
    Q3j                    %   U d Z ddlZddlZddlZddlZddlZddlZddlmZm	Z	m
Z
 ddlZddlZddlZddlmZ ddlmZ ddlmZmZmZ ddlmZ ddlmZ dd	lmZ dd
lmZmZm Z m!Z! ddl"m#Z#m$Z$ ddl%m&Z& ddl'm(Z( ddl)m*Z*m+Z+m,Z,m-Z- ddl.m/Z/m0Z0m1Z1m2Z2 ddl3m4Z4 ddl5m6Z6m7Z7m8Z8m9Z9m:Z:m;Z;m<Z< ddl5m=Z> ddl?m@Z@ ddlAmBZB ddlCmDZDmEZEmFZFmGZGmHZH ddlImJZJmKZKmLZLmMZM ddlNmOZO ddlPmQZQ dZRdZSe*e,dZTe+e-dZU eV       ZWeVeXd<   eWj                  eT       eWj                  eU       g dZZ ej                  g dg dg d g d!g d"g d#g d$g d%g d&g d'g d(g d)g d*g d+g d,g d-g d.g d/g d0g d1g d2g d3g d4g      Z\g d5Z]g d6Z^d7d8gd8d8gd8d7gd9d9gd9d:gd:d9ggZ_g d;Z`d8d8gd:d:gd<d:ggZag d=Zb ej                         Zdej                  j                  d9      Zgegj                  edj                  j                        Zkedj                  ek   ed_l        edj                  ek   ed_i         ej                         Znegj                  enj                  j                        Zkenj                  ek   en_l        enj                  ek   en_i         ej                         Zpegj                  epj                  j                        Zkepj                  ek   ep_l        epj                  ek   ep_i         eQd      Zq ej                  dd>d?@      \  ZsZteqj                  dAB      ZvdCevevdDk  <   eqj                  ddEdFB      Zx e(dGd?dHdI      j                         Zzedj                  edj                  dJenj                  enj                  dJepj                  epj                  dJe_e`dJe\e]dJe\e^dJesetdJevexdJev exdJezexdJ ej                  dK      exdJdLZ|dM Z}dN Z~dO Zej                   j                  dP      ej                   j                  dQeUj                               ej                   j                  dReS      dS                      ZdT ZdU Zej                   j                  dP      ej                   j                  dVeUj                               ej                   j                  dReS      dW                      Zej                   j                  dP      ej                   j                  dQeUj                               ej                   j                  dXdYe!fdZefd[e!fd\e fg      d]                      Zd^ Zd_ Zd` Zda Zdb Zdc Zdd Zde Zdf Zdg Zdh ZddiZej                   j                  djeW      dk        Zej                   j                  djeZ      ej                   j                  dleL      dm               Z	 ddnZej                   j                  djeW      do        Zej                   j                  djeZ      ej                   j                  dleL      dp               Zej                   j                  dP      ej                   j                  dqg  e
eUj                         eS       e
eTj                         eR            dr               Zds Zej                   j                  dP      ej                   j                  dtg  e
eUj                         eS       e
eTj                         eR            du               Zdv Zdw Zdx Zdy Zdz Zej                   j                  djeT      d{        Zej                   j                  djeT      d|        Zd} Zd~ Zd Zd Zej                   j                  deWj                               d        Zd Zd Zd Zd Zd ZddZej                   j                  deZ      ej                   j                  dd      d               Zej                   j                  d e eeZ      jg                  eU                  ej                   j                  dddg      d               Zej                   j                  deZ      ej                   j                  dg d      ej                   j                  dleL      d                      Zej                   j                  dP      ej                   j                  d e e
eZD  cg c]	  } | eUv s|  c} eS             e e
eZD  cg c]	  } | eTv s|  c} eR            z         ej                   j                  dg d      ej                   j                  dleL      d                             Zej                   j                  deZ      ej                   j                  d eeLeM            d               Zd Zej                   j                  djeW      d        Zej                   j                  djeW      ej                   j                  ddgeLz         d               Zej                   j                  djeW      d        Zej                   j                  djeZ      ej                   j                  deM      d               Zd Zej                   j                  djeW      d        Zej                   j                  djeW      ej                   j                  deM      d               Zd Zd Zej                   j                  ddgeLz         d        Zej                   j                  d e ee|j                               ddhz
              ej                   j                  de*e,g      d               Zej                   j                  de|j                               ej                   j                  de+e-g      d               Zd Zd Zd Zej                   j                  djeW      ej                   j                  dddg      ej                   j                  ddgeLz   eMz         d                      Zej                   j                  dP      ej                   j                  dRg d      ej                   j                  dQeUj                               d                      Zej                   j                  d ed<            d        Zd Zej                   j                  dQe*e,g      ej                   j                  dd:dEg      d               Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zej                   j                  d e e1j                          e2j                                     d        Zd Zej                   j                  dQeWj                               d        Zej                   j                  dP      ej                   j                  dReS      d               Zej                   j                  dP      ej                   j                  d ed<            ej                   j                  dReS      d                      Zej                   j                  dRddg      dń        Zej                   j                  dRddg      dƄ        Zej                   j                  dRddg      dǄ        Zej                   j                  dRddg      dȄ        Zej                   j                  deM      ej                   j                  d e+        e-       g      dʄ               Zej                   j                  dQeUj                               d˄        Zd̄ Zej                   j                  dej                  e+dfej                  e-dfee*dfee,dfg      ej                   j                  dddg      dӄ               Zej                   j                  d eeTj                         ddg            dׄ        Zej                   j                  dej                  e+fej                  e*fg      dل        Zdڄ Zej                   j                  dP      ej                   j                  dQe+e-g      ej                   j                  d ej                  ej                  d:ej                  dEddg       ej                  ej                  ej                  d<dEddg       ej                  d9d:d<dEej                  ej                  g       ej                  d9d:d<ej                  dej                  g      g      ej                   j                  dReS      dބ                             Zd߄ Zd Zd Zd Zd Zd Zd Zd Zd Zej                   j                  d ej                  ej                  gdCgdgdgg       ej                  g d      f ej                  ej                  gdgej                  gg       ej                  g d      fgddg      d        Zyc c} w c c} w )z-
Testing for the tree module (sklearn.tree).
    N)chainpairwiseproduct)NumpyPickler)assert_allclose)clonedatasetstree)DummyRegressor)NotFittedError)SimpleImputer)accuracy_scoremean_absolute_errormean_poisson_deviancemean_squared_error)cross_val_scoretrain_test_split)make_pipeline)_sparse_random_matrix)DecisionTreeClassifierDecisionTreeRegressorExtraTreeClassifierExtraTreeRegressor)CRITERIA_CLFCRITERIA_REGDENSE_SPLITTERSSPARSE_SPLITTERS)_py_precompute_absolute_errors)
NODE_DTYPE	TREE_LEAFTREE_UNDEFINED_build_pruned_tree_py_check_n_classes_check_node_ndarray_check_value_ndarray)Tree)compute_sample_weight)xpx)assert_almost_equalassert_array_almost_equalassert_array_equalcreate_memmap_backed_dataignore_warnings)	_IS_32BITCOO_CONTAINERSCSC_CONTAINERSCSR_CONTAINERS)_weighted_percentile)check_random_state)ginilog_loss)squared_errorabsolute_errorfriedman_msepoisson)r   r   )r   r   	ALL_TREES)r   r      r   r   r      ir   r   r   r   r   )r   r         r   r=   r   r   r<   皙?r   r;   r<   )rB   r   r         r   r    @r<   r   r   rC   r   r<   )rB   rB   r   g333333r   r   r   r   r   r   rA   r   r   r<   )rB   rB   r   r   r   r   r   r?   r   r   r   r   r   r<   )rB   r   r;   
   r;   r   皙	r   r;   r?   r=   r<   )zG @r         r      r   r   rH            ?r   rF   r<   )rI   r   rJ   rK   r   rL   r   r   rH   rM   r   r   rE   r<   )rI      rJ   rK   r   rL   r   r   rH   rM   r   r   rE   r<   )rI   rO   rJ   rK   r   rL   r   r   rH   rM   rN   r   rB   r   )   rO   r>   r<   rN   r=   rG   r   r<   r@   r?   r   rP   r   )rP   r   r<   r<   r<   rB   r<   r   r   rE   r?   r   r<   r   )rP   r   r<   rP   r?   rB   rG   rP   r   rB   r<   rP   rP   r   )r<   r<   r   rP   rP   rB   r<   rP   r   r@   r<   rP   r?   r   )r?   r<   r   r?   r   r=   rG   r   r<   r@   r?   r   r?   r<   )rI   rO   rJ   rK   r   r<   r   r   rH   rM   rN   r   rF   r<   )rI   rO   rJ   rK   r   r<   r   r   rH   rM         ?r<   rB   rB   )rI   rO   rJ   rK   r   rG   r   r   rH   rM   rN   r   rB   rB   )rP   r   r>   r<   rN   rE   rG   r   r<   r@   r?   r<   r   rB   )rP   r   r<   r<   r<   rE   r<   r   r   rE   r   r   r   r<   )rP   r<   r<   r<   rP   rB   rG   rP   r   rB   r   rP   r<   r<   )r<   r<   r   r   r<   rF   r<   rP   r   r@   r<   rP   r<   r<   )r?   r<   r   r<   r   r=   r<   r   r<   rE   r   r   r<   r   )r<   r<   r   r   r   r   r<   r<   r<   r<   r<   r<   r   r   r   r<   r   r   r<   r   r   r   r   )      ?rD   333333?皙?rG   g333333@@g)\(?{Gz?gףp=
@rU   g?        rS   rP   rL   r   r         @g|?5^?g(\??r   rE   rB   r<   rP   )rB   rB   rB   r<   r<   r<   r?   )rB   r<   r<      rG   )random_state	n_samples
n_features)   r>   sizerW   g?r;   )r^   r^         ?)densityr[   Xy)r^   r?   )irisdiabetesdigitstoy	clf_small	reg_small
multilabel
sparse-pos
sparse-neg
sparse-mixzerosc                 ~   |j                   | j                   k(  s,J dj                  ||j                   | j                                t        | j                  |j                  |dz          t        | j                  |j                  |dz          | j                  t
        k(  }t        j                  |      }t        | j                  |   |j                  |   |dz          t        | j                  |   |j                  |   |dz          t        | j                  j                         |j                  j                         |dz          t        | j                  |j                  |dz          t        | j                  |j                  |dz   	       t        | j                  |   |j                  |   |d
z   	       y )Nz({0}: inequal number of node ({1} != {2})z: inequal children_rightz: inequal children_leftz: inequal featuresz: inequal thresholdz: inequal sum(n_node_samples)z: inequal n_node_samplesz: inequal impurityerr_msgz: inequal value)
node_countformatr+   children_rightchildren_leftr    nplogical_notfeature	thresholdn_node_samplessumr)   impurityr*   value)dsmessageexternalinternals        I/DATA/.local/lib/python3.12/site-packages/sklearn/tree/tests/test_tree.pyassert_tree_equalr      s   <<1<<' 299Q\\1<<	
' 	!**G6P,P 	'4M*M 9,H~~h'H			(QYYx0'<P2P 	Hq{{84g@U6U 		11
 	!**G6P,P 

AJJBV8VW	1778,g@Q6Q    c                     t         j                         D ]  \  } } |d      }|j                  t        t               t        |j                  t              t        dj                  |               |dd      }|j                  t        t               t        |j                  t              t        dj                  |               y )Nr   r[   Failed with {0}r<   )max_featuresr[   )
	CLF_TREESitemsfitrd   re   r+   predictTtrue_resultru   namer&   clfs      r   test_classification_toyr      s    oo'
d"13;;q>;8I8P8PQU8VW213;;q>;8I8P8PQU8VW (r   c            
         t         j                         D ]  \  } } |d      }|j                  t        t        t        j                  t        t                           t        |j                  t              t        dj                  |              |j                  t        t        t        j                  t        t              d             t        |j                  t              t        dj                  |               y )Nr   r   sample_weightr   rN   )r   r   r   rd   re   rx   oneslenr+   r   r   r   ru   fullr   s      r    test_weighted_classification_toyr     s    oo'
d"1BGGCFO43;;q>;8I8P8PQU8VW1BGGCFC$893;;q>;8I8P8PQU8VW (r   z%ignore:.*friedman_mse.*:FutureWarningr&   	criterionc                    |dk(  rht        j                  t        j                  t                    dz   }t        j                  t              |z   }t        j                  t
              |z   }nt        }t
        } | |d      }|j                  t        |       t        |j                  t              |        | |dd      }|j                  t        |       t        |j                  t              |       y )Nr9   r<   r   r[   r   r   r[   )rx   absminre   arrayr   r   rd   r   r   r   )r&   r   ay_trainy_testregr   s          r   test_regression_toyr     s    
 I FF266!9!((1+/+&*

