U
    3d                     @   s   d Z ddlmZ ddlmZ ddlZddlZddlm	Z
 ddlmZ ddlmZ ddlmZ ddlmZ dd	lmZ d
d Zdd Zdd Zdd Zdd Zdd Zdd ZdS )zTest the 20news downloader, if the data is available,
or if specifically requested via environment variable
(e.g. for travis cron job).    )partial)patchN)check_as_framecheck_pandas_dependency_message)check_return_X_y)assert_allclose_dense_sparse	normalizec           	      C   s,  | ddd}|j dst| d|jddd dd}|j|jdd  ksNtt|j d	d
gksjtt|j	t|jkstt|j	t|j
kst|j
d	 }|j|jd	  }|j|}|j
t|j|kd	 d	  }||kst| dddd\}}t|t|j
kst|j|jjks(td S )NallF)subsetshuffle.. _20newsgroups_dataset:)r   
categoriesr   r      T)r   r   Z
return_X_y)DESCR
startswithAssertionErrorZtarget_namesnpuniquetargettolistlen	filenamesdataindexwhereshape)	fetch_20newsgroups_fxtr   Z	data2catsZentry1categorylabelZentry2Xy r&   F/tmp/pip-unpacked-wheel-zrfo1fqw/sklearn/datasets/tests/test_20news.pytest_20news   s&      
r(   c                 C   s\   | dd}t |d t |jks$tt |d t |jks>tt |d t |jksXtdS )zuChecks the length consistencies within the bunch

    This is a non-regression test for a bug present in 0.16.1.
    r   r   r   r   r   N)r   r   r   r   r   )r!   r   r&   r&   r'   test_20news_length_consistency3   s    
r*   c                 C   sB  | dd}t |jst|jjdks*t|jjd dks>t|jjtjksPt|j	
ds`t| dd}t |jszt|jjdkst|jjd d	kst|jjtjkst|j	
dstt| dd}t|| | d
d}t |jst|jjdkst|jjd dkst|jjtjks,t|j	
ds>td S )NZtrainr)   )2,  ; r   r+   r   test)l  r,   r.   r   )I  r,   r/   )spZisspmatrix_csrr   r   r    r   Zdtyper   Zfloat64r   r   r   r   )!fetch_20newsgroups_vectorized_fxtbunchZ
fetch_funcr&   r&   r'   test_20news_vectorized?   s(    



r3   c                 C   sf   | dd}| dd}|d d d }|d d d }t |t| ttjj| dddsbtd S )NFr	   Tr   d   r   )Zaxis)r   r
   r   ZallcloseZlinalgZnormZtodenser   )r1   r$   ZX_ZX_normr&   r&   r'   test_20news_normalization]   s    

r5   c                    s   t d | dd}t||  |j}|jdks2tt fdd|jjD sPtdD ]}||	 ksTtqTd|	 kszt|j
jdkstd S )	NZpandasTZas_frame)r+   i< c                    s   g | ]}t | jqS r&   )
isinstanceZSparseDtype).0colpdr&   r'   
<listcomp>o   s     z(test_20news_as_frame.<locals>.<listcomp>)ZbeginnerZ	beginnersZ	beginningZ
beginningsZbeginsZbegleyZbegoneZcategory_class)pytestZimportorskipr   framer    r   r   r   Zdtypeskeysr   name)r1   r2   r>   Zexpected_featurer&   r:   r'   test_20news_as_frameg   s    


	rA   c                 C   s   t |  d S )Nr   )r1   Zhide_available_pandasr&   r&   r'   test_as_frame_no_pandas   s    rB   c                 C   s`   t dN}t d:}d|_d|_d}tjt|d | dd W 5 Q R X W 5 Q R X W 5 Q R X d S )Nzos.path.existszjoblib.loadT)r$   r%   zThe cached dataset located in)matchr6   )r   Zreturn_valuer=   Zraises
ValueError)r1   Zmock_is_existZ	mock_loaderr_msgr&   r&   r'   test_outdated_pickle   s    

rF   )__doc__	functoolsr   Zunittest.mockr   r=   Znumpyr   Zscipy.sparsesparser0   Z"sklearn.datasets.tests.test_commonr   r   r   Zsklearn.utils._testingr   Zsklearn.preprocessingr
   r(   r*   r3   r5   rA   rB   rF   r&   r&   r&   r'   <module>   s"    
