U -e0@sddlZddlZddlZddlmZddlmZddlm Z m Z ddl m Z ddl mZmZmZddZejd d d gd d ZddZejd d d gddZejdejejejgejdejejejgddZejdejejejgddZddZddZddZddZd d!Z ejjd"d#d$d%gd&d'd%gge!d(d$d%gd)d'd%ggej!d*d+d,gd-d.d,gge"d/ej!d*d$d,gd-ej#d,gge"d/ej!d*d$d,gd-e$d0d,gge"d/ej!dd$d,gd-d'd,gge"d/ej!dd$dgd-ej#dgge"d/ej!dd$dgd-e$d0dgge"d/gd1d2d3d4d5d6d7d8gd9d:d;Z%ejd d d gejdgejd?dd@gdAdBZ&ejdgejdCd'd%gd$d%gd'd%ggdd$d$gdddgdd$d$ggfdDd-gdEd-gdFd*gdEd-ggdddddgddddd$gdd$dddggfgdGdHZ'dIdJZ(ejd?dKd@dgejdLdKd@dgdMdNZ)ejdOdPdQgejd"d$d'ge!dRdSggdTdUZ*ejdOdPdQgdVdWZ+ejjdXd&d%gd#d%ggd&d#gd%ggej,fe!d$d'gdYd'ggd$dYgd'ggej-fej!d+d,gd.d,gge"d/d+d.gd,ggej,fe!d+d,gd.d,ggd+d.gd,ggej.fe!d$d'gej#d'ggd$ej#gd'ggejfej!d+ej#gdej#gge"d/d+dgej#ggej,fej!d+e$d0gde$d0gge"d/d+dge$d0ggej,fgd1d2d3dZd[d\d]gd9d^d_Z/ejd d d gejjd`ej!d-d*gge"d/j0ej!d-dagge"d/j0d-d*dbggej,fej!d$d'ggdcd/j0ej!d$ddggdcd/j0d$d'dYggej1fej!d-d*gge"d/j0ej!d-dagge"d/j0e!d-d*dbggej,fej!dd-gge"d/j0ej!dd*gge"d/j0dd-degge"fej!d-d*gge"d/j0ej!d-ej#gge"d/j0d-d*degge"fej!d-dgge"d/j0ej!d-ej#gge"d/j0d-ddegge"fej!d-ej#gge"d/j0ej!d-dgge"d/j0d-ej#degge"fgd3d2dfdgdhdidjgd9dkdlZ2dmdnZ3dodpZ4dqdrZ5ejjdsd@dtdugfdKdtdvdugfdbd'd*gdwdxgfgd@dydzgd9d{d|Z6d}d~Z7ejjd"d&d'd%gd#d$d%gge!d(d'd%gdd$d%ggej!d-d.d,gd*d+d,gge"d/gd1d2d3gd9ddZ8ejjd`ej!d-d*gge"d/j0ej!d-dagge"d/j0d-d*dbggej,fej!d$d'ggdcd/j0ej!d$ddggdcd/j0d$d'dYggej1fej!d-d*gge"d/j0ej!d-dagge"d/j0e!d-d*dbggej,fgd3d2dgd9ddZ9ddZ:ddZ;ejde$eddZ?ddZ@ddZAddZBddZCejdej#de$d0gddZDejd?d&dYgd&dYdd-ggddZEejjdd>d=gddgd9ejjd?d@d-d'd*ggd@dzgd9ddZFejde e gddZGejddd'iddiddid'ddddddgejddd-d*dbdagggddZHejd?dKd@d*ggddZIejd?d-gdaggddZJejdddYiddiddiddiddidYddddddgddZKejd?d@d*ggddZLejd?d-gdaggddZMddZNejddYd$ddddigddÄZOddńZPddDŽZQddɄZRdd˄ZSdd̈́ZTejddd$dϜgddфZUejdd'dYdϜgddӄZVddՄZWejddddddddgejddddgddZXddZYejdej#dgddZZddZ[ejd d d gejdddgddZ\ejd d d gddZ]ejd d d gddZ^ejd d d gddZ_ddZ`ddZaejdej#dgddZbejdddgejdej#dgddZcejjd`ej!d-ej#gge"d/j0ej!d-d*gge"d/j0ej!d-ej#dage"d/gej,fej!d-ej#gge"d/j0ej!d-d*gge"d/j0ej!d-ej#dage"d/gej,fej!dej#ggejd/j0ej!dRggejd/j0e!ddSej#ggejfgdddgd9ddZdejde!dej#dRggj0e!dej#dggj0e!dSggfe!ddSdRggj0e!dddggj0e!ej#ggfej!dbej#d*gge"d/j0e!dej#dggj0ej!dagge"d/fej!dbd-d*gge"d/j0e!dddggj0ej!ej#gge"d/fgdd Zed d Zfd d Zgejddd.ggej!dd.ggdd/ej!dd.ggdd/gejdd+d.ggej!d+d.ggdd/ej!d+d.ggdd/gddZhddZiddZjddZkejdd>d=gddZlejdej!d-gdgge"d/dgej#gej#ggejmdgdgdgge"d/fej!ej#gdgd-gge"d/dgej#gej#ggejmdgej#gej#gge"d/fgd d!Znd"d#Zod$d%Zpd&d'Zqd(d)Zrd*d+ZsejdddYiddiddiddiddidYddddddgd,d-Ztd.d/Zud0d1Zvd2d3Zwd4d5Zxejdddidd'igd6d7Zyejddd$idd8igd9d:Zzd;d<Z{d=d>Z|dS(?Nsparse)NotFittedError) OneHotEncoderOrdinalEncoder) is_scalar_nan)_convert_containerassert_allcloseassert_array_equalcCstdddgdddgg}t}tdd}||}||}|jdksLt|jdksZtt|shtt|rvtt| dd dd d gd dd dd ggt| |dS) NrF sparse_outputr ?) nparrayr fit_transformshapeAssertionErrorrissparser toarray)XZ enc_sparseZ enc_denseX_trans_sparseZ X_trans_denserj/var/www/html/Darija-Ai-Train/env/lib/python3.8/site-packages/sklearn/preprocessing/tests/test_encoders.py!test_one_hot_encoder_sparse_denses   rhandle_unknownignoreinfrequent_if_existc Cstdddgdddgdddgg}tdddgg}tdd}||tjtdd ||W5QRXt|d}|||}t || td d d d d d d ggt ||dS) Nrr r r errorr Found unknown categoriesmatchrr) rrrfitpytestraises ValueError transformcopyr rr r rX2ohZ X2_passedrrr#test_one_hot_encoder_handle_unknown(s"     r2c CsLtdgdgg}tddgd}d}tjt|d||W5QRXdS)Nab categorieszqThis OneHotEncoder instance is not fitted yet. Call 'fit' with appropriate arguments before using this estimator.r')rrrr*r+rr-)rencmsgrrrtest_one_hot_encoder_not_fitted@s r9c Cstddddgd}tddgd}t|d}|||}t||tddddgdd