\!hI!TSSKrSSKrSSKJr SSKJrJr SSKJ r "SS\\5r g)N)chi2_contingency) BaseEstimatorClassifierMixin)NotFittedErrorc.\rSrSrSrSSjrSrSrSrg) OneRClassifieraCOneR (One Rule) Classifier. Parameters ---------- resolve_ties : str (default: 'first') Option for how to resolve ties if two or more features have the same error. Options are - 'first' (default): chooses first feature in the list, i.e., feature with the lower column index. - 'chi-squared': performs a chi-squared test for each feature against the target and selects the feature with the lowest p-value. Attributes ---------- self.classes_labels_ : array-like, shape = [n_labels] Array containing the unique class labels found in the training set. self.feature_idx_ : int The index of the rules' feature based on the column in the training set. self.p_value_ : float The p value for a given feature. Only available after calling `fit` when the OneR attribute `resolve_ties = 'chi-squared'` is set. self.prediction_dict_ : dict Dictionary containing information about the feature's (self.feature_idx_) rules and total error. E.g., `{'total error': 37, 'rules (value: class)': {0: 0, 1: 2}}` means the total error is 37, and the rules are "if feature value == 0 classify as 0" and "if feature value == 1 classify as 2". (And classify as class 1 otherwise.) For usage examples, please see https://rasbt.github.io/mlxtend/user_guide/classifier/OneRClassifier/ cJSS1nX;a[SU<SU<S35eXlg)Nfirst chi-squaredzresolve_ties must be in z. Got .) ValueError resolve_ties)selfralloweds jC:\Users\julio\OneDrive\Documentos\Trabajo\Ideas Frescas\venv\Lib\site-packages\mlxtend/classifier/oner.py__init__OneRClassifier.__init__<s/M*  &9@,O )c ^[URS5HWn[R"USS2U45RSURS:Xa[R "SU-5 O [R"U5RSmU4Sjn0n[R "URS5GHn[R"USS2U45HnU"XXg5n[R"U5n [R"U5Ul[R"T[S9n SX'[R"X5n Xe;aS0S.XV'U XVS U'XVS ==U - ss'M [Rn S/n UHnX^S U :dMX^S n XS 'M URS :XaUH+nXS :XaM X^S U :XdMU RU5 M- /nU HynUUS n[R"T[!U545n[#U5H.unn[R$"X!SS2U4U:HTS 9USS2U4'M0 ['U5unn nM{ URW5 [R"U5nU Un UUUlGMURS:XdGMU Sn GM W UlX]UlU$)a<Learn rule from training data. Parameters ---------- X : {array-like, sparse matrix}, 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, shape = [n_samples] Target values. Returns ------- self : object NrzyFeature array likely contains at least one non-categorical column. Column %d appears to have a unique value in every row.cJ>USS2U4U:Hn[R"XTS9$)N minlength)npbincount)Xy feature_index feature_valuemaskn_class_labelss rcompute_class_counts0OneRClassifier.fit..compute_class_countscs*Q %&-7D;;qw.A ArdtypeF) total errorrules (value: class)r(r'r rr )rangeshaperuniquewarningswarnarangeargmax class_labels_onesboolsuminfrappendzeroslen enumeraterrp_value_ feature_idx_prediction_dict_)rrrcr#prediction_dictrr class_countsmost_frequent_class inverse_indexerrorbest_errbest_idxip_values feature_idxrulesaryidxr_p best_p_idxr"s @rfitOneRClassifier.fitDs&qwwqz"Ayy1a4!''*aggaj8 %()) #1++A. B YYqwwqz2M!#1Q -=+>!? 3A-W &(ii &=#%'YYq\"!#d C 5: 2|:;!7'(026O2 ( ./EF! .}=F=-"@2vvHvH$"%m4x?.1-@H#$RL%   M1(ARL( &)-8HD * ) #+K+K89OPE((NCJ#?@C"+E"2Q&(kk;/145'AsF #3 "2#!6JAq!Q-$,."YYx0 #J/ ( 4 ""g-#A;Q3T% / 9 rc[US5(d [S5eURSn[R"UR S[R S9n[5nUHnX%nURU5 M [UR5U- n[U5(a[U5SUSS&UHnUSS2UR4U:HnX%X8'M! U$)a8Predict class labels for X. Parameters ---------- X : {array-like, sparse matrix}, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. Returns ---------- maj : array-like, shape = [n_samples] Predicted class labels. r<z8Estimator not fitted, call `fit` before using the model.r(rr%N) hasattrrr<rr7r+int_setaddr1r8listr;) rrrHy_pred rule_labelsr class_label other_labelr!s rpredictOneRClassifier.predictst/00 M %%&<=!''!*BGG4 e "M.K OOK (#$,,- ; {  [)!,F1I #MQ)))*m;D /FL# r)r1r;r:r<rN)r ) __name__ __module__ __qualname____firstlineno____doc__rrOr[__static_attributes__rrrrs&P)rh*rr) r-numpyr scipy.statsr sklearn.baserrsklearn.exceptionsrrrcrrrhs&(7-O]OOr