Dj ^ddlZddlmZddlZddlZddlZddlmZddl m Z ddl m Z ddl mZmZmZdd l mZmZmZdd lmZdd lmZmZdd l mZdd lmZddlmZgdZ dZ!GddeZ"GddeZ# ddZ$e jJjMe jNe jPe jRe jTe jVe$_,dZ-y)N)LinearRegression) BaseARIMA)_doc) _validation)ndiffs is_constantnsdiffs)diff is_iterable check_endog)if_has_delegate)AbstractContext ContextType) _auto_solvers)DTYPE) statsmodels) auto_arima AutoARIMAStepwiseContextc tdD]2}|j|dstjd|zt4|S)N)solver transparamsz?%s has been deprecated and will be removed in a future version.)popwarningswarnDeprecationWarning)kwargsks WC:\Crop_Prediction\Backend\crop-ai-system\venv\Lib\site-packages\pmdarima/arima/auto.py_warn_for_deprecationsr"#sB &. ::a  MM.012, .. McLeZdZejj ddddej Z d dZd dZ e d d dZ e d ddZ e d dd Z e dd Zy)r)yXfit_argsreturn_valid_fitssarimax_kwargsNc' <||_||_||_||_||_||_||_||_| |_| |_ | |_ | |_ | |_ ||_ ||_||_||_||_||_||_||_||_||_||_||_||_||_||_||_||_||_| |_|!|_ |"|_!|#|_"|$|_#|%|_$|&|_%tMdi|'}'|'|_'y)N)(start_pdstart_qmax_pmax_dmax_qstart_PDstart_Qmax_Pmax_Dmax_Q max_ordermseasonal stationaryinformation_criterionalphatest seasonal_teststepwisen_jobs start_paramstrendmethodmaxiteroffset_test_argsseasonal_test_argssuppress_warnings error_actiontracerandom random_staten_fitsout_of_sample_sizescoring scoring_argswith_interceptr"r)(selfr-r.r/r0r1r2r3r4r5r6r7r8r9r:r;r<r=r>r?r@rArBrCrDrErFrGrHrIrJrKrLrMrNrOrPrQrRrs( r!__init__zAutoARIMA.__init__7s.V          "  $%:"  *   (   0"4!2(  ( "4 (,'1&1 r#c J|jsin |j}t|fid|d|jd|jd|jd|j d|j d|jd|jd |jd |jd |jd |jd |jd|jd|jd|j d|j"d|j$d|j&d|j(d|j*d|j,d|j.d|j0d|j2d|j4d|j6d|j8d|j:d|j<d|j>d |j@d!|jBd"|jDd#|jFd$d%d&|jHd'|jJd(|jLd)|jNd*|||_(|S)+alFit the auto-arima estimator Fit an AutoARIMA to a vector, ``y``, of observations with an optional matrix of ``X`` variables. Parameters ---------- y : array-like or iterable, shape=(n_samples,) The time-series to which to fit the ``ARIMA`` estimator. This may either be a Pandas ``Series`` object (statsmodels can internally use the dates in the index), or a numpy array. This should be a one-dimensional array of floats, and should not contain any ``np.nan`` or ``np.inf`` values. X : array-like, shape=[n_obs, n_vars], optional (default=None) An optional 2-d array of exogenous variables. If provided, these variables are used as additional features in the regression operation. This should not include a constant or trend. Note that if an ``ARIMA`` is fit on exogenous features, it must be provided exogenous features for making predictions. **fit_args : dict or kwargs Any keyword arguments to pass to the auto-arima function. r'r-r.r/r0r1r2r3r4r5r6r7r8r9r:r;r<r=r>r?r@rArBrCrDrErFrGrHrIrJrKrLrMrNr)FrOrPrQrRr*))rrr-r.r/r0r1r2r3r4r5r6r7r8r9r:r;r<r=r>r?r@rArBrCrDrErFrGrHrIrJrKrLrMrNrOrPrQrRmodel_)rSr&r'r(r*s r!fitz AutoARIMA.fits2$(;;DKK ++LL+ff + LL + ** +**+**+LL+ff+LL+**+**+**+nn+ ff!+"]]#+$%+&#'"<"<'+(**)+*++,,,-+.]]/+0;;1+2**3+4**5+6;;7+8LL9+:"22;+< $66=+>#44?+@**A+B**C+D;;E+F**G+H;;I+J$K+L $66M+NLLO+P**Q+R ..S+T*W+ Z r#rVc F|jj|||||||S)N)r'startenddynamicreturn_conf_intr>typ)rVpredict_in_sample)rSr'rYrZr[r\r>r]s r!r^zAutoARIMA.predict_in_samples5{{,,+-  r#c@|jj||||S)N) n_periodsr'r\r>)rVpredict)rSr`r'r\r>s r!razAutoARIMA.predicts,{{""+ #  r#c B|jj|f||d|S)N)r'rF)rVupdate)rSr&r'rFrs r!rczAutoARIMA.updates5"t{{!!      