{Kg dZddlZddlZddlZddlmZddlmZddlm Z m Z ddl m Z m Z ddlmZmZddlZddlZddlmZdd lmZd d lmZd d lmZmZmZmZmZm Z d d l!m"Z"m#Z#m$Z$m%Z%m&Z&d dl'm(Z(m)Z)m*Z*d dl+m,Z,m-Z-d dl.m/Z/m0Z0m1Z1m2Z2m3Z3d dl4m5Z5m6Z6m7Z7d dl8m9Z9m:Z:m;Z;d dlZ>d dl?m@Z@d dlAmBZBd dlCmDZDmEZEd dlFmGZGd dlHmIZImJZJd dlKmLZLmMZMmNZNmOZOd dlKmPZQd dlRmSZSmTZTmUZUddlmVZVddlWmXZXddlYmZZZdd l[m\Z\m]Z]dd!l^m_Z_m`Z`maZambZbmcZcmdZdmeZemfZfmgZgmhZhmiZimjZjmkZkdd"llmmZmmnZnmoZompZpdd#lqmrZrmsZsmtZtdaugd$Zvd%Zwd&Zxeiey'd(Zzd)Z{d*Z|d+Z}d,Z~d-Zd.Zd/Zd0Zd1Zd2Zd3Zd4Zdd6Zd7Zd8ZGd9d:Zd;Zd<Z dd=Z dd>Zd?Zd@ZdAZeiey'dBZeiey'dCZeiey'dDZeiey'dEZeiey'ddFZdGZeieyef'dHZdIZeidJZdKZeiey'dLZeiey'dMZdNZeiey'dOZeiey'dPZeidQZeidRZeidSZeiey'ddTZeiey'dUZeiey'dVZeiey'dWZdXZeidYZeidZZeid[Zd\Zeiey'd]Zeiey'd^Zeid_Zeidd`Zeiey'daZeiey'dbZeiey'dcZeiey'dddZeiey'deZeiey'dfZeiey'dgZei ddhZdiZddjZdkZeiey'dlZeiey'dmZeiey'dnZeiey'doZeiey'dpZeiey'ddqZeidrZeiey'dsZeidtZduZdvZeiey'dwZeiey'd5ejfdxZeidyZeiey'dzZeiey'd{Zeiey'd|Zeiey'd}Zeiey'd~Zeiey'dZeiey'dZeiey'dZeiey'dZdZdZe9fdZeiey'dZeiey'dZeiey'dZeiey'dZeiey'dZeiey'dZdZdZdZdZdZdZeiey'dZdZdZdZdZdZdZdZdZdZdZdZdZdZdZdZeiey'dZy)zQVarious utilities to check the compatibility of estimators with scikit-learn API.N) nullcontext)deepcopy)partialwraps) isfunction signature)IntegralReal)sparse)rankdata)config_context) ClusterMixinRegressorMixinclone is_classifieris_outlier_detector is_regressor) load_iris make_blobsmake_classificationmake_multilabel_classificationmake_regression)DataConversionWarningNotFittedErrorSkipTestWarning)SelectFromModel SelectKBest)LinearRegressionLogisticRegressionRANSACRegressorRidge SGDRegressor)accuracy_scoreadjusted_rand_scoref1_score) linear_kernelpairwise_distances rbf_kernel) ShuffleSplittrain_test_split) _safe_split) make_pipeline)StandardScalerscale)BaseRandomProjection)DecisionTreeClassifierDecisionTreeRegressor)_atol_for_type_convert_to_numpy get_namespace)yield_namespace_device_dtype_combinationsdevice)InvalidParameterErrorgenerate_invalid_param_valmake_constraint)shuffle) is_scalar_nan)Interval) _DEFAULT_TAGS _safe_tags) SkipTest_array_api_for_tests _get_argsassert_allcloseassert_allclose_dense_sparseassert_array_almost_equalassert_array_equalassert_array_lessassert_raise_messagecreate_memmap_backed_dataignore_warningsraisesset_random_state)_IS_PYPYSPARSE_ARRAY_PRESENT parse_version sp_version) _num_samplescheck_is_fittedhas_fit_parameter) PLSCanonical PLSRegressionCCAPLSSVDc#K|jj}t|}ttt t |drSttt|ds6ttttdttdtttd|dstt t"|t$vrt&|d s |dst(|drt*t,t/|d rt0t2t4t6tt6dt8|d rt;|D]}|yyw) N sample_weightpairwiseones)kindzerosTreadonly_memmap no_validation allow_nansparsifyarray_api_support) __class____name__rAcheck_no_attributes_set_in_initcheck_estimators_dtypescheck_fit_score_takes_yrU"check_sample_weights_pandas_series!check_sample_weights_not_an_arraycheck_sample_weights_listcheck_sample_weights_shape$check_sample_weights_not_overwrittenrcheck_sample_weights_invariance!check_estimators_fit_returns_selfcheck_complex_datacheck_dtype_object$check_estimators_empty_data_messagesCROSS_DECOMPOSITIONcheck_pipeline_consistencycheck_estimators_nan_infcheck_nonsquare_error!check_estimators_overwrite_paramshasattrcheck_sparsify_coefficientscheck_estimator_sparse_arraycheck_estimator_sparse_matrixcheck_estimators_pickle%check_estimator_get_tags_default_keys_yield_array_api_checks estimatornametagschecks b/home/alanp/www/video.onchill/myenv/lib/python3.12/site-packages/sklearn/utils/estimator_checks.py _yield_checksr[sA    ' 'D i D )) !! !!O400//''J, ,6 69G G9H H ++ 3T JJ      22 &&((  T/%:&& J## ++y*%)) && '' "! )4 @@ //  ,Y7EK8!sE*E,c#Kt|}ttttt |drt tttdttddt|dr tttt|dst|dst|d rt d |j#j%vrt&t(t*yw) N multioutputTr`float32raX_dtype multilabelrbmultioutput_only requires_fit class_weight)rA"check_classifier_data_not_an_arraycheck_classifiers_one_label*check_classifiers_one_label_sample_weightscheck_classifiers_classes'check_estimators_partial_fit_n_featurescheck_classifier_multioutputcheck_classifiers_trainr#check_classifiers_regression_target6check_classifiers_multilabel_representation_invariance2check_classifiers_multilabel_output_format_predict8check_classifiers_multilabel_output_format_predict_proba =1"0> =s )DD c#nKt|}tttdttddtt|drt t |ds |dstt|jj}|dk7rt|d rttyw) NTr`rrrrbrrXr)rAcheck_regressors_trainr!check_regressor_data_not_an_arrayrcheck_regressor_multioutput%check_regressors_no_decision_functionrrrfrgcheck_regressors_intrr) regressorrrs r_yield_regressor_checksrs i D ! ($ ?? ($ RR ++ 11 M)) //  .@)A## ##    ' 'D u}"" N'' 11sB3B5c#Kt|}|dstt|drtt tdt|ds t nt gd}|jj}||vr tyyw)Nrbpreserves_dtypeTr` statelesskey)Isomap KernelPCALocallyLinearEmbeddingRandomizedLassoLogisticRegressionCVBisectingKMeans) rA#check_transformer_data_not_an_arraycheck_transformer_general!check_transformer_preserve_dtypesrcheck_transformers_unfitted%check_transformers_unfitted_statelessrfrgcheck_transformer_n_iter) transformerrexternal_solverrs r_yield_transformer_checksrs k "D  11 ## // +T BB k{ 3))33O  ) )D ?"&&#sBB c#Kt|jj}|dvr#tt tdt t |ds tyyw)N)WardAgglomerationFeatureAgglomerationTr` transform)&check_clusterer_compute_labels_predictrfrgcheck_clusteringrrrzr) clustererrs r_yield_clustering_checksrsW 00    ' 'D @@&==55 9k *55 +sAAc#Kt|drtt|drtt|dr8tt tdt t |drttyw)N contamination fit_predictpredictTr`rr) rzcheck_outlier_contaminationcheck_outliers_fit_predictcheck_outliers_trainrrrArr)rs r_yield_outliers_checksr!sdy/*))y-(((y)$""*DAA00 i^ 4+ + 11sA6A8c#\KtD]\}}}tt|||yw)N)array_namespace dtype_namer8)r6rcheck_array_api_input)rrr8rs rrr6s: 3 4   !+!    5s*,c#K|jj}t|}d|dvr.tjdj ||dt y|dr*tjdj |t yt|D]}|t|rt|D]}|t|rt|D]}|t|drt|D]}|t|trt!|D]}|t#|rt%|D]}|t&|dst(t*t,t.t0t2t4t6t8t:|ds%t<t>t@|d rtB|d r tDyyw) N2darrayX_typesz8Can't test estimator {} which requires input of type {} _skip_testz2Explicit SKIP via _skip_test tag for estimator {}.rnon_deterministicrb requires_yrequires_positive_X)#rfrgrAwarningswarnformatrrrrrrrzr isinstancerrrr&check_parameters_default_constructible%check_methods_sample_order_invariancecheck_methods_subset_invariancecheck_fit2d_1samplecheck_fit2d_1featurecheck_get_params_invariancecheck_set_paramscheck_dict_unchangedcheck_dont_overwrite_parameterscheck_fit_idempotentcheck_fit_check_is_fittedcheck_n_features_in check_fit1dcheck_fit2d_predict1dcheck_requires_y_nonecheck_fit_non_negativers r_yield_all_checksrDs    ' 'D i DY' F M Md9o     L @ G G M   y) *Y-i8EK9I,Y7EK8y+&.y9EK:)\*-i8EK99%+I6EK7 00 # $33--   %%   ))  ##  !!##  ' ' !"$$#sGGc "t|r |jSt|tr|js|j jSdj |jjDcgc]\}}dj||c}}}dj|j j|St|dr6td5tjddt|cd d d Sy cc}}w#1swYy xYw) aCreate pytest ids for checks. When `obj` is an estimator, this returns the pprint version of the estimator (with `print_changed_only=True`). When `obj` is a function, the name of the function is returned with its keyword arguments. `_get_check_estimator_ids` is designed to be used as the `id` in `pytest.mark.parametrize` where `check_estimator(..., generate_only=True)` is yielding estimators and checks. Parameters ---------- obj : estimator or function Items generated by `check_estimator`. Returns ------- id : str or None See Also -------- check_estimator ,z{}={}z{}({})rT)print_changed_onlyz\sN)rrgrrkeywordsfuncjoinitemsrrzrresubstr)objkvkwstrings r_get_check_estimator_idsr$}s0#||#w||88$$ $88cll>P>P>RS>RdaW^^Aq1>RSTsxx00(;;sL! t 466%SX.5 4"T5 4s2C?  