h$,fO dZddlZddlmZmZddlmZmZmZm Z ddl m Z m Z m Z ddlZddlmZddlmZdd lmZmZmZmZd d lmZmZmZmZej<eZ ed d dZ!edddZ"edddedddedddfZ#d2dZ$dZ% d3d Z&ee'edgd!geeddd"#dgeeddd$#dgd!ge(edgd!gd!gd%d&ddd'dde)d(d)e)d*d+fddd%d,Z* d4d-Z+eehd.ge'edgd!geeddd"#dgd!ge(edgd!gd/d&d0ddd'de)d(d)e)d*d+fdd/d1Z,y)5zLabeled Faces in the Wild (LFW) dataset This dataset is a collection of JPEG pictures of famous people collected over the internet, all details are available on the official website: http://vis-www.cs.umass.edu/lfw/ N)IntegralReal)PathLikelistdirmakedirsremove)existsisdirjoin)Memory)Bunch)HiddenInterval StrOptionsvalidate_params)RemoteFileMetadata _fetch_remote get_data_home load_descrzlfw.tgzz.https://ndownloader.figshare.com/files/5976018@055f7d9c632d7370e6fb4afc7468d40f970c34a80d4c6f50ffec63f5a8d536c0)filenameurlchecksumzlfw-funneled.tgzz.https://ndownloader.figshare.com/files/5976015@b47c8422c8cded889dc5a13418c4bc2abbda121092b3533a83306f90d900100apairsDevTrain.txtz.https://ndownloader.figshare.com/files/5976012@1d454dada7dfeca0e7eab6f65dc4e97a6312d44cf142207be28d688be92aabfapairsDevTest.txtz.https://ndownloader.figshare.com/files/5976009@7cb06600ea8b2814ac26e946201cdb304296262aad67d046a16a7ec85d0ff87c pairs.txtz.https://ndownloader.figshare.com/files/5976006@ea42330c62c92989f9d7c03237ed5d591365e89b3e649747777b70e692dc1592Tct|}t|d}t|s t|tD]a}t||j }t|r%|r.t jd|jt||Utd|z|rt|d}t}nt|d}t}t|st||j }t|s>|r.t jd|jt||ntd|zd d l } t jd || j|d j!| t#|||fS)z0Helper function to download any missing LFW data) data_homelfw_homezDownloading LFW metadata: %s)dirnamez %s is missing lfw_funneledlfwz!Downloading LFW data (~200MB): %srNz$Decompressing the data archive to %szr:gz)path)rr r rTARGETSrloggerinforrOSErrorFUNNELED_ARCHIVEARCHIVEtarfiledebugopen extractallr) r$funneleddownload_if_missingr%targettarget_filepathdata_folder_patharchive archive_pathr0s 5lib/python3.12/site-packages/sklearn/datasets/_lfw.py_check_fetch_lfwr<KsB 2IIz*H ( Ax9o&" :FJJGfh7o?@@A.9"%0 " #Hg&6&67 l#" ?Mgx8o <== ;=MN \6*5585D| % %%cl ddlm}tddtddf}||}nt dt ||D}|\}}|j |jz |jxsdz}|j |jz |jxsdz} |'t|}t||z}t|| z} t|} |s)tj| || ftj} n)tj| || dftj} t|D]\} } | d zdk(rt j#d | dz| |j%| }|j'|j|j|j |j f}||j)| |f}tj*|tj}|j,dk(rt/d | z|d z}|s|j1d }|| | df<| S#t$r tdwxYw)zInternally used to load imagesr)ImagezThe Python Imaging Library (PIL) is required to load data from jpeg files. Please refer to https://pillow.readthedocs.io/en/stable/installation.html for installing PIL.c3.K|] \}}|xs|yw)N).0sdss r; z_load_imgs..sG51bqwBwGsrdtypeizLoading face #%05d / %05dzLFailed to read the image file %s, Please make sure that libjpeg is installedgo@r )axis.)PILr? ImportErrorslicetuplezipstopstartstepfloatintlennpzerosfloat32 enumerater+r1r2cropresizeasarrayndim RuntimeErrormean) file_pathsslice_colorr[r? default_sliceh_slicew_slicehwn_facesfacesi file_pathpil_imgfaces r; _load_imgsrnvs 1c]E!SM2M ~GC ,FGGGW  %7<<+<1=A  %7<<+<1=A v  O  O*oG '1a ;'1a+2::>"*- 9 t8q= LL4a!eW E**Y',, ]]GMM7<< F   nnaV,Gzz'4 99>=?HI   99!9$Daf 58 L}   "   s HH3Fcgg}}tt|D]}t||}t|stt|D cgc]} t|| } } t | } | |k\sW|j dd}|j |g| z|j | t |} | dk(rtd|ztj|} tj| |}t||||}tj| }tjjdj|||||}}||| fScc} w)z~Perform the actual data loading for the lfw people dataset This operation is meant to be cached by a joblib wrapper. _ rz*min_faces_per_person=%d is too restrictive*)sortedrr r rUreplaceextend ValueErrorrVunique searchsortedrnarangerandom RandomStateshuffle)r8rarbr[min_faces_per_person person_namesr` person_name folder_pathfpaths n_picturesrh target_namesr6riindicess r;_fetch_lfw_peoplersS "2*Lg&678 % +[9 [! /5gk6J/KL!k1%LLZ - -%--c37K    : ;   e $ %*oG!