ddggt||dS) NZ11111111Z22Z333Z4444)r Z55555r%rr) rrreshaperr)r.r r-rr/rrr+test_one_hot_encoder_handle_unknown_stringsLs   r< output_dtype input_dtypecCstjddgg|dj}tjddgddgg|d}td|d}t|||t||||td|dd}t|||t||||dS)Nrr dtypeauto)r6r@F)r6r@r) rasarrayTrr rrr)r-)r>r=r X_expectedr1rrrtest_one_hot_encoder_dtype_s rEcCstd}|ddgddgd}tjddddgddddgg|d}t|d}t|||t| | ||t|d d }t|||t| | ||dS) Npandasr3r4r r ABrr?F)r@r) r* importorskip DataFramerrrr rrr)r-)r=pdX_dfrDr1rrr!test_one_hot_encoder_dtype_pandasns "  rNcCst}dddddgdddddgdd d d dgdd dd dgg}|||}tdddddddddddddddg||ddd d!d"g}td#d$d%d&d'd(d)d*d+d,d-d.d/d0d1g|tjtd2d3|ddgW5QRXdS)4NMaler girlr r Female) 3boy [Z x0_FemaleZx0_MaleZx1_1Zx1_41Zx1_51Zx1_91Zx2_boyZx2_girlZx3_1Zx3_2Zx3_12Zx3_21Zx4_3Zx4_10Zx4_30onetwothreefourfiveZ one_FemaleZone_MaleZtwo_1Ztwo_41Ztwo_51Ztwo_91Z three_boyZ three_girlZfour_1Zfour_2Zfour_12Zfour_21Zfive_3Zfive_10Zfive_30z!input_features should have lengthr')rr)get_feature_names_outr r*r+r,)r7r feature_namesZfeature_names2rrr"test_one_hot_encoder_feature_names~sb     racCs\t}tjddggtdj}|||}tddg||jdgd}tdd g|dS) Nuc❤t1Zdat2r?u x0_c❤t1Zx0_dat2un👍meZinput_featuresun👍me_c❤t1u n👍me_dat2)rrrobjectrCr)r_r )r7rr`rrr*test_one_hot_encoder_feature_names_unicodes rdc Csdd}t|d}tjddggtdj}|||}tddg||jd gd }td d g|d d}t|d|}d}tj t |d|W5QRXdS)z=Check the behaviour of `feature_name_combiner` as a callable.cSs|dt|S)N_)reprfeaturecategoryrrr name_combinerszHtest_one_hot_encoder_custom_feature_name_combiner..name_combiner)Zfeature_name_combinerNoneNr?z x0_'None'Zx0_Noner3rbza_'None'Za_NonecSsdS)Nrrrgrrrwrong_combinerszItest_one_hot_encoder_custom_feature_name_combiner..wrong_combinerzMWhen `feature_name_combiner` is a callable, it should return a Python string.r') rrrrcrCr)r_r r*r+ TypeError)rjr7rr`rlerr_msgrrr1test_one_hot_encoder_custom_feature_name_combiners  rocCstddggj}t}|jddddggd|dddddggksLt||j dksdt|jdddddggd||j d kstdS) Nr r rr r5r6)r r#r#r) rrrCr set_params get_paramsrrrr)rr1rrrtest_one_hot_encoder_set_paramssrrcCsXtdd}||}tddd}||}t||t|rL|jdksPt|S)NrAr5Fr6rZcsr)rrr rrrformatr)rr7ZXtr1ZXtr2rrrcheck_categorical_onehots    rurdefr 7abcr rSrr4rHcatr3rIr?nanmixednumericrcz mixed-nanzmixed-float-nanz mixed-Nonezmixed-None-nanzmixed-None-float-nan)ZidscCstt|dddgf}t|ddgddggtt|ddddgf}t|ddddgddddggtdd|}t|dddddgdddddggdS)Nrr rAr5)rurrr rrr)rXtrrrrtest_one_hot_encoders r~sparse_FTdropfirstc Csdddgdddgdddgg}t||d}||}tj|td}t|||ddgddgddgg}t|d |d }||}t|}t||||dkrdddgdddgdddgg}t||ddgddgd dd ggd }||}tj|td}d|d<t|||ddgddgddgg}t|ddgd d gg|d}||}tj|td}d|d<d|dddf<t|||tdddgdddgg}td}t j t |d||W5QRXdS)Nrxr rwrvr r rrr?rA)rr6r68)rr r6)r r )rr6r r rr)Shape of the passed X data is not correctr') rrrrrcr inverse_transformreescaper*r+r,)r rrrr7X_trexpr8rrrtest_one_hot_encoder_inversesH        rz X, X_transrZr[r\c CsJt|d|}d}|r"t|d}tjt|d||W5QRXdS)zCheck that `inverse_transform` raise an error with unknown samples, no dropped feature, and `handle_unknow="error`. Non-regression test for: https://github.com/scikit-learn/scikit-learn/issues/14934 rzqSamples \[(\d )*\d\] can not be inverted when drop=None and handle_unknown='error' because they contain all zerosrr'N)rr)rr*r+r,r)rX_transrr7r8rrr?test_one_hot_encoder_inverse_transform_raise_error_with_unknownKs rcCsJtjddgddgddggtd}tddd }||}t|||dS) NrOr rQr r r? if_binaryFrr)rrrcrrr r)roherrrr&test_one_hot_encoder_inverse_if_binaryks   rr reset_dropcCstjddgddgddggtd}t|dd}||||}|}|j|d t| ||t |||t||dS) NrOr rQr r r?Frr) rrrcrr)r-r_rpr rr )rrrrrr`rrrtest_one_hot_encoder_drop_resetrs     rmethodr)r@@c Cs6t}d}tjt|dt|||W5QRXdS)N'Expected 2D array, got 1D array