feature Added metrics.mean_absolute_percentage_error metric and the associated scorer for regression problems. #10708 fixed with the PR #15007 by Ashutosh Hathidara. The scorer and some practical test cases were taken from PR #10711 by Mohamed Ali Jamaoui.
Feature Added metrics.rand_score implementing the (unadjusted) Rand index. #17412 by Uwe F Mayer.
Feature metrics.plot_confusion_matrix now supports making colorbar optional in the matplotlib plot by setting colorbar=False. #17192 by Avi gupta
Enhancement Add sample_weight parameter to metrics.median_absolute_error. #17225 by Lucy Liu.
Enhancement Add pos_label parameter in metrics.plot_precision_recall_curve in order to specify the positive class to be used when computing the precision and recall statistics. #17569 by guillaume Lemaitre.
Enhancement Add pos_label parameter in metrics.plot_roc_curve in order to specify the positive class to be used when computing the roc auc statistics. #17651 by Clara Matos.
Fix Fixed a bug in metrics.classification_report which was raising AttributeError when called with output_dict=True for 0-length values. #17777 by Shubhanshu Mishra.
Fix Fixed a bug in metrics.classification_report which was raising AttributeError when called with output_dict=True for 0-length values. #17777 by Shubhanshu Mishra.
Fix Fixed a bug in metrics.jaccard_score which recommended the zero_division parameter when called with no true or predicted samples. #17826 by Richard Decal and Joseph Willard
Fix bug in metrics.hinge_loss where error occurs when y_true is missing some labels that are provided explicitly in the labels parameter. #17935 by Cary goltermann.
Fix Fix scorers that accept a pos_label parameter and compute their metrics from values returned by decision_function or predict_proba. Previously, they would return erroneous values when pos_label was not corresponding to classifier.classes_[1]. This is especially important when training classifiers directly with string labeled target classes. #18114 by guillaume Lemaitre.
Fix Fixed bug in metrics.plot_confusion_matrix where error occurs when y_true contains labels that were not previously seen by the classifier while the labels and display_labels parameters are set to None. #18405 by Thomas J. Fan and Yakov Pchelintsev.

`sklearn.model_selection`#

Major Feature Added (experimental) parameter search estimators model_selection.HalvingRandomSearchCV and model_selection.HalvinggridSearchCV which implement Successive Halving, and can be used as a drop-in replacements for model_selection.RandomizedSearchCV and model_selection.gridSearchCV. #13900 by Nicolas Hug, Joel Nothman and Andreas Müller.
Feature model_selection.RandomizedSearchCV and model_selection.gridSearchCV now have the method score_samples #17478 by Teon Brooks and Mohamed Maskani.
Enhancement model_selection.TimeSeriesSplit has two new keyword arguments test_size and gap. test_size allows the out-of-sample time series length to be fixed for all folds. gap removes a fixed number of samples between the train and test set on each fold. #13204 by Kyle Kosic.
Enhancement model_selection.permutation_test_score and model_selection.validation_curve now accept fit_params to pass additional estimator parameters. #18527 by gaurav Dhingra, Julien Jerphanion and Amanda Dsouza.
Enhancement model_selection.cross_val_score, model_selection.cross_validate, model_selection.gridSearchCV, and model_selection.RandomizedSearchCV allows estimator to fail scoring and replace the score with error_score. If error_score="raise", the error will be raised. #18343 by guillaume Lemaitre and Devi Sandeep.
Enhancement model_selection.learning_curve now accept fit_params to pass additional estimator parameters. #18595 by Amanda Dsouza.
Fix Fixed the len of model_selection.ParameterSampler when all distributions are lists and n_iter is more than the number of unique parameter combinations. #18222 by Nicolas Hug.
Fix A fix to raise warning when one or more CV splits of model_selection.gridSearchCV and model_selection.RandomizedSearchCV results in non-finite scores. #18266 by Subrat Sahu, Nirvan and Arthur Book.
Enhancement model_selection.gridSearchCV, model_selection.RandomizedSearchCV and model_selection.cross_validate support scoring being a callable returning a dictionary of of multiple metric names/values association. #15126 by Thomas Fan.

`sklearn.multiclass`#

Enhancement multiclass.OneVsOneClassifier now accepts the inputs with missing values. Hence, estimators which can handle missing values (may be a pipeline with imputation step) can be used as a estimator for multiclass wrappers. #17987 by Venkatachalam N.
Fix A fix to allow multiclass.OutputCodeClassifier to accept sparse input data in its fit and predict methods. The check for validity of the input is now delegated to the base estimator. #17233 by Zolisa Bleki.