DDct|dg}t|r|dgdgfvr|t|tr|t }|St|t r|t }|St|tr|td}|S|td}|S|dgfvr^t|t r'|d td d fd td d fg }|S|d td d fd td d fg }|Sd|jd|}tj|tt!||}|S)z)Construct Estimator instance if possible._required_parametersrbase_estimatorr random_stater<)C estimatorsest1) max_depthr)est2)r+zCan't instantiate estimator z parameters )getattrlen issubclassr!rrr"rr#r r2r1rgrrrrB) Estimatorrequired_parametersrmsgs r_construct_instancer6s!)-CRH  K=3C2D"E E )_5%&6&89 D CI~6%eg. @ ?I7%l&BC : 7&&81&=> 6 5!l^$5 5)^4%!6QR!ST!6QR!ST  . !&!7!RS!TU!7!RS!TU  /y/A/A.BC124  MM# /3- K ct||\}}|s||fS|j|||jj|S)N)reason)marks)_should_be_skipped_or_markedparammarkxfail)rrpytestshould_be_markedr9s r_maybe_mark_xfailrAsK .wrappeds3 |5)<)<)E)E(Fb Q  r7)r;rrrrgr)rrshould_be_skippedrHrGr9s` @@r _maybe_skiprJsZ !=Y Nv  (25'(B$$J 5\  Nr7ct|tr|jjn |j}t |dxsi}||vrd||fSy)N _xfail_checksrT)Fz*placeholder reason that will never be used)rrrrgrA)rrrG xfail_checkss rr;r;sN)35'(B$$Ji_=CL\!\*--- >r7cddltdDr d}t|fd}jj d|t S)aPytest specific decorator for parametrizing estimator checks. The `id` of each check is set to be a pprint version of the estimator and the name of the check with its keyword arguments. This allows to use `pytest -k` to specify which tests to run:: pytest test_check_estimators.py -k check_estimators_fit_returns_self Parameters ---------- estimators : list of estimators instances Estimators to generated checks for. .. versionchanged:: 0.24 Passing a class was deprecated in version 0.23, and support for classes was removed in 0.24. Pass an instance instead. .. versionadded:: 0.24 Returns ------- decorator : `pytest.mark.parametrize` See Also -------- check_estimator : Check if estimator adheres to scikit-learn conventions. Examples -------- >>> from sklearn.utils.estimator_checks import parametrize_with_checks >>> from sklearn.linear_model import LogisticRegression >>> from sklearn.tree import DecisionTreeRegressor >>> @parametrize_with_checks([LogisticRegression(), ... DecisionTreeRegressor()]) ... def test_sklearn_compatible_estimator(estimator, check): ... check(estimator) rNc3<K|]}t|tywN)rtype).0ests r z*parametrize_with_checks..,s 7JS:c4 JuPassing a class was deprecated in version 0.23 and isn't supported anymore from 0.24.Please pass an instance instead.c3KD]B}t|j}t|D]}t||}t ||DywrP)rQrgrrrA)rrrr+r?s rchecks_generatorz1parametrize_with_checks..checks_generator4sJ#I ?++D*95t,' 5&AA6$sA A zestimator, check)ids)r?any TypeErrorr= parametrizer$)r+r5rXr?s` @rparametrize_with_checksr]s]P 7J 77 / nB ;; " ",.4L # r7Fc,ttr d}t|tjfd}|r|S|D]\} |y#t$r-}t j t|tYd}~Cd}~wwxYw)aCheck if estimator adheres to scikit-learn conventions. This function will run an extensive test-suite for input validation, shapes, etc, making sure that the estimator complies with `scikit-learn` conventions as detailed in :ref:`rolling_your_own_estimator`. Additional tests for classifiers, regressors, clustering or transformers will be run if the Estimator class inherits from the corresponding mixin from sklearn.base. Setting `generate_only=True` returns a generator that yields (estimator, check) tuples where the check can be called independently from each other, i.e. `check(estimator)`. This allows all checks to be run independently and report the checks that are failing. scikit-learn provides a pytest specific decorator, :func:`~sklearn.utils.estimator_checks.parametrize_with_checks`, making it easier to test multiple estimators. Parameters ---------- estimator : estimator object Estimator instance to check. .. versionadded:: 1.1 Passing a class was deprecated in version 0.23, and support for classes was removed in 0.24. generate_only : bool, default=False When `False`, checks are evaluated when `check_estimator` is called. When `True`, `check_estimator` returns a generator that yields (estimator, check) tuples. The check is run by calling `check(estimator)`. .. versionadded:: 0.22 Returns ------- checks_generator : generator Generator that yields (estimator, check) tuples. Returned when `generate_only=True`. See Also -------- parametrize_with_checks : Pytest specific decorator for parametrizing estimator checks. Examples -------- >>> from sklearn.utils.estimator_checks import check_estimator >>> from sklearn.linear_model import LogisticRegression >>> check_estimator(LogisticRegression(), generate_only=True) rVc3fKtD]}t|}t|f ywrP)rrJr)rrrs rrXz)check_estimator..checks_generators4&y1E 51EWUD11 12s.1N) rrQr[rgrBrrrr)r generate_onlyr5rXr exceptionrs` @rcheck_estimatorrb@sl)T" / n  ? # #D2 !!,. 5 ; ) / ; MM#i./ : : ;sA B&#BBc|t1tdddddd\}}tj|}||fatS)Nrr<@*) n_samples n_features n_informativebiasnoiser))REGRESSION_DATASETrr. fit_transform)rrs r_regression_datasetrosM!  1   * *1 -T r7c4|j}|jj}|dk(r|jdd|vr|dk7r|jdd|vr|j&|jt d|j|d vr|jd |d k(r|jd |d k(r|jd d|jjdk(r|jd d|dvr|jd|dk(r|jdd|vr|jdd|vr&|jt d|j d|vr|jdd|vr|jdd|vr#|jd s|jd!|d"k(r|jd#$|d%k(rd&|_|d'k(r|jd#(t|d)rt |jd|_ t|d*rd&|_ |d+k(r|jd,-|d.k(rd|_ t|tr|jd/t|tr|jd&0|d1vr|jd2|d3k(r|jd45d6d7g}||vrt|d8r|jd9:t|d;r|jd9<|d=k(r|jd>?|d@k(r(t t#dAk\rdBndC}|j|D|t$vr|jd&/|dEk(r|jdFG|dHk(r|jd&Iyy)JNTSNEr  perplexityn_iterrrtmax_iterrv) LinearSVR LinearSVCrfNMFiDictionaryLearning lasso_lars)rvtransform_algorithm MiniBatchNMFT)rvfresh_restarts) MLPClassifier MLPRegressordMiniBatchDictionaryLearning n_resampling)r n_estimators)r max_trialsr)rn_init)r batch_sizeMLP)r MeanShift?) bandwidth TruncatedSVDr< LassoLarsIC)noise_variance n_clustersn_best SelectFdr?alphaTheilSenRegressor) n_components)r!)HistGradientBoostingClassifierHistGradientBoostingRegressor)min_samples_leafDummyClassifier stratified)strategyRidgeCVRidgeClassifierCVcvr-rn_splits)r OneHotEncoderignore)handle_unknownQuantileRegressorz1.6.0highszinterior-point)solverSpectralEmbeddinggh㈵>) eigen_tolHDBSCAN min_samples)rrfrg set_paramsrvminrrrrzrrmax_subpopulationrr0rrRrQru)rparamsrloo_cvrs r_set_checking_parametersrs ! ! #F    ' 'D v~*6dfnA&V    )  #a1C1C*D E - -  " - 5=  # . ' '  ", O    ' '> 9  "T B 4 4  # . 0 0  ! ,!,#a1G1G*HIv+6A&vdooe&<+ { s+ ~"#  } C0y,'"9#7#7; y(#  {3' ""&) #)12 !,)[) q! RR a0    l3, -F 6gi6"y*%a( H5 ""&-*@@FVF+ ""!, ""t, y+r7c$eZdZdZdZddZdZy) _NotAnArrayzvAn object that is convertible to an array. Parameters ---------- data : array-like The data. c8tj||_yrP)rasarraydata)selfrs r__init__z_NotAnArray.__init__%sJJt$ r7Nc|jSrP)r)rdtypecopys r __array__z_NotAnArray.__array__(s yyr7cj|jdk(rytdj|j)Nmay_share_memoryTz%Don't want to call array_function {}!)rgr[r)rrtypesrErFs r__array_function__z_NotAnArray.__array_function__+s- ==. .?FFt}}UVVr7)NN)rgr __qualname____doc__rrrr7rrrs%Wr7rc8t|dd}t|dk(S)zReturns True if estimator accepts pairwise metric. Parameters ---------- estimator : object Estimator object to test. Returns ------- out : bool True if _pairwise is set to True and False otherwise. metricN precomputed)r0bool)rrs r_is_pairwise_metricr1s"Y$ /F -' ((r7c#K|jdk(sJd|jfdD]}||j|f|jd}|jj d|_|j j d|_d|fdD]\}|j|}|j j d|_|jj d|_|dz|f^yw) auGenerate sparse matrices or arrays with {32,64}bit indices of diverse format. Parameters ---------- X_csr: scipy.sparse.csr_matrix or scipy.sparse.csr_array Input in CSR format. Returns ------- out: iter(Matrices) or iter(Arrays) In format['dok', 'lil', 'dia', 'bsr', 'csr', 'csc', 'coo', 'coo_64', 'csc_64', 'csr_64'] csr)doklildiabsrcsccoorint64coo_64)rr_64N)rrasformatrowastypecolindicesindptr)X_csr sparse_formatX_coors r_generate_sparse_datarCs <<5   C U^^M:::D NN5 !E   )EI   )EI E/' NN= )II$$W- 88??7+e#Q&& (sC;C=c  t||}td\}}|j|d}t||}t ||}t |} |j ||} |j ||} | j||t| jD cic]#\} } t| tjs!