insteadr')rr*r+r,getattr)rrr1r8rrrtest_X_is_not_1Dsrc CsRtd}|ddddg}t}d}tjt|dt|||W5QRXdS)NrFr r#rr')r*rJSeriesrr+r,r)rrLrr1r8rrrtest_X_is_not_1D_pandass  rzX, cat_exp, cat_dtyper stringz missing-floatzmissing-np.nan-objectzmissing-float-nan-objectcCs||dddfD]}tdd}||t|jts:tt|j|D]l\}}|}t|drt|dsrt|dd|ddkstn||kstt |j |sFtqFqdS)Nr:rAr5) rr) isinstance categories_listrziptolistrr issubdtyper@)rZcat_exp cat_dtypeXir7resrZres_listrrrtest_one_hot_encoder_categoriess#   rzX, X2, cats, cat_dtypedcint64r#zz object-stringzobject-string-nonezobject-string-nanzobject-None-and-nanzobject-nan-and-Nonec Cst|d}tdddgdddgg}t|||t|jdt|dksVt|j d t|dkstt|j dj |kstt|d}t j tdd||W5QRXt||d}tdddgdddgg}t||||dS)Nr5rrrr&r'r6r )rrrr rrrr6rrrr@r*r+r,r)r-)rr0catsrr r7rrrr)test_one_hot_encoder_specified_categoriess:   rc CsDtjddggtdj}tdddggd}tdddgdddgg}t||||t| |||j d dddgkst t |j djtjst td d ggj}td d d ggd}d }tjt|d | |W5QRXtd d tjggj}td tjd ggd}tjt|d | |W5QRXdS)Nr3r4r?rr5rrrr r r z%Unsorted categories are not supportedr')rrrcrCrr r)r-rrrrrrr@object_r*r+r,rz)rr7rr8rrr(test_one_hot_encoder_unsorted_categoriess rc Cstjddgddggtdj}tdddgdddggd}td d d d d d gd d d d d d gg}t||||jd dddgkst t |jdj tj st |jd dddgkst t |jdj tj st dS) Nr3r4rr r?rr r5rr)rrrcrCrr rrrrrrr@rrr7rrrr7test_one_hot_encoder_specified_categories_mixed_columns-s&rcCsLtd}|ddgddgd}t|}t|ddddgddddggdS)NrFr3r4r r rGr)r*rJrKrur )rLrMr}rrrtest_one_hot_encoder_pandas:s rzdrop, expected_namesx0_cZx2_bZx1_2x0_bZx2_abinaryZmanualcCs>dddgdddgg}t|d}|||}t||dS)Nrr r3r4r)rr)r_r )rZexpected_namesrrr`rrr'test_one_hot_encoder_feature_names_dropCs   rcCsddgddgddgg}tddddgddddgddddgg}tddg}td d d }||}t|j|t||d d gdd gdd gg}tddgddgddgg}tddg}td d d }||}t|j|t||dS)NrSyesnorYrrrrFrtruer3false)rrrrr drop_idx_r )rexpectedZexpected_drop_idxrresultrrr*test_one_hot_encoder_drop_equals_if_binaryUs         rrcCsXt}tjdddgdddggdd}t|||dtdd}t|||dS)Nrr rr?float64)rrrr rZastyperrrrtest_ordinal_encoderms  rzobject-string-catc Cst|d}tdgdgg}t|||t|jdt|dksJt|jd t|dksht|jdj |ks|tt|d}t j t dd||W5QRXdS)Nr5rrrr&r')rrrr rrr6rrrr@r*r+r,r))rr0rrr7rrrr)test_ordinal_encoder_specified_categories~s  rc Csdddgdddgg}t}||}tj|td}t|||tddddgddddgg}td}t j t |d ||W5QRXdS) Nrxr rwrvr r?rrr') rrrrrcr rrrr*r+r,)rr7rrr8rrrtest_ordinal_encoder_inverses  rcCstddd}tjddgddgdd ggtd }tjdd gd dgddggtd }||||}tjd dgddgddggdd }t||||}tjddgddgddggtd }t||dS)Nuse_encoded_valuer unknown_valuer3xr4yrrr?ZxyZblar r rr)rrrrcr)r-r r)r7X_fitr X_trans_encr X_trans_invinv_exprrr+test_ordinal_encoder_handle_unknowns_strings         rr@cCstddd}tjddgddgdd gg|d }tjdd gd dgddgg|d }||||}tjddgddgd d ggdd }t||||}tjddgddgddggtd }t||dS)Nrrr r r r?rVrr)rrrr)r-r rrc)r@r7rrrrrrrrr,test_ordinal_encoder_handle_unknowns_numerics         rcCs`tdtjd}tdgdgdgg}|||dgdgdgg}t|dgdgtjggdS)Nrrr r r r#r)rrrzrr)r-r )r7rrrrr(test_ordinal_encoder_handle_unknowns_nans  rc CsNtdtjtd}tdgdgdgg}tjtdd||W5QRXdS)Nr)r rr@r r r z'dtype parameter should be a float dtyper') rrrzintrr*r+r,r))r7rrrr8test_ordinal_encoder_handle_unknowns_nan_non_float_dtypesrc Cs\tjdddddggtdj}dddg}t|d}d}tjt|d||W5QRXdS)NZLowZMediumZHighr?r5z*Shape mismatch: if categories is an array,r') rrrcrCrr*r+r,r))rrr7r8rrr+test_ordinal_encoder_raise_categories_shapes   rcstddtjddddgddddggdd}tjddgd d ggd dtjddgd d ggddtd d gddggtddgddggtjdd gd d ggddfD]BtfddtdDstt |qddgd d ggtfddtdDstt |dd gd d ggtfddtdDsltt |dS)NrAr5rrrr?r r r r#rr3r4rrabcdrccsg|]}j|jjkqSrrr@.0irr7rr sz'test_encoder_dtypes..cs"g|]}tj|jtjqSr)rrrr@integerrr7rrrscsg|]}j|jdkqS)rcrrrrrr s) rrrr)allrangerr r-r)rrrrtest_encoder_dtypess& "      rcstd}tddtjddddddgddddddggdd}|jdd