ה9iRRdZddlZddlmZmZddlmZddlmZm Z ddl m Z ddl Z ddl mZddlmZmZmZdd lmZdd lmZdd lmZdd lmZmZmZmZmZmZdd l m!Z!m"Z"ddl#m$Z$m%Z%dZ&GddeeeZ'y)zBase class for mixture models.N)ABCMetaabstractmethod) nullcontext)IntegralReal)time)cluster) BaseEstimator DensityMixin _fit_context)kmeans_plusplus)ConvergenceWarning)check_random_state)_convert_to_numpy_is_numpy_namespace _logsumexp_max_precision_float_dtype get_namespaceget_namespace_and_device)Interval StrOptions)check_is_fitted validate_datac^|j|k7rtd|d|d|jy)zValidate the shape of the input parameter 'param'. Parameters ---------- param : array param_shape : tuple name : str zThe parameter 'z' should have the shape of z , but got N)shape ValueError)param param_shapenames _/var/www/html/backtest/airagagent/rag_env/lib/python3.12/site-packages/sklearn/mixture/_base.py _check_shaper!s2 {{k![%++ /  "ceZdZUdZeedddgeedddgeedddgeedddgeedddgehdgd gd gd geedddgd Ze e d <dZ e d&dZ d&dZe dZd&dZedd&dZd&dZe dZe dZe dZdZd&dZdZdZd'dZd&dZe d&d Ze d&d!Zd&d"Zd#Z d$Z!d%Z"y)( BaseMixturezBase class for mixture models. This abstract class specifies an interface for all mixture classes and provides basic common methods for mixture models. Nleft)closedgr>kmeansrandomrandom_from_data k-means++ random_statebooleanverbose n_componentstol reg_covarmax_itern_init init_paramsr, warm_startr.verbose_interval_parameter_constraintsc ||_||_||_||_||_||_||_||_| |_| |_ yNr/) selfr0r1r2r3r4r5r,r6r.r7s r __init__zBaseMixture.__init__GsN)"   &($ 0r"cy)zCheck initial parameters of the derived class. Parameters ---------- X : array-like of shape (n_samples, n_features) Nr;Xxps r _check_parameterszBaseMixture._check_parameters_ r"cJt||\}}}|j\}}|jdk(rtj||j f|j }tj|j d|j|j}d|tj||f<n^|jdk(rb|j|j||j f|j |}||j|d d d |jfz}n|jd k(rc|j ||j f|j |}|j!||j d } t#| D] \} } d|| | f<n{|jdk(rltj||j f|j }t%||j |\}} d|| tj|j f<|j'|y )a?Initialize the model parameters. Parameters ---------- X : array-like of shape (n_samples, n_features) random_state : RandomState A random number generator instance that controls the random seed used for the method chosen to initialize the parameters. rAr()dtyper%) n_clustersr4r,r)sizerFdeviceaxisNr*F)rIreplacer+)r,)rrr5npzerosr0rFr KMeansfitlabels_arangeasarrayuniformsumnewaxischoice enumerater _initialize) r;r@r,rA_rK n_samplesresplabelindicescolindexs r _initialize_parametersz"BaseMixture._initialize_parametersis1r: Avww 1   x '88Y(9(9:!''JD#00Q 12D9%u, -    )::$$9d6G6G*H$IggD BFF4aF(BJJ7 7D   !3 388D--.aggfD#)) 1 15*G(0 % U#$UCZ  %    ,88Y(9(9:!''JD(!!)JAw ;)r;r@r^s r r[zBaseMixture._initializes r"c*|j|||S)aEstimate model parameters with the EM algorithm. The method fits the model ``n_init`` times and sets the parameters with which the model has the largest likelihood or lower bound. Within each trial, the method iterates between E-step and M-step for ``max_iter`` times until the change of likelihood or lower bound is less than ``tol``, otherwise, a ``ConvergenceWarning`` is raised. If ``warm_start`` is ``True``, then ``n_init`` is ignored and a single initialization is performed upon the first call. Upon consecutive calls, training starts where it left off. Parameters ---------- X : array-like of shape (n_samples, n_features) List of n_features-dimensional data points. Each row corresponds to a single data point. y : Ignored Not used, present for API consistency by convention. Returns ------- self : object The fitted mixture. ) fit_predict)r;r@ys r rRzBaseMixture.fits6 A r"T)prefer_skip_nested_validationct|\}}t|||j|jgd}|jd|j kr(t d|j d|jd|j|||jxr t|d }|r |jnd}|j }g}d |_ t|j} |j\} }t|D]Q} |j!| |r|j#|| ||r |j n |j$} g} |j&dk(r|j)}d}jd }td|j&dzD]}| }|j+||\}}|j-||||j/||} | j1| | |z }|j3||t5||j6ksd }n|j9| || |kDs||j k(s5| }|j)}}| }||_ T|js)|j&dkDrt;j<d t>|jA||_!