-igQ|SrSSKrSSKJr SSKJr SSKrSSK J r SSK J r SSKJr SSKJrJrJrJr S S KJrJr \"\\"\SSS S 9\R2\"S 5/\"\S SS S 9S/SS/S\"S15/S/\"\SSS S 9/\/S.SS9S\R4SSSSSS.Sj5rSrS#Sjr\"S\\R2\"S 5\"\SSS S 9/\"\S SS S 9S/SS/S\"S15/S/\"\SSS S 9/\/S.SS9S\R4SSSSSS.Sj5rSr\"SS/SS/\\"S/5/S/\\ S/SS/S/S .SS9SSSSS!.S"j5r!g)$aThis module implements a loader and dumper for the svmlight format This format is a text-based format, with one sample per line. It does not store zero valued features hence is suitable for sparse dataset. The first element of each line can be used to store a target variable to predict. This format is used as the default format for both svmlight and the libsvm command line programs. N)closing)Integral) __version__) check_array) HasMethodsInterval StrOptionsvalidate_params)_dump_svmlight_file_load_svmlight_fileleft)closedread no_validationbooleanautof n_featuresdtype multilabel zero_basedquery_idoffsetlengthT)prefer_skip_nested_validationFrrrrrrrc 6[[U/UUUUUUUS95$)aeLoad datasets in the svmlight / libsvm format into sparse CSR matrix. This format is a text-based format, with one sample per line. It does not store zero valued features hence is suitable for sparse dataset. The first element of each line can be used to store a target variable to predict. This format is used as the default format for both svmlight and the libsvm command line programs. Parsing a text based source can be expensive. When repeatedly working on the same dataset, it is recommended to wrap this loader with joblib.Memory.cache to store a memmapped backup of the CSR results of the first call and benefit from the near instantaneous loading of memmapped structures for the subsequent calls. In case the file contains a pairwise preference constraint (known as "qid" in the svmlight format) these are ignored unless the query_id parameter is set to True. These pairwise preference constraints can be used to constraint the combination of samples when using pairwise loss functions (as is the case in some learning to rank problems) so that only pairs with the same query_id value are considered. This implementation is written in Cython and is reasonably fast. However, a faster API-compatible loader is also available at: https://github.com/mblondel/svmlight-loader Parameters ---------- f : str, path-like, file-like or int (Path to) a file to load. If a path ends in ".gz" or ".bz2", it will be uncompressed on the fly. If an integer is passed, it is assumed to be a file descriptor. A file-like or file descriptor will not be closed by this function. A file-like object must be opened in binary mode. .. versionchanged:: 1.2 Path-like objects are now accepted. n_features : int, default=None The number of features to use. If None, it will be inferred. This argument is useful to load several files that are subsets of a bigger sliced dataset: each subset might not have examples of every feature, hence the inferred shape might vary from one slice to another. n_features is only required if ``offset`` or ``length`` are passed a non-default value. dtype : numpy data type, default=np.float64 Data type of dataset to be loaded. This will be the data type of the output numpy arrays ``X`` and ``y``. multilabel : bool, default=False Samples may have several labels each (see https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/multilabel.html). zero_based : bool or "auto", default="auto" Whether column indices in f are zero-based (True) or one-based (False). If column indices are one-based, they are transformed to zero-based to match Python/NumPy conventions. If set to "auto", a heuristic check is applied to determine this from the file contents. Both kinds of files occur "in the wild", but they are unfortunately not self-identifying. Using "auto" or True should always be safe when no ``offset`` or ``length`` is passed. If ``offset`` or ``length`` are passed, the "auto" mode falls back to ``zero_based=True`` to avoid having the heuristic check yield inconsistent results on different segments of the file. query_id : bool, default=False If True, will return the query_id array for each file. offset : int, default=0 Ignore the offset first bytes by seeking forward, then discarding the following bytes up until the next new line character. length : int, default=-1 If strictly positive, stop reading any new line of data once the position in the file has reached the (offset + length) bytes threshold. Returns ------- X : scipy.sparse matrix of shape (n_samples, n_features) The data matrix. y : ndarray of shape (n_samples,), or a list of tuples of length n_samples The target. It is a list of tuples when ``multilabel=True``, else a ndarray. query_id : array of shape (n_samples,) The query_id for each sample. Only