gd d gd d gddd}|tfddtd Dst t  | ||dd gddgddgd}|dj |dj |dj g|tfddtd Dst t  | |dS)NrFrAr5rrrr?r r r r#rr)rHrICrcsg|]}j|jdkqS)rrrrrrrsz.test_encoder_dtypes_pandas..r3r4rrrHrIrcs g|]}j|j|kqSrrrZX_typer7rrr"s)r*rJrrrrKr)rrrr r-rr@)rLrrrrrtest_encoder_dtypes_pandass  "  "rcCs*t}ddgddgg}tj|j|dS)NrOr rQr )rrtestingZassert_no_warningsr)r7rrrrtest_one_hot_encoder_warning&sr missing_valuec Csdddd|g}t|d}ddddd gddd dd gdddd|gg}||}d d d d d gd d d d d gd d d d d gg}t|||j|kstd d t|j|jD}| |}t j |t d}t |drzt|dd|ddt |dstt |dstt|ddddf|ddddft|dddf|dddft |dsftt |dstnt||t||dS)NrvrVr rrrxr rwr3r rcSsg|]\}}||qSrr)rryrhrrrr:sz4test_one_hot_encoder_drop_manual..r?r:)r:r:)rrrr rrrrrrrrrcr) rZ cats_to_dropr7rZtransrZ dropped_catsZ X_inv_transZX_arrayrrr test_one_hot_encoder_drop_manual,s2    (   *" rrRc CsNt|d}d}tjt|d(|dddgdddgdd d ggW5QRXdS) Nrz-`drop` should have length equal to the numberr'rxr rwrvr r ;)rr*r+r,r))rr7rnrrrtest_invalid_drop_lengthQs rdensityrZdensecCst|d}t||d}dddgdddgg}||||t|j|j|dkrbt|jd n0t||j|jD]\}}}|t||ksrtqrt|jt j st|jj t kstdS) Nrrrr r3r r4rr) rr)r rrrrrrrZndarrayr@rc)rrZohe_baseZohe_testrZdrop_catZdrop_idxZcat_listrrrtest_categoriesYs     rEncodercCsd|dkstdS)N categoricalZX_types)Z _get_tagsr)rrrr"test_encoders_has_categorical_tagsmsrkwargsmax_categories min_frequency g(\?r)rrrVr6rAc Cstdgddgddgddgdgj}tf|d d d ||}t|jdddggdgdgdgdgd gg}td dgdd gdd gdd gdd gg}||}t||dddgdgdD}| |}t||| } tddg| dS)zpTest that different parameters for combine 'a', 'c', and 'd' into the infrequent category works as expected.r3rr4rrrSrr r"F)r6r rer rcSsg|] }|gqSrrrcolrrrrsz2test_ohe_infrequent_two_levels..infrequent_sklearnr#rx0_infrequent_sklearnN rrrCrr)r infrequent_categories_r-r rr_) rr6X_trainrX_testrr expected_invX_invr`rrrtest_ohe_infrequent_two_levelsrs(2(    rcCstdgddgddgddgdgj}td d d |d |}|jd |jd dksdttdgdgg}||}t d gdgg|| }t dg|| |}t dgdgg|dS)z3Test two levels and dropping the frequent category.r3rr4rrrSrr r"Fr r rrrrr rrN) rrrCrr)rrrr-r r_r r)rrrrrr` X_inverserrr,test_ohe_infrequent_two_levels_drop_frequents"2   r c Csztdgddgddgddgdgj}td d d |d }d |dd}tjt|d||W5QRXdS)z_Test two levels and dropping any infrequent category removes the whole infrequent category.r3rr4rrrSrr r"Fr r Unable to drop category r( from feature 0 because it is infrequentr'NrrrCrr*r+r,r)rrrr8rrr5test_ohe_infrequent_two_levels_drop_infrequent_errorss2rrgQ?g{Gz?rc Cstdgddgddgddgdgj}tfd d d ||}t|jddggdgdgdgdgd gg}td ddgddd gdd dgddd gddd gg}||}t||dgdgdgdgdgg}| |}t||| }tdddg|dS)zkTest that different parameters for combing 'a', and 'd' into the infrequent category works as expected.r3rr4rrrSrr r"Fr rrr rrrrrNr) rrrrrrrrr`rrr test_ohe_infrequent_three_levelss022    rc Cstdgddgddgddgdgj}td d d|d |}tdgdgdgg}td d gd d gd d gg|||jdd|d}tj t |d|dgdgg}W5QRXtd d gd d gg|dS)z5Test three levels and dropping the frequent category.r3rr4rrrSrr r"Fr rr r!r%r&r'rN) rrrCrr)r r-rpr*warns UserWarning)rrrrr8rrrr.test_ohe_infrequent_three_levels_drop_frequents 2"rc Csztdgddgddgddgdgj}td d d|d }d |d d}tjt|d||W5QRXdS)z7Test three levels and dropping the infrequent category.r3rr4rrrSrr r"Fr r rr r'Nrrrrr7test_ohe_infrequent_three_levels_drop_infrequent_errorss2rc Cstdgddgddgddgdgj}td d dd |}t|jddggdgdgdgdgg}td d d gd d d gd d d gd d d gg}||}t||dgg}d}t j t |d||W5QRXdS)zmTest that different parameters for combining 'a', and 'd' into the infrequent category works as expected.r3rr4rrrSrr r$F)r rrr rbadz.Found unknown categories \['bad'\] in column 0r'N) rrrCrr)r rr-r r*r+r,)rrrrrr8rrr(test_ohe_infrequent_handle_unknown_errors"2*  rcCstjdgddgdgtdj}tfddddggd d d ||}dgdgdgdgdgg}td d gd d gd d gd d gd d gg}||}t||dddgg}dgdgg}|D].