3CGGAwCKKNF+

CCGGAwCKKNF+r   c                     t        j                  d      } d| d dd df<   d| dd dd f<   t        j                  | j                        \  }}t        j                  |j                         |j                         g      j                  }| j                         } t        j                         D ]  \  }} |d      }|j                  ||        |j                  ||       dk(  sJ dj                  |              |dd      }|j                  ||        |j                  ||       dk(  r~J dj                  |              y )	N)rG   rG   r<   r>   r   r   rR   r   r[   r   )rx   rp   indicesshapevstackravelr   r   r   r   scoreru   )re   gridxgridyrd   r   r&   r   s          r   test_xorr   &  s   
AAbqb"1"fIAab!"fI::agg&LE5
		5;;=%++-0133A		Aoo'
d"1yyA#%E'8'?'?'EE%21yyA#%E'8'?'?'EE% (r   c                     t        t        j                         t              D ]"  \  \  } }} ||d      }|j	                  t
        j                  t
        j                         t        |j                  t
        j                        t
        j                        }|dkD  sJ dj                  | ||              ||dd      }|j	                  t
        j                  t
        j                         t        |j                  t
        j                        t
        j                        }|dkD  rJ dj                  | ||              y )Nr   r   rY   z0Failed with {0}, criterion = {1} and score = {2}rP   r   rN   )r   r   r   CLF_CRITERIONSr   rf   datatargetr   r   ru   )r   r&   r   r   r   s        r   	test_irisr   ;  s    #*9??+<n#MtiYQ7		4;;'s{{4995t{{Cs{ 	
NUU)U
 	
{ YQQG		4;;'s{{4995t{{Cs{ 	
NUU)U
 	
{ $Nr   z
name, Treec                 2    ||d      }|j                  t        j                  t        j                         t	        t        j                  |j                  t        j                              }|t        j                  d      k(  sJ d|  d| d|        y )Nr   r   zFailed with z, criterion = z and score = )r   rg   r   r   r   r   pytestapprox)r   r&   r   r   r   s        r   test_diabetes_overfitr   N  sw     
3CGGHMM8??+xHMM0JKEFMM!$$ 
tfN9+]5'J$r   zcriterion, metricr6   r7   r8   r9   c           	         t        |d|      }t        j                  t        j                  }} || | di |ddij	                  ||      j                  |            } || | di |ddij	                  ||      j                  |            } || | di |ddij	                  ||      j                  |            }	d|	cxk  r|cxk  r|k  sn J |	||f       y )	NrM   r   	max_depthr<   r;      r    )dictrg   r   r   r   r   )
r&   r   metricglobal_random_seedkwargsrd   re   loss1loss4loss7s
             r   test_diabetes_underfitr   ]  s     IADVWF==(//qA1d1V1q155a;CCAFGE1d1V1q155a;CCAFGE1d1V1q155a;CCAFGE u$u$u$;ueU&;;$r   c            	          t         j                         D ]v  \  } } |ddd      }|j                  t        j                  t        j
                         |j                  t        j                        }t        t        j                  |d      t        j                  t        j                  j                  d         dj                  |              t        t        j                  |d      |j                  t        j                        dj                  |              t!        |j                  t        j                        t        j"                  |j%                  t        j                              ddj                  |              y y )Nr<   *   r   r   r[   r   r   rr   rO   )r   r   r   rf   r   r   predict_probar*   rx   r}   r   r   ru   r+   argmaxr   r)   exppredict_log_proba)r   r&   r   prob_predicts       r   test_probabilityr   u  s     oo'
dQQR@		4;;'((3!FF<#GGDIIOOA&'%,,T2	

 	IIlA&KK		"%,,T2	

 	dii(FF3((34%,,T2		
 (r   c                      t        j                  d      d d t         j                  f   } t        j                  d      }t        j	                         D ]!  \  }} |d d      }|j                  | |       # y )Ni'  r   r   r[   )rx   arangenewaxis	REG_TREESr   r   rd   re   r   r&   r   s        r   test_arrayreprr     s[     			%BJJ'A
		%Aoo'
dT21 (r   c                     ddgddgddgddgddgddgg} g d}t         j                         D ]L  \  }} |d      }|j                  | |       t        |j	                  |       |dj                  |      	       N t        j                         D ]L  \  }} |d      }|j                  | |       t        |j	                  |       |dj                  |      	       N y )
NrE   rB   r<   rP   )r<   r<   r<   r<   r<   r<   r   r   r   rr   )r   r   r   r+   r   ru   r   r)   )rd   re   r   TreeClassifierr   TreeRegressorr   s          r   test_pure_setr     s    
bB8b"X1v1v1v>AA ) 1n!,13;;q>16G6N6Nt6TU !2
  )0m+1CKKNA7H7O7OPT7UV  1r   c            
         t        j                  g dg dg dg dg dg dg dg      } t        j                  g d      }t        j                  d	
      5  t        j	                         D ]Z  \  }} |d      }|j                  | |       |j                  | |        |j                  |  |       |j                  |  |        \ 	 d d d        y # 1 sw Y   y xY w)N)gs_c@d	a@籛 `8`@?c@)g_9a@g 8`@g-Vu]@g    @Xd@)gSW j_@r   r   r   )g ً`@4Ta@	lKa@{c@)g|@Y@g~G`a@gwI?lKa@g/"c@)g_@r   r   r   )g:^@r   r   r   )rR   gAw?gtQ?5??rW   g7G?gۺ?gb'?raise)allr   r   )rx   r   errstater   r   r   r   s        r   test_numerical_stabilityr     s    
DDDDDDD	

	A 	WXA		!#//+JD$A&CGGAqMGGArNGGQBNGGQBO , 
"	!	!s   A2CCc            	         t        j                  ddddddd      \  } }t        j                         D ]  \  }} |d      }|j	                  | |       |j
                  }t        j                  |dkD        }|j                  d   dk(  sJ d	j                  |             |dk(  rsJ d	j                  |              t        d      }|j	                  t        j                  t        j                         t        dt        t        j                        
      }|j	                  t        j                  t        j                         t        |j
                  |j
                         y )N  rG   r?   r   Fr\   r]   n_informativen_redundant
n_repeatedshuffler[   r   皙?r   r[   max_leaf_nodes)r	   make_classificationr   r   r   feature_importances_rx   r}   r   ru   r   rf   r   r   r   r+   )rd   re   r   r&   r   importancesn_importantclf2s           r   test_importancesr     s"   ''DAq  oo'
d"1..ff[3./  #r)I+<+C+CD+II)a?!2!9!9$!?? ( !a
0CGGDIIt{{#!qTYYPDHHTYY$s//1J1JKr   c                      t               } t        j                  t              5  t	        | d       d d d        y # 1 sw Y   y xY w)Nr   )r   r   raises
ValueErrorgetattr)r   s    r   test_importances_raisesr     s-    
 
"C	z	"+, 
#	"	"s	   :Ac            	         t        j                  ddddddd      \  } }t        ddd	      j                  | |      }t	        d
dd	      j                  | |      }t        |j                  |j                         t        |j                  j                  |j                  j                         t        |j                  j                  |j                  j                         t        |j                  j                  |j                  j                         t        |j                  j                  |j                  j                         y )Ni  rG   r?   r   Fr   r4   r>   )r   r   r[   r6   )r	   r   r   r   r   r)   r   r+   tree_rz   rw   rv   r|   )rd   re   r   r   s       r   )test_importances_gini_equal_squared_errorr     s     ''DAq !6QQ
O
S
S	1C  !QQ	c!Qi  00#2J2JKsyy((#))*;*;<syy..		0G0GHsyy//1I1IJsyy//1I1IJr   c                  V   t         j                         D ]  \  } } |d      }|j                  t        j                  t        j
                         |j                  t        t        j                  t        j                  j                  d               k(  sJ  |d      }|j                  t        j                  t        j
                         |j                  t        t        j                  t        j                  j                  d               k(  sJ  |d      }|j                  t        j                  t        j
                         |j                  dk(  sJ  |d      }|j                  t        j                  t        j
                         |j                  dk(  sJ  |d      }|j                  t        j                  t        j
                         |j                  dk(  sJ  |d      }|j                  t        j                  t        j
                         |j                  t        dt        j                  j                  d   z        k(  sJ  |d      }|j                  t        j                  t        j
                         |j                  t        j                  j                  d   k(  sJ  |d       }|j                  t        j                  t        j
                         |j                  t        j                  j                  d   k(  rJ  y )	Nsqrt)r   r<   log2r?   rV   rN   rR   )r:   r   r   rf   r   r   max_features_intrx   r   r   r  )r   TreeEstimatorests      r   test_max_featuresr    s   (0m0		4;;'  C		0B(C$DDDD0		4;;'  C		0B(C$DDDD+		4;;'  A%%%+		4;;'  A%%%.		4;;'  A%%%-		4;;'  Cdiiooa.@(@$AAAA-		4;;'  DIIOOA$6666.		4;;'  DIIOOA$6666?  1r   c                  J	   t         j                         D ]o  \  } } |       }t        j                  t              5  |j                  t               d d d        |j                  t        t               g dg}t        j                  t              5  |j                  |       d d d         |       }t        d d }t        j                  t              5  |j                  t        |       d d d        t        j                  t              } |       }|j                  |t               t        |j                  t              t                |       }t        j                  t              5  |j                  t               d d d        |j                  t        t               t        j                   t              }t        j                  t              5  |j                  |d d dd f          d d d        t        j"                  t              j                  } |       }|j                  t        j$                  t        |      t               t        j                  t              5  |j                  t               d d d        t        j                  t              5  |j'                  t               d d d         |       }|j                  t        t               t        j                  t              5  |j                  |       d d d        t        j                  t              5  |j'                  |       d d d         |       }t        j                  t              5  |j'                  t               d d d        r t)        d      }t        j                  t        d      5  |j                  g dgg d	       d d d        t        j                  t        d
      5  |j                  g dgg d       d d d        y # 1 sw Y   xY w# 1 sw Y   vxY w# 1 sw Y   ;xY w# 1 sw Y   xY w# 1 sw Y   OxY w# 1 sw Y   xY w# 1 sw Y   xY w# 1 sw Y   \xY w# 1 sw Y   6xY w# 1 sw Y   uxY w# 1 sw Y   xY w# 1 sw Y   y xY w)N)rE   rB   r<   rB   r<   r9   r   zy is not positive.*Poissonmatchr   r<   rP   )r   r   r   zSome.*y are negative.*Poisson)r>   grP   )r   r   r   r   r   r   rd   r   re   r   rx   asfortranarrayr)   r   r   r   asarrayr   dotapplyr   )	r   r  r  X2y2XftXtr   s	            r   
test_errorr  +  s   (0mo]]>*a  + 	1]]]:&b! ' osV]]:&GGArN ' q!oACKKNK8 o]]>*KKN + 	1JJqM]]:&KK!QR%! ' XXa[]]oq"q!]]:&KKN ']]:&IIaL ' o1]]:&KKO ']]:&IIbM ' o]]>*IIaL +*k  1r  )
4C	z)E	FY' 
G	z)H	I\* 
J	Is +*
 '& '& +* '& '&&&
 '&&&
 +*
 
G	F	I	Is   PPP%-P2P?+Q"Q:Q&-Q3'R /R*RP	P"	%P/	2P<	?Q		Q	Q#	&Q0	3Q=	 R
	RR"c                     t        j                  t        j                  t         j                        } t        j
                  }t        dt        j                               D ]  \  }}t        |   } |d|d      }|j                  | |       |j                  j                  |j                  j                  dk7     }t        j                  |      dkD  sJ dj                  |              |d	|d      }|j                  | |       |j                  j                  |j                  j                  dk7     }t        j                  |      dkD  rJ dj                  |              y
)z Test min_samples_split parameterdtypeN  rG   r   )min_samples_splitr   r[   rB   	   r   rA   N)rx   r  rf   r   float32r   r   r:   keysr   r   r|   rw   r   ru   )rd   re   r   r   r  r  node_sampless          r   test_min_samples_splitr   m  s(   
$))2::6AA !(inn6F G!$  a
 	1yy//		0G0G20MNvvl#a'G):)A)A$)GG' !.q
 	1yy//		0G0G20MNvvl#a'G):)A)A$)GG'+ !Hr   c                     t        j                  t        j                  t         j                        } t        j
                  }t        dt        j                               D ]  \  }}t        |   } |d|d      }|j                  | |       |j                  j                  |       }t        j                  |      }||dk7     }t        j                  |      dkD  sJ dj                  |              |d|d      }|j                  | |       |j                  j                  |       }t        j                  |      }||dk7     }t        j                  |      dkD  rJ dj                  |              y )	Nr  r  r>   r   )min_samples_leafr   r[   r;   r   r   )rx   r  rf   r   r  r   r   r:   r  r   r   r  bincountr   ru   )	rd   re   r   r   r  r  outnode_counts
leaf_counts	            r   test_min_samples_leafr'    s@   
$))2::6AA !(inn6F G!$ ~A
 	1iiooa kk#& !12
vvj!A%E'8'?'?'EE%  a
 	1iiooa kk#& !12
vvj!A%E'8'?'?'EE%/ !Hr   c                    t         |   d   j                  t        j                        }| ||      }t         |   d   }t        j                  |j                  d         }t        j                  |      }t        |    }t        dt        j                  ddd            D ]  \  }}	 ||	|d      }
|
j                  |||	       |*|
j                  j                  |j                               }n|
j                  j                  |      }t        j                  ||
      }||dk7     }t        j                   |      ||
j"                  z  k\  rJ dj%                  | |
j"                                |j                  d   }t        dt        j                  ddd            D ]  \  }}	 ||	|d      }
|
j                  ||       |*|
j                  j                  |j                               }n|
j                  j                  |      }t        j                  |      }||dk7     }t        j                   |      ||
j"                  z  k\  rJ dj%                  | |
j"                                y)zPTest if leaves contain at least min_weight_fraction_leaf of the
    training setrd   Nre   r   r  rN   rM   )min_weight_fraction_leafr   r[   r   )weightsz,Failed with {0} min_weight_fraction_leaf={1})DATASETSastyperx   r  rngrandr   r}   r:   r   linspacer   r   r  tocsrr#  r   r)  ru   )r   r	   sparse_containerrd   re   r*  total_weightr  r   fracr  r$  node_weightsleaf_weightss                 r   check_min_weight_fraction_leafr6    s-    	3&&rzz2A#Q3Ahhqwwqz"G66'?LdOM !(bkk!S!6L M%).WX
 	1G,'))//!''),C))//!$C{{38#LA$56vvl#|c6R6R'RR 	
:AAc22	
R !N* 771:L 'bkk!S!6L M%).WX
 	1'))//!''),C))//!$C{{3'#LA$56vvl#|c6R6R'RR 	
:AAc22	
R !Nr   r   c                     t        | d       y Nrf   r6  r   s    r   ,test_min_weight_fraction_leaf_on_dense_inputr;    s    "40r   csc_containerc                      t        | d|       y Nrl   )r1  r9  r   r<  s     r   -test_min_weight_fraction_leaf_on_sparse_inputr@    s     #4Vr   c                    t         |   d   j                  t        j                        }| ||      }t         |   d   }|j                  d   }t
        |    }t        dt        j                  ddd            D ]  \  }} |||dd	      }	|	j                  ||       |*|	j                  j                  |j                               }
n|	j                  j                  |      }
t        j                  |
      }||dk7     }t        j                  |      t        ||	j                  z  d      k\  rJ d
j!                  | |	j                  |	j"                                t        dt        j                  ddd            D ]  \  }} |||dd	      }	|	j                  ||       |*|	j                  j                  |j                               }
n|	j                  j                  |      }
t        j                  |
      }||dk7     }t        j                  |      t        ||	j                  z  ||	j"                  z        k\  rJ d
j!                  | |	j                  |	j"                                y)zzTest the interaction between min_weight_fraction_leaf and
    min_samples_leaf when sample_weights is not provided in fit.rd   Nre   r   r  rN   r?   r>   )r)  r   r"  r[   zBFailed with {0} min_weight_fraction_leaf={1}, min_samples_leaf={2}r   )r+  r,  rx   r  r   r:   r   r/  r   r   r  r0  r#  r   maxr)  ru   r"  )r   r	   r1  rd   re   r2  r  r   r3  r  r$  r4  r5  s                r   4check_min_weight_fraction_leaf_with_min_samples_leafrC    sD   
 	3&&rzz2A#Q3A771:LdOM 'bkk!S!6L M%))	
 	1'))//!''),C))//!$C{{3'#LA$56vvl#sC8881(
 
