B 0dSq@sbdZddlmZmZddlmZddlZddlm Z ddl m Z ddl m Z ddl mZmZmZdd l mZmZdd lmZdd lmZd d lmZd dlmZddlmZddlmZddlmZddlmZddl m!Z!ddl"m#Z#ddl$m%Z%ddl&m'Z'ddl(m)Z)ddl(m*Z*ddl+m,Z,GdddeeedZ-Gdddee-Z.Gdd d ee-Z/dS)!z"Stacking classifier and regressor.)ABCMetaabstractmethod)deepcopyN)Parallel)clone)ClassifierMixinRegressorMixinTransformerMixin) is_classifier is_regressor)NotFittedError) _VisualBlock)_fit_single_estimator)_BaseHeterogeneousEnsemble)LogisticRegression)RidgeCV)cross_val_predict)check_cv) LabelEncoder)Bunch)if_delegate_has_method)check_classification_targets)check_is_fitted) column_or_1d)delayedcseZdZdZedddddddfdd Zd d Zd d Zed dZ dddZ e ddZ ddZ eddddZddZZS) _BaseStackingzBase class for stacking method.NautorF)cv stack_methodn_jobsverbose passthroughcs6tj|d||_||_||_||_||_||_dS)N) estimators)super__init__final_estimatorrr r!r"r#)selfr$r'rr r!r"r#) __class__L/var/www/html/venv/lib/python3.7/site-packages/sklearn/ensemble/_stacking.pyr&)s z_BaseStacking.__init__cCs&|jdk rt|j|_n t||_dS)N)r'rfinal_estimator_)r(defaultr*r*r+_clone_final_estimator=s z$_BaseStacking._clone_final_estimatorcCsg}xtt|D]h\}}|jdkr4||ddq|j|dkrlt|jdkrl||ddddfq||qW|jr||t |rtj ||j dSt |S)a1Concatenate the predictions of each first layer learner and possibly the input dataset `X`. If `X` is sparse and `self.passthrough` is False, the output of `transform` will be dense (the predictions). If `X` is sparse and `self.passthrough` is True, the output of `transform` will be sparse. This helper is in charge of ensuring the predictions are 2D arrays and it will drop one of the probability column when using probabilities in the binary case. Indeed, the p(y|c=0) = 1 - p(y|c=1) r predict_probarN)format) enumeratendimappendZreshape stack_method_lenclasses_r#sparseissparseZhstackr1np)r(X predictionsX_metaZest_idxpredsr*r*r+_concatenate_predictionsCs    z&_BaseStacking._concatenate_predictionscCs\|dkr dS|dkr:t|ddr$dSt|ddr4dSdSnt||sTtd|||SdS)Ndroprr0decision_functionpredictz9Underlying estimator {} does not implement the method {}.)getattrhasattr ValueErrorr1)nameZ estimatormethodr*r*r+ _method_namegs    z_BaseStacking._method_namec s|\}}jgt|}tjdfdd|D_t_d}xXt ||D]J\}} | dkrj|} | j|<|d7}t | dr| j _ q`dj|<q`Wt j tdt d rjd krtj_fd d t |||D_d k rd ind tjdfddt |jD} dd t j|D_| } tj| dS)aFit the estimators. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. y : array-like of shape (n_samples,) Target values. sample_weight : array-like of shape (n_samples,) or default=None Sample weights. If None, then samples are equally weighted. Note that this is supported only if all underlying estimators support sample weights. .. versionchanged:: 0.23 when not None, `sample_weight` is passed to all underlying estimators Returns ------- self : object )r!c3s,|]$}|dkrttt|VqdS)r@N)rrr).0est)r; sample_weightyr*r+ sz$_BaseStacking.fit..rr@rfeature_names_in_)rL classifier random_stateNcs g|]\}}}|||qSr*)rH)rIrFrJmeth)r(r*r+ sz%_BaseStacking.fit..rKc 3sB|]:\}}|dkrttt|t|jjdVqdS)r@)rrGr! fit_paramsr"N)rrrrr!r")rIrJrQ)r;rrSr(rLr*r+rMs cSsg|]\}}|dkr|qS)r@r*)rIrQrJr*r*r+rRs)rK)Z_validate_estimators_validate_final_estimatorr r6rr! estimators_rZnamed_estimators_ziprDrNrrr rPr:randomZ RandomStater5r?rr,) r(r;rLrKnamesZall_estimatorsr Zest_fitted_idxZname_estZorg_estZcurrent_estimatorr<r=r*)r;rrSrKr(rLr+fit{s>            z_BaseStacking.fitc CsPy t|Wn6tk rB}zt|jjd|Wdd}~XYnX|jdjS)z+Number of features seen during :term:`fit`.z' object has no attribute n_features_in_Nr)rr AttributeErrorr)__name__rUn_features_in_)r(Znfer*r*r+r\s z_BaseStacking.n_features_in_cs0t|fddt|j|jD}||S)z9Concatenate and return the predictions of the estimators.cs&g|]\}}|dkrt||qS)r@)rC)rIrJrQ)r;r*r+rRsz,_BaseStacking._transform..)rrVrUr5r?)r(r;r<r*)r;r+ _transforms z_BaseStacking._transformr,)delegatecKst||jj||f|S)aPredict target for X. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. **predict_params : dict of str -> obj Parameters to the `predict` called by the `final_estimator`. Note that this may be used to return uncertainties from some estimators with `return_std` or `return_cov`. Be aware that it will only accounts for uncertainty in the final estimator. Returns ------- y_pred : ndarray of shape (n_samples,) or (n_samples, n_output) Predicted targets. )rr,rB transform)r(r;predict_paramsr*r*r+rBsz_BaseStacking.predictcCsDt|j\}}td||dd}td|gdgdd}td||fddS)NparallelF)rX dash_wrappedr'serial)rb)rVr$r)r(r'rXr$raZ final_blockr*r*r+_sk_visual_block_ s z_BaseStacking._sk_visual_block_)N)N)r[ __module__ __qualname____doc__rr&r.r? staticmethodrHrYpropertyr\r]rrBrd __classcell__r*r*)r)r+r&s  $  d  r) metaclasscseZdZdZdddddddfdd Zd d Zdfd d Zed dfddZed dddZ ed dddZ ddZ fddZ Z S)StackingClassifieraStack of estimators with a final classifier. Stacked generalization consists in stacking the output of individual estimator and use a classifier to compute the final prediction. Stacking allows to use the strength of each individual estimator by using their output as input of a final estimator. Note that `estimators_` are fitted on the full `X` while `final_estimator_` is trained using cross-validated predictions of the base estimators using `cross_val_predict`. Read more in the :ref:`User Guide `. .. versionadded:: 0.22 Parameters ---------- estimators : list of (str, estimator) Base estimators which will be stacked together. Each element of the list is defined as a tuple of string (i.e. name) and an estimator instance. An estimator can be set to 'drop' using `set_params`. final_estimator : estimator, default=None A classifier which will be used to combine the base estimators. The default classifier is a :class:`~sklearn.linear_model.LogisticRegression`. cv : int, cross-validation generator or an iterable, default=None Determines the cross-validation splitting strategy used in `cross_val_predict` to train `final_estimator`. Possible inputs for cv are: * None, to use the default 5-fold cross validation, * integer, to specify the number of folds in a (Stratified) KFold, * An object to be used as a cross-validation generator, * An iterable yielding train, test splits. For integer/None inputs, if the estimator is a classifier and y is either binary or multiclass, :class:`~sklearn.model_selection.StratifiedKFold` is used. In all other cases, :class:`~sklearn.model_selection.KFold` is used. These splitters are instantiated with `shuffle=False` so the splits will be the same across calls. Refer :ref:`User Guide ` for the various cross-validation strategies that can be used here. .. note:: A larger number of split will provide no benefits if the number of training samples is large enough. Indeed, the training time will increase. ``cv`` is not used for model evaluation but for prediction. stack_method : {'auto', 'predict_proba', 'decision_function', 'predict'}, default='auto' Methods called for each base estimator. It can be: * if 'auto', it will try to invoke, for each estimator, `'predict_proba'`, `'decision_function'` or `'predict'` in that order. * otherwise, one of `'predict_proba'`, `'decision_function'` or `'predict'`. If the method is not implemented by the estimator, it will raise an error. n_jobs : int, default=None The number of jobs to run in parallel all `estimators` `fit`. `None` means 1 unless in a `joblib.parallel_backend` context. -1 means using all processors. See Glossary for more details. passthrough : bool, default=False When False, only the predictions of estimators will be used as training data for `final_estimator`. When True, the `final_estimator` is trained on the predictions as well as the original training data. verbose : int, default=0 Verbosity level. Attributes ---------- classes_ : ndarray of shape (n_classes,) Class labels. estimators_ : list of estimators The elements of the estimators parameter, having been fitted on the training data. If an estimator has been set to `'drop'`, it will not appear in `estimators_`. named_estimators_ : :class:`~sklearn.utils.Bunch` Attribute to access any fitted sub-estimators by name. n_features_in_ : int Number of features seen during :term:`fit`. Only defined if the underlying classifier exposes such an attribute when fit. .. versionadded:: 0.24 feature_names_in_ : ndarray of shape (`n_features_in_`,) Names of features seen during :term:`fit`. Only defined if the underlying estimators expose such an attribute when fit. .. versionadded:: 1.0 final_estimator_ : estimator The classifier which predicts given the output of `estimators_`. stack_method_ : list of str The method used by each base estimator. See Also -------- StackingRegressor : Stack of estimators with a final regressor. Notes ----- When `predict_proba` is used by each estimator (i.e. most of the time for `stack_method='auto'` or specifically for `stack_method='predict_proba'`), The first column predicted by each estimator will be dropped in the case of a binary classification problem. Indeed, both feature will be perfectly collinear. References ---------- .. [1] Wolpert, David H. "Stacked generalization." Neural networks 5.2 (1992): 241-259. Examples -------- >>> from sklearn.datasets import load_iris >>> from sklearn.ensemble import RandomForestClassifier >>> from sklearn.svm import LinearSVC >>> from sklearn.linear_model import LogisticRegression >>> from sklearn.preprocessing import StandardScaler >>> from sklearn.pipeline import make_pipeline >>> from sklearn.ensemble import StackingClassifier >>> X, y = load_iris(return_X_y=True) >>> estimators = [ ... ('rf', RandomForestClassifier(n_estimators=10, random_state=42)), ... ('svr', make_pipeline(StandardScaler(), ... LinearSVC(random_state=42))) ... ] >>> clf = StackingClassifier( ... estimators=estimators, final_estimator=LogisticRegression() ... ) >>> from sklearn.model_selection import train_test_split >>> X_train, X_test, y_train, y_test = train_test_split( ... X, y, stratify=y, random_state=42 ... ) >>> clf.fit(X_train, y_train).score(X_test, y_test) 0.9... NrFr)rr r!r#r"c stj|||||||ddS)N)r$r'rr r!r#r")r%r&)r(r$r'rr r!r#r")r)r*r+r&s zStackingClassifier.__init__cCs,|jtdt|js(td|jdS)N)r-z:'final_estimator' parameter should be a classifier. Got {})r.rr r,rEr1)r(r*r*r+rTs  z,StackingClassifier._validate_final_estimatorcs8t|t||_|jj|_t||j||S)aFit the estimators. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. y : array-like of shape (n_samples,) Target values. sample_weight : array-like of shape (n_samples,), default=None Sample weights. If None, then samples are equally weighted. Note that this is supported only if all underlying estimators support sample weights. Returns ------- self : object Returns a fitted instance of estimator. )rrrY_ler7r%r_)r(r;rLrK)r)r*r+rYs zStackingClassifier.fitr,)r^c stj|f|}|j|S)aPredict target for X. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. **predict_params : dict of str -> obj Parameters to the `predict` called by the `final_estimator`. Note that this may be used to return uncertainties from some estimators with `return_std` or `return_cov`. Be aware that it will only accounts for uncertainty in the final estimator. Returns ------- y_pred : ndarray of shape (n_samples,) or (n_samples, n_output) Predicted targets. )r%rBrmZinverse_transform)r(r;r`Zy_pred)r)r*r+rBszStackingClassifier.predictcCst||j||S)aPredict class probabilities for `X` using the final estimator. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. Returns ------- probabilities : ndarray of shape (n_samples, n_classes) or list of ndarray of shape (n_output,) The class probabilities of the input samples. )rr,r0r_)r(r;r*r*r+r0sz StackingClassifier.predict_probacCst||j||S)aDecision function for samples in `X` using the final estimator. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. Returns ------- decisions : ndarray of shape (n_samples,), (n_samples, n_classes), or (n_samples, n_classes * (n_classes-1) / 2) The decision function computed the final estimator. )rr,rAr_)r(r;r*r*r+rAsz$StackingClassifier.decision_functioncCs ||S)aReturn class labels or probabilities for X for each estimator. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. Returns ------- y_preds : ndarray of shape (n_samples, n_estimators) or (n_samples, n_classes * n_estimators) Prediction outputs for each estimator. )r])r(r;r*r*r+r_(szStackingClassifier.transformcs$|jdkrt}n|j}t|S)N)r'rr%rd)r(r')r)r*r+rd9s z$StackingClassifier._sk_visual_block_)N)N)r[rerfrgr&rTrYrrBr0rAr_rdrjr*r*)r)r+rls  rlcsXeZdZdZddddddfdd Zdd Zdfd d Zd d ZfddZZ S)StackingRegressoralStack of estimators with a final regressor. Stacked generalization consists in stacking the output of individual estimator and use a regressor to compute the final prediction. Stacking allows to use the strength of each individual estimator by using their output as input of a final estimator. Note that `estimators_` are fitted on the full `X` while `final_estimator_` is trained using cross-validated predictions of the base estimators using `cross_val_predict`. Read more in the :ref:`User Guide `. .. versionadded:: 0.22 Parameters ---------- estimators : list of (str, estimator) Base estimators which will be stacked together. Each element of the list is defined as a tuple of string (i.e. name) and an estimator instance. An estimator can be set to 'drop' using `set_params`. final_estimator : estimator, default=None A regressor which will be used to combine the base estimators. The default regressor is a :class:`~sklearn.linear_model.RidgeCV`. cv : int, cross-validation generator or an iterable, default=None Determines the cross-validation splitting strategy used in `cross_val_predict` to train `final_estimator`. Possible inputs for cv are: * None, to use the default 5-fold cross validation, * integer, to specify the number of folds in a (Stratified) KFold, * An object to be used as a cross-validation generator, * An iterable yielding train, test splits. For integer/None inputs, if the estimator is a classifier and y is either binary or multiclass, :class:`~sklearn.model_selection.StratifiedKFold` is used. In all other cases, :class:`~sklearn.model_selection.KFold` is used. These splitters are instantiated with `shuffle=False` so the splits will be the same across calls. Refer :ref:`User Guide ` for the various cross-validation strategies that can be used here. .. note:: A larger number of split will provide no benefits if the number of training samples is large enough. Indeed, the training time will increase. ``cv`` is not used for model evaluation but for prediction. n_jobs : int, default=None