h@ hUddlZddlZddlmZmZddlZddlZddlm Z ddl m Z ddl mZddlmZmZmZmZmZddlmZmZmZmZmZegZeeed<egZeeed<eezZegZeeed <egZ eeed <ee zZ!d:d Z"d Z#d \Z$Z%e&e$e%zdz Z'e"e$e%e'\Z(Z)ej*+dddgddgfddgddgfddgddgfgdZ,dZ-dZ.dZ/dZ0dZ1ej*+d ed!Z2ej*+d ed"Z3d#Z4d$Z5d%Z6d&Z7d'Z8d(Z9d)Z:d*Z;d+Zej*+d.e!d/Z?ej*+d0d1ej*+d2d1ej*+d.e!ej*+d3d4d5gd6Z@ej*+d.e!ej*+d7ejAejAfejBejBfejCejBfejDejBffd8ZEej*+d.e!d9ZFdS);N)AnyList)DataDimensionalityWarning)euclidean_distances)GaussianRandomProjectionSparseRandomProjection_gaussian_random_matrix_sparse_random_matrixjohnson_lindenstrauss_min_dim)assert_allcloseassert_allclose_dense_sparseassert_almost_equalassert_array_almost_equalassert_array_equalall_sparse_random_matrixall_dense_random_matrixall_SparseRandomProjectionall_DenseRandomProjectioncBtj|}tj||||||||ff||f}||fS)Nsizeshape) nprandom RandomStatesp coo_matrixrandnrandinttoarraytocsr) n_samples n_features n_nonzeros random_staterngdata_coos l/var/www/html/services/rag-service/venv/lib/python3.11/site-packages/sklearn/tests/test_random_projection.pymake_sparse_random_datar*%s )   - -C} IIj ! ! IJ 77 JZ 88  *%   H     x~~// //cVtj|s|S|SN)rissparser!)matrixs r)densifyr04s) ;v    ~~r+) gY@zn_samples, epsdn?皙?Z皙?2i皙?ctjt5t||ddddS#1swxYwYdS)Neps)pytestraises ValueErrorr )r#r>s r)test_invalid_jl_domainrBEs z " "::%iS9999::::::::::::::::::s 9==c"tjt5tddgzddgzdddn #1swxYwYttjddd t jdd dS) Nr3r5r=r1)r1r1r?)r?r@rAr rrr fullr+r)test_input_size_jl_min_dimrJRs z " "@@%a3%iQ#Y????@@@@@@@@@@@@@@@" !Rh//RWXs5K5KsAAAcgd}|D]B\}}tjt5|||dddn #1swxYwYCdS)N))rr)rF)rFrL)rFr)rLrr?r@rA random_matrixinputs n_componentsr$s r)check_input_size_random_matrixrR^s 8 8 8F$*44 j ]: & & 4 4 M, 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4 444s >A A cPgd}|D]\}}|||j||fksJdS)N))rF)rTrF)rTrT)rFrFrrNs r)check_size_generatedrUesb - - -F$*   j}\:66<  A        r+ct|ddd}tdtj|dtdtj|ddS)Ni'rFrr&rD?)r0rrmeanlinalgnorm)rOAs r)check_zero_mean_and_unit_normr]nsb  eQQ77788AaQ///c29>>!#4#4a88888r+cd\}}dD]A}tjt5||||dddn #1swxYwYBdS)N)rTr1)gr9r6densityrM)rOrQr$r`s r)%check_input_with_sparse_random_matrixraxs$L*#EE ]: & & E E M, G D D D D E E E E E E E E E E E E E E EEEs>A A rOc`t|t|t|dSr-)rRrUr])rOs r)$test_basic_property_of_random_matrixrcs4#=111'''!-00000r+cnt|tj|d}t|dS)NrXr_)ra functoolspartialr])rOrandom_matrix_denses r)+test_basic_property_of_sparse_random_matrixrhs;)-888#+M3GGG!"566666r+cd}d}t||d}tdtj|dttj|dd|z ddS) Nr3r2rrWr9rErFddof)r rrrYvar)rQr$r\s r)test_gaussian_random_matrixrmsh LJ jqIIIAc271::q111bfQQ///\1A1EEEEEr+c vd}d}dD]}d|z }t|||d}t|}tj|}tj|tj|z |vsJtj| tj|z |vsJ|dkrtj|dksJn d |vsJtj|d ksJt tj|d kdd|z z d t tj|tj|tj|z kdd|zz d t tj|tj| tj|z kdd|zz d t tj|d kd dd|z z dz|z d t tj|tj|tj|z kd ddd|zz z dzd|zz d t tj|tj| tj|z kd ddd|zz z dzd|zz d dS) Nr3i)g333333?rXrFr)r`r&rXrEr9rD)decimalrj) r r0runiquesqrtrrrYrl)rQr$r`sr\valuess r)test_sparse_random_matrixrtsLJ, ,  K ! *gA    AJJ1wqzzBGL111V;;;; {RW\222f<<<< c>>76??a'''''&====76??a'''' BGAH--q1q5y!DDDD GAbgl&;&;;; < z~eps=0.100000 and n_samples=1000 lead to a target dimension of 5920 which is larger than the original space with n_features=100)match)r*rxr?r@rAry)r_r{rp expected_msgs r).test_too_many_samples_to_find_a_safe_embeddingrs%dC66GD!0  6s ; ; ;   ]:\ : : :   FF4LLL               sA**A. 1A. ctddd\}}d}t|d}|}|dk}||}tD]}|d|d }||}t|d}|}||}||z } | d |zksJd |z | ksJdS) Nii:r;T)squaredr9rvr)rQr>r&rF)r*rravelrx fit_transformmaxmin) rrr>original_distances non_identicalr{r projectedprojected_distancesdistances_ratios r)(test_random_projection_embedding_qualityrs %au55GD! C,T4@@@+1133&#-M,M:0//  6s K K K$$T** 1)TJJJ177992-@-0BB ""$$q3w....3w,,.......!