| | %}} } t | }td5|j| | t| dj}d d d |jD]\} }t!|| }td5t|dj}d d d k(sJd | d |d |t#|t#| k(sJt%|| }|r&t'||| dt)|j*|j,|j,k(sJ|j*|j*k(rJd}|D]}t!| |d }||dk(r~|||}td5t!||| | }d d d t|t.sJtt.sJ|r&t1||z t)|j*ksJ||}td5t!||| }d d d td5tdj}d d d k(sJd |d|d|dt#t#| k(sJt%|| }|r&t'|||dt)|j*nBt3|dr6|j,|j,k(sJ|j*|j*k(sJ|dk(st3| ds| j5|}td5|j5|}t|dj}d d d |k(sJd|d|dt#t#| k(sJt%|| }|r$t'||dt)|j*\|j,|j,k(sJ|j*|j*k(rJy cc} } w#1swYxYw#1swYRxYw#1swYvxYw#1swYxYw#1swYxYw#1swYxYw)a_Check that the estimator can work consistently with the Array API By default, this just checks that the types and shapes of the arrays are consistent with calling the same estimator with numpy arrays. When check_values is True, it also checks that calling the estimator on the array_api Array gives the same results as ndarrays. rgr(Frr7T)array_api_dispatchrN'z,' attribute is in wrong namespace, expected z got )xpz not the sameratol)score score_samplesdecision_functionrpredict_log_proba predict_probarrz)' output is in wrong namespace, expected z, got .z# did not the return the same resultshaperinverse_transformz;'inverse_transform' output is in wrong namespace, expected z4inverse_transform did not the return the same result)rCrr_enforce_estimator_tags_Xrrrrvarsrrrndarrayrr5rgr0 array_devicer4rEr3rrfloatabsrzr) rrrr8r check_valuesrrrrSX_xpy_xprrarray_attributesest_xpinput_ns attribute est_xp_param attribute_nsest_xp_param_npmethods method_namemethodresult result_xp result_ns result_xp_npinverse_resultinvese_result_xpinverse_result_nsinvese_result_xp_nps rrrds ov 6B B /DAq %(A!.!4A!.!4A  C ::a: 'D ::a: 'DGGAqM&*#Y__%6%6zsE*UBJJ:WU %63ZF 4 0 4 &q)22 1 +002Yvs+ t 4(6q9BBL5x' u@ K. " ' L)\$-????+LR@  %}-#AGG,  ??o&;&;; ;;??o&;&;; ;;+30G k40 >  ' !Aq\F488GFK8tD 9fe, ,,i/ //6I-.1HHHH AYF488GFK8> 9t 4%i03<s)$@P@PQr7ct|} ddl}tjddgddgddgddgddgddgddgddgddgddgddgddgg }|j t ||d}|j gd}|j dgd z}t|d r|j |d} |j||| y#t$rtd j|wxYw#t$r tdwxYw)Nrr<r r-rFr r<r<r<r<r r r r r<r<r r rrr[zPEstimator {0} raises error if 'sample_weight' parameter is of type pandas.SerieszUpandas is not installed: not testing for input of type pandas.Series to class weight.) rpandasrarray DataFramerSeriesrArrr ImportErrorrB)rrrpdrrweightss rrkrkCs<n%I$  HHAAAAAAAAAAAA   LL2>1EEL R II: ;))QC"H% i%7 8 QU +A  MM!QgM 6 %%+VD\     ;   sB"C,0C$C))C,,Dczt|}tjddgddgddgddgddgddgddgddgddgddgddgddgg }tt ||}tgd}tdgdz}t |dr%t|j jd d}|j||| y) Nr<r r-rr%r&rrr') rrr)rrrArreshaper)rrrrrr.s rrlrlosn%I  F F F F F F F F F F F F  A -na@AA89A1#(#G)!34 r1- . MM!QgM.r7ct|}tjjd}d}t ||j |df}tj |dz}t||}dg|z}|j|||y)Nrr-rr') rrrrrrarangerr)rrrrndrhrrr[s rrmrms~n%I ))   "CI!.#++Iq>+2RSA )q A!)Q/AC)OM MM!QmM4r7ct|}tjddgddgddgddgddgddgddgddgddgddgddgddgddgddgddgddgg}tjgd}t||}|j ||tj t |tt5|j ||tj dt |zdddtt5|j ||tj t |dfdddy#1swYTxYw#1swYyxYw)Nr<r-r rr<r<r<r<r r r r r<r<r<r<r r r r r') rrr)rrr]r1rMr)rrrrrs rrnrnsCn%I  F F F F F F F F F F F F F F F F!  A( ABA!)Q/A MM!Qbggc!foM6   a"''!c!f**= >    a"''3q61+*> ?      s34D:>3E:EEct|}t|}t|dt|dtjddgddgddgddgddgddgddgddgddgddgddgddgddgddgddgddggtj}tjgdt }|d k(r*|}|}tj t| } d |d } n|d k(r~tj||dzg}tj|d|z g}tj t|dz } d| t|dt||| d\}}} d |d} ntt||}t||}|j||d|j||| dD]A} t|| st|| |} t|| |} t!| | | Cy)Nrr(r<r-r rrr7r])rzFor z; sample_weight=None is not equivalent to sample_weight=onesr_z?, a zero sample_weight is not equivalent to removing the sample)rr[)rrrrr)rrNrr)float64intr]r1vstackhstackr=rrrrzr0rF)rrr^ estimator1 estimator2X1y1X2y2sw2rrX_pred1X_pred2s rrprps ~&J~&JZa0Za0  F F F F F F F F F F F F F F F F! $jj' B* B# NB v~  ggCG$4&S T  YYBF| $ YYAF| $ggCGaK(CGIb"c: B4&W X  ":r 2B ":r 2BNN24N0NN23N/P >6 *1gj&1"5G1gj&1"5G ('7 K Qr7ct|}t|dtjddgddgddgddgddgddgddgddgddgddgddgddgddgddgddgddggtj}tjgdt }t ||}tj|jd}d |d<|j}|j||| |d }t||| y) Nrr(r<r-r rr9r7g$@r'z8 overwrote the original `sample_weight` given during fitr:) rrNrr)r;r<rr]rrrrE)rrrrrsample_weight_originalsample_weight_fitrs rroros n%IYQ/  F F F F F F F F F F F F F F F F! $jj' A* AMA!)Q/AWWQWWQZ0 $1.335 MM!Q&7M8NOG%'=wOr7c tjjd}t||j d}|j t }t|}|dddfdzj t}t|}t||}|j||t|dr|j|t|dr|j|tt dd 5|j||j t dddd |d vr6d d i|d<d}tt"|5|j||dddy|j||y#1swYYxYw#1swYyxYw)Nr)r rrrrrzUnknown label typeTrrstringrfoobarrrz#argument must be .* string.* numberr)rrrrrrobjectrAr<rrrrzrrrM Exceptionr[)rrrrrrrr5s rrsrs-sF ))   "C!.#++8+2LMA A n %D 1a41S!An%I!)Q/A MM!Qy)$!y+&A  !5 E a&)* FtI&%.$4 IS ) MM!Q * )  a! F E* )s0&E.?E:.E7:Fctjjd}|jdd|jdzz}|j dd}|j ddd dz}t |}t|d ttd 5|j||dddy#1swYyxYw) Nrgrry?r0r<rr lowhighrr(zComplex data not supportedrQ) rrrrr1randintrrNrMrr)rrrrrrs rrrrrRs ))   #C  rCKKRK$888A "aA  +b0An%IYQ/  "> ? a @ ? ?s B66B?c|dvrytjjd}|dvrd|jdz}nd|jdz}t ||}|dddfj t }t|}t||}t|drd |_ t|d rd |_ t|d rd |_ t|d |j||d D]T}t||s|jj!}t#||||j|k(rLJd |zy)N)SpectralCoclusteringr)r!r-rfr-rr rr<rrrrrrz$Estimator changes __dict__ during %s)rrrrrrr<rrrzrrrrNr__dict__rr0)rrr5rrrr dict_befores rr r `s:  '' ))   "C ""   ) )   ) )!.!4A !Q$sAn%I!)Q/Ay.)!" y,' y(# Y" MM!QP 9f %#,,113K &GIv &q )%%4 6? 4 Qr7cL|jdxs|jd S)N_)rr)attrs r_is_public_parameterrbs"$: c(: ;;r7c>t|jdryt|}tjj d}d|j dz}t||}|dddfjt}t||}t|drd|_ t|drd|_ t|d|jj}|j!|||j}|j#Dcgc]}t%|s|} }| Dcgc]}||j#vs|} }| rJd d j'| z| Dcgc]}||||ur|} }| rJd d j'| zycc}wcc}wcc}w) Ndeprecated_originalrr-r[rrr<rzEstimator adds public attribute(s) during the fit method. Estimators are only allowed to add private attributes either started with _ or ended with _ but %s added, zEstimator changes public attribute(s) during the fit method. Estimators are only allowed to change attributes started or ended with _, but %s changed)rzrrrrrrrrr<rrrrNr]rrrrbr) rrrr5rrdict_before_fitdict_after_fitrpublic_keys_after_fitattrs_added_by_fitattrs_changed_by_fits rr r s~&&(=>n%I ))   "C CKKWK %%A!.!4A !Q$sA!)Q/Ay.)!" y,' Y"((--/O MM!Q''N&**,,0DS0I, -,?;O;O;Q0Q, " "&+=!>  ? !)(C C s(; ; ( $  "67  8 ##1sFF"F:FFc tjjd}d|jdz}t ||}|dddfj t }t|}t||}t|drd|_ t|drd|_ t|d|j||dD].}t||sttd t!|||d0y) Nrr-r[rrr<rr\zReshape your data)rrrrrrr<rrrzrrrNrrJrr0)rrr5rrrrs rrrs ))   "C CKKWK %%A!.!4A !Q$sAn%I!)Q/Ay.)!" y,' Y" MM!QP 9f % /F1KQqT Qr7c ||}|jd}|Dcgc]}||jd|}}t|tr|d}t t d|}t j|r-|j}|Dcgc]}|j}}tj|tj|fScc}wcc}w)Nr<rc |dS)Nrr)xs rz#_apply_on_subsets..sQqTr7) rr1rtuplelistmapr issparsetoarrayrravel)rr result_fullribatchresult_by_batchrns r_apply_on_subsetsrysq'KJGHIqetEMM!Z89qOI+u%!!n s>?CD {#!))+ 0?@1199;@ 88K "((?"; ;;JAs CCctjjd}d|jdz}t ||}|dddfj t }t|}t||}t|drd|_ t|drd|_ t|d|j||dD]J}d j|| }t||s#tt!|||\}} t#|| d | Ly) Nrr-r[rrr<rrrrrrz={method} of {name} is not invariant when applied to a subset.rrHz>rr)rrrrrrr<rrrzrrrNrrryr0rE) rrr5rrrrr5rvrxs rrrs ))   "C CKKWK %%A!.!4A !Q$sAn%I!)Q/Ay.)!" y,' Y" MM!QOVVW  9f %+< 6*A, (K KtS Qr7c tjjd}d|jdz}t ||}|dddfj tj }t|drd||dk(<t|}t||}t|d rd|_ t|d rd|_ t|d|j||tjj|j d}d D]W}d j#|| }t||s#t%t'||||t'||||d|Yy)Nrr-r[rrrr<r rrr{zY{method} of {name} is not invariant when applied to a datasetwith different sample order.r|& .>r~)rrrrrrrrArrrzrrrNr permutationrrrFr0) rrr5rrridxrr5s rrrsA ))   "C CKKWK %%A!.!4A !Q$rxx A.m4!q& n%I!)Q/Ay.)!" y,' Y" MM!Q ))   +C + &T& * 9f % (* 6*1-c2* 6*1S62  r7c$tjjd}d|jdz}t ||}|dddfj t }t|}t||}t|drd|_ t|drd|_ t|d|dk(r|jd |d k(r|jd gd}tt|d5|j!||dddy#1swYyxYw)Nrr-)r<rrrr<rOPTICSrrrqrrr)z1 samplez n_samples = 1z n_samples=1z one samplez1 classz one classTrL)rrrrrrr<rrrzrrrNrrMrrrrr5rrrmsgss rrrAs ))   "C CKKWK %%A!.!4A !Q$sAn%I!)Q/Ay.)