 	
OVV#..0D0D
	
 
% !N. !(bkk!S!6L M%)) 	
 	1'))//!''),C))//!$C{{3'#LA$56vvl#sC888C000(
 
 	
 PVV#..0D0D
	
 
% !Nr   c                     t        | d       y r8  rC  r:  s    r   Btest_min_weight_fraction_leaf_with_min_samples_leaf_on_dense_inputrF  0  s    8vFr   c                      t        | d|       y r>  rE  r?  s     r   Ctest_min_weight_fraction_leaf_with_min_samples_leaf_on_sparse_inputrH  5  s    
 9l]r   zTreeEstimator, criterionc                    t        j                  d|      \  }}dD ]  }dD ]  } | ||||      }|j                  ||       |j                  }|j                  |j
                  z  |j                  d   z  }	t        |j                        D ]R  }
|j                  |
   t        k7  s|j                  |
   }|j                  |
   }|	|
   |	|   |	|   z   z
  }||dz
  kD  rRJ    y )Nd   r\   r[   r  )rT   g-C6?r   )r   r   min_impurity_decreaser[   r   g|=)r	   r   r   r   r~   weighted_n_node_samplesr   rangert   rw   r    rv   )r  r   r   rd   re   r   expected_decreaser  r
   weighted_impuritynodeleftrightactual_decreases                 r   test_min_impurity_decreaserU  @  s    ''#DVWDAq '!4#-&7/	C GGAqM99D < <<qwwqzI  doo. %%d+y8--d3D //5E&7&=)$/2CE2JJ'O
 +->-FFFF / "5 'r   c            
         t         j                         D ]C  \  } }d| v r!t        j                  t        j                  }}n t
        j                  t
        j                  }} |d      }|j                  ||       |j                  ||      }g d}|D ci c]  }|t        |j                  |       }}t        j                  |      }	t        j                  |	      }
t        |
      |j                  k(  sJ |
j                  ||      }||k(  sJ dj                  |              |D ]-  }t!        t        |
j                  |      ||   d| d|         / F y	c c}w )
z8Test pickling preserves Tree properties and performance.
Classifierr   r   )r   rt   capacity	n_classesrw   rv   n_leavesrz   r{   r~   r|   rM  r   z6Failed to generate same score  after pickling with {0}z"Failed to generate same attribute z after pickling with rr   N)r:   r   rf   r   r   rg   r   r   r   r   pickledumpsloadstype	__class__ru   r+   )r   r  rd   re   r  r   
attributes	attributefitted_attributeserialized_objectest2score2s               r   test_picklerf  l  sM   (0m499dkkqA==(//qA+1		!Q

  GQ
FPIwsyy)44j 	 
 #LL-||-.DzS]]***Aq! 	
DKKDQ	
 *I

I. +8 Dv	 *M  14
s   EzTree, criterionc                    ddgddgddgddgddgddgddgddgddgddgddgddgg}t        j                  ddgddgddgddgddgddgddgddgddgddgddgddgg      }ddgddgddgddgg}t        j                  ddgddgddgddgg      }|t        v }|dk(  r ||dk  xx   dz  cc<   ||dk  xx   dz  cc<   |r | d|	      }|j                  ||      j	                  |      }t        ||       |j                  d
k(  sJ |j                  |      }	t        |	      dk(  sJ |	d   j                  d
k(  sJ |	d   j                  dk(  sJ |j                  |      }
t        |
      dk(  sJ |
d   j                  d
k(  sJ |
d   j                  dk(  sJ y  | d|	      }|j                  ||      j	                  |      }t        ||       |j                  d
k(  sJ y )NrE   rB   r<   rP   r   r?   r9   r;   )r[   r   r;   rP   )r;   r;   )rx   r   r   r   r   r+   r   r   r   r   r)   )r&   r   rd   re   r   y_trueis_clfr   y_hatproba	log_probar   s               r   test_multioutputrn    sm    
R	R	R	
A	
A	
A	Q	Q	Q	
B	
B	
B	A 	GGGFFFGGGFFF	
	A" bAq6B7QG,AXXAwAQ!Q89F.(FI	!q&	Q	v{q Y71%%a(5&){{f$$$!!!$5zQQx~~'''Qx~~'''))!,	9~"""|!!V+++|!!V+++ Y71%%a(E6*{{f$$$r   c                  d   t         j                         D ]  \  } } |d      }|j                  t        t               |j
                  dk(  sJ t        |j                  ddg       t        j                  t        t        j                  t              dz  f      j                  } |d      }|j                  t        |       t        |j
                        dk(  sJ t        |j                        dk(  sJ t        |j
                  ddg       t        |j                  ddgddgg        y )Nr   r   rP   rB   r<   rE   )r   r   r   rd   re   
n_classes_r+   classes_rx   r   r   r   r   )r   r   r   _ys       r   test_classes_shapers    s     ) 1n!,1~~"""3<<"a1 YY288A;?+,..!,23>>"a'''3<< A%%%3>>Aq623<<2q'B7);< !2r   c                     t         j                  d d } t         j                  d d }t        d|      }t        j                         D ]=  \  }} |d      }|j                  | ||       t        |j                  |       |       ? y )N}   balancedr   r   r   )	rf   r   r   r'   r   r   r   r)   r   )unbalanced_Xunbalanced_yr   r   r   r   s         r   test_unbalanced_irisry    sr    99Tc?L;;t$L)*lCM ) 1n!,l-HCKK5|D !2r   c                     t        t        j                         t        j                  t        j
                  g      D ]  \  \  } }} |d      }t        j                  t        j                  |      }t        j                  }t        |j                  ||      j                  |      |       t        j                  t        j                  d|      }t        j                  }t        |j                  ||      j                  |      |       t        j                  t        j                  d|      }t        j                  }t        |j                  ||      j                  |      |       t        j                  t        j                  |      }t        j                  }t        |j                  ||      j                  |      |       t        D ]U  } |t        j                  |      }t        j                  }t        |j                  ||      j                  |      |       W t        D ]U  } |t        j                  |      }t        j                  }t        |j                  ||      j                  |      |       W t        j                  t        j                  d d d   |      }t        j                  d d d   }t        |j                  ||      j                  |      |        y )Nr   r   r  C)orderr  Fr?   )r   r:   r   rx   float64r  r  rf   r   r   r+   r   r   ascontiguousarrayr1   r0   )r   r  r  r  rd   re   csr_containerr<  s           r   test_memory_layoutr    s   (/BJJ

3)$}u + JJtyy.KK3771a=003Q7 JJtyy59KK3771a=003Q7 JJtyy59KK3771a=003Q7   %8KK3771a=003Q7 ,Mdiiu5AAswwq!}44Q7; , ,Mdiiu5AAswwq!}44Q7; , JJtyy1~U3KK!3771a=003Q7Q)r   c                  z   t        j                  d      d d t         j                  f   } t        j                  d      }d|d d t        j                  d      }d||dk(  <   t	        d      }|j                  | ||       t        |j                  |       t        j                  d             t        j                  d      d d t         j                  f   } t        j                  d      }d|dd d	|dd d| dddf<   t        j                  d      }d
||d	k(  <   t	        dd      }|j                  | ||       |j                  j                  d   dk(  sJ d||d	k(  <   t	        dd      }|j                  | ||       |j                  j                  d   dk(  sJ t        j                  } t        j                  }t        j                  d| j                   d   d      }t	        d      }|j                  | |   ||          t        j"                  || j                   d         }t	        d      }|j                  | ||       |j                  j$                  t&        j(                  j*                  k7  }t-        |j                  j                  |   |j                  j                  |          y )NrJ  rW   2   r   r   r      r<   rP   gRQ?r   g     b@rN   g     H@)	minlength)rx   r   r   r   r   r   r+   r   rp   r   r{   rf   r   r   r-  randintr   r#  rw   r
   _treer    r*   )rd   re   r   r   
duplicatesr   r   s          r   test_sample_weightr  8  sA    			#q"**}%A
AAcrFGGCLMM!q&
 a
0CGGAqG.s{{1~rwws|4 			#q"**}%A
AAbIAc#JAc#gqjMGGCLM M!q&
 11
=CGGAqG.99q!U***M!q&
 11
=CGGAqG.99q!T))) 			AAQ
C0J
 a
0CGGAjM1Z=)KK
aggajAM!q1DHHQH/yy&&$***>*>>H		H%tzz';';H'Er   c                  D   t        j                  d      d d t         j                  f   } t        j                  d      }d|d d t	        d      }t         j
                  j                  dd      }t        j                  t              5  |j                  | ||       d d d        t        j                  d      }t        j                  d      }t        j                  t        |	      5  |j                  | ||       d d d        y # 1 sw Y   lxY w# 1 sw Y   y xY w)
NrJ  rW   r  r   r   r<   r   zgInput should have at least 1 dimension i.e. satisfy `len(x.shape) > 0`, got scalar `array(0.)` instead.r	  )rx   r   r   r   r   randomr.  r   r   r   r   r   reescape	TypeError)rd   re   r   r   expected_errs        r   test_sample_weight_invalidr  l  s    
		#q"**}%A
AAcrF
 a
0CIINN3*M	z	"1M2 
# HHQKM99BL 
y	51M2 
6	5 
#	" 
6	5s   
D
,D
DDc                    t         |    } |d      }|j                  t        j                  t        j                          |dd      }|j                  t        j                  t        j                         t        |j                  |j                         t        j                  t        j                  t        j                  t        j                  f      j                  } |ddddddddddddgd      }|j                  t        j                  |       t        |j                  |j                          |dd      }|j                  t        j                  |       t        |j                  |j                         t        j                  t        j                  j                        }|t        j                  dk(  xx   d	z  cc<   dd
dd} |d      }|j                  t        j                  t        j                  |        ||d      }|j                  t        j                  t        j                         t        |j                  |j                          |d      }|j                  t        j                  t        j                  |dz          ||d      }|j                  t        j                  t        j                  |       t        |j                  |j                         y )Nr   r   rv  class_weightr[          @rR   r  r<   rJ  g      Y@rP   )r   r   rf   r   r   r)   r   rx   r   r   r   r   )	r   r   clf1r   
iris_multiclf3clf4r   r  s	            r   test_class_weightsr    s    t_N q)DHHTYY$zBDHHTYY$1143L3LM DKKdkkBCEEJ$$$