}|j|d|td gd gg||qdS)zG'a' is the only frequent category, all other categories are infrequent.r3rrrYr?rrr4Fr"r6rr r rrrrN) rrrcrCrr)r-r rp)rrrrrrZdropsrrrr5test_ohe_infrequent_two_levels_user_cats_one_frequent,s&" (    rcCstjdgddgddgddgdgtd j}tddddggd d d d |}t|jdddggdgdgdgdgdgg}tddgddgddgddgddgg}||}t ||dddgdgdD}| |}t||dS)zFTest that the order of the categories provided by a user is respected.r3rr4rrrSrr r?Fr"r r6rr rrr rcSsg|] }|gqSrrrrrrr^sz.rr#N rrrcrCrr)r rr-r rrrrrrrrrrr(test_ohe_infrequent_two_levels_user_catsHs(( (   rc Cstjdgddgddgddgdgtd j}tddddggd d dd |}t|jddggdgdgdgdgd gg}tdddgdddgdddgdddgdddgg}||}t ||dgdgdgdgdgg}| |}t||dS)zTest that the order of the categories provided by a user is respected. In this case 'c' is encoded as the first category and 'b' is encoded as the second one.r3rr4rrrSrr r?Fr"rrrr rNrrrrr*test_ohe_infrequent_three_levels_user_catscs2( 2   rc Cstjdddddddddg dddddddddg f}tdddd}||ddgddgg}||}t|ddddgddddggdS) zaTest infrequent categories where feature 0 has infrequent categories, and feature 1 does not.rr r r rF)rrrN)rc_rr)r-r )rrrrrrrtest_ohe_infrequent_mixeds 2  r!c Cstjdddddddddg dddddddddg dddddddddg f}tdddd }||}t|jdddgt|jdddgt|jdd |}td d d dddddg|ddddddddgddddddddgddddddddgddddddddgddddddddgddddddddgddddddddgddddddddgddddddddgg }t||dddgdddgg}| |}ddddddddgddddddddgg}t||| |}tj ddd gddd ggt d}t||tdddd  |}tjtdd| |W5QRXdddgdddgg}| |}ddddddddgddddddddgg}t||| |}tj dddgdddggt d}t||d S)z?Test infrequent categories with feature matrix with 3 features.rr r r rrSrAr"r6rr NZx0_0Zx0_3rZx1_0Zx1_5Zx1_infrequent_sklearnZx2_0Zx2_1r#rr?r$r&r')rr rrrr rr_r r-rrrcr)r*r+r,) rrrr`rr X_test_transrrrrr'test_ohe_infrequent_multiple_categoriess   (   ( r$c Cs(td}|jdddddddddg dddd d d dddg d d d gd}tdddd}||}t|jdddgt|jdddd gddddddgddddddgddddddgddddddgddddddgddddddgddddddgddddddgddddddgg }t|||jddgdd gd d d gd}ddddddgddddddgg}| |}t||| |}t j ddgddggt d}t|||jddgd dgd d d gd}| |}ddddddgddddddgg}t||| |}t j ddgddggt d}t||dS)zHTest infrequent categories with a pandas dataframe with multiple dtypes.rFr3frr4rr rrSrV)strrr&rcolumnsrAr"r"r rr?N)r*rJrKrrrr rr r-rrrrc) rLrrrrrr#rrrrr.test_ohe_infrequent_multiple_categories_dtypess\            r*rX)rrcCsptdgddgddgddgdgj}tfd d d |}|||dgg}t|d ggd S),All user provided categories are infrequent.r3rr4rrrSrr r"Frr N)rrrCrr)r-r rrrrrrr$test_ohe_infrequent_one_level_errors5s2 r-cCsftjdgdgtdj}tfddddggdd d ||}|dgdgg}t|d gd ggd S) r+rr r?rrr3r4Fr"rr N)rrrcrCrr)r-r r,rrr5test_ohe_infrequent_user_cats_unknown_training_errorsCs r.c CsHddgddgddgg}d}tjt|dtdd |W5QRXdS) NrOr rQr r z'`sparse` was renamed to `sparse_output`r'Fr)r*r FutureWarningrr))rr8rrr&test_one_hot_encoder_sparse_deprecatedTsr0zinput_dtype, category_dtypeZOOZOUZUOZUUSOZSUZSS array_typerrZ dataframec Cstjdgdgg|d}tjddg|dg}t|dd|}tdgdgdgdgg||d}||}tddgddgddgddgg}t||t|d|} | |}tdgdgdgdgg}t||d S) a"Check that encoding work with object, unicode, and byte string dtypes. Non-regression test for: https://github.com/scikit-learn/scikit-learn/issues/15616 https://github.com/scikit-learn/scikit-learn/issues/15726 https://github.com/scikit-learn/scikit-learn/issues/19677 r4r3r?Frsrr r5N) rrrr)rr-r rr ) r>Zcategory_dtyper2rr6rrrroerrrtest_encoders_string_categories]s  "  r4c Cshtjdgdggdd}tjddgddg}t|dd}td}tjt|d ||W5QRXd S) zCheck that this mixture of predefined categories and X raises an error. Categories defined as bytes can not easily be compared to data that is a string. r4r3Ur?SFrszjIn column 0, the predefined categories have type 