!" y,' Y" x- v~, D  $ 6 a 7 6 6s *DDctjjd}d|jdz}t ||}|dddfj t }t|}t||}t|drd|_ t|drd|_ |dk(rd|_ |d k(rd |_ t||}t|dgd }tt |d 5|j#||dddy#1swYyxYw)Nrr-)rr<rrr<rRandomizedLogisticRegressionr!r)z1 feature\(s\)zn_features = 1z n_features=1TrL)rrrrrrr<rrrzrrsample_fractionresidual_thresholdrNrMrrrs rrrjs ))   "C CKKWK %%A!.!4A !Q$sAn%I!)Q/Ay.)!" y,'  --$% !   '* $!)Q/AY" @D  $ 6 a 7 6 6s C<<Dctjjd}d|jdz}|j t }t |}t||}t|drd|_ t|drd|_ t|dtt5|j||dddy#1swYyxYw)Nrr-rfrrr<r)rrrrrr<rrrzrrrNrMrr)rrr5rrrs rrrs ))   "C CKKbK ""A  An%I!)Q/Ay.)!" y,' Y"   a   s "B>>Cctdgdgdgddd\}}tj|}t||}|rt ||g\}}t ||||yNr3rrrr<r<r<rr 皙?rhcentersr)ri cluster_std)rr.rnrrK_check_transformer)rrrarrs rrrsh I&  DAq &&q)A!+q1A(!Q01t[!Q/r7cBtdgdgdgddd\}}tj|}t||}t |}t t j |}t||||t|||j|jyr) rr.rnrrrrrtolist)rrrrthis_Xthis_ys rrrs I&  DAq &&q)A!+q1A ^F A 'Ft[&&9t[!((*ahhjAr7ct\}}t|}tttfd|d5|j |dddy#1swYyxYw)NzThe unfitted transformer z\ does not raise an error when transform is called. Perhaps use check_is_fitted in transform.r:)rorrMAttributeErrorrr)rrrrs rrrs\  DAq $K  $ &!, ,  a    s AActjjd}|jd}t ||}t |}|j |}|jd|jdk(sJy)ztCheck that using transform without prior fitting doesn't raise a NotFittedError for stateless transformers. rrfrrN)rrrrrrrr)rrrrX_transs rrrsm ))   "C  !A!+q1A $K##A&G == qwwqz )) )r7cvtj|j\}}t|}t ||t vritj tj|tj|f}|ddddfxxdzcc<t|trt|}n|}|j||t|}|j||} t| tr| D]} | jd|k(rJn| jd|k(sJt|dr|t vr&|j||} |j||} n$|j|} |j||} t|dr|dz} t| t| trKt| tr;t!| | | D]*\} }}t#| |d d |z t#| |d d |z ,nDt#| | d |zd t#| | d d |z t%| |k(sJt%| |k(sJt|drjt|ds\|j&dk(rL|jddkDr9t)t*d|d5|j|ddddfdddyyyyyy#1swYyxYw)Nr r<rrrrr is non deterministic{Gz?z9fit_transform and transform outcomes not consistent in %sr~z7consecutive fit_transform outcomes not consistent in %srrrzThe transformer zr does not raise an error when the number of features in transform is different from the number of features in fit.r:r0)rrrrrNruc_rrrrnrprzrrArBziprFrSndimrMr)rtransformer_origrrrhriry_transformer_cloneX_predx_predrGX_pred3r5x_pred2x_pred3s rrrsJJqM//Iz()K[! "" UU2::a="**Q-/ 0 3Q36 a a %RB OOArk*  , ,Q" , 5F&% F<<?i/ //||A)+++{K( & &!++Ar2G!//R/8G!++A.G!//R/8G &,? @00C3-  fe $GU)C,/,I(,W!" -U!" -J )S   )Q   (I5 55(I5 55 Aw { <!  Q&tf-55%%a3B3i0= _)js J//J8ct|dr|dz}t|tdgdgdgddd \}}t||t }t |}t ||}t|t|}|j|||j||d d g}|D]<}t||d} | t||} | ||} | ||} t| | >y)Nrrrr3rrrr rrkernelrrn) rArBrrr)rrrNr-rr0rF) rrr5rrrpipelinefuncs func_namer func_pipeliner result_pipes rrvrv;s.&9:,,sm I&  DAq ".!JGAn%I!)Q/AYY'H MM!Q LLA o &E y)T2  #Hi8M!QZF'1-K ( = r7c 0tjjd}d}|j|df}t ||}tj |dz}t |}t||}t|gd}|D]}t||d} | | ||t| jjD cgc]} | j} } | ddk(r| dd} | ddvrkJd |d t|jd | d ycc} w) Nrr3r-r)rr partial_fitrrnrr<)rYz.Expected y or Y as second argument for method  of z. Got arguments: r)rrrrrr4rrrNr0r parametersvaluesrrQrg) rrr5rhrrrrrrprEs rrjrj\s ))   "CI )Q (A!.!4A )q An%I!)Q/AY KE y)T2   AJ$-dO$>$>$E$E$GH$GqAFF$GDHAw& ABx7j( d9o66> (IsDctjjd}d|jdj tj z}t ||}|j tj}|j tj}|j tj}tjddgdztj}t||}gd }||||fD]Q} t|} t| d| j| ||D]!} t| | st!| | | #Sy) Nrr-rrr<r rr9r\)rrrrrrrr;rrr)rrrNrrzr0) rrr5 X_train_32 X_train_64X_train_int_64X_train_int_32rrX_trainrrs rririzs  ))   "CS[[g[.55bjjAAJ*>:FJ""2::.J&&rxx0N&&rxx0N !Q"BHH-A!.!4ALG NNK.) A& gq!Fy&)* 6*73 Lr7c tdgdgdgdd\}}tj|}t||}t |dD]}|j |}t |}t||j||}|j||j|}t||gd d gD]P\} } t| tr| d} | j|k(r+J|d | d |jd | jdy)Nr3rrrr)rhrr)rrrrnrz (method=z3) does not preserve dtype. Original/Expected dtype=z , got dtype=r)rr.rnrrArrrNrrrrrprrg) rrrrrX_castrX_trans1X_trans2Xtrs rrrs# I&  DAq &&q)A!"2A6A,2CD%,- %,,VQ7??61-77?x2_k4RSJB"e$U88u$ & &*++0>>*:,rxxjPQS $TEr7ct|}t|dtjdj dd}d|d}t t |5|j|gdddtjdj dd}t|tjgd}d }t t | 5|j||dddy#1swY~xYw#1swYyxYw) Nr<rr-The estimator zc does not raise a ValueError when an empty data is used to train. Perhaps use check_array in train.r:r&) r<rr<rr<rr<rr<rr<rzG0 feature\(s\) \(shape=\(\d*, 0\)\) while a minimum of \d* is required.rQ) rrNremptyr1rMrrrr))rreX_zero_samplesrX_zero_featuresrr5s rrtrts nAQXXa[((A.N I I   G , nb! -hhqk))"a0O "!RXX.R%STA WC  # & oq! ' & - , ' &sC$C0$C-0C9ctjjd}t||j d}|j d}tj |d<|j d}tj |d<tjd}d|ddt||}d|d}d|d }d|d } ||fD]} tt 5t|} t| d ttd dg|5| j| |ddd| j||t!| dr.ttd dg|5| j#| dddt!| dr.ttd dg| 5| j%| ddddddy#1swYxYw#1swYZxYw#1swY,xYw#1swYxYw)Nrrr-rrPrrrz& doesn't check for NaN and inf in fit.z* doesn't check for NaN and inf in predict.z, doesn't check for NaN and inf in transform.rr<rNaNrrr)rrrrrrrr]rrLrrrNrMrrrzrr) rrr5X_train_finite X_train_nan X_train_infrerror_string_fiterror_string_predicterror_string_transformrrs rrwrws ))   "C.  1N++7++KK++7++KK  A AbqE!.!4A#D6)OP'v-WX TFFG - m 4n-I Y * 5%.BRS gq)T MM.! ,y), %.0 %%g. y+. %.2 ''0 )5 4. TS)5 4sT,GF-:GF9 (GGG-F6 2G9G >GG GG ctdd\}}t|}ttd|d5|j ||dddy#1swYyxYw)z8Test that error is thrown when non-square data provided.rfr)rhrizThe pairwise estimator z+ does not raise an error on non-square datar:N)rrrMrrrrrrrs rrxrxsV r 2DAqn%I %dV+V W   a    s A  Ac>gd}tdgdgdgddd\}}t||t }t|}|d r_tj j d }|j|jd d }tj|jd|<t|} t| |}t| | j|||r t| } nbt!j"| } | j$} | j'drd| vs| j)dsd| vsJt!j*| } t-} |D]$}t/| |st1| ||| |<&| D]#}t1| ||}t3| ||%y)z'Test that we can pickle all estimators.r\r3rrrr rrrrcrgrF)replacer0rrrs_sklearn_versionN)rrr)rArrrchoicerrr1rrrNrrKpickledumpsrrrloadsdictrzr0rF)rrra check_methodsrrrrmaskrunpickled_estimatorpickled_estimatorrrrunpickled_results rr~r~ sSM I&  DAq ".!JGA n %D Kii##B'zz!&&"ez4 ff " dn%I!)Q/AY MM!Q7 B#LL3**  ! !* - { "k&:&::&F'*;; ;;$ll+<= VF 9f %7WY7:F6N ?7#6?B$VF^5EFr7ct|dsyt|}tdd\}}t||}t ||} t |r*t j|}|j|||n|j||ttd|d5|j|dddd f|dddy#t$rYywxYw#1swYyxYw) Nr2r<rhr)classesrzZ does not raise an error when the number of features changes between calls to partial_fit.r:r0) rzrrrrrruniquerNotImplementedErrorrMr)rrrrrrs rrr@s >= 1n%I  3DAq!.!4A!.!4A  #iilG  ! !!Q ! 8  ! !!Q ' TF#G G  a3B3i+     sACC CCCc d\}}}t|}t|}td|||\}}|j|||j |}|j ||fk(s#Jdj ||f|j |jjdk(sJt|dr|j|} t| tjsJ| j ||fk(s#Jdj ||f| j | dkDjt} |j | } t#| |t|d r|j%|} t| t&r|d st)|D]z} | | j |d fk(s&Jd j |d f| | j t#tj*| | d jt|dd| f|nf|d sa| j ||fk(s#Jd j ||f| j t#| j-jt|t|drlt|d r_t)|D]P} |j%|dd| f}|j|}t#t/|t/|dd| fRyyy)N)rgrr-rg)r)rhn_labels n_classesSThe shape of the prediction for multioutput data is incorrect. Expected {}, got {}.irzaThe shape of the decision function output for multioutput data is incorrect. Expected {}, got {}.rr poor_scorer zTThe shape of the probability for multioutput data is incorrect. Expected {}, got {}.r<axis)rArrrrrrrr^rzrrrrrr<classes_rHrrqrangeargmaxroundr )rrrhrrrrry_preddecisiondec_preddec_expy_probry_proba y_decisions rrr]s%-"Ix i Di I )9x9 DAqMM!Q   q !F <9%ay9a.877=v"Aq 88 #IIfQia077? ?? ? ;;288G,bhhw.?@ @@ @ " "7 +F IIf%Mryyq!1=3Dq3HI  w && &y,'Y 5 A~r!2222(- . -s J&&J0c&tdd\}}t|}t|t|dr_|j |j |}|j d|j |j |}t||yy)z2Check that predict is invariant of compute_labels.rfrrcompute_labelsF)rN)rrrNrzrrrrH)rr rrrrFrGs rrrs  3DAqn%IYy*+--"**1-E2--"**1-7G, ,r7cd}d}tjjd}|jd}|jd}tjd}t t 5t|}ttdd | 5} |j||ddd jr dddyt|j||| dddy#1swYExYw#1swYyxYw) Nz6Classifier can't train when only one class is present.z8Classifier can't predict when only one class is present.rrrrrclassTrr:)rrrrr]rLrrrMrrraised_and_matchedrHr) rclassifier_origrrr5rX_testrrcms rrrsOU ))   "Ckkwk'G [[g[ &F  A - 0?