 D 	HHTYY
#1143L3LMzBDHHTYY
#1143L3LM GGDKK--.M$++"#s*#u-Lq)DHHTYY]3|!DDHHTYY$1143L3LM q)DHHTYY]A%56|!DDHHTYY]31143L3LMr   c                 @   t         |    }t        j                  t        t        j                  t              dz  f      j
                  } |dddgd      }d}t        j                  t        |      5  |j                  t        |       d d d        y # 1 sw Y   y xY w)	NrP   rN   rR   rB   r<   r   r  zBnumber of elements in class_weight should match number of outputs.r	  )r   rx   r   re   r   r   r   r   r   r   rd   )r   r   rr  r   rs   s        r   test_class_weight_errorsr    st     t_N	Arxx{Q'	(	*	*B CC'8&9
JCRG	z	12 
2	1	1s   4BBc                      t        j                  dd      \  } }d}t        j                         D ]:  \  }} |d |dz         j	                  | |      }|j                         |dz   k(  r:J  y NrJ  r<   rK  r;   )r   r   )r	   make_hastie_10_2r:   r   r   get_n_leavesrd   re   kr   r  r  s         r   test_max_leaf_nodesr    sk    $$sCDAq	A(0md1q5AEEaK!QU***  1r   c                      t        j                  dd      \  } }d}t        j                         D ]4  \  }} |d|      j	                  | |      }|j                         dk(  r4J  y r  )r	   r  r:   r   r   	get_depthr  s         r   test_max_leaf_nodes_max_depthr    s`    $$sCDAq	A(0ma:>>q!D}}!###  1r   c                      dD ]]  } t        t               j                  dgdggddg      j                  |       }d|j                  d   cxk  rdk  rPJ d        J d        y )N)rY  r   rw   rv   r{   r~   rz   r|   r   r<   rF   r?   z Array points to arbitrary memory)r   r   r   r   flat)attrr   s     r   test_arrays_persistr    sm    	 .044qcA3Z!QHNNPTUUZZ]&Q&J(JJ&J(JJ&	r   c                     t        d      } t        j                  d      }| j                  ddd      }t        j                         D ];  \  }} |d      }|j                  ||       |j                  j                  dk(  r;J  y )Nr   )rG   r^   rP   rG   r   )	r3   rx   rp   r  r:   r   r   r   r   )r[   rd   re   r   r  r  s         r   test_only_constant_featuresr    ss    %a(L
AQ5)A(0m+1yy""a'''  1r   tree_clsc                    t        d      }|j                  dd      }d}|d d dfxx   |z  cc<   |j                  ddd      } | d      }|j                  ||       |j                  d   dk(  sJ |j                  d   dkD  sJ y )Nr   rG   rP   gHz>r  r   r<   )r3   r.  r  r   r   )r  r[   rd   feature_thresholdre   r  s         r   test_almost_constant_featurer    s    
 &a(L"a A adG  GQ5)A

"CGGAqM##A&!+++##A&***r   c                  r   t        j                  t        j                  g dgt        j                  d      f            } g d}t        j                         D ]\  \  }}d|vs |dd      }|j                  | |       |j                  j                  dk(  sJ |j                  j                  d	k(  r\J  y )
N)r   r   r   r   r   r<   rP   r;   r>   rM   r   )r;   rL   )r   r   r   r<   r<   rP   rP   rP   r?   r?   r?   	ExtraTreer   r<   r   rP   r>   )
rx   	transposer   rp   r:   r   r   r   r   rt   rd   re   r   r  r  s        r   ,test_behaviour_constant_feature_after_splitsr    s    

		568IJK	A 	*A(0md"QQ?CGGAqM99&&!+++99''1,,,  1r   c                     t        j                  t        j                  dgdgdgdgg      t        j                  d      g      } t        j                  g d      }t        j                         D ]k  \  }} |dd      }|j                  | |       |j                  j                  dk(  sJ t        |j                  |       t        j                  dd	             m t        j                         D ]k  \  }} |dd      }|j                  | |       |j                  j                  dk(  sJ t        |j                  |       t        j                  d
d	             m y )NrR   rW   )r;   r  )rW   rR   rW   rR   r   r<   r   rh  rN   )r;   )rx   hstackr   rp   r   r   r   r   r   r+   r   r   r   r   r  s        r   (test_with_only_one_non_constant_featuresr    s   
		288cUSEC53%89288I;NOPA
%&A(0m;1yy""a'''3,,Q/1EF	  1  )0m;1yy""a'''3;;q>2774+=>	  1r   c                  &   t        j                  dd      j                  t         j                        j	                  dd      } t               }t        j                  t        d      5  |j                  | g d       d d d        y # 1 sw Y   y xY w)Ng\)c=Hr;   rB   r<   r  r	  )r   r<   r   r<   )
rx   repeatr,  r~  reshaper   r   r   r   r   )rd   r   s     r   test_big_inputr  #  s^    
		(A%%bjj199"a@A
 
"C	z	3<  
4	3	3s   )BBc                  z    ddl m}  t        j                  t              5   |         d d d        y # 1 sw Y   y xY w)Nr   _realloc_test)sklearn.tree._utilsr  r   r   MemoryErrorr  s    r   test_reallocr  +  s"    1	{	# 
$	#	#s   1:c                     dt        j                  d      z  } t        j                  j	                  dd      }t        j                  j                  ddd      }d| dz   z  }t        d|      }t        j                  t              5  |j                  ||       d d d        d| dz
  z  dz
  }t        d|      }t        j                  t              5  |j                  ||       d d d        y # 1 sw Y   VxY w# 1 sw Y   y xY w)	NrO   PrG   rP   r   r<   best)splitterr   )structcalcsizerx   r  randnr  r   r   r   	Exceptionr   r  )n_bitsrd   re   huger   s        r   test_huge_allocationsr  2  s    %%F
		AA
		!Q#A !D
 &
FC	y	!1 
"
 !q D
 &
FC	{	#1 
$	# 
"	! 
$	#s   C0C<0C9<Dc                     t         |    }t        |   d   }t        |   d   }|dv r|j                  d   dz  }|d | }|d | }t        t        z   t
        z   D ]5  } ||      } |d|      j                  ||      }	 |d|      j                  ||      }
t        |	j                  |
j                  dj                  |              |	j                  |      }| t        v r"|	j                  |      }|	j                  |      }t        t
        z   t        z   D ]t  } ||t        j                        }t!        |
j                  |      |       | t        v s?t!        |
j                  |             t!        |
j                  |             v 8 y )	Nrd   re   )rh   rg   r   r>   r[   r   5{0} with dense and sparse format gave different treesr  )r:   r+  r   r/   r0   r1   r   r   r   ru   r   r   r   r   rx   r  r*   )r
   datasetr   r  rd   re   r\   r1  X_sparser   r   y_predy_probay_log_probasparse_container_testX_sparse_tests                   r   check_sparse_inputr  G  st   dOM#A#A ((GGAJ!O	jyMjyM*^;nL#A& qI>BB1aHqI>BB8QOGGGGCJJ4P	
 19ooa(G--a0K%3n%D~%U!1("**MM%aii&>Gy )!//-*H'R)''6 &V% Mr   	tree_typer  )rj   ri   rh   rl   rm   rn   ro   rp   c                 0    |dk(  rdnd }t        | ||       y )Nrh   r?   r  )r  r  r   s      r   test_sparse_inputr  p  s     (dIy'95r   rg   rk   c                     t        | |d       y )NrP   r  )r  r  s     r   test_sparse_input_reg_treesr    s    
 y'1-r   )rm   rn   ro   rp   c                    t         |    }t        |   d   } ||      }t        |   d   } |ddd      j                  ||      } |ddd      j                  ||      }t        |j                  |j                  dj                  |              t        |j                  |      |j                  |              |ddd	      j                  ||      } |ddd	      j                  ||      }t        |j                  |j                  dj                  |              t        |j                  |      |j                  |              |d|j                  d   dz  
      j                  ||      } |d|j                  d   dz  
      j                  ||      }t        |j                  |j                  dj                  |              t        |j                  |      |j                  |              |dd      j                  ||      } |dd      j                  ||      }t        |j                  |j                  dj                  |              t        |j                  |      |j                  |             y )Nrd   re   r   r<   rP   )r[   r   r   r  rG   )r[   r   r  )r[   r"  r?   r   )	r:   r+  r   r   r   ru   r*   r   r   )	r  r  r<  r  rd   r  re   r   r   s	            r   test_sparse_parametersr    s9    i(M#AQH#A 	11BFFq!LA11BFFxQRSA		?FFyQ
 aiilAIIaL9 	11KOOPQSTUA11KOO!	A 		?FFyQ
 aiilAIIaL9 	1x~~a7HA7MNRRSTVWXA1x~~a7HA7MNRR!	A 		?FFyQ
 aiilAIIaL9 	1Q7;;AqAA1Q7;;HaHA		?FFyQ
 aiilAIIaL9r   ztree_type, criterionc                 v   t         |    }t        |   d   } ||      }t        |   d   } |dd|      j                  ||      } |dd|      j                  ||      }	t        |j                  |	j                  dj                  |              t        |	j                  |      |j                  |             y )Nrd   re   r   r?   r[   r   r   r  )r:   r+  r   r   r   ru   r*   r   )
r  r  r<  r   r  rd   r  re   r   r   s
             r   test_sparse_criteriar    s     i(M#AQH#A1YGKKAqQA1YGKKHVWXA		?FFyQ
 aiilAIIaL9r   zcsc_container,csr_containerc                    t         |    }d}d}|}t        j                  |      }t        d      }g }	g }
d}|g}t	        |      D ]x  }|j                  |d      }|j                  |      d | }|	j                  |       |j                  dd|f      dz
  }|
j                  |       ||z  }|j                  |       z t        j                  |	      j                  t        j                        }	t        j                  |t        j                        }t        j                  t        j                  |
      t        j                        }
 ||
|	|f||f      }|j                         } ||
|	|f||f      }|j                         }|j                  dd|f      }|j                         }|j                   d	k(  j#                         dkD  sJ |j                   d	k(  j#                         dkD  sJ  |d|
      j%                  ||      } |d|
      j%                  ||      }t'        |j(                  |j(                  dj+                  t,                     ||f}t/        ||      D ]  \  }}t1        |j(                  j3                  |      |j(                  j3                  |             t1        |j3                  |      |j3                  |             t1        |j3                  |      |j(                  j3                  |             t1        |j(                  j5                  |      j                         |j(                  j5                  |      j                                t1        |j5                  |      j                         |j5                  |      j                                t1        |j5                  |      j                         |j(                  j5                  |      j                                t1        |j7                  |      |j7                  |             t,        t8        v st1        |j;                  |      |j;                  |              y )Nr?   rG   r   rN   r_   r<   r  r   rW   r  r  )r:   rx   r   r3   rN  binomialpermutationappendconcatenater,  int32r   r  toarrayr  copyr   r}   r   r   r   ru   r
   r   r*   r  decision_pathr   r   r   )r  r<  r  r  r   r]   r\   samplesr[   r   r   offsetindptrin_nonzero_i	indices_idata_ir  rd   r  X_testre   r   r   XsX1r  s                              r   test_explicit_sparse_zerosr    s   
 i(MIJ Iii	"G &a(LGDFXF:"++Is; ,,W5l{C	y!&&q#[N&CaGF+f  nnW%,,RXX6GXXfBHH-F88BNN4(

;DdGV4Y
<STHA!	w	:'>M ""$FQ5A "&&(M MMS %%'!+++#%**,q000 	1	:>>q!DA1	:>>xKA		?FFtL -	 B"b/B!!''--"3QWW]]25FG!!''"+qwwr{;!!''"+qww}}R/@A!GG!!"%--/1F1Fr1J1R1R1T	
 	"OOB'')1??2+>+F+F+H	
 	"OOB'')177+@+@+D+L+L+N	
 	"!))B-2?9%aoob&91??2;NO% "r   c                    t         |    }t        j                  d d df   j                         }t        j                  d d df   j	                  d      }t        j
                  }t        j                  t              5   |d      j                  ||       d d d         |d      }|j                  ||       t        j                  t              5  |j                  |g       d d d        y # 1 sw Y   YxY w# 1 sw Y   y xY w)Nr   r  r   )r:   rf   r   r   r  r   r   r   r   r   r   )r   r  rd   X_2dre   r  s         r   check_raise_error_on_1d_inputr  ,  s    dOM		!Q$A99QT?""7+DA	z	"1%))!Q/ 
# Q
'CGGD!	z	"QC 
#	" 
#	"
 
#	"s   >C0C<0C9<Dc                 X    t               5  t        |        d d d        y # 1 sw Y   y xY wN)r-   r  r:  s    r   test_1d_inputr  <  s    		%d+ 
		s    )r1  c                 Z   t         |    }t        j                  dgdgdgdgdgg      }g d}g d}| ||      } |d      }|j                  |||       |j                  j
                  dk(  sJ  |dd      }|j                  |||       |j                  j
                  dk(  sJ y )	Nr   r<   )r   r   r   r   r<   )rA   rA   rA   rA   rA   r   r   g?)r[   r)  )r:   rx   r   r   r   r   )r   r1  r  rd   re   r   r  s          r    test_min_weight_leaf_split_levelr  B  s     dOM
1#sQC!qc*+AA-M#Q
Q
'CGGAqG.99!###
Q
ECGGAqG.99!###r   c                    t         j                  t        j                  d      }t	        |           }|j                  t         t               t        |j                  t               |j                  j                  |             y NFr  
X_smallr,  rx   r  r:   r   y_smallr+   r  r   )r   	X_small32r  s      r   test_public_apply_all_treesr	  V  sR    rzz6I
D/
CGGGWsyy)399??9+EFr   r  c                     |t         j                  t        j                  d            }t	        |           }|j                  t         t               t        |j                  t               |j                  j                  |             y r  r  )r   r  r  r  s       r   test_public_apply_sparse_treesr  _  sY     gnnRZZenDEI
D/
CGGGWsyy)399??9+EFr   c                  V   t         j                  } t         j                  }t        dd      j	                  | |      }|j                  | d d       j                         }t        |g dg dg       dt        j                  t        j                  ddg} g d}t        j                  |       j                  d	d      } t        d
      j	                  | |      } |j
                  |       j                         j                  d      }t        ||j                  j                         y )Nr   r<   r  rP   )r<   r<   r   r<   r   r<   r?   r   r   r   r<   r<   rB   r   axis)rf   r   r   r   r   r  r  r+   rx   nanr   r  r   r}   r   r|   )rd   re   r  node_indicatorr
   r|   s         r   test_decision_path_hardcodedr  i  s    		AA
 a1
=
A
A!Q
GC&&q!u-557N~	9'=>
 