'bytes' which is incompatible with values of type 'str_'.r'N) rrrrrr*r+r,r))rr6rr8rrr$test_mixed_string_bytes_categoricals|s r7cCsPtjdd|d|ggtdj}tddd|}|}t|ddd |gdS) Nr3r4r?Fr!rr Zx0_arZx0_)rrrcrCrr)r_r )rrrnamesrrr)test_ohe_missing_values_get_feature_namessr:c Cstd}|jddddgtjdddtjgtddd d gd }tdd ddd ddgd ddd dddgddd ddd dgd dddddd gg}t|}t||dS) NrFdogryr rr#r?)col1col2r<r=r'r ) r*rJrKrrrzfloatrur )rLdfexpected_df_transr}rrr%test_ohe_missing_value_support_pandass    rA pd_nan_typepd.NAznp.nanc Cstd}|dkr|jntj}|d|jdd|ddgddi}td d d d gd d d d gd d d d gd d d d gd d d d gg}td |d }| |}t ||t |j d kst t|j d dd dddgt|j d d st dS)NrFrCr<rr3r4rir?rr Fr8r:)r*rJNArrzrKrrrrr lenrrr isnan)rBr rLpd_missing_valuer?r@rdf_transrrr1test_ohe_missing_value_support_pandas_categoricals*         rIc Csddgddgddgg}tdd|d}||}tdddgdddgdddgg}t||d d gg}tdddgg}d }tjt|d ||}W5QRXt||| |}t |tjddggt d dS)zZCheck drop='first' and handle_unknown='ignore'/'infrequent_if_exist' during transform.r3rr4r r rFrrr rr tFound unknown categories in columns \[0, 1\] during transform. These unknown categories will be encoded as all zerosr'r?N rrrrr r*rrr-rr rcr rrrrDrwarn_msgrrrr/test_ohe_drop_first_handle_unknown_ignore_warnss.     rOc Csddgddgddgg}tdd|d}||}tddddgddddgddddgg}t||d d gg}tddddgg}d }tjt|d ||}W5QRXt||| |}t |tjdd ggt dd S)zDCheck drop='if_binary' and handle_unknown='ignore' during transform.r3rr4r r rFrJrr rKr'Nr?rLrMrrr3test_ohe_drop_if_binary_handle_unknown_ignore_warnss.        rPc Csddgddgddgg}tdd|ddgddggd}||d dgg}tddgg}d }tjt|d ||}W5QRXt||d S) znCheck drop='first' and handle_unknown='ignore'/'infrequent_if_exist' during fit with categories passed in.r3rr4r r rF)rrr r6rzqFound unknown categories in columns \[0\] during transform. These unknown categories will be encoded as all zerosr'N) rr)rrr*rrr-r )r rrrrDrNrrrr'test_ohe_drop_first_explicit_categoriess  rQc Cstd}|jdddgdddgdddgd }td d }|jdd d }tjt|d||W5QRX||tjt|d| |W5QRXdS)zJRaise informative error message when pandas output and sparse_output=True.rFr r r rr4)r3r4r3r'Trr-zxPandas output does not support sparse data. Set sparse_output=False to output pandas DataFrames or disable pandas outputr'N) r*rJrKr set_outputr+r,rr)r-)rLr?rr8rrr'test_ohe_more_informative_error_message7s $   rTc CsXttjdddggj}ttjd}dtj}tjt|d| |W5QRXdS)zDTest ordinal encoder with nan passthrough fails when dtype=np.int32.rrr?zdThere are missing values in features \[0\]. For OrdinalEncoder to encode missing values with dtype: r'N) rrrzrCrint32r*r+r,r))rr3r8rrrBtest_ordinal_encoder_passthrough_missing_values_float_errors_dtypeKs   rVencoded_missing_valuercCstjtjdddggtjdj}t|d|}t|jdks@t t |jdddtjg| |}t ||gdgdgdgg| |}t ||dS) z.Test ordinal encoder with nan on float dtypes.rrr?rWr rrN) rrrzrrCrr)rErrr r-r)rWrr3rr rrr5test_ordinal_encoder_passthrough_missing_values_floatYs  rYc Cs$td}|dkr|jntj}|d|jdd|ddgddi}t|d |}t |j d ksbt t |j d d d dddgt |j d dst ||}t|dgdg|gdgdgg||}|jdkst t |d dd fddgt |d d d fddgt |ds t d S)z0Check ordinal encoder is compatible with pandas.rFrCr<rr3r4rir?rXr rNr r:@rr)rr r r)r*rJrDrrzrKrrr)rErrr rFr-r rr)rBrWrLrGr?r3rHr rrr=test_ordinal_encoder_missing_value_support_pandas_categoricalks$   r[rZzobject-None-missing-valuezobject-nan-missing_valueznumeric-missing-valuec Cstt|d}tdgtjgg}t||||jdj|ksBtt|d}t j t dd| |W5QRXdS)z.Test ordinal encoder for specified categories.r5rrr&r'N) rrrrzr rrr@rr*r+r,r))rr0rrr3rrrr=test_ordinal_encoder_specified_categories_missing_passthroughs&  r\zX, expected_X_trans, X_testrrcCs8tddd}||}t||t||dggdS)z>Test the interaction between missing values and handle_unknownrr:rgN)rrr