+  g>N  NN7A &   1 0 :--f5qBVW 1 0   1 0s*-C/ C#C/=C/#C, (C//C8c^|d}|d}tjjd}|jd}|jd}tjddz}|j }t |} t| drd |g} ttf|} } nd } ttfd } } t| | d | 5} | j|||| jr d d d y t| j|tj d|d d d y #1swYy xYw)zCheck that classifiers accepting sample_weight fit or throws a ValueError with an explicit message if the problem is reduced to one class. zu failed when fitted on one label after sample_weight trimming. Error message is not explicit, it should have 'class'.z; prediction results should only output the remaining class.r)rrrrr r[z\bclass(es)?\bz\bsample_weight\bNTrr'r:)rrrrr4rrrUAssertionErrorrr[rMrrrHrr])rr error_fit error_predictr5rrrr[rrerr_typerrs rrr s &; ;fWXM ))   "Ckkxk(G [[h[ 'F " AFFHM'J_5"M2+Z8)'$& 3T' g F"w?  G F    v & ]  G F Fs?"D#*0D##D,c tdd\}}|j|}t||d\}}tj |}||dk7}||dk7}|dvr&||j z}||j z}|rt ||||g\}}}}||fg}t|} | ds|j||f|D]\} } tj| } t| } | j\}}t|}t|| } t|| } t!|| d s3t#t$d |d  5|j'| | dd ddd|j'| | |j'| j)| j)t+|dsJ|j-| }|j|fk(sJ| dst/| |dkDsJd}d}| d s| drLt#t$|j1|d 5|j-| j3d ddddnEt#t$|j1|d 5|j-| j4dddt+|drC |j7| }| dk(r]| ds|j|fk(sJ|j|dfk(sJ|j9dkDjt:}t=||n4|j|| fk(sJt=tj>|d|| d s| drLt#t$|j1|d 5|j7| j3d ddddnEt#t$|j1|d 5|j7| j4dddt+|ds |jC| }|j|| fk(sJt=tj>|d|tEtjF|dtjH|| d s| drLt#t$|j1|d 5|jC| j3d ddddnEt#t$|j1|d 5|jC| j4dddt+|dsB|jK| }tM|tjN|ddt=tjP|tjP|y#1swYxYw#1swYxYw#1swY xYw#1swYxYw#1swYxYw#t@$rYwxYw#1swYxYw#1swYxYw)N,rrrr(r ) BernoulliNB MultinomialNB ComplementNB CategoricalNBrrbThe classifier does not raise an error when incorrect/malformed input data for fit is passed. The number of training examples is not the same as the number of labels. Perhaps use check_X_y in fit.r:r0rrg(\?zuThe classifier {} does not raise an error when shape of X in {} is not equal to (n_test_samples, n_training_samples)z|The classifier {} does not raise an error when the number of features in {} is different from the number of features in fit.r\rr<rrrrrr)r))rrr=r.rnrrKrAappendrrr1rrrrrNrMrrrrzrr$rr1Trrur<rHrrrrGsumr]rrElogargsort)rrrarX_my_my_bX_bproblemsrrrrrrhrirr msg_pairwiser5rrr y_log_probs rrr; sPCa8HC **W CsCa0HC   ( ( -C cQh-C cQh-C PP swwy swwy6S#s7KLS#sc |H o &D  c #1))A,L ! :?+ %j! 4 %j! 4$O$%dV,<<q!CR&) q!qxxz188:.z:...##A&|| |+++L!!!V,t3 33 G    O$J(//i@&&qyyQ'78  J 40KL&&qss+M :2 3 %77:> 23'~~)==='~~)Q??? ( 01 4<>i-CCCC&ryy'BFKO,J'#&$0$7$7>Q$R'882q9IJ  $&$'JJt5H$I'88=  : /--a0F<P Q( #" , 3 3D/ J#002q1AB   " # 4 A#005  z#67'99!<  BFF6NADI"2::j#92::f;MNaH ML(  '   sU1"U>$V CV2("V *V24V%V2 "WW1U; >V V V" V2%V/ *V22 V?>V?W W c||kr|}|dz}n|}|dz}tj|}d}ttj|||dk(sJ|y)Nr<zThe number of predicted outliers is not equal to the expected number of outliers and this difference is not explained by the number of ties in the decision_function values)rsortr1r) num_outliersexpected_outliersrstartendsorted_decisionr5s rcheck_outlier_corruptionr8 so''!#!Q ggh'O 9 ryys34 5 :?C? :r7cd}t|d\}}t|d}|r t|}|j\}}t |}t ||j ||j |j|j|}|j|fk(sJ|jjdk(sJttj|tjddg|j|} |j!|} | | fD]8} | jtjd k(sJ| j|fk(r8Jt#t$5|j|j&ddd| dk\j)t*} d| | dk(<t| |t#t$5|j|j&ddd| |j,z } t/| | t#t$5|j!|j&dddt1|d rt1|d sxd }||z }|j3| |j ||j|}tj4|dk7}||k7r|j|} t7||| yyyy#1swYIxYw#1swYxYw#1swYxYw)Nrrrrr(rr0r<rrnoveltyr3r)rr=rKrrrNrrrrr^rHrrr)rrrMrr&rr<offset_rErzrr'r8)rrrarhrr`rirrrscoresoutputry_decr4rr3s rrr sI   :DAq"A %a (GGIzn%IYMM! MM!((*   q !F <rightneitherz:contamination constraint should be an interval in (0, 0.5]) rzrArZrr?rrQr leftrCclosed)rrcontamination_constraintsc constraints rrr$ s >#; <nCCC . E Eo V 1JK1JA 1h'1JK L R  0 j( +4'OOs*$$+__q(J,=,=AU,U  L L  LW0 LsC c tddddddd\}}t|}|dd |dd }}|d d}|j}t|}t |} t | | j ||j|} | j ||j|} | j ||j|} t| | t| | | j| jk(sJ| j| jk(sJt| t| k(sJt| t| k(sJy) Nrr rr-rTrrhrirrlengthallow_unlabeledr)P) rr/rrqrrNrrrHrrQ) rrrrry_trainry_train_list_of_listsy_train_list_of_arraysrry_pred_list_of_listsy_pred_list_of_arrayss rrr? sP ) DAq aA"vq"vWG rsVF#NN,!']'JZ ^^GW - 5 5f =F%>>'3HIQQ'NN74JKSSv45v34 <<066 66 6 <</55 55 5 <4 56 66 6 <4 45 55 5r7c xt|}t|d\}}}t|d|dddd\}}t|}|d| || d} }|d| || d} } |j || d } t || d} | t |d | d | | }t|tjsJ|d t|d |j| jk(s$J|d|jd| jd|j| jk(s$J|d|jd| jdy)zeCheck the output of the `predict` method for classifiers supporting multilabel-indicator targets.rrr r-rTrrKNr does not have a  method.z2.predict is expected to output a NumPy array. Got instead.z(.predict outputs a NumPy array of shape instead of rz>.predict does not output the same dtype than the targets. Got ) rrNrr/rr0rBrrrrQrr)rrrrh test_size n_outputsrrrrrOy_testresponse_method_namepredict_methodrs rrrh s'JZ &0#Iy) ) DAq aA )na nVG )na nVGNN7G$$Z)=tDN$01E0FhOPP F #F fbjj ) &B <.  # ) <<6<< ' &8Gll^1 & ' <<6<< ' &||nLa 9 'r7c  t|}t|d\}}}t|d|dddd\}}t|}|d| || d} }|d| } |j || d } t || d} | t |d | d | | } t| trt| |k(sJd |d t| d|d| D]}|j|dfk(sJd |d|jd|dfd|jjdk(sJd |d|jdd |d}t|jdd|yt| tj r| j||fk(sJd |d| jd||fd| jjdk(sJd |d| jdd |d}t#d| |t#| d|yt%dt'| d|d)zkCheck the output of the `predict_proba` method for classifiers supporting multilabel-indicator targets.rUr r-rTrrKNrrWrXzWhen zn.predict_proba returns a list, the list should be of length n_outputs and contain NumPy arrays. Got length of rZrzx.predict_proba returns a list, this list should contain NumPy arrays of shape (n_samples, 2). Got NumPy arrays of shape fzW.predict_proba returns a list, it should contain NumPy arrays with floating dtype. Got rYz.predict_proba returns a list, each NumPy array should contain probabilities for each class and thus each row should sum to 1 (or close to 1 due to numerical errors).r<rr:zX.predict_proba returns a NumPy array, the expected shape is (n_samples, n_outputs). Got zN.predict_proba returns a NumPy array, the expected data type is floating. Got z.predict_proba returns a NumPy array, this array is expected to provide probabilities of the positive class and should therefore contain values between 0 and 1.zUnknown returned type z by z4.predict_proba. A list or a Numpy array is expected.)rrNrr/rr0rBrrqr1rrr^rEr'rrrIrrQ)rrrrhr[r\rrrrrOr^predict_proba_methodrrrs rrr s'JZ &0#Iy) ) DAq aA )na nVG )nGNN7G$*":/CTJ#$01E0FhOPP !& )F &$6{i' D6f+l9+Q 8 ' D::)Q/ v))- Lq>"!% / ::??c) vA::,i) )v%%  DHH!H,a A'( FBJJ '|| 955 D6==C\\NIy12! 5 5 ||  C' D6339<<.  K ' D6C C  !VW5&!W5$T&\N$tf=B B  r7c 2t|}t|d\}}}t|d|dddd\}}t|}|d| || d} }|d| } |j || d } t || d} | t |d | d | | } t| tjsJ|d t| d | j||fk(sJ|d| jd||fd| jjdk(sJ|d| jd y)zoCheck the output of the `decision_function` method for classifiers supporting multilabel-indicator targets.rUr r-rTrrKNrrWrXz<.decision_function is expected to output a NumPy array. Got rYz].decision_function is expected to provide a NumPy array of shape (n_samples, n_outputs). Got rZrraz?.decision_function is expected to output a floating dtype. Got )rrNrr/rr0rBrrrrQrrr^)rrrrhr[r\rrrrrOr^decision_function_methodrs rrr s~'JZ &0#Iy) ) DAq aA )na nVG )nGNN7G$.&z3GN'$01E0FhOPP %f -F fbjj ) &V ~Y ( ) <+2RSA )q A!.!4An%IY MM!Q   q !FY   -h(=>h7 a1bjj=)* .!!!$I 5 QcSQ9 C  1vz3z 7:= >  >FLLNIOO$56 . - s AF2F> 2F;ctj|}t|}|dk(r||jkD}t ||j |||j |}t|dr^|j|}t|tjsJt|dk(r|jdkDjt}|j|} t!| |d|ddj#t%t&| ddj#t%t&|d  nt)|d d d k(rtj*|d jt} |j| } t!| |d|ddj#t%t&| ddj#t%t&|d  |dk7r2t!tj|tj|t!||jd|ddj#t%t&|ddj#t%t&|jd  y)Nrrr rz0decision_function does not match classifier for z : expected 'rez', got 'rr:decision_function_shapeovrr<rr z"Unexpected classes_ attribute for )rrrmeanrNrrrzrrrr1rurr<rrHrrrrr0r) rrrrrrrrrr decision_yy_exps rcheck_classifiers_predictionsrzo siilG'J } LZ NN1a    "Fz.