| 8;O O  99\*L __\< 8F z65& 9Eii GII"%%g.'NF7O6E &, &&5Ms Ebooleanneither)closedleft)r$r4r[r}rbrar5 return_X_y)prefer_skip_nested_validationg?FNc0t|||\}} tjd|t|dd} | j t } | | ||||\} } }| j t| d}td}|r|| fSt|| | || S) a Load the Labeled Faces in the Wild (LFW) people dataset (classification). Download it if necessary. ================= ======================= Classes 5749 Samples total 13233 Dimensionality 5828 Features real, between 0 and 255 ================= ======================= Read more in the :ref:`User Guide `. Parameters ---------- data_home : str or path-like, default=None Specify another download and cache folder for the datasets. By default all scikit-learn data is stored in '~/scikit_learn_data' subfolders. funneled : bool, default=True Download and use the funneled variant of the dataset. resize : float or None, default=0.5 Ratio used to resize the each face picture. If `None`, no resizing is performed. min_faces_per_person : int, default=None The extracted dataset will only retain pictures of people that have at least `min_faces_per_person` different pictures. color : bool, default=False Keep the 3 RGB channels instead of averaging them to a single gray level channel. If color is True the shape of the data has one more dimension than the shape with color = False. slice_ : tuple of slice, default=(slice(70, 195), slice(78, 172)) Provide a custom 2D slice (height, width) to extract the 'interesting' part of the jpeg files and avoid use statistical correlation from the background. download_if_missing : bool, default=True If False, raise an OSError if the data is not locally available instead of trying to download the data from the source site. return_X_y : bool, default=False If True, returns ``(dataset.data, dataset.target)`` instead of a Bunch object. See below for more information about the `dataset.data` and `dataset.target` object. .. versionadded:: 0.20 Returns ------- dataset : :class:`~sklearn.utils.Bunch` Dictionary-like object, with the following attributes. data : numpy array of shape (13233, 2914) Each row corresponds to a ravelled face image of original size 62 x 47 pixels. Changing the ``slice_`` or resize parameters will change the shape of the output. images : numpy array of shape (13233, 62, 47) Each row is a face image corresponding to one of the 5749 people in the dataset. Changing the ``slice_`` or resize parameters will change the shape of the output. target : numpy array of shape (13233,) Labels associated to each face image. Those labels range from 0-5748 and correspond to the person IDs. target_names : numpy array of shape (5749,) Names of all persons in the dataset. Position in array corresponds to the person ID in the target array. DESCR : str Description of the Labeled Faces in the Wild (LFW) dataset. (data, target) : tuple if ``return_X_y`` is True A tuple of two ndarray. The first containing a 2D array of shape (n_samples, n_features) with each row representing one sample and each column representing the features. The second ndarray of shape (n_samples,) containing the target samples. .. versionadded:: 0.20 r$r4r5z Loading LFW people faces from %srlocationcompressverbose)r[r}rbralfw.rst)dataimagesr6rDESCR) r<r+r1r cacherreshaperUrr)r$r4r[r}rbrar5rr%r8m load_funcrir6rXfdescrs r;fetch_lfw_peoplersV"2hDW"H LL3X> 1a8A)*I#,1 #E6<  c%j"%A  "F&y  uV,f r=c t|d5}|Dcgc]/}|jjjd1}}dddDcgc]}t |dkDs|} }t | } t j | t} t} t| D]\} }t |dk(r2d| | <|dt|ddz f|dt|ddz ff}nSt |d k(r2d| | <|dt|ddz f|dt|ddz ff}ntd | dz|fzt|D]R\}\}} t||}ttt|}t|||}| j!