BFFBFFAq!AA
B"A a044Q:D'T''*22488a8@N~tzz'@'@Ar   c                    t         j                  }t         j                  }|j                  d   }t        |    } |dd      }|j                  ||       |j                  |      }|j                         }|j                  ||j                  j                  fk(  sJ |j                  |      }t        |      D 	
cg c]  \  }	}
||	|
f    }}	}
t        |t        j                  |             |j                  j                  t         k(  }t        t        j"                  ||      t        j                  |             |j%                  d      j'                         }|j                  j(                  |k  sJ y c c}
}	w )Nr   rP   r  r  r<   r  )rf   r   r   r   r:   r   r  r  r   rt   r  	enumerater*   rx   r   rw   r    r  r}   rB  r   )r   rd   re   r\   r  r  node_indicator_csrr  leavesr  jleave_indicator
all_leavesr   s                 r   test_decision_pathr  |  sB   		AA
IdOM
Q!
4CGGAqM**1-'//1NIsyy/C/C#DDDD YYq\F8A&8IJ8I1~ad+8IOJorwwY/GH ((I5J
~z*BGG),D
 """*..0I99)+++ Ks   <E=c                     t          |t              }}t        |    }t        j                  t
              5   |d      j                  ||       d d d        y # 1 sw Y   y xY wNr   r   )X_multilabely_multilabelr:   r   r   r  r   )r   r  rd   re   r  s        r   test_no_sparse_y_supportr     sH     |4qAdOM	y	!1%))!Q/ 
"	!	!s   AA!c                     t        ddd      } | j                  dgdgdgdgdggg dg d	
       t        | j                  j                  g d       t        | j                  j                  j                  g d       | j                  dgdgdgdgdggg dt        j                  d      
       t        | j                  j                  g d       t        | j                  j                  j                  g d       | j                  dgdgdgdgdggg d       t        | j                  j                  g d       t        | j                  j                  j                  g d       t        ddd      } dgdgdgdgdgg}| j                  |g dg d
       t        | j                  |      g d       t        | j                  j                  g d       t        | j                  j                  j                  g d       y)aP  Check MAE criterion produces correct results on small toy datasets:

    ## First toy dataset
    ------------------
    | X | y | weight |
    ------------------
    | 3 | 3 |  0.1   |
    | 5 | 3 |  0.3   |
    | 8 | 4 |  1.0   |
    | 3 | 6 |  0.6   |
    | 5 | 7 |  0.3   |
    ------------------
    |sum wt:|  2.3   |
    ------------------

    Because we are dealing with sample weights, we cannot find the median by
    simply choosing/averaging the centre value(s), instead we consider the
    median where 50% of the cumulative weight is found (in a y sorted data set)
    . Therefore with regards to this test data, the cumulative weight is >= 50%
    when y = 4.  Therefore:
    Median = 4

    For all the samples, we can get the total error by summing:
    Absolute(Median - y) * weight

    I.e., total error = (Absolute(4 - 3) * 0.1)
                      + (Absolute(4 - 3) * 0.3)
                      + (Absolute(4 - 4) * 1.0)
                      + (Absolute(4 - 6) * 0.6)
                      + (Absolute(4 - 7) * 0.3)
                      = 2.5

    Impurity = Total error / total weight
             = 2.5 / 2.3
             = 1.08695652173913
             ------------------

    From this root node, the next best split is between X values of 3 and 5.
    Thus, we have left and right child nodes:

    LEFT                    RIGHT
    ------------------      ------------------
    | X | y | weight |      | X | y | weight |
    ------------------      ------------------
    | 3 | 3 |  0.1   |      | 5 | 3 |  0.3   |
    | 3 | 6 |  0.6   |      | 8 | 4 |  1.0   |
    ------------------      | 5 | 7 |  0.3   |
    |sum wt:|  0.7   |      ------------------
    ------------------      |sum wt:|  1.6   |
                            ------------------

    Impurity is found in the same way:
    Left node Median = 6
    Total error = (Absolute(6 - 3) * 0.1)
                + (Absolute(6 - 6) * 0.6)
                = 0.3

    Left Impurity = Total error / total weight
            = 0.3 / 0.7
            = 0.428571428571429
            -------------------

    Likewise for Right node:
    Right node Median = 4
    Total error = (Absolute(4 - 3) * 0.3)
                + (Absolute(4 - 4) * 1.0)
                + (Absolute(4 - 7) * 0.3)
                = 1.2

    Right Impurity = Total error / total weight
            = 1.2 / 1.6
            = 0.75
            ------

    ## Second toy dataset:
    ------------------
    | X | y | weight |
    ------------------
    | 1 | 1 |   3    |
    | 2 | 1 |   3    |
    | 3 | 3 |   2    |
    | 4 | 1 |   1    |
    | 5 | 2 |   2    |
    ------------------
    |sum wt:|   11   |
    ------------------

    The weighted median is 1
    Total error = Absolute(1 - 3) * 2 + Absolute(1 - 2) * 2 = 6

    The best split is between X values of 2 and 3, with:
    - left node being the first 2 data points, both with y=1
      => AE and impurity is 0
    - right node being the last 3 data points, weighted median is 2.
      Total error = (Absolute(2 - 3) * 2)
                  + (Absolute(2 - 1) * 1)
                  + (Absolute(2 - 2) * 2)
                  = 3
    r   r7   rP   )r[   r   r   r?   r>   rO   )rM   r   r?   r;   r?   )333333?333333?r   rR   r#  )rd   re   r   )g,d?gܶm۶m?g?)      @g      @r$  )ffffff?rQ   gUUUUUU?)r;   rX   r$  rc   r<   )r[   r   r   r;   )r<   r<   r?   r<   rP   )r?   r?   rP   r<   rP   )r<   r<   rP   rP   rP   )gtE]t?r   r"  )r<   r<   rP   N)r   r   r   r   r~   r+   r   r  rx   r   r   )dt_maerd   s     r   test_maer'    s   H #"21F
 JJ3aS1#s
#
/  
 FLL))+LMv||))..@ JJ1#sQC!qc*oRWWUVZJXv||,,.CDv||))..>
 JJ1#sQC!qc*oJ>v||,,.CDv||))..>""F
 qcA3aS!A
JJ

%  
 FNN1%7FLL))+=>v||))..	:r   c                     d} t        j                  dt         j                        }d}d }t        j                  t        j                  |fD ]  }t        j                         D ]G  \  }} || |      } ||      j                         }|\  }	\  }
}}||	k(  sJ | |
k(  sJ t        ||       I t        j                         D ]B  \  }} || |      } ||      j                         }|\  }	\  }
}}||	k(  sJ | |
k(  sJ ||k(  rBJ   y )Nr?   r  rJ  c                 R    t        j                  t        j                  |             S r  )r[  r]  r\  )objs    r   _pickle_copyz)test_criterion_copy.<locals>._pickle_copy8  s    ||FLL-..r   )
rx   r   intpr  deepcopyr   r   
__reduce__r+   r   )	n_outputsrY  r\   r+  	copy_func_typenamecriteriaresult	typename_
n_outputs_rp  
n_samples_s                r   test_criterion_copyr8  1  s"    I		!277+II/ ii=	'--/KAx	95Hx(335F5;2I/
Jy(((
***y*5 0 (--/KAx	95Hx(335F5;2I/
Jy(((
***
*** 0 >r   c                    t         j                  j                  d      j                  dd      dz  }t	        j
                  |j                  d            }|d d d df   }|  | |      }|d d df   }t        d      j                  ||      } |j                  |      }t        t        j                  |j                  j                  t        k(        d         }|j                  |      }t        j                  t        j                   |j                  j"                               d   }t%        |      dk(  sJ t%        |      dk(  sJ y )Nr   rJ  rL   g*Gr  rB   r   )rx   r  RandomStater  r(   
nan_to_numr,  r   r   r  setwherer   rw   r    
differenceisfiniter{   r   )	r1  r   rd   re   r
   terminal_regions	left_leaf
empty_leafinfinite_thresholds	            r   "test_empty_leaf_infinite_thresholdrD  M  s    99  #))#r2T9D>>$++i01DQVA#QQUA a044Q:D!tzz!}BHHTZZ55BCAFGI%%&67J2;;tzz/C/C#D"DEaH!"a'''z?ar   c                 b   t         |    } | d   | d   }} |dd      }|j                  ||      }|j                  }|j                  }t	        j
                  t	        j                  |      dk\        sJ t	        j
                  t	        j                  |      dk\        sJ t        ||||       y Nrd   re   r^   r   )r   r[   r+  cost_complexity_pruning_path
ccp_alphas
impuritiesrx   r   diffassert_pruning_creates_subtreer  r  rd   re   r  infopruning_pathrJ  s           r   'test_prune_tree_classifier_are_subtreesrP  `  s    
 wG3<qA
"1
5C++Aq1D??LJ66"'','1,---66"''*%*+++"8Q<@r   c                 b   t         |    } | d   | d   }} |dd      }|j                  ||      }|j                  }|j                  }t	        j
                  t	        j                  |      dk\        sJ t	        j
                  t	        j                  |      dk\        sJ t        ||||       y rF  rG  rM  s           r   'test_prune_tree_regression_are_subtreesrR  r  s     wG3<qA
"1
5C++Aq1D??LJ66"'','1,---66"''*%*+++"8Q<@r   c                      t        d      } | j                  dgdggddg       t        dd      }|j                  dgdggddg       t        | j                  |j                         y )Nr   r   r<   rG   )r[   	ccp_alpha)r   r   assert_is_subtreer   )r  r   s     r   test_prune_single_node_treerV    s`    !q1DHHqcA3Z!Q  "qB?DHHqcA3Z!Q djj$**-r   c                     g }|D ].  } | d|d      j                  ||      }|j                  |       0 t        |      D ]%  \  }}t        |j                  |j                         ' y )Nr^   r   )r   rT  r[   )r   r  r   rU  r   )	estimator_clsrd   re   rO  
estimatorsrT  r  prev_estnext_ests	            r   rL  rL    sh    J!	2QRSWWq
 	#	 " 'z2((..(..9 3r   c                 >   | j                   |j                   k\  sJ | j                  |j                  k\  sJ | j                  }| j                  }|j                  }|j                  }dg}|r1|j	                         \  }}t        | j                  |   |j                  |          t        | j                  |   |j                  |          t        | j                  |   |j                  |          t        | j                  |   |j                  |          ||   ||   k(  rt        t        |j                  |          nXt        | j                  |   |j                  |          |j                  ||   ||   f       |j                  ||   ||   f       |r0y y )N)r   r   )rt   r   rw   rv   popr*   r   r)   r~   r|   rM  r!   r{   r  )	r
   subtreetree_c_lefttree_c_rightsubtree_c_leftsubtree_c_rightstacktree_node_idxsubtree_node_idxs	            r   rU  rU    s   ??g00000>>W.....$$K&&L**N,,OHE
*/))+''!JJ}%w}}5E'F	
 	MM-('*:*:;K*L	
 	.0F0FGW0X	
 	((7++,<=	

 *+?O/PP0A0ABR0ST  }-w/@/@AQ/R LL+m4nEU6VWXLLm,o>N.OP3 r   r  r  r  c                    t         d   }|d   j                  t        j                  d      }|t	        |      }n~ ||d         }t        j
                  |j                  t        j                        |_        t	        |j                  |j                  |j                  f      \  |_        |_        |_        t	        t        j
                  t        t        j                              }t        |    |      }|j                  ||       t        |j                  |      |j                  |             t        |j                  |      j                         |j                  |      j                                y )Nrj   rd   Fr  r  )r  )r+  r,  rx   r  r,   r   r   r   r  r  r:   r   r+   r   r  todense)r   r  r1  r  r  
X_readonly
y_readonlyr  s           r   "test_apply_path_readonly_all_treesrj    s%    {#Gcl!!"**5!9G.w7
%gcl3
((:??"**E

 &__j00*2C2CD
		
O
 +288G2::+NOJ
D/8
,CGGJ
#s{{:.G0DE*%--/1B1B71K1S1S1Ur   )r6   r8   r9   c                    t         j                  t         j                  }} ||       }|j                  ||       t	        j
                  |j                  |            t        j                  t	        j
                  |            k(  sJ y )Nr  )	rg   r   r   r   rx   r}   r   r   r   )r   r&   rd   re   r   s        r   test_balance_propertyrl    s\     ==(//qA