///2(BJJ/// w<1  (1,44S9H ))(3G  IIc#w/0IIc#v./ Z!:E Be K8!4;;C@J'' 3E  IIc#uo.IIc#v./  ~ 299Q<6):;  IIc#w' ( IIc#z223 4   r7c|dvr|S|S)N)LabelPropagationLabelSpreadingSelfTrainingClassifierr)rry_namess r _choose_check_classifiers_labelsr s! S S r7cftddd\}}t||d\}}tj|}||dk7}||dk7}t ||}t ||}gd}d d g}t j ||}t j ||} ||| fg} t|d s| j|||f| D]9\} } } | | jd fD]}t|| |}t| |||;ddg}t j ||} t||| }t||||y)Nr3rr)rhr)rrr(r )onetwothreerrrrOr0r<) rr=r.rnrrtakerAr%rrrz)rr X_multiclass y_multiclassX_binaryy_binarylabels_multiclass labels_binaryy_names_multiclassy_names_binaryr.rrr y_names_irs rrr sP!+1#"L,")|RS!TL,!#11,?LLA-.HLA-.H,_lKL((CH/ENM!2LAWW]H5N8^45H o= 9|5GHI! 1g!7>>##67I1$9EB )!R G8" GMWW]H5N/hOH!(HdOLr7c t\}}t||dd}tjj d}|j d|j d}t||}tjj d}t|}t|}t|t||tvrDtj|d|z|j dt|zg}|j}n|}|j|||j|} |j||j!t"|j|} t%| | d|y)Nrrr-rr rr~)rorrrrrXrrrrNrur=r1r&rrrrrE) rregressor_origrr`r5r regressor_1 regressor_2rpred1r s rrr s+  DAq!.!CR&9A ))   "C AAGGAJ 'A!.!4A ))   "C'K'K[![! "" YY1q53;;qs1v;#>>? @ TT OOAr    "EOOAryy'(    "EE5tT:r7c t\}}|j|}t|}t|}t ||}t ||}|t vrctjjd}tj|d|z|jdt|zg}|j}n|}|rt|||g\}}}t|dst|drd|_|dk(rd|_t%t&d|d  5|j)||dd dddt+||j)|||j)|j-|j-|j/|} | j0|j0k(sJt3|d s|j5||dkDsJyy#1swYxYw)Nrr ralphasrrPassiveAggressiveRegressorr"r#r:r0rrr)rorr/rrrrurrrr=rXr1r&rKrzrr*rMrrrNrrrrAr) rrrarrrrr5rrs rrr s  DAq A aAn%I!)Q/A!)Q/A ""ii##A& YY1q53;;qs1v;#>>? @ TT ,aBZ81b 9h 'GIw,G  ++  dV$$ $   a3B  Y MM!R MM!((*biik*   q !F <<288 ## # i\ 2q"%+++ 3)  s G  Gctjjd}t|}|j d}t ||}t ||dddf}|j||gd}|D]}t||sJy)Nr)rrr)rrr) rrrrnormalrrrrz)rrrrrrrrs rrr+ s ))   "Cn%I   A!.!4A!)Qq!tW5A MM!Q GE 9i000r7c t|drdg}nddg}|D]i}t|dd\}}t||dd \}}}} t|d rt||}t||}t t j |}|dk(rd d d } nd d d d} t|j| } t| dr| jdt| dr| jd t| dr| jdt| dr| jdt| | j||| j|} t|drMt j| dk(dkDrjJy)Nrrr r-rrf)rr)rrr[r)r\g-C6?)rr<)rr<r rrtrrurvrwmin_weight_fraction_leafr)rn_iter_no_change)rrgףp= ?)rArr+r)r1rrrrrzrNrrrw) rrr. n_centersrrrrrOr]rrrs rrr< s|/}53q6 )!L1+; qCa, (& o: 60F '2G '*+ >#/L#6:L?+66L6Q :x (  ! ! ! - :z *  ! !4 ! 0 :9 :  ! !4 ! @ :1 2  ! !2 ! 6$w(##F+/|<776Q;'$. ..Gr7ct|}t|dr|jdt|dr|jdt||j |||j |}|jd|j |||j |} t || d t ||d kDsJy) Nrtrrurvrrwbalancedrweighted)average)rrzrrNrrr&) rrrrOrr]r.rry_pred_balanceds r'check_class_weight_balanced_classifiersri s'Jz8$S)z:&t,Z NN7G$    'Fz2NN7G$ ((0O FOZ @8 D  r7cBtjddgddgddgddgddgg}tjgd}|}t|dr|jd t|d r|jd t|d r|jd t ||jd|j ||j j}t|}tttj|}|tj|dk(|zz |tj|dk(|zz d}|j||j ||j j} t|| d|zy)z4Test class weights with non-contiguous class labels.rgrrB)r<r<r<r0r0rtrrurvrwrr-rrrr<r0)r<r0z>Classifier %s is not computing class_weight=balanced properly.r:N) rr)rzrrNrcoef_rr1rrr'rE) r Classifierrrr coef_balancedrhrr coef_manuals r-check_class_weight_balanced_linear_classifierr sz 4,q D$<#sc3ZPQA "#AJz8$ T*z:&t,z4 #Z z2NN1a(..335MAIc"))A,'(I qAv2 3 R94 5L|4..A&,,113KPSWWr7c tdd\}}t||t}t|}t ||}t ||j }t|}|j|||j }|jD]I\}} ||} tj| tj| k(r7Jd|d|d| d| d y) Nrrirjrrz, should not change or mutate the parameter z from z to z during fit.) rrr)rrrNrrrrjoblibhash) rrrrrroriginal_params new_params param_nameoriginal_value new_values rryry s 1 3DAq!.!JGAn%I!)Q/AY ! ! #Fv&OMM!Q%%'J&5&;&;&=" Nz* {{9%^)DD Z < D'>r7cp t|}tt|jdryt t|j}t rdD]}||vs|j|dt|jDDcgc] }|D]}| }}}tt|t|z t|z }t|D cgc]} | jdr| c} }|rJd|dt|d y#t$rtd|dwxYwcc}}wcc} w) zCheck setting during init.rz9 should store all parameters as an attribute during init.rdN)r c32K|]}t|ywrP)rD)rRparents rrTz2check_no_attributes_set_in_init.. sV>UFi/>Usr`zR should not set any attribute apart from parameters during init. Found attributes r) rrrzrQrrDrOremove__mro__setrrsorted) rrr init_paramsr params_parentr<parents_init_params invalid_attrras rrhrh sD .) tI'')>?DO445KCk!""3' Wd9o>U>UVVM"E " VtI'#k*::SAT=UULRT__S=QRSL  % ' |1  W X   Ss D D-D3,D3D*c ntjddgddgddgddgddgddgddgddgddgg }tjgd}t||}t|}|j |||j |}|j tj|jsJ|j |}t||tjtj|}tj|jsJ|j |}t||y)Nr0r<r ) r<r<r<r r r r-r-r-)rr)rrrrrdr rsrrHrrr)rrrrrS pred_origrs rr{r{ s  H H H F F F H F H  A ,-A!.!4A  CGGAqM AILLN ??399 %% % ;;q>DtY' ,,v||C( )C ??399 %% % ;;q>DtY'r7ctjddgddgddgddgddgddgddgddgddgddgddgddgg }t||}tjgd}t||}dD]}t |||||y)Nr-rr<r r) r<r<r<r r r r<r<r<r r r  NotAnArrayPandasDataframe)rr)rr"check_estimators_data_not_an_arrayrrrrobj_types rrr s  F F F F F F F F F F F F  A ".!4A 56A!.!4A5*4AxP6r7czt\}}t||}t||}dD]}t|||||y)Nr)rorrrrs rrr, s@  DAq!.!4A!.!4A5*4AxP6r7c|tvr tdt|}t|}t|t||dvrt dj ||dk(r=t tj|}t tj|}nv ddl } tj|}|jdk(r| j|d}n| j|d}| jtj|d}|j|||j|} |j|||j|} t!| | d | y#t$r td wxYw) NzoSkipping check_estimators_data_not_an_array for cross decomposition module as estimators are not deterministic.rzData type {0} not supportedrrr<FrzDpandas is not installed: not checking estimators for pandas objects.rr~)rurBrrNrrrrrr(rr+r*r,rrrE) rrrrr estimator_1 estimator_2rX_r-rr s rrr5 sZ "" %  'K'K[![!886==hGHH< A ' A '  ABww!|YYrY.\\"5\1bjjm%8BOOB    #EOOAq    "EE5tT: V  s !A5E,,Fc:|j}tt5t|}t |t ||j |usJt|jd|j} d}t|jjDcgc] }||s |}}|j}|tt|dgd}|D]}|j |j"k7s+Jd|j$dt'|j(dt*t,t.t0t2t'dt&h} | j5t6j8jt'|j | vxst;|j } | sbJd|j$d |j(d t'|j j(d |j$d t=d | Dd |j$|j?vr0|j "Jd|j$d |j(d}||j$} tA| t6jBrtE| |j d|j$d} tG| r| |j urJ| | |j k(rJ|  dddycc}w#ttf$r YdddywxYw#1swYyxYw)Nrrdc|jdk7xr4|j|jk7xr|j|jk7S)z*Identify hyper parameters of an estimator.r)rr^ VAR_KEYWORDVAR_POSITIONAL)rs r param_filterz.param_filter} sAFFf$3!--/3!"2"22r7r&z parameter rDz has no default valuez Parameter 'z' of estimator 'z ' is of type z which is not allowed. 'z(' must be a callable or must be of type c34K|]}|jywrP)rg)rRrQs rrTz9check_parameters_default_constructible.. s?t}}srzEstimator parameter 'zT' is not returned by get_params. If it is deprecated, set its default value to None.z Parameter z> was mutated on init. All parameters must be stored unchanged.)$rfrLrr6rreprrr0rrrrr[rrr1defaultrrrQrgrr<rrrpupdater sctypeDictcallablerrrrrHr>) rr3rinitrrrr init_param allowed_types allowed_value param_value failure_texts rrrd s##I - 0' 2  i Y##%222y))+@)BTBTU  %T?55<<>>a,q/> %%'!#gi9OQS&T"U"WX %J""j&6&66 Y(( 6 T M  !5!5!7 8Z''(M9- ++, ! joo./&&'} **+4456OO$$L???@ C =fkkm3!))1+JOO+<=!**+,JJ1  1K+rzz2"; 0B0BC! 12;;!-&**<*<<JlJ<&**<*<<JlJ ? 1 0sUAL5(K4 K/+K//K41HLLL/K44LL LLLcJt|dr|dt|jzz }t|drF|jdkDr7t j ||j dk(||j ddz}t|drt j|dS|S)Nrequires_positive_yrr<rrr)r0r<)rArrrrwhereflatr1)rrs rrr s)!67 QQUUW )/AFFQJ HHQ!&&)^Qq A 6)!34zz!W%% Hr7cdt|dvr |dddf}t|dr||jz }dt|dvrbt|drtjntj}tj ||jz j |}|jjdk(r||jz }t|rt|d }|St|d r |||}|S) N1darrayrrrr categoricalrcSkewedChi2Sampler euclidean)rr\) rArrr;rrrrfrgrr()rrrrs rrr sJyi88 adG)!67 K 9)<<( D "(( HHa!%%'k # * *5 1##':: K9% q 5 H I: . 1aL Hr7cgd}|tz }||vry|dk(rt|jd}n t|}t|dr}t }|j |j }}t||}t|dt||}|j||tj|jdk\sJyy)N)r"RidgeClassifierrrryr r~ LassoLarsrBrrvrr<)rurrrzrrtargetrrNrrralln_iter_)rrnot_run_check_n_iterririsrrs rrr s// ## {.)4434? .) y*%{ 4;;2 &y" 5A& %na 8 avvi''1,---&r7cht|}t|dr|tvrgdgdgdgdg}ddgdd gdd gd dgg}n%td gd gdgddd\}}t ||}t |d|j |||tvr|jD] }|dk\r Jy|jdk\sJyy)Nrv)rBrBr)rrBrB)@rr)rreg@rgɿg?g?gg333333?r3rrrr rr<)rrzrurrrNrr)rrrrriter_s rrr"sn%Iy*% & & /?OTA+SzC;d DB"I. EAr*.!.Is"N7Mtt{((**7Ms)rrrr)rrrshallow_paramsrs @rr r AsL nA\\u\-N,,D,)K N~7K7K7MN NN Nr7c t|}|jd}d}|jdi||jd}t|j t|j k(sJ||j D]\}}|||urJ|t j t jdg}t|} |j D]} || } |D]} | | | < |jdi| |jd}t| j t|j k(sJ||j D]\}}| ||urJ|| | | <y#ttf$r} | jj}tjdj|| |dj|}|}|jd} t|j t|j k(sJ|j D]\}}|||urJn$#t $rtj|YnwxYwYd} ~ d} ~ wwxYw)NFrz>get_params result does not match what was passed to set_paramszn{0} occurred during set_params of param {1} on {2}. It is recommended to delay parameter validation until fit.z9Estimator's parameters changed after set_params raised {}r)rrrrrrrrrr[rrfrgrrrr)rrr orig_paramsr5 curr_paramsr!r" test_values test_paramsr default_valuerre_typechange_warning_msgparams_before_exceptions rr r Lsn%I&&E&2K JCI';'&&E&2K {! "c+*:*:*<&= =BsB =!!