//r+cptD],}|ddd}|tt|tt jsJtjt}t||t jsJ|ddd}|t}t|tt jsJtj ||sJ.dS)Nr1Tr)rQ dense_outputr&F) rryr isinstancer~rndarrayr csr_matrixr.)SparseRandomProjr sparse_datas r)+test_SparseRandomProj_output_representationrs666 2Dq Q Q Q t ",,t,,bj99999mD)) ",,{33RZ@@@@@ 2EPQ R R R VVD\\",,t,,bj99999{2<< 44555555#66r+ctD]}|dddt}|jdksJ|jdksJ|t vr#|jdksJt|jdd|j j dtfksJ| t}|j tdfksJ| t}t|||dd}|t}t||t!jt$5| tddd d fdddn #1swxYwY|t vry|d d d }|t}|j td fksJ|j j d tfksJ|j jdksJd|j jksJdS)NrvrrG)rQr&r>r4gQ?rE)r&r>rFrTr3gMbP?)rQr`r&sU)rxryrrQ n_components_rr`rdensity_ components_rr$r~r#rrr?r@rAnnz)r{r projected_1 projected_2rp2 projected_3rs r)2test_correct_RandomProjection_dimensions_embeddingr'sc0&+&+  6s K K K O OPT U U&((((3&&&& 9 9 9:''''  T1 5 5 5~#Z'88888ll4((  Y$44444ll4(( ; 444A3777''-- ; 444]: & & ' ' LLaaa1f & & & ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' 9 9 9!!sEPQRRRB((..I?y#&66666>'C+<<<<<>%++++*****M&+&+s8'E++E/ 2E/ c d}td|t|dz \}}tD]U}tjt 5||dz|dddn #1swxYwYVdS)NrTrFrw)r*intrxr?warnsrry)r$rrr{s r)1test_warning_n_components_greater_than_n_featuresrQsJ%aSa5H5HIIGD!0DD \3 4 4 D D  *q. 9 9 9 = =d C C C D D D D D D D D D D D D D D DDDs #A88A< ?A< crd}td|t|dz \}}tD]}|dd|}|ddt j|}t t|jt|jdS)NrrTrrDrF)rQr&) r*rrxryrrrr0r)r$rrr{rp_dense rp_sparses r)test_works_with_sparse_datarZsJ%aSa5H5HIIGD!0  ##CCCGGMM$$!!DDDHH M$     " H( ) )793H+I+I      r+c4tdddksJdS)zyTest Johnson-Lindenstrauss for small eps. Regression test for #17111: before #19374, 32-bit systems would fail. r3h㈵>r=lJWN)r rIr+r)"test_johnson_lindenstrauss_min_dimrhs( )$ 7 7 7< G G G G G Gr+random_projection_clsc@|d}|t|}|jt jfdt|jDt}t||dS)NrErwcg|]}| SrIrI).0iclass_name_lowers r) z.ws%RRRa  !a ! !RRRr+)dtype) ryrget_feature_names_out__name__lowerrarrayrangerobjectr)rrandom_projection names_outexpected_names_outrs @r)(test_random_projection_feature_names_outrps--1===$!7799I,5;;==RRRR51B1P+Q+QRRR y"455555r+r#)rE r1 r2r$compute_inverse_componentsTFcdd}||||}t||||zdzdz|\}}||fD]} tj5tjddt|| } dddn #1swxYwY|r(t |d sJ|j} | j||fksJ| | } | j| jksJ| | } t | d r| } t| | d d dS)Nr1)rQrr&r3rFrWignorez>The number of components is higher than the number of features)messagecategoryinverse_components_r!gHz>g|=)rtolatol) r*warningscatch_warningsfilterwarningsrrhasattrrrinverse_transformr~r!r )r#r$rrglobal_random_seedrQrX_denseX_csrXrinv_componentsprojected_backprojected_agains r)test_inverse_transformr~sL--!#=' -J#%)' NGUu KK  $ & & ; ;  #T2     *77::I ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; & F,.CDD D D D.BN!'J +EEEEE*< > 1 1 1 1   . . . . . .r+cd}tjd}|dd}|d}|d}||tj}||tj}t|||t|j |j dS)NrrrrrrW)r) rrrrrrfloat32float64r r r)rrr'rrp_32rp_64 projection_32 projection_64s r),test_random_projection_numerical_consistencyrs D )   # #C TA ! !q 1 1 1E ! !q 1 1 1E''(<(<==M''(<(<==MM=t<<<< !2E4EFFFFFr+)r)Grertypingrrnumpyrr? scipy.sparsesparsersklearn.exceptionsrsklearn.metricsrsklearn.random_projectionrrr r r sklearn.utils._testingr r rrrr__annotations__rall_random_matrixrrrxr*r0r#r$rr%rdata_csrmark parametrizerBrJrRrUr]rarcrhrmrtr|rrrrrrrrrrrrint32int64rrrIr+r)rs> 888888//////(=&=$s)===&=%>c>>>,/FF)?(@DI@@@(@'A49AAA14MM 0 0 0 0   # : SZ'%/ 0 0 ((J KKh sc3Z cS#J cS#J::: 444   999EEE*;<<11=<1*BCC77DC7 F F F1 1 1 rFFFBBB   ///<666*'+'+'+TDDD    HHH02FGG 6 6HG 6&:;;';<<02FGG5e}EE,K,KFEHG=<<;,K^02FGG! RZ RZ 2: 2:  / /HG /02FGG G GHG G G Gr+