r#c6|jjS)z Get a summary of the ARIMA model)rVsummaryrSs r!rezAutoARIMA.summary s{{""$$r#)&rNrrrgrNrrrrrgrTFaic皙?kpssocsbTrNNlbfgs2NNTrKFFN rmseNautoN)NNNFFrilevels)rnNFriNN)__name__ __module__ __qualname__r_AUTO_ARIMA_DOCSTRformat_KWARGS_DOCSTR__doc__rTrWrr^rarcrer,r#r!rr-sF%%,, ** -,G   # OSjHTX     (X     X     X%%r#rc*eZdZdZdfd ZdZxZS)raContext manager to capture runtime context for stepwise mode. ``StepwiseContext`` allows one to call :func:`auto_arima` in the context of a runtime configuration that offers additional level of control required in certain scenarios. Use cases that are either sensitive to duration and/or the number of attempts to find the best fit can use ``StepwiseContext`` to control them. Parameters ---------- max_steps : int, optional (default=100) The maximum number of steps to try to find a best fit. When the number of tries exceed this number, the stepwise process will stop and the best fit model at that time will be returned. max_dur : int, optional (default=None) The maximum duration in seconds to try to find a best fit. When the cumulative fit duration exceeds this number, the stepwise process will stop and the best fit model at that time will be returned. Please note that this is a soft limit. Notes ----- Although the ``max_steps`` parameter is set to a default value of None here, the stepwise search is limited to 100 tries to find a best fit model. Defaulting the parameter to None here preserves the intention of the caller and properly handles the nested contexts, like: >>> with StepwiseContext(max_steps=10): ... with StepwiseContext(max_dur=30): ... auto_arima(sample, stepwise=True, ...) In the above example, the stepwise search will be limited to either a maximum of 10 steps or a maximum duration of 30 seconds, whichever occurs first and the best fit model at that time will be returned c|"d|cxkrdkstdtd||dkr td||d}tt|di|y)Nriz&max_steps should be between 1 and 1000z#max_dur should be greater than zero) max_stepsmax_durr,) ValueErrorsuperrrT)rSr}r~r __class__s r!rTzStepwiseContext.__init__7sm  Y)>$)>EF F*?EF F  7a<BC C#  ot-77r#c"tjSrq)rSTEPWISErfs r!get_typezStepwiseContext.get_typeFs###r#rs)rtrurvrzrTr __classcell__)rs@r!rrs#J 8$r#rc* tdAi|*}*tj|}tj|}tj|'}'tj|)})tj||}tj| } tj ||}tj ||d\}}tj ||d\}}tj || d\}} tj | | d\} } ||f| | ffD](\}+},|,dkr td|+|+dkstd|!r|#dkr tdhd }-||-vrtd |-d |tj}.t|td }|jd}/tjtj|||||*|| ||&|%|'| }0t!|rVt#j$dt'tj(|0|f|ddtj*|(dd|)|$|.| Stj,||%}t/t1|t3j4|/dz }t/t1|t3j4|/dz }t1||}t1||}|sdx} }|j7}1|.t9j;||}2||2j=|z }1|rdx}} |dk(rdx} x} } nV| Tt?|1f||| d|} | dkDr?|=tA|| |}3t3jBt |3djEr| dz} | dkDrtA|1| |}4n|1}4|4jddk(rtd| z|| dkDrtA|| |}3n|}3nd}3|TtG|4f|||d|}|dkDr?|=tA|3|d}3t3jBt |3djEr|dz}|stjH|| |dkDrtA|4|d}4t!|4r|sdntKjLd| d|f}5| dkDr|dk(rtj*|(d }(n8| dkDr|dkDrn-|dk(rn'|dkrtj*|(d }(n tdt'tj(|0|f|d|df|5|(d|)|$|.| S|dkDr(| dkDrt1||dz }| dkDrt1||dz }|(dk(r|| zd v}(|sp|t2jN}n|dkr td!tjPdAid"|d#|d$|0d%|d&| d'|d(|d)|d*|d+| d,| d-|!d.