returned when query_id is set to True. See Also -------- load_svmlight_files : Similar function for loading multiple files in this format, enforcing the same number of features/columns on all of them. Examples -------- To use joblib.Memory to cache the svmlight file:: from joblib import Memory from sklearn.datasets import load_svmlight_file mem = Memory("./mycache") @mem.cache def get_data(): data = load_svmlight_file("mysvmlightfile") return data[0], data[1] X, y = get_data() r )tupleload_svmlight_filesrs W/var/www/html/venv/lib/python3.13/site-packages/sklearn/datasets/_svmlight_format_io.pyload_svmlight_filer% s4^  C!!!   c[U[5(a [USSS9$[U[R5(a[R "U5nO,[U[ 5(d[S[U5-5e[RRU5upUS:XaSSK nURUS5$US:XaSSK J n U"US5$[US5$) NrbF)closefdz1expected {str, int, path-like, file-like}, got %sz.gzrz.bz2)BZ2File) isinstanceintopenosPathLikefspathstr TypeErrortypepathsplitextgzipbz2r*)r_extr6r*s r$ _gen_openr:s!SAtU++ Ar{{ # # IIaL 3  KdSTgUVV WW  a FA e|yyD!! q$At}r&c ^[US5(a[XX#XEU5upxppO/[[U55n[XX#XEU5upxppSSS5 U(d%[R "W [R 5n [R "WW5n[R "W [R5n [R "W [RS9n [R "W [R5n [R"XS9nXU W U 4$!,(df  N=f)Nrr) hasattrrrr:np frombufferfloat64longlongint64asarray)rrrrrrr actual_dtypedataindindptrlabelsqueryindicess r$_open_and_loadrKsq&9L jh: 6 CYq\ "a=P*(F> :LV# vrzz2 ==| ,DmmC-G ]]6 5F MM% *E ::d (D &&% //# "s D D,z array-like)filesrrrrrrrc US:wdUS:aUS:XaSnUS:wdUS:aUc [S5eUVs/sH$n[UUU[U5[U5UUS9PM& n nUSLdUS:Xa*[SU 55(aU H up n U S-n M [ S U 55S-n UcU nOX:a[S R X55e/n U HfupnnnUR SS- U4n[R"XU4U5nUR5 U UU4- n U(dMUU RU5 Mh U $s snf) aLoad dataset from multiple files in SVMlight format. This function is equivalent to mapping load_svmlight_file over a list of files, except that the results are concatenated into a single, flat list and the samples vectors are constrained to all have the same number of features. In case the file contains a pairwise preference constraint (known as "qid" in the svmlight format) these are ignored unless the query_id parameter is set to True. These pairwise preference constraints can be used to constraint the combination of samples when using pairwise loss functions (as is the case in some learning to rank problems) so that only pairs with the same query_id value are considered. Parameters ---------- files : array-like, dtype=str, path-like, file-like or int (Paths of) files to load. If a path ends in ".gz" or ".bz2", it will be uncompressed on the fly. If an integer is passed, it is assumed to be a file descriptor. File-likes and file descriptors will not be closed by this function. File-like objects must be opened in binary mode. .. versionchanged:: 1.2 Path-like objects are now accepted. n_features : int, default=None The number of features to use. If None, it will be inferred from the maximum column index occurring in any of the files. This can be set to a higher value than the actual number of features in any of the input files, but setting it to a lower value will cause an exception to be raised. dtype : numpy data type, default=np.float64 Data type of dataset to be loaded. This will be the data type of the output numpy arrays ``X`` and ``y``. multilabel : bool, default=False Samples may have several labels each (see https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/multilabel.html). zero_based : bool or "auto", default="auto" Whether column indices in f are zero-based (True) or one-based (False). If column indices are one-based, they are transformed to zero-based to match Python/NumPy conventions. If set to "auto", a heuristic check is applied to determine this from the file contents. Both kinds of files occur "in the wild", but they are unfortunately not self-identifying. Using "auto" or True should always be safe when no offset or length is passed. If offset or length are passed, the "auto" mode falls back to zero_based=True to avoid having the heuristic check yield inconsistent results on different segments of the file. query_id : bool, default=False If True, will return the query_id array for each file. offset : int, default=0 Ignore the offset first bytes by seeking forward, then discarding the following bytes up until the next new line character. length : int, default=-1 If strictly positive, stop reading any new line of data once the position in the file has reached the (offset + length) bytes threshold. Returns ------- [X1, y1, ..., Xn, yn] or [X1, y1, q1, ..., Xn, yn, qn]: list of arrays Each (Xi, yi) pair is the result from load_svmlight_file(files[i]). If query_id is set to True, this will return instead (Xi, yi, qi) triplets. See Also -------- load_svmlight_file: Similar function for loading a single file in this format. Notes ----- When fitting a model to a matrix X_train and evaluating it against a matrix