r-)rZexpected_X_transrr3rrrr/test_ordinal_encoder_handle_missing_and_unknowns   r]c Cstdddgdddgg}t|}t}d}tjt|d||W5QRXtjt|d| |W5QRX| |}t|}tjt|d| |W5QRXdS)zCheck that we raise proper error with sparse input in OrdinalEncoder. Non-regression test for: https://github.com/scikit-learn/scikit-learn/issues/19878 r r r rz6A sparse matrix was passed, but dense data is requiredr'N) rrrZ csr_matrixrr*r+rmr)rr)rZX_sparseencoderrnrrrrrtest_ordinal_encoder_sparses   r_c Cstddddddgddtjf}tdddggddd }||tdddggd d }tjtd d ||W5QRXdS)zCheck OrdinalEncoder.fit works with unseen category when `handle_unknown="use_encoded_value"`. Non-regression test for: https://github.com/scikit-learn/scikit-learn/issues/19872 rr r rNr:rr)r6r rr$rr&r')rrZnewaxisrr)r*r+r,)rr3rrr-test_ordinal_encoder_fit_with_unseen_categorys$  r`rZAAOr5rcCs4tddd}||||}t|ddggdS)zChecks that `OrdinalEncoder` transforms string dtypes. Non-regression test for: https://github.com/scikit-learn/scikit-learn/issues/19872 rirrN)rr)r-r )rrr7rrrr1test_ordinal_encoder_handle_unknown_string_dtypes s   rbcCsftddddgdd}t|}t|jtj|ddj| |}t|dgd gd gdggd S) zCheck that `OrdinalEncoder` accepts Python integers that are potentially larger than 64 bits. Non-regression test for: https://github.com/scikit-learn/scikit-learn/issues/20721 l HP 1&l H]viel :?i}GalIRK2e6kr:r r)Zaxisr r N) rrr;rr)r rsortrCr-)rr^rrrr#test_ordinal_encoder_python_integer's  rdcCsLtd}dddg}|jdddgg|d}t|}|}t||d S) z-Check feature names out is same as the input.rFr4rr3r r r r'N)r*rJrKrr)r_r )rLr9rr7Zfeature_names_outrrr.test_ordinal_encoder_features_names_out_pandas;s    recCstjdgdgtjggtd}tdtjdd|}||}t|dgdgdggtjd gtjggtd}||}t|tjgdgg||}|ddd kst t |ddst d S) zECheck interactions between encode_unknown and missing value encoding.r3r4r?rr rrWrr rN) rrrzrcrr)r-r rrrF)rr3rrr# X_roundtriprrr0test_ordinal_encoder_unknown_missing_interactionGs    ri with_pandasc Cstjddgddgdtjggtd}d}|rPtd}|j|d d gd }|d }n|d }tdd}tjt |d| |W5QRXdS)zXCheck OrdinalEncoder errors when encoded_missing_value is used by an known category.r3r;r4ryrr?zTencoded_missing_value \(1\) is already used to encode a known category in features: rFletterZpetr'z \['pet'\]z\[1\]r rXr'N) rrrzrcr*rJrKrr+r,r))rjr error_msgrLr3rrr0test_ordinal_encoder_encoded_missing_value_errores"   rmz4X_train, X_test_trans_expected, X_roundtrip_expected1c Cstdtjtjd|}tdgtjgdgg}||}t||||}|jd}t |D]V}||df} ||df} | dkr| dkst q`t | rt | st q`| | ks`t q`dS)znCheck transform when unknown_value and encoded_missing_value is nan. Non-regression test for #24082. rrgrnr4rN) rrrzr)rr-r rrrrrrF) rZX_test_trans_expectedZX_roundtrip_expectedr3rr#rhZ n_samplesrZ expected_valvalrrr9test_ordinal_encoder_unknown_missing_interaction_both_nans(       rpc Cstd}|ddgddgd}t}|jddd}tjt|d ||W5QRXtd d jd d}td d jdd}||}||}t| |t | |j d S)z*Check OneHotEncoder works with set_output.rFr3r4r r rGrRz*Pandas output does not support sparse datar'FrdefaultN) r*rJrKrrSr+r,rr to_numpyr r_r()rLrMrr(Z ohe_defaultZ ohe_pandas X_defaultX_pandasrrrtest_one_hot_encoder_set_outputs    rucCsttd}|ddgddgd}tjdd}tjdd}||}||}t||t| |j d S) z+Check OrdinalEncoder works with set_output.rFr3r4r r rGrqrRN) r*rJrKrrSrr rrr r_r()rLrMZ ord_defaultZ ord_pandasrsrtrrrtest_ordinal_set_outputs   rvcCszdddddgddgg}t|d}|ddggt|t|jksFtt|jD]$\}}|jtksftt|||qPd S) zjCheck that the categories_ dtype is `object` for string categories Regression test for gh-25171. asZmmasZeasZrasZacsrn2r5N) rr)rErr enumerater@rcr )r6r7nryrrr test_predefined_categories_dtypes r{cCs~tjdgdgtjggtd}tdd|}t|dgdgdggtddd |}td gg}||}t|dggd S) zBCheck missing