#11~"'C'"$FF7BFFD)K;'K!&&( #J/  E&+K # 4$ $$3{34(222> ;++-.#k6F6F6H2IIN3NI'--/DAq&q>Q.33.0?!B#0 JG) z* 6-- ,,2F6:t,LPVV# +6''222> 66;;=>##((*C!, 1 1 316q9Q>>>!4%6MM"456- 6sD4FI9A!I43AI I  I4 I+ (I4*I+ +I44I9ct\}}t||}t|}d}t|ds-t t |5|j ||dddyy#1swYyxYw)NzUnknown label type: rbrrQ)rorrrArMrr)rrrrrr5s rrrs`  DAq!.!4A nA C a_ - Jc * EE!QK+ * . * *s A""A+c ddg}tddd|dd\}}t||d d \}}}}t|} t| d rt| d r| j ||| j |ddd fj d} | j|j d} t| t| } } t| | yyy#t$rtjtj| Dcgc]}| | |k(jncc}wc}}tj|}t!|tj"t%|YywxYw)N)r r )rrrrrrT)rhr)rirrr=皙?rrrr<r)decimals)rr+rrzrrrrr rGrrr)rrwr)rHr4r1)rrrrrrrrOr]rab rank_proba rank_scoregroupgrouped_y_score sorted_idxs rrrse vG   DAq(8 1!($GVWfn%Iy-.79o3V gw'  # #F +AqD 1 7 7 7 D  ' ' / 5 5r 5 B!)!hqkJ  G %j* =4W. G !hh<>IIjNFFMM'7#$#F 9f % * 6*622# Y MM'7# 9f %3F3F;J}}Z--r{{;"((:#3#34888"((2::.222 (vd^d^@GGO   s.&Ictjjd}t|}t |d|j vr|j dd}|jd|df}t||}t|r|j|}n|jd d| }t||}t|jd ds. t|t|j j"d |j'|| t|y#t$$rY*wxYw#t$$r}t%d |d}~wwxYw)NrgrFrrr rrrrUrz) passes check_is_fitted before being fit!zFEstimator fails to pass `check_is_fitted` even though it has been fit.)rrrrrNrrrrrrXrrAgetrTrrfrgrr)rrrrrhrrrs rrr/sK ))   #Cn%IYy++--.I s)Q 0A!)Q/AN# JJIJ & KKAAIK 6!)Q/A i $ $[% 8  I & &&//01  MM!Q "      T  s*-D# D2# D/.D/2 E ; EE ctjjd}t|}t |d|j vr|j dd}|jd|df}t||}t|r|j|}n|jdd| }t||}t|d rJ|j||t|d sJ|j|jd k(sJy) NrrFrrr rrrUn_features_in_r<)rrrrrNrrrrrrXrrzrr'r)rrrrrhrrs rrrVs ))   "Cn%IYy++--.I s)Q 0A!)Q/AN# JJIJ & KKAAIK 6!)Q/Ay"233 3 MM!Q 9. // /  # #qwwqz 11 1r7c8tjjd}t|}t |d}|j d|df}t ||}d} |j|dy#t$r tfd|DsYdydwwxYw)Nrrr r)z1requires y to be passed, but the target y is Nonez.s?->c3#b'>->s) rrrrrNrrrrrZ)rrrrrhrexpected_err_msgsr*s @rrrps ))   "Cn%IYI s)Q 0A!)Q/A  a ?->??H@sA00 B9BBct|}d|dvxsd|dv}|r|drytjjd}t |}t |d|j vr|jdd }|j|d f }t||}t|r|j| }n|jdd | }t||}|j|||j|jdk(sJgd} |dddgf} d|jdd} | D]O} t!|| st#|| } | dk(r t%| |} t't(| 5| | dddQt!|dsyt |}t+|r(|j-||tj.|n|j-|||j|jdk(sJt't(| 5|j-| |dddy#1swYxYw#1swYyxYw)NrrrrbrrFrrrrr rUr<)rrrrrz'X has 1 features, but \w+ is expecting z features as inputrrrQrr)rArrrrrNrrrrrrXrrr'rrzr0rrMrrrr)rrris_supported_X_typesrrrhrrrX_badr5rcallable_methods r!check_n_features_in_after_fittingr0s+ n %D T)_$H i(H 4#8 ))   "Cn%IYy++--.I A 'A!)Q/AI JJIJ & KKAAIK 6!)Q/A MM!Q  # #qwwqz 11 1M a!fIE 4QWWQZL@R SCy&) !)V4 W %o;O Jc * E "+ *  9m ,n%IYaBIIaL9a#  # #qwwqz 11 1  # &eQ' ' &+ * ' &s/ H*H7*H4 7Ic t|}t|dsyt|jj }tt j}|j ||k(sJ|d||j |z y)N _get_tagsz@._get_tags() is missing entries for the following default tags: )rrzrr2rr@ intersection)rrr tags_keysdefault_tags_keyss rrrsn%I 9k *I'')..01IM..01  ! !"3 48I I & !7!78I!J J K M Ir7c  ddl}t|}d|dvxsd|dv}|r|drytj j d}t|}t||jd}t||}|j\}} t jt| D cgc]} d |  c} } |j|| d } t|r|j|} n|j!dd | } t#|| } t%j&5t%j(ddt*d|j-| | dddt/|ds t1dt3|j4tj6sJ|j4j8t:k(sJt=|j4| |j>}|jAdr2d|vs.|jCdsd|jDvrt1d|dg}dD]@}t/||stG||}|dk(r tI|| }|jK||fB|D]G\}}t%j&5t%j(ddt*d|| dddI| ddddft| D cgc]} d|  c} df| dd d!tM| d dd"fg}|jOjQDcic] \}}d#|vr||}}}tSd$|jUD}|D]\}}|j| |d }tWjXd%|}|D]+\}}t[t0||d&'5||ddd-t/|d(r|rqt|}t]|r*t j^| }|ja| | |)n|ja| | t[t0|*5|ja|| dddy#t$r tdwxYwcc} w#1swYxYw#1swYxYwcc} wcc}}w#1swY xYw#1swYKxYw)+NrHpandas is not installed: not checking column name consistency for pandasrrrrb)r$rcol_Fcolumnsrr rUerrorz#X does not have valid feature namessklearn)messagermodulefeature_names_in_zTEstimator does not have a feature_names_in_ attribute after fitting with a dataframerrrrz2 does not document its feature_names_in_ attribute)rrrrrrrrrr0z.Ns LO5$OszBThe feature names should match those that were passed during fit. z did not raiserrrrQ)1r(r,rBrArrrrrNrrrr)rr*rrXrrr filterwarnings UserWarningrrzrrr@rrrRrHrrrrr0rr%rrrrZrrescaperMrrr)rrr-rr-rrX_origrhrirnamesrrrrrr/r` invalid_namesrrrearly_stopping_enabled invalid_nameadditional_messager. expected_msgrs r(check_dataframe_column_names_consistencyrOs  n %DT)_$H i(H 4#8 ))   "Cn%IY ZZXZ &F &y& 9F"LLIz HH% *;<*;QQCj*;< =E VU 7AI JJIJ & KKAAIK 6!)Q/A  " 9    a # 91 2 7   i112:: >> >  & & , , 66 6y22E:!++Kz*K';+?+? +K )?)? @P Q  My&) !)V4 W %o;Ofo67"# 6  $ $ &  # #=$   1I' &# ttTU,1*,= >,=qqc ",= >&  "1IB3uQRy>BRRT U  M $..06688JC s " U 8  ! LFMMO LL,9( ( Q 5 Ayy Q!" $ *LD&,4&8Ou *y-04J .)  #iilG  ! !!Q ! 8  ! !!Q ' Jl 3  ! !% +4 33-:_   V   .= # "`' & ?"4 3sMP3' Q  0Q<&Q Q* Q/ Q5R3QQQ' 5Q? R c|j}d|dvs|drytdgdgdgddd \}}tj|}t |}t ||}|j d }t||}|tvrMtjtj|tj|f}|dddd fxxdzcc<|j|| }t|D cgc]} d |  } } ttd5|j| dddddd|j| } | Jt!| tj"sJ| j$t&k(sJt)d| DsJt!|t*r|dj d } n|j d } t-| | k(sJd| dt-| ycc} w#1swYxYw)Nrrrbr3rrrr rrr<rfeaturez'input_features should have length equalrQc3<K|]}t|tywrP)rr)rRrs rrTz:check_transformer_get_feature_names_out..sC1Bz$$1BrU Expected  feature names, got )r2rr.rnrrrrNrurrrrrMrrfrrrrRrrpr1) rrrrrrrir X_transformrinput_featuresfeature_names_outn_features_outs r'check_transformer_get_feature_names_outrYms  % % 'DY'4+@ I&  DAq &&q)A()K!+q1AJ[! B "" UU2::a="**Q-/ 0 3Q36 a ++A+4K-2:->?->sm->N?  "K L)).1*=> M$99.I  (( ( ' 44 4  " "f ,, , C1BC CC C+u%$Q--a0$**1- .0P >""6s;L7M6NOP0%@ M Ls8 G(G--G6c. ddl}|j}d|dvs|dryt dgdgdgdd d \}}t j |}t|}t||}|jd }t||}|tvrMtjtj|tj|f}|ddd d fxxd zcc<t|D cgc]} d |  } } |j!|| d} |j | |} t|D cgc]} d|  } } t#t$d5|j'| ddd|j'}|j'| }t)||t+| t,r| djd }n| jd }t/||k(sJd|dt/|y#t$r tdwxYwcc} wcc} w#1swYxYw)Nrr7rrrbr3rrr rrr<rFr:rbadz0input_features is not equal to feature_names_in_rQrSrT)r(r,rBr2rr.rnrrrrNrurrrrr*rMrrfrHrrpr1)rrr-rrrrrirrfeature_names_indfrUinvalid_feature_namesfeature_names_out_defaultfeature_names_in_explicit_namesrXs r.check_transformer_get_feature_names_out_pandasras5   % % 'DY'4+@ I&  DAq &&q)A()K!+q1AJ[! B "" UU2::a="**Q-/ 0 3Q36 a +0+<=+B++B"+5K16j0AB0A1s1#Y0AB  "T U))*?@ V!, A A C&1&G&G'#02QR+u%$Q--a0$**1- %&.8X >""6s;T7U6VWX8e   V   :> C U Us#G)* H, H H )G> Hc tjjd}|jd}|j ddd}t ||}|j dj}|r_|jj}t|t|z }t|t|z }d|d |d |} ||k(sJ| td d i} gd } |D]} |j| } | dk(rtd| Dr$td| Drtd| d|dd| d|d}|d| d} t|}|jd i| | i| D]u}t||s|d| d} t!t"|| 5tdt%|dDrt'|||nt'||||dddw| Dcgc] }t)|} }| D]} t+|}|jd i| |i| D]u}t||s|d| d} t!t"|| 5tdt%|dDrt'|||nt'||||dddwy#1swY?xYwcc}w#t,$rYwxYw#1swYxYw) Nrrrr rfFrz>Mismatch between _parameter_constraints and the parameters of z%. Consider the unexpected parameters z% and expected but missing parameters BadTyper)rrrnrrbc3fK|])}t|txr|jtk(+ywrP)rr?rQr rRrIs rrTz)check_param_validation..s0 )  z8 , LH1L L)/1c3fK|])}t|txr|jtk(+ywrP)rr?rQr res rrTz)check_param_validation..s0 )  z8 , HD1H H)rfzThe constraint for parameter rzc can't have a mix of intervals of Integral and Real types. Use the type RealNotInt instead of Real.zThe 'z' parameter of z must be .* Got .* instead.z@ does not raise an informative error message when the parameter z% does not have a valid type or value.zc does not have a valid type. If any Python type is valid, the constraint should be 'no_validation'.rc3bK|]'}t|txr|jd)ywr N)rrrrRX_types rrTz)check_param_validation..s0"Fvs+I0II"Fs-/rra  does not have a valid value. Constraints should be disjoint. For instance [StrOptions({'a_string'}), str] is not a acceptable set of constraint because generating an invalid string for the first constraint will always produce a valid string for the second constraint.c3>K|]}|jdywri)rrjs rrTz)check_param_validation..Cs!