The number of jobs to run in parallel for `fit` of all `estimators`. `None` means 1 unless in a `joblib.parallel_backend` context. -1 means using all processors. See Glossary for more details. passthrough : bool, default=False When False, only the predictions of estimators will be used as training data for `final_estimator`. When True, the `final_estimator` is trained on the predictions as well as the original training data. verbose : int, default=0 Verbosity level. Attributes ---------- estimators_ : list of estimator The elements of the estimators parameter, having been fitted on the training data. If an estimator has been set to `'drop'`, it will not appear in `estimators_`. named_estimators_ : :class:`~sklearn.utils.Bunch` Attribute to access any fitted sub-estimators by name. n_features_in_ : int Number of features seen during :term:`fit`. Only defined if the underlying regressor exposes such an attribute when fit. .. versionadded:: 0.24 feature_names_in_ : ndarray of shape (`n_features_in_`,) Names of features seen during :term:`fit`. Only defined if the underlying estimators expose such an attribute when fit. .. versionadded:: 1.0 final_estimator_ : estimator The regressor to stacked the base estimators fitted. stack_method_ : list of str The method used by each base estimator. See Also -------- StackingClassifier : Stack of estimators with a final classifier. References ---------- .. [1] Wolpert, David H. "Stacked generalization." Neural networks 5.2 (1992): 241-259. Examples -------- >>> from sklearn.datasets import load_diabetes >>> from sklearn.linear_model import RidgeCV >>> from sklearn.svm import LinearSVR >>> from sklearn.ensemble import RandomForestRegressor >>> from sklearn.ensemble import StackingRegressor >>> X, y = load_diabetes(return_X_y=True) >>> estimators = [ ... ('lr', RidgeCV()), ... ('svr', LinearSVR(random_state=42)) ... ] >>> reg = StackingRegressor( ... estimators=estimators, ... final_estimator=RandomForestRegressor(n_estimators=10, ... random_state=42) ... ) >>> from sklearn.model_selection import train_test_split >>> X_train, X_test, y_train, y_test = train_test_split( ... X, y, random_state=42 ... ) >>> reg.fit(X_train, y_train).score(X_test, y_test) 0.3... NFr)rr!r#r"c stj|||d|||ddS)NrB)r$r'rr r!r#r")r%r&)r(r$r'rr!r#r")r)r*r+r&s zStackingRegressor.__init__cCs,|jtdt|js(td|jdS)N)r-z9'final_estimator' parameter should be a regressor. Got {})r.rr r,rEr1)r(r*r*r+rTs  z+StackingRegressor._validate_final_estimatorcst|dd}t|||S)aFit the estimators. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. y : array-like of shape (n_samples,) Target values. sample_weight : array-like of shape (n_samples,), default=None Sample weights. If None, then samples are equally weighted. Note that this is supported only if all underlying estimators support sample weights. Returns ------- self : object Returns a fitted instance. T)warn)rr%rY)r(r;rLrK)r)r*r+rYs zStackingRegressor.fitcCs ||S)aReturn the predictions for X for each estimator. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. Returns ------- y_preds : ndarray of shape (n_samples, n_estimators) Prediction outputs for each estimator. )r])r(r;r*r*r+r_szStackingRegressor.transformcs$|jdkrt}n|j}t|S)N)r'rr%rd)r(r')r)r*r+rd s z#StackingRegressor._sk_visual_block_)N)N) r[rerfrgr&rTrYr_rdrjr*r*)r)r+rnCs~  rn)0rgabcrrcopyrnumpyr:ZjoblibrZ scipy.sparser8baserrr r r r exceptionsr Zutils._estimator_html_reprr_baserrZ linear_modelrrZmodel_selectionrrZ preprocessingrutilsrZutils.metaestimatorsrZutils.multiclassrZutils.validationrrZ utils.fixesrrrlrnr*r*r*r+s:                   t,