|Tt#| |||}t|j$}|j'd}|j)dd|j)d|dz||_|| t j*d d gfScc}w#1swYxYwcc}w#t$rt|t|d }YwxYw)z}Perform the actual data loading for the LFW pairs dataset This operation is meant to be cached by a joblib wrapper. rb Nr rGrIrrzinvalid line %d: %rzUTF-8zDifferent personsz Same person)r2decodestripsplitrUrVrWrTlistrYrvr TypeErrorstrrsrappendrnshapepopinsertarray)index_file_pathr8rarbr[ index_fileln split_linessl pair_specsn_pairsr6r`rj componentspairjnameidx person_folder filenamesrkpairsrrhs r;_fetch_lfw_pairsr{sm ot $M AKL2ryy{((*006L LM*:c"gk":J:*oGXXgS )FJ":.) : z?a F1IAJqM 2Q 67AJqM 2Q 67D_ !F1IAJqM 2Q 67AJqM 2Q 67D 2a!eZ5HHI I'o )NA{c K $%5t < VGM$:;testtrain10_folds)subsetr$r4r[rbrar5rc t|||\}}tjd||t|dd} | j t } dddd } || vr1t d |d tt| jt|| |} | | |||| \} }}td }t| jt| d| |||S)a Load the Labeled Faces in the Wild (LFW) pairs dataset (classification). Download it if necessary. ================= ======================= Classes 2 Samples total 13233 Dimensionality 5828 Features real, between 0 and 255 ================= ======================= In the official `README.txt`_ this task is described as the "Restricted" task. As I am not sure as to implement the "Unrestricted" variant correctly, I left it as unsupported for now. .. _`README.txt`: http://vis-www.cs.umass.edu/lfw/README.txt The original images are 250 x 250 pixels, but the default slice and resize arguments reduce them to 62 x 47. Read more in the :ref:`User Guide `. Parameters ---------- subset : {'train', 'test', '10_folds'}, default='train' Select the dataset to load: 'train' for the development training set, 'test' for the development test set, and '10_folds' for the official evaluation set that is meant to be used with a 10-folds cross validation. data_home : str or path-like, default=None Specify another download and cache folder for the datasets. By default all scikit-learn data is stored in '~/scikit_learn_data' subfolders. funneled : bool, default=True Download and use the funneled variant of the dataset. resize : float, default=0.5 Ratio used to resize the each face picture. color : bool, default=False Keep the 3 RGB channels instead of averaging them to a single gray level channel. If color is True the shape of the data has one more dimension than the shape with color = False. slice_ : tuple of slice, default=(slice(70, 195), slice(78, 172)) Provide a custom 2D slice (height, width) to extract the 'interesting' part of the jpeg files and avoid use statistical correlation from the background. download_if_missing : bool, default=True If False, raise an OSError if the data is not locally available instead of trying to download the data from the source site. Returns ------- data : :class:`~sklearn.utils.Bunch` Dictionary-like object, with the following attributes. data : ndarray of shape (2200, 5828). Shape depends on ``subset``. Each row corresponds to 2 ravel'd face images of original size 62 x 47 pixels. Changing the ``slice_``, ``resize`` or ``subset`` parameters will change the shape of the output. pairs : ndarray of shape (2200, 2, 62, 47). Shape depends on ``subset`` Each row has 2 face images corresponding to same or different person from the dataset containing 5749 people. Changing the ``slice_``, ``resize`` or ``subset`` parameters will change the shape of the output. target : numpy array of shape (2200,). Shape depends on ``subset``. Labels associated to each pair of images. The two label values being different persons or the same person. target_names : numpy array of shape (2,) Explains the target values of the target array. 0 corresponds to "Different person", 1 corresponds to "same person". DESCR : str Description of the Labeled Faces in the Wild (LFW) dataset. rzLoading %s LFW pairs from %srrrrrr!)rrrzsubset='z' is invalid: should be one of )r[rbrarr)rrr6rr)r<r+r1r rrrvrrskeysr rrrrU)rr$r4r[rbrar5r%r8rrlabel_filenamesrrr6rrs r;fetch_lfw_pairsrsL"2hDW"H LL/B 1a8A()I%"O _$tF?#7#7#9:; =  8_V%<=O#,)&f#E6< "F  ]]3u:r *!  r=)NTT)NFNr)NFN)-__doc__loggingnumbersrrosrrrros.pathr r r numpyrVjoblibr utilsrutils._param_validationrrrr_baserrrr getLogger__name__r+r/r.r*r<rnrrrNrMrrrrBr=r;rs"22''SS   8 $  8 O & 8 O$ <S # <S  <S 2(&VBVST('V8T*KD!T)