&JF1&Js)rrrrrXrrrrArrQrZrrrrzrMr9rAr0r;r:r)rrrrrestimator_paramsvalidation_paramsunexpected_paramsmissing_paramsrparam_with_bad_type fit_methodsr constraintsrrrrI bad_values rcheck_param_validationrus~ ))   "C  !A Aqr "A!.!4A%00e0<AAC*AAFFH 12S9I5JJ-.5F1GGLTFS44E3FG##1"2 4  !$44=g=42$y"b13HK& $;;JG / )   )   )  / |4vF$$   OD69TUf# $I K  .)   A ,?@A!F9f-&< DD  -UGL",YI"F /GIv.q1.GIv.q!4ML"0FQQ[zz2[ Q%J 6zB  !I ;J #: ;%y&1f!!+ -""1P&0 &J 3 62152 621a8QP&&u'VMLR '  (QPs2AJ!(J. J3 AK !J+ 3 J?>J?K c> |j}d|dvs|drytjjd}t |}|j d t |  |jddd t|  t| fd } fd }||d }|jD]e\}t |}t|s||} tvr| d} |jd ||} tvr| d} t| | gy)Nrrrbrrrr rfctvr"|jjS|jjSrP)rurr)rSrrrs rfit_then_transformz6check_set_output_transform..fit_then_transform]sD & &771a=**1a0 0wwq!}&&q))r7c(|jSrP)rn)rSrrs rrnz1check_set_output_transform..fit_transformbs  A&&r7)rrnrr)r2rrrrrrrXrrNrrzru set_outputrF) rrrrrrxrntransform_methodstransform_methodX_trans_no_settingX_trans_defaultrrs ` @@rcheck_set_output_transformrMs6  % % 'DY'4+@ ))   "C()K  !A!"2A6A Aqr "A!"2A6A[!* '(&#4"9"9";K( {D) -k: & &!3A!6 3*;7 & &-a0O %%7I##.s G2F$7; %2Fs)rrzfit_transform/dfzfit_transform/arrayrn)rrrurrfrzrn) rrrr]routputscasescasedata_fitdata_transformrr`rs ` r_output_from_fit_transformr}s:G B' !2q) !1b) $a+  E  G2F GG      OOHa (**(22>1E %//?$k&G&G&IJGDM  R " E{O,JD$**(66tQ? %33D!<$k&G&G&IJGDM   Nr7c|\}} |\} } || sJ|jdr|nd} ||| | } || | y#t$r}t|d|d||d}~wwxYw)aCheck if the generated DataFrame by the transformer is valid. The DataFrame implementation is specified through the parameters of this function. Parameters ---------- name : str The name of the transformer. case : str A single case from the cases generated by `_output_from_fit_transform`. index : index or None The index of the DataFrame. `None` if the library does not implement a DataFrame with an index. outputs_default : tuple A tuple containing the output data and feature names for the default output. outputs_dataframe_lib : tuple A tuple containing the output data and feature names for the pandas case. is_supported_dataframe : callable A callable that takes a DataFrame instance as input and return whether or E.g. `lambda X: isintance(X, pd.DataFrame)`. create_dataframe : callable A callable taking as parameters `data`, `columns`, and `index` and returns a callable. Be aware that `index` can be ignored. For example, polars dataframes would ignore the idnex. assert_frame_equal : callable A callable taking 2 dataframes to compare if they are equal. r]Nr;indexz, does not generate a valid dataframe in the z] case. The generated dataframe is not equal to the expected dataframe. The error message is: )rr)rrroutputs_defaultoutputs_dataframe_libis_supported_dataframecreate_dataframeassert_frame_equalrfeature_names_defaultdf_transfeature_names_dataframe_libexpected_indexexpected_dataframers r_check_generated_dataframersJ&5"G ",A)H) !( ++ + #mmD1UtN)4N8%78 f@G001s 4   s A A" AA"c |j}d|dvs|drytjjd}t |} |j d} t || } |jddd } t|| } t| t| jd D cgc]} d |  } } t| jdD cgc]} d |  }} || | | }t | jd}t||| || }|dk(r&t | j|}t}nt | }t|} |5t||| || }ddd|D]}t'|||||||||ycc} wcc} w#1swY3xYw#t $r7}|j#}t%|}|d|vs d|vsJ|Yd}~yd}~wwxYw)aCheck that a transformer can output a DataFrame when requested. The DataFrame implementation is specified through the parameters of this function. Parameters ---------- name : str The name of the transformer. transformer_orig : estimator The original transformer instance. dataframe_lib : str The name of the library implementing the DataFrame. is_supported_dataframe : callable A callable that takes a DataFrame instance as input and returns whether or not it is supported by the dataframe library. E.g. `lambda X: isintance(X, pd.DataFrame)`. create_dataframe : callable A callable taking as parameters `data`, `columns`, and `index` and returns a callable. Be aware that `index` can be ignored. For example, polars dataframes will ignore the index. assert_frame_equal : callable A callable taking 2 dataframes to compare if they are equal. context : {"local", "global"} Whether to use a local context by setting `set_output(...)` on the transformer or a global context by using the `with config_context(...)` rrrbNrrrr rfr<rrrrrzlocal)transform_outputz% output does not support sparse data.z.The transformer outputs a scipy sparse matrix.)r2rrrrrrrXrrNrrr{rrrr capitalizerr)rr dataframe_librrrcontextrrrrrrr\rr]transformer_defaultrtransformer_dfcontext_to_use outputs_dfrcapitalized_lib error_messagers r%_check_set_output_transform_dataframersJ  % % 'DY'4+@ ))   "C()K  !A!"2A6A Aqr "A!"2A6A[!+0+<=+ 4^ '224A D E V?=P    Q s<+ F FF)F+F)F&"F)) G)2-G$$G)c  ddlt||dfdfdjj |y#t$r tdwxYw)Nrz0pandas is not installed: not checking set outputr(c0t|jSrPrr*)rr-s rroz<_check_set_output_transform_pandas_context..KAr||)Dr7c.j||d|S)NF)r;rr)r*)rr;rr-s rroz<_check_set_output_transform_pandas_context..Ls2<< wU%4@4 r7rrrrr)r(r,rBrtestingr)rrrr-s @r*_check_set_output_transform_pandas_contextrAsXK* D ::88  KIJJKs 4A ct||dyNrrrrs r!check_set_output_transform_pandasrT.t5EwOr7ct||dyNglobalrrs r$check_global_output_transform_pandasrX.t5ExPr7c  ddlddlm}fd}t ||dfd|||y#t$r t dwxYw)Nr)rz0polars is not installed: not checking set outputct|tjr|j}j ||dS)Nr)schemaorient)rrrrr*)rr;rpls rrzD_check_set_output_transform_polars_context..create_dataframecs2 grzz *nn&G||Age|<.mrr7r)rpolars.testingrr,rBr)rrrrrrs @r*_check_set_output_transform_polars_contextr\sRK5= * D)- KIJJKs (=ct||dyrrrs r!check_set_output_transform_polarsrtrr7ct||dyrrrs r(check_global_set_output_transform_polarsrxrr7ctjjd}t|}t |d}t |d|\}}t ||}|dvrtj|}t|drtj|d<t|r|j|}n|jdd | }t||}|j}|jd |j!||t|d r|j#||j$j&rJt)||y)aCheck that estimators able to do inplace operations can work on read-only input data even if a copy is not explicitly requested by the user. Make sure that a copy is made and consequently that the input array and its writeability are not modified by the estimator. rrr-)rhrir))Lasso ElasticNetMultiTaskElasticNetMultiTaskLassomissing_valuesrPrr rUF)writerN)rrrrrNrrasfortranarrayrzrrrrXrrsetflagsrrflags writeablerE) rrrrrhrr`rX_copys rcheck_inplace_ensure_writeabler|s ))   "Cn%IYI  ac JDAq!)Q/A OO   a y*+&&$I JJIJ & KKAAIK 6!)Q/A VVXFJJUJ MM!Qy+&Aww   Avr7)NF)Nr;F)Nr;)r])F)Fr;)T)rrrr contextlibrrr functoolsrrinspectrrnumbersr r rnumpyrscipyr scipy.statsr rrbaserrrrrrdatasetsrrrrr exceptionsrrrfeature_selectionrr linear_modelrr r!r"r#metricsr$r%r&metrics.pairwiser'r(r)model_selectionr*r+model_selection._validationr,rr- preprocessingr.r/random_projectionr0treer1r2utils._array_apir3r4r5r6r8rutils._param_validationr9r:r;r=_missingr>_param_validationr?_tagsr@rArrBrCrDrErFrGrHrIrJrKrLrMrNfixesrOrPrQrR validationrSrTrUrmrurrrrrrrrrrr$r6rArJr;r]rbrorrrrrr rr}r|rkrlrmrnrprorFrsrrr rbr rryrrrrrrrrrrrvrjrirrtrwrxr~rrrrrrrrr8rrrrrrrgrqrrrzrrrr;rrrrrryrhr{rrrrrrrrr r rrrrr rrrr0rrOrYrarurrrrrrrrrrrrr7rrsW "$)"  PO<DCLL<5$14@ 6 #'MLHHH7t!+H-( ) D22'> 62*  6%r!/H,^ V* ? ;|N;b  x,vWW*)$'J  LIf  "3;lOR -(( )( V=*/+/8=* 5+ 5-(!@)!@H-(AL)ALH%PP=+67!8!H $$N<-(5)5p-()0<$-(!R)!RH-(()(V%%P<$-(0)0"-(B)B"-( !) ! -( *) *X1v>>@:44,@-(")".-(-1)-1`  2G2Gj-(,),8-(9N)9Nx-()2-(73)73v-( -) --(X)X,-(! )! H:CGOGOT@,EPPL6-(%6)%6P-()))X-(P )P f-(*)*Z-( *) *-( ,) ,..(-(7)7D==@MD-(;);6-(*//,)/,d 1 1 -()/))/X-()*-(%)%P-( ) B-(")"J-(()(D-(Q)Q0-(Q)Q-(+;)+;\fKR  4A 4-('.)'.T-(*)*<-(O)O-(50)50p-( ) -()G))GX(PV7t$N244-(A()A(H L,^-P`6Xrt9n-J`*Z:zW t&PQ0PQ-(*)*r7