|"d/|#d0|d1|d2| d3|(d4|)}6n|/d5krt1|d}t1|d}dx}} t1||}7t1||}8t1|| }9t1| | }:tjR|fid#|d6|d7|d8|d9|d:|*d;|d2| d<|d=|%d>|&d?|'d|7d%|d|8d|9d&| d|:d'|d)|d*|d+| d,| d1|d@|d3|(|)}6|6jU};t'|;|$|.| S)BNpqPQrz.max_d & max_D must be positive integers (>= 0)z/d & D must be None or a positive integer (>= 0)z5n_fits must be a positive integer for a random search>NrraiserKignorezerror_action must be one of z , but got T)dtypepreserve_series) rCrDrErF fit_paramsrIrKrJrPrOrQr=zEInput time-series is completely constant; returning a (0, 0, 0) ARMA.)rrr)rrrrF)r'orderseasonal_orderrRr)r:r?r7) differenceslag)arraxisa)The seasonal differencing order, D=%i, was too large for your time series, and after differencing, there are no samples remaining in your data. Try a smaller value for D, or if you didn't set D to begin with, try setting it explicitly. This can also occur in seasonal settings when m is too large.)r?r>r1)r.r4rzGdata follow a simple polynomial and are not suitable for ARIMA modelingrp)rrz3max_order must be None or a positive integer (>= 0)r&r' fit_partialr.r4r:r9r0r2r6r8rLrMrNrBr;rKrRr*rnrCrDrErFrrIrJrOrPrQr=r,)+r"val check_kwargscheck_m check_trace check_n_jobscheck_start_max_valuesrtimerrshape functoolspartialsolvers_fit_candidate_modelr rr_return_wrapper_sort_and_filter_fitsauto_interceptcheck_information_criterionintminnpfloorcopyrrWrar r apply_along_axisanyr warn_for_D sm_compatcheck_seasonal_orderinf_RandomFitWrapper_StepwiseFitWrappersolve)r?r@rArBrCrDrErFrGrHrIrJrKrLrMrNr)rOrPrQrRr*r(_d_max_dactionsrY n_samplesrxxlmdiffxregdxssnsearchrrrr sorted_ress< r!rrJskb&11H''(89))*<=##L1L%%n5N Ax A OOE "E   h /F//DNGU//DNGU//DNGU//DNGU5zAu:.3 F A:MN N >Av "233 3&1*/0 09G7"#\34 4 IIKE AUD9A I##$$!+!-!3K"1~ 4 5  ) )##/##5&$/ %  ue - - ''(=(: < E288IM23 4E E288IM23 4E'5!G'5!G  A B}   # #Aq ) A  A AvEE  B *!-u *( * q5Q]A1!4H"";H1EIIKQ 1u "! +  xx{aP    } q5A1!4HH  y        q5Q]H!;H"";H1EIIKQ  a 1u "! +2"*l//Aq! = q5Q!V //ENUq1u  !V  U //EN;< <  ) )a)#  %   ue   1u 19q1u%E 19q1u%Ea%F*   I ]./ /**   $              &   ! "# $*% &*' . r>'1oG'1oG ! !Gg        ,,   &         0  &  2  &   ! "# $% &' () *+ ,- ./ 01 23 4#85 6*9 >J :'8% GGr#)r&r'r(r*r)czt|s|g}|r#tdtj|z z|s|dS|S)aIf the user wants to get all of the models back, this will return a list of the ARIMA models, otherwise it will just return the model. If this is called from the end of the function, ``fits`` will already be a list. We *know* that if a function call makes it here, ``fits`` is NOT None or it would have thrown an exception in :func:`_post_ppc_arima`. Parameters ---------- fits : iterable or ARIMA The ARIMA(s) return_all : bool Whether to return all. zTotal fit time: %.3f secondsr)r printr)fits return_allrYrKs r!rrsD$ t v  , e0CDE Aw Kr#))NrNrrgrrgrNrrrrrgrTFrhrirjrkTrNNrlrmNNTrKFFNrnFrroNrpN).numpyrsklearn.linear_modelrrrrbaserr%rrrutilsr r r r r rutils.metaestimatorsr_contextrrrr compat.numpyrcompatrr__all__r"rrrrwrx _Y_DOCSTR _EXOG_DOCSTR_FIT_ARGS_DOCSTR_SARIMAX_ARGS_DOCSTR_VALID_FITS_DOCSTRrzrr,r#r!rs:1  //2222' - _% _%H6$o6$v               UtHp ,,33 nn   " ",,-- 4 r#