U f{jd@sddlZddlmZddlmZmZddlmZddl m Z ddl m Z m Z mZmZeGd d d eeZed d Zd dZddZddZddZeGdddeZddZddZddZddZdS)N)TransformerMixin) BaseEstimator_check_estimator)_get_check_scoring) parallel_func)array_split_idx ProgressBarverbosefill_doccseZdZdZedddddZfddZdd Zd d Zd d Z ddZ ddZ ddZ ddZ dddZddZeddZZS)SlidingEstimatoraSearch Light. Fit, predict and score a series of models to each subset of the dataset along the last dimension. Each entry in the last dimension is referred to as a task. Parameters ---------- %(base_estimator)s %(scoring)s %(n_jobs)s %(verbose)s Attributes ---------- estimators_ : array-like, shape (n_tasks,) List of fitted scikit-learn estimators (one per task). N)r cCs,t|t|dd|_||_||_||_dS)N_estimator_type)rgetattrrbase_estimatorn_jobsscoring)selfrrrr rX/var/www/html/bwave/python/venv/lib/python3.8/site-packages/mne/decoding/search_light.py__init__$s zSlidingEstimator.__init__csBdtt|}t|dr:|dd}|dt|j7}|dS)N< estimators_z, fitted with %i estimators>)superr __repr__hasattrlenrrrepr_str __class__rrr-s   zSlidingEstimator.__repr__c s|ttj|jddd\}}t__djj f}t |jd|d.|fddt ||ddD}W5QRXt j |jdtd _d }|D] } | D]} | j|<|d 7}qqS) aFit a series of independent estimators to the dataset. Parameters ---------- X : array, shape (n_samples, nd_features, n_tasks) The training input samples. For each data slice, a clone estimator is fitted independently. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_tasks). y : array, shape (n_samples,) | (n_samples, n_targets) The target values. **fit_params : dict of string -> object Parameters to pass to the fit method of the estimator. Returns ------- self : object Return self. rFmax_jobsr z Fitting %smesgc3s,|]$\}}j||fVqdSN)rsubset).0pb_idxsplit fit_paramsp_funcpbryrr Psz'SlidingEstimator.fit..axis)dtyperr) _check_Xyr_sl_fitrshapelistrr-r"__name__r r npemptyobject) rXr0r-parallelrr& estimatorsidxZjob_estimatorsestrr,rfit4s*     zSlidingEstimator.fitcKs|j||f||S)a,Fit and transform a series of independent estimators to the dataset. Parameters ---------- X : array, shape (n_samples, nd_features, n_tasks) The training input samples. For each task, a clone estimator is fitted independently. The feature dimension can be multidimensional, e.g.:: X.shape = (n_samples, n_features_1, n_features_2, n_estimators) y : array, shape (n_samples,) | (n_samples, n_targets) The target values. **fit_params : dict of string -> object Parameters to pass to the fit method of the estimator. Returns ------- y_pred : array, shape (n_samples, n_tasks) | (n_samples, n_tasks, n_targets) The predicted values for each estimator. )rB transform)rr=r0r-rrr fit_transform`szSlidingEstimator.fit_transformc s||t|j|jdt|jkr2tdtt|j |jddd\}}d|j j f}t j ||dd}tt|j|\}}t|jd|d(|fdd t|||D} W5QRXt j| d d} | S) :Aux. function to make parallel predictions/transformation.r3The number of estimators does not match X.shape[-1]Fr#Transforming %sr2r%c3s(|] \}}}|||VqdSr')r()r)r*rAxmethodr.r/rrr1sz.SlidingEstimator._transform..r)r5 _check_methodrr7rr ValueErrorr _sl_transformrr"r9r: array_splitzipr r concatenate) rr=rJr>rr&X_splitsr@ est_splitsy_predrrIr _transformws.   zSlidingEstimator._transformcCs ||dS)aTransform each data slice/task with a series of independent estimators. The number of tasks in X should match the number of tasks/estimators given at fit time. Parameters ---------- X : array, shape (n_samples, nd_features, n_tasks) The input samples. For each data slice/task, the corresponding estimator makes a transformation of the data, e.g. ``[estimators[ii].transform(X[..., ii]) for ii in range(n_estimators)]``. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_tasks). Returns ------- Xt : array, shape (n_samples, n_estimators) The transformed values generated by each estimator. rCrTrr=rrrrCszSlidingEstimator.transformcCs ||dS)a+Predict each data slice/task with a series of independent estimators. The number of tasks in X should match the number of tasks/estimators given at fit time. Parameters ---------- X : array, shape (n_samples, nd_features, n_tasks) The input samples. For each data slice, the corresponding estimator makes the sample predictions, e.g.: ``[estimators[ii].predict(X[..., ii]) for ii in range(n_estimators)]``. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_tasks). Returns ------- y_pred : array, shape (n_samples, n_estimators) | (n_samples, n_tasks, n_targets) Predicted values for each estimator/data slice. predictrUrVrrrrWszSlidingEstimator.predictcCs ||dS)a*Predict each data slice with a series of independent estimators. The number of tasks in X should match the number of tasks/estimators given at fit time. Parameters ---------- X : array, shape (n_samples, nd_features, n_tasks) The input samples. For each data slice, the corresponding estimator makes the sample probabilistic predictions, e.g.: ``[estimators[ii].predict_proba(X[..., ii]) for ii in range(n_estimators)]``. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_tasks). Returns ------- y_pred : array, shape (n_samples, n_tasks, n_classes) Predicted probabilities for each estimator/data slice/task. predict_probarUrVrrrrXszSlidingEstimator.predict_probacCs ||dS)a:Estimate distances of each data slice to the hyperplanes. Parameters ---------- X : array, shape (n_samples, nd_features, n_tasks) The input samples. For each data slice, the corresponding estimator outputs the distance to the hyperplane, e.g.: ``[estimators[ii].decision_function(X[..., ii]) for ii in range(n_estimators)]``. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_estimators). Returns ------- y_pred : array, shape (n_samples, n_estimators, n_classes * (n_classes-1) // 2) Predicted distances for each estimator/data slice. Notes ----- This requires base_estimator to have a ``decision_function`` method. decision_functionrUrVrrrrYsz"SlidingEstimator.decision_functioncCsB|dk r,t|t|ks$t|dkr,td|jdkr>tddS)z"Aux. function to check input data.Nrz"X and y must have the same length.z"X must have at least 3 dimensions.)rrLndim)rr=r0rrrr5s  zSlidingEstimator._check_Xyc st}|||jdt|jkr,td||j|jtt t |j |jdd\}}t j ||dd}t |j|}|fddt||D}t j|dd}|S)akScore each estimator on each task. The number of tasks in X should match the number of tasks/estimators given at fit time, i.e. we need ``X.shape[-1] == len(self.estimators_)``. Parameters ---------- X : array, shape (n_samples, nd_features, n_tasks) The input samples. For each data slice, the corresponding estimator scores the prediction, e.g.: ``[estimators[ii].score(X[..., ii], y) for ii in range(n_estimators)]``. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_tasks). y : array, shape (n_samples,) | (n_samples, n_targets) The target values. Returns ------- score : array, shape (n_samples, n_estimators) Score for each estimator/task. rrF)r$r2c3s |]\}}||VqdSr'r)r)rArHr.rr0rrr1sz)SlidingEstimator.score..r)rr5r7rrrLrr_fix_aucr _sl_scorerr:rNrOrP) rr=r0 check_scoringr>rrQrRscorerr\rr`s$   zSlidingEstimator.scorecCs0t|jdds$td|jdf|jdjS)Nrclasses_zUclasses_ attribute available only if base_estimator has it, and estimator %s does not)rrAttributeErrorra)rrrrras  zSlidingEstimator.classes_)NN)N)r9 __module__ __qualname____doc__r rrrBrDrTrCrWrXrYr5r`propertyra __classcell__rrr!rr s" , -r c Ksbddlm}t}t|jdD]<}||}|j|d|f|f|||||dq |S)aeAux. function to fit SlidingEstimator in parallel. Fit a clone estimator to each slice of data. Parameters ---------- %(base_estimator)s X : array, shape (n_samples, nd_features, n_estimators) The target data. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_estimators) y : array, shape (n_sample, ) The target values. fit_params : dict | None Parameters to pass to the fit method of the estimator. Returns ------- estimators_ : list of estimators The fitted estimators. r)cloner.r)Z sklearn.baserhr8ranger7rBappendupdate) estimatorr=r0r/r-rhriirArrrr6$s  r6c Csbt|D]T\}}t||}||d|f}|dkr)rrtrrrrrr!rrrs   zGeneralizingEstimator.__repr__c s|tjdjjf}ttj|jddd\}}t |jdt j |d4|fddt ||dt j dD}W5QRXt j|d d }|S) rErGrFr#r%c3s(|] \}}j||VqdSr'rr()r)r*Zx_splitrJr.r/rrrr1sz3GeneralizingEstimator._transform..r3Z n_per_splitrr2)r5rKrr"r9r _gl_transformrr7r rrr r:rP)rr=rJr&r>rrSrrvrrTs(   z GeneralizingEstimator._transformcCs ||dS)aFTransform each data slice with all possible estimators. Parameters ---------- X : array, shape (n_samples, nd_features, n_slices) The input samples. For estimator the corresponding data slice is used to make a transformation. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_estimators). Returns ------- Xt : array, shape (n_samples, n_estimators, n_slices) The transformed values generated by each estimator. rCrUrVrrrrCszGeneralizingEstimator.transformcCs ||dS)aPredict each data slice with all possible estimators. Parameters ---------- X : array, shape (n_samples, nd_features, n_slices) The training input samples. For each data slice, a fitted estimator predicts each slice of the data independently. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_estimators). Returns ------- y_pred : array, shape (n_samples, n_estimators, n_slices) | (n_samples, n_estimators, n_slices, n_targets) The predicted values for each estimator. rWrUrVrrrrWszGeneralizingEstimator.predictcCs ||dS)aEstimate probabilistic estimates of each data slice with all possible estimators. Parameters ---------- X : array, shape (n_samples, nd_features, n_slices) The training input samples. For each data slice, a fitted estimator predicts a slice of the data. The feature dimension can be multidimensional e.g. ``X.shape = (n_samples, n_features_1, n_features_2, n_estimators)``. Returns ------- y_pred : array, shape (n_samples, n_estimators, n_slices, n_classes) The predicted values for each estimator. Notes ----- This requires ``base_estimator`` to have a ``predict_proba`` method. rXrUrVrrrrXsz#GeneralizingEstimator.predict_probacCs ||dS)a2Estimate distances of each data slice to all hyperplanes. Parameters ---------- X : array, shape (n_samples, nd_features, n_slices) The training input samples. Each estimator outputs the distance to its hyperplane, e.g.: ``[estimators[ii].decision_function(X[..., ii]) for ii in range(n_estimators)]``. The feature dimension can be multidimensional e.g. ``X.shape = (n_samples, n_features_1, n_features_2, n_estimators)``. Returns ------- y_pred : array, shape (n_samples, n_estimators, n_slices, n_classes * (n_classes-1) // 2) The predicted values for each estimator. Notes ----- This requires ``base_estimator`` to have a ``decision_function`` method. rYrUrVrrrrYsz'GeneralizingEstimator.decision_functionc st}|djjf}ttj|jddd\}}|jj t t |jdt j |d6|fddt||dt j dD}W5QRXtj|d d }|S) aScore each of the estimators on the tested dimensions. Parameters ---------- X : array, shape (n_samples, nd_features, n_slices) The input samples. For each data slice, the corresponding estimator scores the prediction, e.g.: ``[estimators[ii].score(X[..., ii], y) for ii in range(n_slices)]``. The feature dimension can be multidimensional e.g. ``X.shape = (n_samples, n_features_1, n_features_2, n_estimators)``. y : array, shape (n_samples,) | (n_samples, n_targets) The target values. Returns ------- score : array, shape (n_samples, n_estimators, n_slices) Score for each estimator / data slice couple. z Scoring %srFr#r%c 3s*|]"\}}j||VqdSr'ru)r)r*rHr.r/rrr0rrr1+s z.GeneralizingEstimator.score..rwrr2)rr5r"r9r _gl_scorerr7rrr]r rrr r:rP)rr=r0r_r&r>rr`rryrr`s,   zGeneralizingEstimator.score) r9rcrdrerrTrCrWrXrYr`rgrrr!rrts rtc Cs|jd|jd}}t|D]\}}|tjd|jdtd|jdf}|tj|||jddf}t||} | |} | jdkrt| ||| jdg} n,tj||| jddf t } t| | } |dkrt | |t |} | | dd|df<| |d|q| S)aTransform the dataset. This will apply each estimator to all slices of the data. Parameters ---------- X : array, shape (n_samples, nd_features, n_slices) The training input samples. For each data slice, a clone estimator is fitted independently. The feature dimension can be multidimensional e.g. X.shape = (n_samples, n_features_1, n_features_2, n_estimators) Returns ------- Xt : array, shape (n_samples, n_slices) The transformed values generated by each estimator. rrrrN.)r7rnZ transposer:Zr_r[riZreshaperZastypeint _gl_init_predrrk) r?r=rJr/rrn_iterrmrAZX_stackrCrpr7rSrrrrx5s("    rxcCsV|jd|jd}}|jdkr>t||||jdf|j}nt|||f|j}|S)zrutilsr r r r r r6rMror^rKrtrxr|rzr]rrrrs,     & '