#CGGAqM66#++a.!V]]266!9%====r   seedc           	         ddgddgddgddgddgddgddgddgg}g d}t        d|       }|j                  ||       t        j                  |j	                  |            dk(  sJ t        d|       }|j                  ||       t        j
                  |j	                  |      dkD        sJ d	}t        j                  |dz  dz  d
d||dz  dz  |       \  }}d|d|k  |dk  z  <   t        j                  |      }t        d|       }|j                  ||       t        j
                  |j	                  |      dkD        sJ y )Nr   r<   rP   r?   r   r   r   r   r<   rP   r?   r;   r6   r   r9   rG   r"  r  )effective_ranktail_strengthr\   r]   r   r[   rB   )	r   r   rx   aminr   r   r	   make_regressionr   )rm  rd   re   r   r]   s        r   test_poisson_zero_nodesrt    sN    Q!Q!Q!Q!Q!Q!Q!QHA A  /
MCGGAqM773;;q>"a'''
)$
GCGGAqM66#++a.1$%%% J##!A~* 1n)DAq ArAv!a%
q	A
)$
GCGGAqM66#++a.1$%%%r   c            	      0   t         j                  j                  d      } d\  }}}t        j                  ||z   ||       }| j                  dd|      t        j                  |d      z  }| j                  t        j                  ||z        	      }t        |||| 
      \  }}}	}
t        dd|       }t        dd|       }|j                  ||	       |j                  ||	       t        d      j                  ||	      }||	df||
dffD ]  \  }}}t        ||j                  |            }t        |t        j                  |j                  |      dd             }t        ||j                  |            }|dk(  r
|d|z  k  sJ |d|z  k  rJ  y )Nr   )  rv  rG   r\   r]   r[   rE   rP   )lowhighr`   r   r  )lam)	test_sizer[   r9   rG   )r   r  r[   r6   mean)strategytraintestgV瞯<rN   g      ?)rx   r  r:  r	   make_low_rank_matrixuniformrB  r9   r   r   r   r   r   r   r   clip)r-  n_trainn_testr]   rd   coefre   X_trainr  r   r   tree_poitree_msedummyval
metric_poi
metric_msemetric_dummys                     r   test_poisson_vs_mser    s   
 ))


#C".GVZ%%F"z	A
 ;;2AJ;7"&&:KKDq4x()A'7	1S($GVWf %rH %!RcH LL'"LL'"F+//AE1FFF3KL	1c*1h.>.>q.AB
*1bggh6F6Fq6I5RV.WX
,Qa0@A &=j 0000D<//// Mr   rY  c                 N   d\  }}t        j                  ||||dd      \  }} | dd      j                  ||      } | dd      j                  ||      }t        |j                  |j                  | d	       t        |j                  |      |j                  |             y
)z3Test that criterion=entropy gives same as log_loss.)r  r>   r   r   )rY  r\   r]   r   r   r[   r5   +   r   entropyz> with criterion 'entropy' and 'log_loss' gave different trees.N)r	   r   r   r   r   r   r   )r&   rY  r\   r]   rd   re   tree_log_losstree_entropys           r   'test_criterion_entropy_same_as_log_lossr  ?  s     "Iz'' DAq :B?CCAqIM)"=AA!QGL(PQ
 M))!,l.B.B1.EFr   c                  6   t        j                  d      \  } }t        dd      j                  | |       j	                  | |      }d fd}t        j                   |             }|j	                  | |      }t        j                  ||      sJ y )Nr   r   r?   r  c                     | j                         j                  | j                  j                               j	                         S r  )byteswapviewr  newbyteorderr.  )arrs    r   reduce_ndarrayz8test_different_endianness_pickle.<locals>.reduce_ndarray^  s/    ||~""399#9#9#;<GGIIr   c                     t        j                         } t        j                  |       }t        j
                  j                         |_        |j
                  t        j                  <   |j                         | j                  d       | S Nr   )ioBytesIOr[  Picklercopyregdispatch_tabler  rx   ndarraydumpseek)fpr   r  s     r    get_pickle_non_native_endiannesszJtest_different_endianness_pickle.<locals>.get_pickle_non_native_endiannessa  sb    JJLNN1"11668'5$	s	q	r   )	r	   r   r   r   r   r[  loadrx   isclose)rd   re   r   r  new_clf	new_scorer   r  s         @@r    test_different_endianness_pickler  W  s    ''Q7DAq
 a1
=CGGAqMIIaOEJ kk:<=Ga#I::eY'''r   c                  N   t        j                  d      \  } }t        dd      j                  | |       j	                  | |      } G d dt
              fd}t        j                   |             }|j	                  | |      }t        j                  ||      sJ y )Nr   r   r?   r  c                        e Zd Z fdZ xZS )Ptest_different_endianness_joblib_pickle.<locals>.NonNativeEndiannessNumpyPicklerc                     t        |t        j                        r7|j                         j	                  |j
                  j                               }t        | !  |       y r  )	
isinstancerx   r  r  r  r  r  supersave)selfr*  r_  s     r   r  zUtest_different_endianness_joblib_pickle.<locals>.NonNativeEndiannessNumpyPickler.savex  s@    #rzz*lln))#))*@*@*BCGLr   )__name__
__module____qualname__r  __classcell__)r_  s   @r   NonNativeEndiannessNumpyPicklerr  w  s    	 	r   r  c                      t        j                         }  |       }|j                         | j                  d       | S r  )r  r  r  r  )r  r  r  r   s     r   'get_joblib_pickle_non_native_endiannesszXtest_different_endianness_joblib_pickle.<locals>.get_joblib_pickle_non_native_endianness}  s3    JJL+A.	s	q	r   )
r	   r   r   r   r   r   joblibr  rx   r  )rd   re   r   r  r  r  r  r   s         @@r   'test_different_endianness_joblib_pickler  p  s    ''Q7DAq
 a1
=CGGAqMIIaOE,  kkACDGa#I::eY'''r   c                    t         rt        j                  nt        j                  }g d}| j                  j
                  j                         D ci c]  \  }\  }}|| }}}}|D ]  }|||<   	 t        j                  t        |j                               t        |j                               d      }| j                  |d      S c c}}}w )N)
left_childright_childrz   r|   namesformats	same_kindcasting)r.   rx   int64r  r  fieldsr   listr  valuesr,  )node_ndarraynew_dtype_for_indexing_fieldsindexing_field_namesr   r  r1  new_dtype_dict	new_dtypes           r   "get_different_bitness_node_ndarrayr    s    09BHHrxx! V -9,>,>,E,E,K,K,M,M(juae,M   %<t % ~**,-$~?T?T?V:WXI y+>>s   Cc                    | j                   j                  j                         D ci c]  \  }\  }}|| }}}}| j                   j                  j                         D cg c]  \  }}|	 }}}|D cg c]  }d|z   	 }}t	        j                   t        |j                               t        |j                               |d      }| j                  |d      S c c}}}w c c}}w c c}w )NrO   )r  r  offsetsr  r  )r  r  r   r  rx   r  r  r,  )	r  r   r  r1  r  r  r  shifted_offsetsr  s	            r   $get_different_alignment_node_ndarrayr    s    ,8,>,>,E,E,K,K,M,M(juae,M   ,8+=+=+D+D+K+K+MN+M-%v+MGN078fq6zO8.--/0N1134&	
I y+>> O8s   C$C&7C,c                     t         rt        j                  nt        j                  } | j                         \  }\  }}}}|j                  |d      }|j                         }t        |d         |d<   ||||f|fS )Nr  r  nodes)r.   rx   r  r  r.  r,  r  r  )	r
   r  r  r]   rY  r/  statenew_n_classes	new_states	            r   "reduce_tree_with_different_bitnessr    sw    %288I:I$//:K7H0z9i%$$Y$DM

I;Ig<NOIgz=)<iHHr   c                  0   t        j                  d      \  } }t        dd      j                  | |       j	                  | |      }fd}t        j                   |             }|j	                  | |      }|t        j                  |      k(  sJ y )Nr   r   r?   r  c                     t        j                         } t        j                  |       }t        j
                  j                         |_        t        |j
                  t        <   |j                         | j                  d       | S r  )r  r  r[  r  r  r  r  r  
CythonTreer  r  r  r  r   s     r   "pickle_dump_with_different_bitnesszItest_different_bitness_pickle.<locals>.pickle_dump_with_different_bitness  s^    JJLNN1"11668'I$	s	q	r   )	r	   r   r   r   r   r[  r  r   r   )rd   re   r   r  r  r  r   s         @r   test_different_bitness_pickler    s    ''Q7DAq
 a1
=CGGAqMIIaOE kk<>?Ga#IFMM),,,,r   c                  0   t        j                  d      \  } }t        dd      j                  | |       j	                  | |      }fd}t        j                   |             }|j	                  | |      }|t        j                  |      k(  sJ y )Nr   r   r?   r  c                      t        j                         } t        |       }t        j                  j                         |_        t        |j                  t        <   |j                         | j                  d       | S r  )
r  r  r   r  r  r  r  r  r  r  r  s     r   "joblib_dump_with_different_bitnesszPtest_different_bitness_joblib_pickle.<locals>.joblib_dump_with_different_bitness  sY    JJLO"11668'I$	s	q	r   )	r	   r   r   r   r   r  r  r   r   )rd   re   r   r  r  r  r   s         @r   $test_different_bitness_joblib_pickler    s     ''Q7DAq
 a1
=CGGAqMIIaOE kk<>?Ga#IFMM),,,,r   c                  L   t         r#t        j                  t        j                        n"t        j                  t        j                        } t        j                  t        j                        t        j                  t        j                        g}||D cg c]  }|j                          c}z  }t        j                  ddg|       }|D ]  }t        |j                  |      |         t        j                  t        d      5  t        j                  ddgg|       }t        ||        d d d        t        j                  t        d      5  |j                  t        j                        }t        ||        d d d        y c c}w # 1 sw Y   ^xY w# 1 sw Y   y xY w)Nr   r<   r  zWrong dimensions.+n_classesr	  zn_classes.+incompatible dtype)r.   rx   r  r  r  r  r   r#   r,  r   r   r   r~  )expected_dtypeallowed_dtypesdtrY  wrong_dim_n_classeswrong_dtype_n_classess         r   test_check_n_classesr    s)   +4RXXbhh'"((288:LNhhrxx("((288*<=N>B>Rr(>BBN!Q~6I))"-~>  
z)F	G hhAx~F,n= 
H 
z)H	I ) 0 0 <.? 
J	I C 
H	G 
J	Is   F	
'F,FFF#c                     t        j                  t         j                        } d}t        j                  ||       }| | j	                         g}|D ]  }t        |||        t        j                  t        d      5  t        || d       d d d        |d d d d d df   t        j                  |      fD ]>  }t        j                  t        d      5  t        || |j                         d d d        @ t        j                  t        d	      5  t        |j                  t         j                        | |       d d d        y # 1 sw Y   xY w# 1 sw Y   xY w# 1 sw Y   y xY w)
N)r>   r<   rP   r  )r  expected_shapezWrong shape.+value arrayr	  )r<   rP   r<   zvalue array.+C-contiguouszvalue array.+incompatible dtype)rx   r  r~  rp   r  r%   r   r   r   r  r   r,  r  )r  r  value_ndarrayr  r  problematic_arrs         r   test_check_value_ndarrayr    s.   XXbjj)NNHH^>BM$n&A&A&CDN"^	
 
 
z)C	D.	
 
E
 *!Q(3R5F5F}5UV]]:-HI -.44 JI W 
z)J	K  ,))	
 
L	K 
E	D JI 
L	Ks$   ?E
E,E"
EE	"E+c                     t         } t        j                  d|       }|t        |      t	        |      g}||D cg c]+  }|j                  |j                  j                               - c}z  }|D ]  }t        ||         t        j                  t        d      5  t        j                  d|       }t        ||        d d d        t        j                  t        d      5  |d d d   }t        ||        d d d        |j                  j                  j                         D ci c]  \  }\  }}|| }}}}|j                         }	t        j                  |	d	<   t        j                  t!        |	j#                               t!        |	j%                               d
      }
|j                  |
      }t        j                  t        d      5  t        ||        d d d        |j                         }	t        j&                  |	d<   t        j                  t!        |	j#                               t!        |	j%                               d
      }
|j                  |
      }t        j                  t        d      5  t        ||        d d d        y c c}w # 1 sw Y   xY w# 1 sw Y   xY wc c}}}w # 1 sw Y   xY w# 1 sw Y   y xY w)N)r>   r  )r  zWrong dimensions.+node arrayr	  )r>   rP   znode array.+C-contiguousrP   r{   r  znode array.+incompatible dtyper  )r   rx   rp   r  r  r,  r  r  r$   r   r   r   r  r   r  r  r  r  r  r~  )r  r  valid_node_ndarraysr  problematic_node_ndarrayr   r  r1  
dtype_dictr  r  s              r   test_check_node_ndarrayr  	  s\   N88D7L 	*<8,\:
 8K8K

399))+,8K  #LH # 
z)G	H#%88F.#I 4^T 
I 
z)C	D#/!#4 4^T 
E 7C6H6H6O6O6U6U6WX6W"2$
$+6WJX  __&N"$((N;~**,-$~?T?T?V:WXI  ,229=	z)I	J4^T 
K  __&N#%::N< ~**,-$~?T?T?V:WXI  ,229=	z)I	J4^T 
K	JM 
I	H 
E	D Y 
K	J 
K	Js;   0J%J&J),J6J= K	J&)J3=K	KSplitterc                 d   t         j                  j                  d      }d}dt        j                  ddgt         j                        }}t        d   ||      } | ||dd|d	
      }t        j                  |      }t        j                  |      }|j                  |k(  sJ t        ||       sJ y	)z&Check that splitters are serializable.r   rG   rP   r?   r  r4   r>   rN   N)monotonic_cst)rx   r  r:  r   r,  r   r[  r\  r]  r   r  )	r  r-  r   r/  rY  r   r  splitter_serializesplitter_backs	            r   test_splitter_serializabler  P	  s    
 ))


#CLbhh1vRWW=yIV$Y	:I	<CDQHh/LL!34M%%555mX...r   c                     t        | j                  d            }t        d      }|j                  t        t
               t        j                  ||       t        j                  |d      }t        |j                  |j                  d       y)zhCheck that Trees can be deserialized with read only buffers.