X_test, it is essential that X_train and X_test have the same number of features (X_train.shape[1] == X_test.shape[1]). This may not be the case if you load the files individually with load_svmlight_file. Examples -------- To use joblib.Memory to cache the svmlight file:: from joblib import Memory from sklearn.datasets import load_svmlight_file mem = Memory("./mycache") @mem.cache def get_data(): data_train, target_train, data_test, target_test = load_svmlight_files( ["svmlight_file_train", "svmlight_file_test"] ) return data_train, target_train, data_test, target_test X_train, y_train, X_test, y_test = get_data() rrTz:n_features is required when offset or length is specified.)rrFc3# UH6n[US5=(a [R"US5S:v M8 g7fr rN)lenr>min).0tmps r$ &load_svmlight_files..s3$USTCSQ[%GRVVCF^a5G%GSTs>Ar c3r# UH-n[US5(aUSR5OSv M/ g7frO)rPmax)rRrFs r$rTrUs+ >ASc#a&kkc!fjjlq0As57z=n_features was set to {}, but input file contains {} features) ValueErrorrKboolallrWformatshapesp csr_matrix sort_indicesappend)rLrrrrrrrrrr8rJn_fresultrErGy query_valuesr\Xs r$r#r#sJ ! vzzV'; ! vzz'9UVV A      N  Uf$UST$U!U!U#$ A1a qLG$%  >A > > BC   K R R   F23.vq,a1$j1 MM4&15 9 !Q$ 8 MM, ' 34 MO s+D?c U(aURS[-R55 URSSS/U-R55 URS5 URSUR 555 [ R "U5n[ R "U5nU(dU(dUSS2[R4n[UUUUUUUU5 g)Nz7# Generated by dump_svmlight_file from scikit-learn %s z# Column indices are %s-based zeroones# c3,# UH nSU-v M g7f)s# %s N)rRlines r$rT!_dump_svmlight..sG2F$Y%2Fs) writerencode writelines splitlinesr]issparser>newaxisr ) rfrdrr one_basedcommentrX_is_spy_is_sps r$_dump_svmlightrxs J[Xfh  .&%1K K S S U   G'2D2D2FGGkk!nGkk!nG g am     r&z sparse matrixrn)rfrdrrrurr)rrurrc &UbI[U[5(aURS5 OURS5nSU;a [ S5e[ USSS9n[ R"U5(a4URS S :wa U(d[ S UR<35eO0URS :wa U(d[ S UR<35e[ USS 9nURS URS :wa.[ SURS <SURS <S35eXqLa"[US5(aUR5nO#Un[US5(aUR5 XLa"[US5(aUR5nO#Un[US5(aUR5 Uc$[R"/[RS9nOO[R "U5nURS URS :wa[ SUR<35eU(+n [US5(a[#XX&XU5 g[%US5n[#XX&XU5 SSS5 g!,(df  g=f)aDump the dataset in svmlight / libsvm file format. This format is a text-based format, with one sample per line. It does not store zero valued features hence is suitable for sparse dataset. The first element of each line can be used to store a target variable to predict. Parameters ---------- X : {array-like, sparse matrix} of shape (n_samples, n_features) Training vectors, where `n_samples` is the number of samples and `n_features` is the number of features. y : {array-like, sparse matrix}, shape = (n_samples,) or (n_samples, n_labels) Target values. Class labels must be an integer or float, or array-like objects of integer or float for multilabel classifications. f : str or file-like in binary mode If string, specifies the path that will contain the data. If file-like, data will be written to f. f should be opened in binary mode. zero_based : bool, default=True Whether column indices should be written zero-based (True) or one-based (False). comment : str or bytes, default=None Comment to insert at the top of the file. This should be either a Unicode string, which will be encoded as UTF-8, or an ASCII byte string. If a comment is given, then it will be preceded by one that identifies the file as having been dumped by scikit-learn. Note that not all tools grok comments in SVMlight files. query_id : array-like of shape (n_samples,), default=None Array containing pairwise preference constraints (qid in svmlight format). multilabel : bool, default=False Samples may have several labels each (see https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/multilabel.html). .. versionadded:: 0.17 parameter `multilabel` to support multilabel datasets. Examples -------- >>> from sklearn.datasets import dump_svmlight_file, make_classification >>> X, y = make_classification(random_state=0) >>> output_file = "my_dataset.svmlight" >>> dump_svmlight_file(X, y, output_file) # doctest: +SKIP Nasciizutf-8z comment string contains NUL bytecsrF) accept_sparse ensure_2dr z(expected y of shape (n_samples, 1), got z&expected y of shape (n_samples,), got )r}rz2X.shape[0] and y.shape[0] should be the same, got z and z instead.sorted_indicesr_r<z-expected query_id of shape (n_samples,), got rnwb)r+bytesdecoderorXrr]rrr\ndimr=rr_r>arrayint32rCrxr-) rfrdrrrurryvalXvalrts r$dump_svmlight_filersX gu % % NN7 #nnW-G G ?@ @ q ?D {{4 ::a=A j@D L  99>*4::WX X q .D zz!} 1 %zz!}djjm -   yWT#344    !  1n % % NN  yWT#344    !  1n % % NN  88Bbhh/::h' >>!  *EM^^U Iq'qQII !T]a 1 H M]]s *J J)rr)"__doc__os.pathr. contextlibrnumbersrnumpyr> scipy.sparsesparser]rutilsrutils._param_validationrr r r _svmlight_format_fastr rr1r/r@r%r:rKr#rxrrrkr&r$rs  WW   Xq$v 6 KK v    !T&A4H  k *fX"67KHaf=>*#'!* **  H%$HV*0.   KK v  Xq$v 6   !T&A4H  k *fX"67KHaf=>* #'#, **  a'&aH:O ,O ,:wi( ) k%!4( k#' "  @N @Nr&