`sklearn.multioutput`#

Enhancement multioutput.MultiOutputClassifier and multioutput.MultiOutputRegressor now accepts the inputs with missing values. Hence, estimators which can handle missing values (may be a pipeline with imputation step, HistgradientBoosting estimators) can be used as a estimator for multiclass wrappers. #17987 by Venkatachalam N.
Fix A fix to accept tuples for the order parameter in multioutput.ClassifierChain. #18124 by gus Brocchini and Amanda Dsouza.

`sklearn.naive_bayes`#

Enhancement Adds a parameter min_categories to naive_bayes.CategoricalNB that allows a minimum number of categories per feature to be specified. This allows categories unseen during training to be accounted for. #16326 by george Armstrong.
API Change The attributes coef_ and intercept_ are now deprecated in naive_bayes.MultinomialNB, naive_bayes.ComplementNB, naive_bayes.BernoulliNB and naive_bayes.CategoricalNB, and will be removed in v1.1 (renaming of 0.26). #17427 by Juan Carlos Alfaro Jiménez.

`sklearn.neighbors`#

Efficiency Speed up seuclidean, wminkowski, mahalanobis and haversine metrics in neighbors.DistanceMetric by avoiding unexpected gIL acquiring in Cython when setting n_jobs>1 in neighbors.KNeighborsClassifier, neighbors.KNeighborsRegressor, neighbors.RadiusNeighborsClassifier, neighbors.RadiusNeighborsRegressor, metrics.pairwise_distances and by validating data out of loops. #17038 by Wenbo Zhao.
Efficiency neighbors.NeighborsBase benefits of an improved algorithm = 'auto' heuristic. In addition to the previous set of rules, now, when the number of features exceeds 15, brute is selected, assuming the data intrinsic dimensionality is too high for tree-based methods. #17148 by geoffrey Bolmier.
Fix neighbors.BinaryTree will raise a ValueError when fitting on data array having points with different dimensions. #18691 by Chiara Marmo.
Fix neighbors.NearestCentroid with a numerical shrink_threshold will raise a ValueError when fitting on data with all constant features. #18370 by Trevor Waite.
Fix In methods radius_neighbors and radius_neighbors_graph of neighbors.NearestNeighbors, neighbors.RadiusNeighborsClassifier, neighbors.RadiusNeighborsRegressor, and neighbors.RadiusNeighborsTransformer, using sort_results=True now correctly sorts the results even when fitting with the “brute” algorithm. #18612 by Tom Dupre la Tour.

`sklearn.neural_network`#

Efficiency Neural net training and prediction are now a little faster. #17603, #17604, #17606, #17608, #17609, #17633, #17661, #17932 by Alex Henrie.
Enhancement Avoid converting float32 input to float64 in neural_network.BernoulliRBM. #16352 by Arthur Imbert.
Enhancement Support 32-bit computations in neural_network.MLPClassifier and neural_network.MLPRegressor. #17759 by Srimukh Sripada.
Fix Fix method neural_network.MLPClassifier.fit not iterating to max_iter if warm started. #18269 by Norbert Preining and guillaume Lemaitre.

`sklearn.pipeline`#

Enhancement References to transformers passed through transformer_weights to pipeline.FeatureUnion that aren’t present in transformer_list will raise a ValueError. #17876 by Cary goltermann.
Fix A slice of a pipeline.Pipeline now inherits the parameters of the original pipeline (memory and verbose). #18429 by Albert Villanova del Moral and Paweł Biernat.