||_||_"|j+||\}}|jG|d S) aEstimate model parameters using X and predict the labels for X. The method fits the model ``n_init`` times and sets the parameters with which the model has the largest likelihood or lower bound. Within each trial, the method iterates between E-step and M-step for `max_iter` times until the change of likelihood or lower bound is less than `tol`, otherwise, a :class:`~sklearn.exceptions.ConvergenceWarning` is raised. After fitting, it predicts the most probable label for the input data points. .. versionadded:: 0.20 Parameters ---------- X : array-like of shape (n_samples, n_features) List of n_features-dimensional data points. Each row corresponds to a single data point. y : Ignored Not used, present for API consistency by convention. Returns ------- labels : array, shape (n_samples,) Component labels. )rFensure_min_samplesrz:Expected n_samples >= n_components but got n_components = z, n_samples = rE converged_r%FTzBest performing initialization did not converge. Try different init parameters, or increase max_iter, tol, or check for degenerate data.rL)$rrfloat64float32rr0rrBr6hasattrr4infrlrr,range_print_verbose_msg_init_begrc lower_bound_r3_get_parameters_e_step_m_step_compute_lower_boundappend_print_verbose_msg_iter_endabsr1_print_verbose_msg_init_endwarningswarnr_set_parametersn_iter_ lower_bounds_argmax)r;r@rgrAr\do_initr4max_lower_boundbest_lower_boundsr,r]init lower_boundcurrent_lower_bounds best_params best_n_iter convergedn_iterprev_lower_bound log_prob_normlog_respchanges r rfzBaseMixture.fit_predicts8a A $"**bjj)AVW X 771:)) )**.*;*;)<= wwqzl,  qR(F74+FG 'Q66')$*;*;< ww 1&M$ 0D  , ,T 2++A|+C%,266'$2C2CK#% }}!"224  ! #At}}q'89 F'2$.2ll1l.D+M8LLHL4"&";";Hm"TK(// <(+;;F44VVD6{TXX-$(  00iH0Ow4N&1O"&"6"6"8K"(K(<%&/DOI$ 0R4==1#4 MM9#   [R0" +. ll1l, 8yyy**r"ctt||\}}|j||\}}|j||fS)aE step. Parameters ---------- X : array-like of shape (n_samples, n_features) Returns ------- log_prob_norm : float Mean of the logarithms of the probabilities of each sample in X log_responsibility : array, shape (n_samples, n_components) Logarithm of the posterior probabilities (or responsibilities) of the point of each sample in X. rE)r_estimate_log_prob_respmean)r;r@rAr\rrs r ruzBaseMixture._e_step:sB aB'A"&">">qR">"H xww}%x//r"cy)a*M step. Parameters ---------- X : array-like of shape (n_samples, n_features) log_resp : array-like of shape (n_samples, n_components) Logarithm of the posterior probabilities (or responsibilities) of the point of each sample in X. Nr>)r;r@rs r rvzBaseMixture._m_stepNs r"cyr:r>)r;s r rtzBaseMixture._get_parameters\ r"cyr:r>)r;paramss r r~zBaseMixture._set_parameters`rr"clt|t||d}t|j|dS)aCompute the log-likelihood of each sample. Parameters ---------- X : array-like of shape (n_samples, n_features) List of n_features-dimensional data points. Each row corresponds to a single data point. Returns ------- log_prob : array, shape (n_samples,) Log-likelihood of each sample in `X` under the current model. Fresetr%rL)rrr_estimate_weighted_log_prob)r;r@s r score_sampleszBaseMixture.score_samplesds2  $ /$::1=AFFr"cpt|\}}t|j|j|S)aCompute the per-sample average log-likelihood of the given data X. Parameters ---------- X : array-like of shape (n_samples, n_dimensions) List of n_features-dimensional data points. Each row corresponds to a single data point. y : Ignored Not used, present for API consistency by convention. Returns ------- log_likelihood : float Log-likelihood of `X` under the Gaussian mixture model. )rfloatrr)r;r@rgrAr\s r scorezBaseMixture.scorews1"a ARWWT//2344r"ct|t|\}}t||d}|j|j |dS)aPredict the labels for the data samples in X using trained model. Parameters ---------- X : array-like of shape (n_samples, n_features) List of n_features-dimensional data points. Each row corresponds to a single data point. Returns ------- labels : array, shape (n_samples,) Component labels. Frr%rL)rrrrr)r;r@rAr\s r predictzBaseMixture.predictsF a A $ /yy99!