    Non-regression test for gh-25584.
    z
clf.joblibr   r   r)	mmap_modez?The trees of the original and loaded classifiers are not equal.N)strjoinr   r   r  r  r  r  r  r   r   )tmpdirpickle_pathr   
loaded_clfs       r   /test_tree_deserialization_from_read_only_bufferr  b	  sh    
 fkk,/0K
 a
0CGGGW
KK[![C8J		Ir   c                 6   t        j                  ddgddgg      }t        j                  ddg      } | d      j                  ||        | d      }d}t        j                  t
        |      5   |j                  ||       ddd       y# 1 sw Y   yxY w)zhCheck that an error is raised when min_sample_split=1.

    non-regression test for issue gh-25481.
    r   r<   rR   )r  zb'min_samples_split' .* must be an int in the range \[2, inf\) or a float in the range \(0.0, 1.0\]r	  N)rx   r   r   r   r   r   )r&   rd   re   r
   msgs        r   test_min_sample_split_1_errorr  u	  s     	1a&1a&!"A
!QA 	3##Aq) !$D	0  
z	-A 
.	-	-s   2BBc                 0   t        j                  g dg      j                  }t        j                  g d      }| dk(  rt         j                  nt         j                  }t        dd|       }|j                  ||       |j                  t         j                  gg      }t        | ||dd       g       |dd	 }|dd	 }t        dd|       }|j                  ||       |j                  t         j                  gg      }t        | ||d
d       g       y)z=Check missing values goes to correct node during predictions.	r   r<   rP   r?   rO   r  rL         	r   rA   r#  rA   r%  r%  rQ   g?g@r7   r   r<   r  r@   NrB   r=   )
rx   r   r   medianr|  r   r   r   r  r   )r   rd   re   node_value_funcdtcr  X_equaly_equals           r   ;test_missing_values_best_splitter_on_equal_nodes_no_missingr  	  s     	01244A
>?A#,0@#@biibggO
R1	
RCGGAqM [[266($FF_QrsV456 fGfG
R1	
RCGGGW [[266($FF_WRS\:;<r   c                 p   t        j                  g dg      j                  }t        j                  g d      }t        |d|       }|j	                  ||       |j
                  j                  d   }|j
                  j                  d   }|j
                  j                  |   }|j
                  j                  |   }||kD  }	|j
                  j                  |   d   }
|j
                  j                  |   d   }|j                  t         j                  gg      }|	rt        |
|       yt        ||       y)zCheck missing values go to the correct node during predictions for ExtraTree.

    Since ETC use random splits, we use different seeds to verify that the
    left/right node is chosen correctly when the splits occur.
    r
  r  r<   r  r   N)rx   r   r   r   r   r   rw   rv   rM  r   r   r  r   )r   rm  rd   re   etrr  r  left_samplesright_samples	went_lefty_pred_lefty_pred_rightr  s                r   =test_missing_values_random_splitter_on_equal_nodes_no_missingr  	  s    	01244A
>?A
$!y
QCGGAqM ((+J))**1-K 9944Z@LII55kBM},I ))//*-a0K99??;/2L [[266($FV,f-r   r  r4   c                    d}t        j                  t         j                  gdz  g dz   g      j                  }t        j                  |gdz  dgdz  z   dgdz  z         }t	        dd|       }|j                  ||       t        j                  t         j                  dd	gg      j                  }|j                  |      }t        ||ddg       y
)zITest when missing values are uniquely present in a class among 3 classes.r   r;   )r   r<   rP   r?   rO   r  rL   r  r<   rP   r   r  r?   r  Nrx   r   r  r   r   r   r   r+   )r   missing_values_classrd   re   r  r  
y_nan_preds          r   /test_missing_values_best_splitter_three_classesr   	  s     
266(Q,!;;<=??A
&'!+qcAg5a?@A
 bA
SCGGAqMXX2'(**FV$Jz$8!Q#?@r   c                    t        j                  t         j                  gdz  g dz   g      j                  }t        j                  dgdz  dgdz  z         }t	        dd|       }|j                  ||       t        j                  t         j                  d	t         j                  gg      j                  }|j                  |      }t        |g d
       y)zMissing values spanning only one class at fit-time must make missing
    values at predict-time be classified has belonging to this class.r;   r   r<   rP   r?   r;   r>   r   r<   rM   r   rP   r  r>   )r   r<   r   Nr  r   rd   re   r  r  r  s         r   )test_missing_values_best_splitter_to_leftr$  	  s     	266(Q,!334577A
!qA37"#A
 bA
SCGGAqMXX266*+,..F[[ Fvy)r   c                    t        j                  t         j                  gdz  g dz   g      j                  }t        j                  dgdz  dgdz  z   dgdz  z         }t	        dd|       }|j                  ||       t        j                  t         j                  dd	gg      j                  }|j                  |      }t        |g d
       y)zMissing values and non-missing values sharing one class at fit-time
    must make missing values at predict-time be classified has belonging
    to this class.r;   r"  r<   r   rP   r   r  rS   g333333@r  Nr  r#  s         r   *test_missing_values_best_splitter_to_rightr&  	  s    
 	266(Q,!334577A
!qA37"aS1W,-A
 bA
SCGGAqMXXS)*+--F[[ Fvy)r   c                    t        j                  ddddt         j                  ddddt         j                  g
g      j                  }t        j                  d	gdz  dgdz  z         }t	        d
d|       }|j                  ||       t        j                  t         j                  ddgg      j                  }|j                  |      }t        |g d       y)zNCheck behavior of missing value when there is one missing value in each class.r<   rP   r?   r>   rG   r^   rZ   <   r   r   r  gffffff@gA@r  Nr  r#  s         r   >test_missing_values_best_splitter_missing_both_classes_has_nanr)  	  s     	1aArvvr2r2rvv>?@BBA
!qA37"#A
 bA
SCGGAqMXXT*+,..F[[ F vy)r   r
   c                 f   t        j                  ddddt         j                  ddddt         j                  g
g      j                  }t        j                  d	gdz  dgdz  z         } | |      }t	        j
                  t        d
      5   |j                  ||       ddd       y# 1 sw Y   yxY w)z4Check unsupported configurations for missing values.r<   rP   r?   r>   rG   r^   rZ   r(  r   zInput X contains NaNr	  N)rx   r   r  r   r   r   r   r   )r1  r
   rd   re   s       r   test_missing_value_errorsr+  
  s     	1aArvvr2r2rvv>?@BBA
!qA37"#AA	z)?	@A 
A	@	@s   
B''B0c                 D   t         j                  j                         t         j                  }}t        j
                  |ddddf<   t        j
                  |ddddf<    | dd      }|j                  ||       |j                  |      }|d	k\  j                         sJ y)
z5Smoke test for poisson regression and missing values.Nr>   r   rM   rB   r9   r   r   rW   )	rg   r   r  r   rx   r  r   r   r   )r&   rd   re   r   r  s        r   test_missing_values_poissonr-  !
  s     ==qA Acc1fIAcc2gJ

4CGGAqM[[^FcM   r   c                  D    t        j                  | i |\  }}|dkD  }||fS )N   )r	   make_friedman1)argsr   rd   re   s       r   make_friedman1_classificationr2  1
  s-    ""D3F3DAq	BAa4Kr   zmake_data, Tree, tolerancegQ?gQ?gQ?sample_weight_trainr   c                 ~   d\  }} | ||d|      \  }}|j                         }	t        j                  j                  |      }
t        j                  |	|
j                  ddg|j                  ddg      <   t        |	||	      \  }}}}|d
k(  r#t        j                  |j                  d         }nd}d} |||      }|j                  |||       |j                  ||      }t        t                |||            }|j                  ||       |j                  ||      }||z   |kD  sJ d|d| d|        y)zFCheck that trees can deal with missing values have decent performance.)r   rG   rR   )r\   r]   noiser[   FTrY   r   r`   r  r   r   r   NrG   r   r   zscore_native_tree=z + z! should be strictly greater than )r  rx   r  r:  r  choicer   r   r   r   r   r   r   )	make_datar&   r3  r   	tolerancer\   r]   rd   re   	X_missingr-  X_missing_trainX_missing_testr   r   r   r   native_treescore_native_treetree_with_imputerscore_tree_with_imputers                        r   !test_missing_values_is_resiliencerA  7
  sh   ( &Iz'	DAq I
))

 2
3CGIvvIcjj%QWWc
jCD7G1#584O^Wf f$ 5 5a 89 I9KLKOOOWMOJ#)).&A%	@RS /73/55nfMy(+BB 
c) -#$	&Br   zTree, expected_scoreg333333?g(\?c                 H   t         j                  j                  d      }d}|j                  |df      }t        j                  t        j
                  |dz        t        j                  |dz        g      }|j                  ddg|dd	g
      }|j                         j                  t              }||    ||<   |j                  |      }	t         j                  |	|<   |	|dddf<    | |      }
t        |
||d      j                         }||k\  sJ d| d|        y)z@Check the tree learns when only the missing value is predictive.r   rv  r^   r_   rP   FTgffffff?rT   r6  Nr>   r   )cvzExpected CV score: z	 but got )rx   r  r:  standard_normalr  rp   r   r7  r  r,  boolr  r   r|  )r&   expected_scorer   r-  r\   rd   re   X_random_masky_maskX_predictiver
   tree_cv_scores               r    test_missing_value_is_predictiverK  t
  s    ))


"CI)R1A
a0"'')q.2IJKA JJt}9tJMMVVX__T"F#M22F=&&I&6L66LAadG/0D $D!Q15::<MN* 
n-Y}oF*r   zmake_data, Treec                    t         j                  j                  d      }d\  }} | |||      \  }}t         j                  ||j	                  ddg|j
                  ddg      <   t        j                  |j
                  d         }d	|d
d
d<    |d      }|j                  |||        |d      }	|	j                  |dd
dd
d
f   |dd
d          t        |	j                  |      |j                  |             y
)z=Check sample weight is correctly handled with missing values.r   )r  rG   rw  FTrY   r   r6  rW   NrP   r   r   r<   )
rx   r  r:  r  r7  r   r   r   r   r   )
r8  r&   r-  r\   r]   rd   re   r   tree_with_swtree_samples_removeds
             r   test_sample_weight_non_uniformrO  
  s     ))


"C$IzyZcRDAq @BvvAcjj%QWWc
j;< GGAGGAJ'MM#A#Q'LQ7Q/Qqt!tQwZ14a41(003\5I5I!5LMr   c                  F   t        d      j                  t        j                  t        j                        } t        d      j                  t        j                  t        j                        }t        j                  |       }t        j                  |      }||k(  sJ y r  )r   r   rf   r   r   r[  r\  )tree1tree2pickle1pickle2s       r   test_deterministic_picklerU  
  sl     #266tyy$++NE"266tyy$++NEll5!Gll5!Ggr   rd   r>   rM   c                    |j                  dd      }t        j                  dd      } | ||      j                  ||      }t	        |      j                  |j                  dd      |      }|j
                  j                  }t        |dk\        sJ |j                                | t        u r:t        |j
                  j                  dd |j
                  j                  dd        t        j                  |j
                  j                  dk(  |j
                  j                  dk(  z        }t        |j
                  j                  |   d       y)	a  Check that regression trees correctly handle missing values in impurity
    calculations.

    This test verifies that:
    1. Impurity is always non-negative
    2. Impurity is zero for leaves with a single sample
    3. For decision trees, impurity matches reference trees after the first split

    Non-regression test for:
    - Missing values handling in regression criteria:
      https://github.com/scikit-learn/scikit-learn/issues/28254
      https://github.com/scikit-learn/scikit-learn/issues/28316
    - Incorrect/negative impurites for the Poisson criterion with missing values:
      https://github.com/scikit-learn/scikit-learn/issues/32870
    rB   r<   r   r   r   NrP   rW   )r  rx   r   r   r   r   r~   r   r   r   r   flatnonzerorw   r|   )	r&   rd   r   r   re   r
   tree_refr~   
leaves_idxs	            r   'test_regression_tree_missing_values_toyrZ  
  s   < 	
		"aA
		!QA)2DEII!QODT{qyyQ/3Hzz""Hx1}-x||~- $$

++BQ/1H1H!1LM 		!	!R	'DJJ,E,E,JKJ DJJ''
3S9r   c                    t         j                  j                  |       }d}t        j                  |t         j                        j                  dd      }t         j                  |dd d d f<   |j                  |       t        j                  |      }t        | d      j                  ||      }|j                  j                  }t        |dk\        sJ |       y )	NrJ  r  rB   r<   ir>   r  r   )rx   r  r:  r   r~  r  r  r   r   r   r   r~   r   )r   r-  r\   rd   re   r
   r~   s          r   -test_regression_extra_tree_missing_values_toyr\  
  s    
))

 2
3CI
		)2::.66r1=AAcdAgJKKN
		)A+=KOOPQSTUDzz""Hx1}'x'r   c                  >   t        j                  d      \  } }t        j                  j	                  d      }| j                         }|j                  t        j                  dt        j                        | dddgf   dz  	      j                  t              }t        j                  ||<   t        ||d
      \  }}}}t        j                  g dt        j                        }t        ddd      }	 |	j                  ||   ||          t!        |	j"                  j$                  dk\        sJ t        j&                  |	j"                  j(                  dk(  |	j"                  j*                  dk(  z        }
t-        |	j"                  j$                  |
   d       y)a  Check that we properly handle missing values in classification trees using a toy
    dataset.

    The test is more involved because we use a case where we detected a regression
    in a random forest. We therefore define the seed and bootstrap indices to detect
    one of the non-frequent regression.

    Here, we check that the impurity is null or positive in the leaves.