`sklearn.preprocessing`#

Feature preprocessing.OneHotEncoder now supports missing values by treating them as a category. #17317 by Thomas Fan.
Feature Add a new handle_unknown parameter with a use_encoded_value option, along with a new unknown_value parameter, to preprocessing.OrdinalEncoder to allow unknown categories during transform and set the encoded value of the unknown categories. #17406 by Felix Wick and #18406 by Nicolas Hug.
Feature Add clip parameter to preprocessing.MinMaxScaler, which clips the transformed values of test data to feature_range. #17833 by Yashika Sharma.
Feature Add sample_weight parameter to preprocessing.StandardScaler. Allows setting individual weights for each sample. #18510 and #18447 and #16066 and #18682 by Maria Telenczuk and Albert Villanova and @panpiort8 and Alex gramfort.
Enhancement Verbose output of model_selection.gridSearchCV has been improved for readability. #16935 by Raghav Rajagopalan and Chiara Marmo.
Enhancement Add unit_variance to preprocessing.RobustScaler, which scales output data such that normally distributed features have a variance of 1. #17193 by Lucy Liu and Mabel Villalba.
Enhancement Add dtype parameter to preprocessing.KBinsDiscretizer. #16335 by Arthur Imbert.
Fix Raise error on sklearn.preprocessing.OneHotEncoder.inverse_transform when handle_unknown='error' and drop=None for samples encoded as all zeros. #14982 by Kevin Winata.

`sklearn.semi_supervised`#

Major Feature Added semi_supervised.SelfTrainingClassifier, a meta-classifier that allows any supervised classifier to function as a semi-supervised classifier that can learn from unlabeled data. #11682 by Oliver Rausch and Patrice Becker.
Fix Fix incorrect encoding when using unicode string dtypes in preprocessing.OneHotEncoder and preprocessing.OrdinalEncoder. #15763 by Thomas Fan.

`sklearn.svm`#

Enhancement invoke SciPy BLAS API for SVM kernel function in fit, predict and related methods of svm.SVC, svm.NuSVC, svm.SVR, svm.NuSVR, svm.OneClassSVM. #16530 by Shuhua Fan.

`sklearn.tree`#

Feature tree.DecisionTreeRegressor now supports the new splitting criterion 'poisson' useful for modeling count data. #17386 by Christian Lorentzen.
Enhancement tree.plot_tree now uses colors from the matplotlib configuration settings. #17187 by Andreas Müller.
API Change The parameter X_idx_sorted is now deprecated in tree.DecisionTreeClassifier.fit and tree.DecisionTreeRegressor.fit, and has not effect. #17614 by Juan Carlos Alfaro Jiménez.

`sklearn.utils`#

Enhancement Add check_methods_sample_order_invariance to check_estimator, which checks that estimator methods are invariant if applied to the same dataset with different sample order #17598 by Jason Ngo.
Enhancement Add support for weights in utils.sparse_func.incr_mean_variance_axis. By Maria Telenczuk and Alex gramfort.
Fix Raise ValueError with clear error message in utils.check_array for sparse DataFrames with mixed types. #17992 by Thomas J. Fan and Alex Shacked.
Fix Allow serialized tree based models to be unpickled on a machine with different endianness. #17644 by Qi Zhang.
Fix Check that we raise proper error when axis=1 and the dimensions do not match in utils.sparse_func.incr_mean_variance_axis. By Alex gramfort.

Miscellaneous#

Enhancement Calls to repr are now faster when print_changed_only=True, especially with meta-estimators. #18508 by Nathan C..

Code and documentation contributors

Thanks to everyone who has contributed to the maintenance and improvement of the project since version 0.23, including:

Abo7atm, Adam Spannbauer, Adrin Jalali, adrinjalali, Agamemnon Krasoulis, Akshay Deodhar, Albert Villanova del Moral, Alessandro gentile, Alex Henrie, Alex Itkes, Alex Liang, Alexander Lenail, alexandracraciun, Alexandre gramfort, alexshacked, Allan D Butler, Amanda Dsouza, amy12xx, Anand Tiwari, Anderson Nelson, Andreas Mueller, Ankit Choraria, Archana Subramaniyan, Arthur Imbert, Ashutosh Hathidara, Ashutosh Kushwaha, Atsushi Nukariya, Aura Munoz, AutoViz and Auto_ViML, Avi gupta, Avinash Anakal, Ayako YAgI, barankarakus, barberogaston, beatrizsmg, Ben Mainye, Benjamin Bossan, Benjamin Pedigo, Bharat Raghunathan, Bhavika Devnani, Biprateep Dey, bmaisonn, Bo Chang, Boris Villazón-Terrazas, brigi, Brigitta Sipőcz, Bruno Charron, Byron Smith, Cary goltermann, Cat Chenal, CeeThinwa, chaitanyamogal, Charles Patel, Chiara Marmo, Christian Kastner, Christian Lorentzen, Christoph Deil, Christos Aridas, Clara Matos, clmbst, Coelhudo, crispinlogan, Cristina Mulas, Daniel López, Daniel Mohns, darioka, Darshan N, david-cortes, Declan O’Neill, Deeksha Madan, Elizabeth DuPre, Eric Fiegel, Eric Larson, Erich Schubert, Erin Khoo, Erin R Hoffman, eschibli, Felix Wick, fhaselbeck, Forrest Koch, Francesco Casalegno, Frans Larsson, gael Varoquaux, gaurav Desai, gaurav Sheni, genvalen, geoffrey Bolmier, george Armstrong, george Kiragu, gesa Stupperich, ghislain Antony Vaillant, gim Seng, gordon Walsh, gregory R. Lee, guillaume Chevalier, guillaume Lemaitre, Haesun Park, Hannah Bohle, Hao Chun Chang, Harry Scholes, Harsh Soni, Henry, Hirofumi Suzuki, Hitesh Somani, Hoda1394, Hugo Le Moine, hugorichard, indecisiveuser, Isuru Fernando, Ivan Wiryadi, j0rd1smit, Jaehyun Ahn, Jake Tae, James Hoctor, Jan Vesely, Jeevan Anand Anne, JeroenPeterBos, JHayes, Jiaxiang, Jie Zheng, Jigna Panchal, jim0421, Jin Li, Joaquin Vanschoren, Joel Nothman, Jona Sassenhagen, Jonathan, Jorge gorbe Moya, Joseph Lucas, Joshua Newton, Juan Carlos Alfaro Jiménez, Julien Jerphanion, Justin Huber, Jérémie du Boisberranger, Kartik Chugh, Katarina Slama, kaylani2, Kendrick Cetina, Kenny Huynh, Kevin Markham, Kevin Winata, Kiril Isakov, kishimoto, Koki Nishihara, Krum Arnaudov, Kyle Kosic, Lauren Oldja, Laurenz Reitsam, Lisa Schwetlick, Louis Douge, Louis guitton, Lucy Liu, Madhura Jayaratne, maikia, Manimaran, Manuel López-Ibáñez, Maren Westermann, Maria Telenczuk, Mariam-ke, Marijn van Vliet, Markus Löning, Martin Scheubrein, Martina g. Vilas, Martina Megasari, Mateusz górski, mathschy, mathurinm, Matthias Bussonnier, Max Del giudice, Michael, Milan Straka, Muoki Caleb, N. Haiat, Nadia Tahiri, Ph. D, Naoki Hamada, Neil Botelho, Nicolas Hug, Nils Werner, noelano, Norbert Preining, oj_lappi, Oleh Kozynets, Olivier grisel, Pankaj Jindal, Pardeep Singh, Parthiv Chigurupati, Patrice Becker, Pete green, pgithubs, Poorna Kumar, Prabakaran Kumaresshan, Probinette4, pspachtholz, pwalchessen, Qi Zhang, rachel fischoff, Rachit Toshniwal, Rafey Iqbal Rahman, Rahul Jakhar, Ram Rachum, RamyaNP, rauwuckl, Ravi Kiran Boggavarapu, Ray Bell, Reshama Shaikh, Richard Decal, Rishi Advani, Rithvik Rao, Rob Romijnders, roei, Romain Tavenard, Roman Yurchak, Ruby Werman, Ryotaro Tsukada, sadak, Saket Khandelwal, Sam, Sam Ezebunandu, Sam Kimbinyi, Sarah Brown, Saurabh Jain, Sean O. Stalley, Sergio, Shail Shah, Shane Keller, Shao Yang Hong, Shashank Singh, Shooter23, Shubhanshu Mishra, simonamaggio, Soledad galli, Srimukh Sripada, Stephan Steinfurt, subrat93, Sunitha Selvan, Swier, Sylvain Marié, SylvainLan, t-kusanagi2, Teon L Brooks, Terence Honles, Thijs van den Berg, Thomas J Fan, Thomas J. Fan, Thomas S Benjamin, Thomas9292, Thorben Jensen, tijanajovanovic, Timo Kaufmann, tnwei, Tom Dupré la Tour, Trevor Waite, ufmayer, Umberto Lupo, Venkatachalam N, Vikas Pandey, Vinicius Rios Fuck, Violeta, watchtheblur, Wenbo Zhao, willpeppo, xavier dupré, Xethan, Xue Qianming, xun-tang, yagi-3, Yakov Pchelintsev, Yashika Sharma, Yi-Yan ge, Yue Wu, Yutaro Ikeda, Zaccharie Ramzi, zoj613, Zhao Feng.