value or unknown encoding can equal the cardinality.r;ryr?r rXr rrrsnakeN) rrrzrcrrr r)r-)rrr7rrrr1test_ordinal_encoder_missing_unknown_encoding_maxs r}cCstjdgddgddgddgddgdgtdj}tdd d d |}t|d d ddg|jd|j ddkst tjdgddgddgdgtdj}tdd dd |}t|dg|jd|j ddkst tjdgddgddgddgddgdgtdj}tdd dgd |}t|dd ddg|jd|j ddks|t tdd dd |}t|dd d ddg|j dkst dS)zkCheck drop_idx is defined correctly with infrequent categories. Non-regression test for gh-25550. r3r r4r#rrrr?Fr)rrrrZx0_dZx0_errrSrrN) rrrcrCrr)r r_rrr)rrrrr#test_drop_idx_infrequent_categoriess<2 ,2  r~cCstdgddgddgddgdgj}tfd d d ||}t|jddddggt|jddggdgdgdgdgd gg}d gdgdgd gd gg}||}t ||| |}dgdgdgdgdgg}t||dS)zGTest parameters for grouping 'a', and 'd' into the infrequent category.r3rr4rrrSrr rr:rrr rr rN) rrrCrr)r rrr-r r)rrordinalrexpected_transrr expected_inverserrr,test_ordinal_encoder_infrequent_three_levelss.2   rcCstjdgddgddgddgdgtd j}tddddggdd d d |}t|jddddggt|jddggdgdgdgdgd gg}dgdgdgdgd gg}| |}t ||| |}dgdgdgdgdgg}t||dS)zTest that the order of the categories provided by a user is respected. In this case 'c' is encoded as the first category and 'b' is encoded as the second one. r3rr4rrrSrr r?rr:)r6rr rrr r rrN) rrrcrCrr)r rrr-r r)rrrrrr rrrr6test_ordinal_encoder_infrequent_three_levels_user_catsBs4(    rc Cstdddddddddg dddddddddg f}tdd|}t|jdddg|jddkshtddgddgg}ddgddgg}||}t||| |}tj ddgddggt d}t||dS) zETest when feature 0 has infrequent categories and feature 1 does not.rr r r rNrr?) rZ column_stackrr)r rrr-r rrrc)rrrrrr rrrr%test_ordinal_encoder_infrequent_mixedfs2   rc CsVtd}|ddddg}|jdddddddd d g d d d d d dd d d g |jdgddgd dgdg|dddddgd}td d|}t|jd dd gt|jdd d dgt|jdddg|jdd ddgdd d d g|jdgdgdgdg|dddddgd}ddd gdddgdddgd d dgg}| |}t ||dS)zHTest infrequent categories with a pandas DataFrame with multiple dtypes.rFbirdryr;r|r3r%rr4rr rrSrVr#r?)r&rrr&rrr'rr r N) r*rJZCategoricalDtyperKrrr)r rr-r )rLZcategorical_dtyperrrrrrrr:test_ordinal_encoder_infrequent_multiple_categories_dtypes{s:      $ rcCstjdgddgddgddgdtjggtd j}td d d dd |}t|jdddggtjdgdgdgdgd gtjggtd }dgdgdgdgd gdgg}| |}t ||dS)zJCheck behavior of unknown_value and encoded_missing_value with infrequent.r3rr4rrrSrr r?rr )r rrrWrr rN) rrrzrcrCrr)r rr-r )rrrrrrrr.test_ordinal_encoder_infrequent_custom_mappings"0( rcCstjdgddgddgddgdgtd j}tf|d d d |}td d d |}dgdgdgdgd gg}t||||dS)zMAll categories are considered frequent have same encoding as default encoder.r3rr4rrrSrr r?rr:rrNrrrcrCrr)r r-)rrZadjusted_encoderZdefault_encoderrrrr!test_ordinal_encoder_all_frequents, ( rdcCstjdgddgddgddgdgtd j}tf|d d d |}dgdgdgdgd gg}t||dgdgdgdgd ggdS)zAWhen all categories are infrequent, they are all encoded as zero.r3rr4rrrSrr r?rr:rrrNr)rrr^rrrr#test_ordinal_encoder_all_infrequents ( rcCstjtjgddgddgddgdggtdj}td d |}tjdddtjggtdj}||}t|d gd gd gtjggdS)z5Check behavior when missing value appears frequently.rr;rSryrr|deerr?r rr rr N rrrzrcrCrr)r-r rrrrrrr-test_ordinal_encoder_missing_appears_frequents, rcCstjtjgdgddgddgdgdgdd gdgtd j}td d |}tjddgdd gtjd gdd gddggtd }||}t|d dgd dgtjdgddgddggdS)z7Check behavior when missing value appears infrequently.r;rSryrr|rredrgreenr?r#)rr r rNrrrrr/test_ordinal_encoder_missing_appears_infrequents$& r)}rnumpyrr*ZscipyrZsklearn.exceptionsrZsklearn.preprocessingrrZ sklearn.utilsrZsklearn.utils._testingrr r rmarkZ parametrizer2r9r<rUZfloat32rrErNrardrorrrurrcrzr>r~rrrrrrrrZstr_rrCrrrrrrrrrrrrrrrrrrrrrrrrr rrrrrrrrr!r$r*r-r.r0r4r7r:rArIrOrPrQrTrVrYr[r\r]r_r`rbrdrerirmrBrprurvr{r}r~rrrrrrrrrrrrrs        <       /0&     & &* !       -8               $           $[A       % $     "       !    $  $  0