    Non-regression test for:
    https://github.com/scikit-learn/scikit-learn/issues/28254
    T)
return_X_yr   )r<   r;   )r   r  NrP   rO   )nr     r   )prP   Q   '   a   [   &   .      e   r`  Y   R   rJ  r   E      ra     I   J   3   /   k      K   n   r^   r  h   9      r  rt  O   #   M   Z   rp  rf  r`  ^   rd     rO   ]   r  rn  r{  r  r`  ro  m   ru     rG   r~  rv  rl  \   4   r^   r  rO   rO      rl  rz  r  r  r   r  r   rZ   rg  N   r  r  i   r  r   rn  r  f   r  r`  rg  r<   rk  rL       rt  r|  j   r}  r(  8   rz  rs  >   U   ra  rb  P   rm  ?   rM   r  T   r?   r?   L   r  r  r?   r   iHnr   r   rB   r<   rW   )r	   	load_irisrx   r  r:  r  r  r   r  r,  rE  r  r   r   r   r   r   r   r~   rW  rw   r|   r   )rd   re   r-  r:  maskr  r1  r   r   r
   rY  s              r   +test_classification_tree_missing_values_toyr  
  sW    .DAq
))


#CI<<
''bhh
/1QV9q=  fTl 	 ffIdO-iLGQ hh  XXG "&zD DHHWWww/0tzz""a'(((		!	!R	'DJJ,E,E,JKJ DJJ''
3S9r   c                     t        dd      }  | j                  t        j                  t        j                         t        j                  | j                        }t        | j                  || j                        }t        j                  | j                  j                  t
        j                        }d|d<   t        || j                  |       | j                  j                  dk(  sJ |j                  dk(  sJ t!        j"                  t$              5  t'        | j                  j(                  |j(                         ddd       t'        | j                  j(                  d   |j(                  d          t        | j                  || j                        }t        j                  | j                  j                  t
        j                        }d|dd t        || j                  |       | j                  j                  dk(  sJ |j                  dk(  sJ |j                         t'        | j                  j(                  |j(                         y# 1 sw Y   xY w)zHTest pruning a tree with the Python caller of the Cythonized prune tree.r   r<   r  r  r?   N)r   r   rf   r   r   rx   
atleast_1drp  r  n_features_in_r6  rp   r   rt   uint8r"   r   r   AssertionErrorr+   r   r
   rY  pruned_treeleave_in_subtrees       r   test_build_pruned_tree_pyr  .  s   !qA>DDHHTYY$doo.IT00)T__MK xx

 5 5RXXFQ+tzz3CD::  A%%%!!Q&&&	~	&4::++[->->? 
'tzz''*K,=,=a,@A T00)T__MKxx

 5 5RXXFQR +tzz3CD::  A%%%!!Q&>(>(>>&tzz''):):; 
'	&s   +II c                     t        dd      }  | j                  t        j                  t        j                         t        j                  | j                        }t        | j                  || j                        }t        j                  | j                  j                  t
        j                        }d|d<   t        j                   t"        d      5  t%        || j                  |       ddd       y# 1 sw Y   yxY w)z8Test pruning a tree does not result in an infinite loop.r   r<   r  r  z,Node has reached a leaf in the original treer	  N)r   r   rf   r   r   rx   r  rp  r  r  r6  rp   r   rt   r  r   r   r   r"   r  s       r   $test_build_pruned_tree_infinite_loopr  M  s     "qA>DDHHTYY$doo.IT00)T__MK xx

 5 5RXXFQ	H
 	k4::7GH
 
 
s   C77D c                    d dfd	}t         j                  j                  |       }dD ]  }|j                  |df      }|j                  |      }|d|j                  dd      z  z  }t        j                  |      } ||||        ||t        j
                  |      |        |||j                         dz   |        ||||d	
       |j                  |      } ||||        ||||d	
        y)a  
    Test the main bit of logic of the MAE(RegressionCriterion) class
    (used by DecisionTreeRegressor(criterion="absolute_error")).

    The implementation of the criterion relies on an efficient precomputation
    of left/right children absolute error for each split. This test verifies this
    part of the computation, in case of major refactor of the MAE class,
    it can be safely removed.
    c           	         | j                         j                         } t        d| j                  dz         D cg c]  }t	        | d | |d | dd       }}t        t        d| j                  dz         |      D cg c]4  \  }}t        j                  | d | |z
        |d | z  j                         6 }}}t        j                  |      t        j                  |      fS c c}w c c}}w )Nr<   r  Taverage)
r   r  rN  r`   r2   ziprx   r   r}   r   )re   wr  mediansmerrorss         r   compute_prefix_abs_errors_naivezUtest_absolute_errors_precomputation_function.<locals>.compute_prefix_abs_errors_naivej  s    GGINN 1affqj)
) !2A"1r4@) 	 
 E!QVVaZ0':
:1 VVAbqEAI2A&++-: 	 
 xx'!222

s   C$>9C)c                    | j                   d   }|r|dz
  dfnd|f}t        | ||g|| \  }}| |   }||   }	|r|d d d   }|	d d d   }	 ||	      \  }
}|r|
d d d   }
|d d d   }t        ||
d       t        ||d       y )Nr   r<   rB   gdy=)atol)r   r   r   )re   r  r   reverser_  r1  
abs_errorsr  y_sortedw_sortedabs_errors_medians_r  s               r   assert_same_resultszItest_absolute_errors_precomputation_function.<locals>.assert_same_resultsv  s    GGAJ%Ar{Aq6<Q7UTUSTU
GW:W:"~H"~H ?( SX%dd+K"~H
Ke<6r   )r?   r>   rG   r^   r  rJ  r<   r_   g      $@r@   r>   T)r  N)F)rx   r  default_rngr  r   r   roundr  )r   r  r-  r_  re   r  r   r  s          @r   ,test_absolute_errors_precomputation_functionr  _  s    
37  ))

 2
3C$KKaVK$JJqM	TS[[Q'''))A,Aq'*Arwwqz73Aqwwy1}g6Aq'48//!$Aq'*Aq'48 %r   c                    t         j                  j                  |       }t        d      }|j	                  |      }|j                  dd|      |j                  dd|      z  }t        dd      j                  t        j                  |j                  d   df      ||	      j                  j                  j                         d   }t        ||d
d      }t        ||       y)z
    Test that the weighted-median computed under-the-hood when
    building a tree with criterion="absolute_error" is correct.
    g     j@r_   r   r?   r<   r7   )r   r   r  r   r  Tr  N)rx   r  r  r  	lognormalintegersr  r   r   r   r   r   r   r   r2   r   )r   r-  r_  r   r*  tree_leaf_weighted_medianweighted_medians          r   7test_absolute_error_accurately_predicts_weighted_medianr    s    
 ))

 2
3CCA==a= Dll1aal(3;;q!!;+DDG 	(8AF	RWWDJJqM1-.G	L	uuUUWQ	  
 +4"dKO-?r   c                  x   t        j                  g dg dg      j                  dd      j                  t              } g d}t         j
                  | | dk(  <   t        d      D ]T  }t        d|      j                  | |      }t        |j                  j                  t        j                  g d             V y )	Nro  )r<   rP   r<   rP   r<   rP   r<   rP   r   r<   )r   r   r   r   r<   r<   r<   r<   r^   r   )ra   rW   rW   )rx   r   swapaxesr,  floatr  rN  r   r   r+   r   r~   r   )rd   re   r  r
   s       r   "test_splitting_with_missing_valuesr    s     			+-EFG	!Q	 
 	!AAa1fI 2Y$qqAEEaK4::..9I0JK r   c                  >   dddt         j                  t         j                  g} g d}t        j                  |       j                  dd      } t	               j                  | |      }t         |j                  |       |       |j                  j                  dk(  sJ y )Nr   r  rB   r<   r?   )
rx   r  r   r  r   r   r+   r   r   rt   )rd   re   r
   s      r   $test_missing_values_and_constant_toyr    s~     
Aq"&&"&&!AA
B"A!#''1-D |t||A*::  A%%%r   c                  |    t        j                  t        d      5  t        d      } d d d        y # 1 sw Y   y xY w)Nr8   r	  r  )r   warnsFutureWarningr   )r1  s    r   test_friedman_mse_deprecationr    s&    	m>	:!N; 
;	:	:s   2;zX,yrR   r  )rW   rR   r  g      @)rW   rR   rW   zmultiple-non-missingzsingle-non-missing)idsc                 r   t        dd      } |j                  | |       |j                  j                  d   t        k7  sJ |j                  j
                  d   }| t        j                  |           }t        j                  |      sJ |j                         |cxk  r|j                         k  sJ  J y)z
    Check random-split thresholds are finite when the first sample is missing.

    Non-regression test for a subtle bug, see
    https://github.com/scikit-learn/scikit-learn/pull/32119#issuecomment-3765288780
    r<   r   r   N)r   r   r   rw   r    r{   rx   isnanr?  r   rB  )rd   re   r
   r{   non_missings        r   <test_random_splitter_missing_values_uses_non_missing_min_maxr    s     :DDHHQN::##A&)333

$$Q'IRXXa[L/K;;y!!!??	>[__->>>>>>r   r  )__doc__r  r  r  r[  r  r  	itertoolsr   r   r   r  numpyrx   r   joblib.numpy_pickler   numpy.testingr   sklearnr   r	   r
   sklearn.dummyr   sklearn.exceptionsr   sklearn.imputer   sklearn.metricsr   r   r   r   sklearn.model_selectionr   r   sklearn.pipeliner   sklearn.random_projectionr   sklearn.treer   r   r   r   sklearn.tree._classesr   r   r   r   sklearn.tree._criterionr   sklearn.tree._treer   r    r!   r"   r#   r$   r%   r&   r  sklearn.utilsr'   sklearn.utils._array_apir(   sklearn.utils._testingr)   r*   r+   r,   r-   sklearn.utils.fixesr.   r/   r0   r1   sklearn.utils.statsr2   sklearn.utils.validationr3   r   REG_CRITERIONSr   r   r   r:   __annotations__updateSPARSE_TREESr   r  r  y_small_regrd   re   r   r   r  rf   r  r:  r-  r  r   r`   permr   load_diabetesrg   load_digitsrh   r[   make_multilabel_classificationr  r  r  X_sparse_posr  y_randomr  X_sparse_mixrp   r+  r   r   r   markfilterwarningsparametrizer  r   r   r   r   r   r   r   r   r   r   r   r   r   r  r  r   r'  r6  r;  r@  rC  rF  rH  rU  rf  rn  rs  ry  r  r  r  r  r  r  r  r  r  r  r  r  r  r  r  r  r  sortedr<  intersectionr  r  r  r  r  r  r  r  r  r	  r  r  r  r   r'  r8  rD  r  rP  rR  rV  rL  rU  rj  rl  rN  rt  r  r  r  r  r  r  r  r  r  r  r  r  r  r  r  r  r  r   r$  r&  r)  r+  r-  r2  r0  rA  rK  rs  r   rO  rU  r  rZ  r\  r  r  r  r  r  r  r  r  r  )r
   s   0r   <module>r     s     	  	  . .    , ) ) ) ( - (  F * ;   C   2 / (   5 7%O 5.	 3,	
 &	4  	    	    "((56<94:@>>@74545?@A84544/8 P6 	"XBx"bAq6Aq6Aq6:"X1v1v xiiA
t{{''(IIdO	kk$ "8!!#
x++,d#//$'				
v}}))*kk$d#!!$DXDDbR l
 ###1$'\S  !151$RTJRRT ))$++.mm(//:KKfmm4W-[1$<8$84%H5$84288G$84$N	X	X CD!1!1!34n5, 6 5 E,*F*
& CDy'89n5 6 : E CD!1!1!34	,-	./	+,	)*	
< 5 E
<
4W 2L>-K<!7H?+DH>FB8
v +1 ,1 ..9W : /W
 &*:
z +G ,G ..9 : / CD	!!#^	4	!!#^	4!G E!GH0h CD	!!#^	4	!!#^	4@% E@%F=(	E*8Z1h30 +,N ,,N^ +	 ,	+$K$( Y%5%5%78+ 9+(-?"!*&R l3	6 46
 fS->-K-KI-V&WXZ$=>. ? Y. l3$WX.90: : Y 40:h CD<E<4493D$<E~	VW
,D,$$)2C,DnU $WX.9: : Y E:" l3!3~~#FHP 4HPV  +, ,,
 ++dVn-DE$ F ,$$ +G ,G ..9G : /GB& +, ,,: +.90 : ,0J;Z+8 +dVn-DE  F $ vc(--/*k:-FFG &<>Q%RSA TA HMMO4&;=O%PQA R 5A	.:%P +fh%78+dVn-D~-UV W 9 ,6 CD&RS!1!1!34	> 5 T E	> q*& +&B'0T "8:M!NOq!f-G . PG,(2(4?$?"I-,-6@$
B1Uh ,o,,.0G0@0G0G0IJ//& !1!1!34 5, CDn5= 6 E=6 CDq*n5. 6 + E.B y&&9:A ;A y&&9:* ;* y&&9:* ;*  y&&9:* ;* +^<
	 =	 !1!1!34! 5!   
	 	 "7;		 	 "4d;	&(>E	&(;TB .v?' @ 'Z /Y5E5E5G$PT1VW X: 		!	!#89		%	%'=>NN,
 CD"79K!LM 	"&&!RVVQ1-."&&"&&!Q1-.!Q1bffbff-.!Q2661bff-.
 n5#: 6
 N E#:L(,:^<>I$49n@*L(&<
 		BFF8cUSEC51	2HBHH=Q4RS	BFF8cURVVH-	.0IJ 
 !56  ??m0 FDs   	AKAK5	AK?AK