<1yEEr"ct|t||d}t|\}}|j||\}}|j |S)aEvaluate the components' density for each sample. Parameters ---------- X : array-like of shape (n_samples, n_features) List of n_features-dimensional data points. Each row corresponds to a single data point. Returns ------- resp : array, shape (n_samples, n_components) Density of each Gaussian component for each sample in X. FrrE)rrrrexp)r;r@rAr\rs r predict_probazBaseMixture.predict_probasP  $ /a A2212< 8vvhr"ct|t|j\}}}|dkrtd|jz|jj \}}t |j}|j|t|j|}|jdk(rztjtt|j|t|j||D cgc]"\}} } |j!|| t#| $c} } }} n |jdk(rutjtt|j||D cgc]5\}} |j!|t|j|t#| 7c} }} ntjtt|j|t|j||D cgc]3\}} } ||j%| |ftj&| zz5c} } }} |j)t+t-|D cgc]-} |j/t#|| | |j0|/c} } t3||}|j5| ||| fScc} } }wcc} }wcc} } }wcc} w)ayGenerate random samples from the fitted Gaussian distribution. Parameters ---------- n_samples : int, default=1 Number of samples to generate. Returns ------- X : array, shape (n_samples, n_features) Randomly generated sample. y : array, shape (nsamples,) Component labels. r%zNInvalid value for 'n_samples': %d . The sampling requires at least one sample.fulltiedrHrJ)rArK)rrmeans_rr0rrr, multinomialrweights_covariance_typerOvstackzip covariances_multivariate_normalintstandard_normalsqrtconcatrqlenrint64rrU)r;r]rAr\device_ n_featuresrngn_samples_compr covariancesampler@irgmax_float_dtypes r rzBaseMixture.samples 1$++>Aw q=$'+'8'8:   )) : !2!23 (;    6 ) 7:)$++r:)$*;*;R@&72z6++D*c&kJ A ! !V + +.)$++r:N+ 'v++/0A0A2FF  A 7:)$++r:)$*;*;R@&7  3z6)) /C)Dggj)**  A IIs>23 N1-.'R  57Kzz!?7zCQFFU  s-'J) :J0 +8J6 2J=cN|j|||j|zS)aEstimate the weighted log-probabilities, log P(X | Z) + log weights. Parameters ---------- X : array-like of shape (n_samples, n_features) Returns ------- weighted_log_prob : array, shape (n_samples, n_component) rE)_estimate_log_prob_estimate_log_weightsr?s r rz'BaseMixture._estimate_weighted_log_probs.&&qR&043M3MQS3M3TTTr"cy)zEstimate log-weights in EM algorithm, E[ log pi ] in VB algorithm. Returns ------- log_weight : array, shape (n_components, ) Nr>)r;rAs r rz!BaseMixture._estimate_log_weights rCr"cy)a9Estimate the log-probabilities log P(X | Z). Compute the log-probabilities per each component for each sample. Parameters ---------- X : array-like of shape (n_samples, n_features) Returns ------- log_prob : array, shape (n_samples, n_component) Nr>r?s r rzBaseMixture._estimate_log_probs r"c"t||\}}|j||}t|d|}t|rt j dn t }|5||dd|jfz }ddd||fS#1swY|fSxYw)a@Estimate log probabilities and responsibilities for each sample. Compute the log probabilities, weighted log probabilities per component and responsibilities for each sample in X with respect to the current state of the model. Parameters ---------- X : array-like of shape (n_samples, n_features) Returns ------- log_prob_norm : array, shape (n_samples,) log p(X) log_responsibilities : array, shape (n_samples, n_components) logarithm of the responsibilities rEr%)rMrAignore)underN)rrrrrOerrstaterrX)r;r@rAr\weighted_log_probrcontext_managerrs r rz#BaseMixture._estimate_log_prob_resp&s&aB'A <*>>HI(0$ # /r"c |rdnd}|jdk(rtd|dy |jdk\r/t|jz }td|d|dd |ddy y ) z.Print verbose message on the end of iteration.rzdid not converger%zInitialization .rjz . time lapse z.5fzs lower bound N)r.rrr)r;lbinit_has_converged converged_msgts r r{z'BaseMixture._print_verbose_msg_init_end\sw'9 ?Q <<1  OM?!4 5 \\Q ---A !- aWEs81 r"r:)r%)#__name__ __module__ __qualname____doc__rrrrr8dict__annotations__r<rrBrcr[rRr rfrurvrtr~rrrrrrrrrrrryr{r>r"r r$r$1s"(AtFCDsD89tS$v>?h4?@Haf=> L M (( k;%h4GH $D 10  5"n    <5m+6m+^0(        G&5(F& (KGZ U      '@8 0 r"r$) metaclass)(rr|abcrr contextlibrnumbersrrrnumpyrOsklearnr sklearn.baser r r sklearn.clusterr sklearn.exceptionsr sklearn.utilsrsklearn.utils._array_apirrrrrrsklearn.utils._param_validationrrsklearn.utils.validationrrr!r$r>r"r rsX$ '""BB+1,AC $u, ur"