M/PhFdZddlZddlmZddlZddlZddlm Z ddl m Z ddl m Z ddlmZddlmcmZddlmZd d lmZmZd d lmZmZmZd d lmZd d lm Z m!Z!m"Z"m#Z#m$Z$m%Z%m&Z&GddeZ'GddeZ(GddeZ)ej*e)e(dS)zs Vector Autoregressive Moving Average with eXogenous regressors model Author: Chad Fulton License: Simplified-BSD N)warn)Appender)Bunch)_is_using_pandas) var_model)EstimationWarning)INVERT_UNIVARIATESOLVE_LU)MLEModel MLEResultsMLEResultsWrapper)Initialization) is_invertibleconcat prepare_exog!constrain_stationary_multivariate#unconstrain_stationary_multivariateprepare_trend_specprepare_trend_dataceZdZdZ dfd Zdd Zed Zed Zed Z dZ dZ fdZ ddZ ejdZeejj dfd ZxZS)VARMAXuM Vector Autoregressive Moving Average with eXogenous regressors model Parameters ---------- endog : array_like The observed time-series process :math:`y`, , shaped nobs x k_endog. exog : array_like, optional Array of exogenous regressors, shaped nobs x k. order : iterable The (p,q) order of the model for the number of AR and MA parameters to use. trend : str{'n','c','t','ct'} or iterable, optional Parameter controlling the deterministic trend polynomial :math:`A(t)`. Can be specified as a string where 'c' indicates a constant (i.e. a degree zero component of the trend polynomial), 't' indicates a linear trend with time, and 'ct' is both. Can also be specified as an iterable defining the non-zero polynomial exponents to include, in increasing order. For example, `[1,1,0,1]` denotes :math:`a + bt + ct^3`. Default is a constant trend component. error_cov_type : {'diagonal', 'unstructured'}, optional The structure of the covariance matrix of the error term, where "unstructured" puts no restrictions on the matrix and "diagonal" requires it to be a diagonal matrix (uncorrelated errors). Default is "unstructured". measurement_error : bool, optional Whether or not to assume the endogenous observations `endog` were measured with error. Default is False. enforce_stationarity : bool, optional Whether or not to transform the AR parameters to enforce stationarity in the autoregressive component of the model. Default is True. enforce_invertibility : bool, optional Whether or not to transform the MA parameters to enforce invertibility in the moving average component of the model. Default is True. trend_offset : int, optional The offset at which to start time trend values. Default is 1, so that if `trend='t'` the trend is equal to 1, 2, ..., nobs. Typically is only set when the model created by extending a previous dataset. **kwargs Keyword arguments may be used to provide default values for state space matrices or for Kalman filtering options. See `Representation`, and `KalmanFilter` for more details. Attributes ---------- order : iterable The (p,q) order of the model for the number of AR and MA parameters to use. trend : str{'n','c','t','ct'} or iterable Parameter controlling the deterministic trend polynomial :math:`A(t)`. Can be specified as a string where 'c' indicates a constant (i.e. a degree zero component of the trend polynomial), 't' indicates a linear trend with time, and 'ct' is both. Can also be specified as an iterable defining the non-zero polynomial exponents to include, in increasing order. For example, `[1,1,0,1]` denotes :math:`a + bt + ct^3`. error_cov_type : {'diagonal', 'unstructured'}, optional The structure of the covariance matrix of the error term, where "unstructured" puts no restrictions on the matrix and "diagonal" requires it to be a diagonal matrix (uncorrelated errors). Default is "unstructured". measurement_error : bool, optional Whether or not to assume the endogenous observations `endog` were measured with error. Default is False. enforce_stationarity : bool, optional Whether or not to transform the AR parameters to enforce stationarity in the autoregressive component of the model. Default is True. enforce_invertibility : bool, optional Whether or not to transform the MA parameters to enforce invertibility in the moving average component of the model. Default is True. Notes ----- Generically, the VARMAX model is specified (see for example chapter 18 of [1]_): .. math:: y_t = A(t) + A_1 y_{t-1} + \dots + A_p y_{t-p} + B x_t + \epsilon_t + M_1 \epsilon_{t-1} + \dots M_q \epsilon_{t-q} where :math:`\epsilon_t \sim N(0, \Omega)`, and where :math:`y_t` is a `k_endog x 1` vector. Additionally, this model allows considering the case where the variables are measured with error. Note that in the full VARMA(p,q) case there is a fundamental identification problem in that the coefficient matrices :math:`\{A_i, M_j\}` are not generally unique, meaning that for a given time series process there may be multiple sets of matrices that equivalently represent it. See Chapter 12 of [1]_ for more information. Although this class can be used to estimate VARMA(p,q) models, a warning is issued to remind users that no steps have been taken to ensure identification in this case. References ---------- .. [1] Lütkepohl, Helmut. 2007. New Introduction to Multiple Time Series Analysis. Berlin: Springer. Nr rc unstructuredFTr c  l |_|_|_|_|_t |d_t |d_|dvrtdjdkrjdkrtdjdkr jdkrtdt|_ | _ tj \__jjdkojddk_t%|\_}jdk_t+|dst-j|}t1jd} | jz_|jd} | } | jz}| dd| d t8t:zt=j|f||| d | jdksjdkrjs d j _!i_"j#jzj"d <j#d zjzj"d<j#d zjzj"d<j#jzj"d<jdkrj#j"d<n5jdkr*t j#j#dzzd z j"d<j#jzj"d<tIj"%_&tOjjj(dzj }|dd_)|dd_*jdkrjr jdkr(t-j+j,j(fj d<t-j-j#}dj d|z<jdkrIt-j-jdz j#z}|dj#z|df}dj d|z<t-j-jdz j#z}|d| dzj#zz|d| j#zzf}dj d|z<t-j-j#}dj d|z<|d| j#zz|df}jdkr dj d|z<jr*jdkrt,j.dd| ddf_/n4jdks jdkrt,j.dd| ddf_/jdkrt,j.dd| ddf_0nt,j.dd| | df_0jdkr"dt-j-j#z_1n)jdkrt-j2j#_3jr!dt-j-j#z_4fd}d}|d |\_5}|d|\_6}|d|\_7}|d|\_8}|d|\_9}|d|\_:}d_;xj<gdt{| >zz c_<dS) Nrr )diagonalrz3Invalid error covariance matrix type specification.zNInvalid VARMAX(p,q) specification; at least one p,q must be greater than zero.zcEstimation of VARMA(p,q) models is not generically robust, due especially to identification issues.initialization stationaryinversion_method)exogk_statesk_posdefFtrendarma regressionr state_covrobs_cov)offsetstate_intercept)design) transition) selectionr/)r))r*c^j|}tj|||z}||z }||fSN) parametersnps_)keyr+length param_sliceselfs a/var/www/html/test/jupyter/venv/lib/python3.11/site-packages/statsmodels/tsa/statespace/varmax.py_slicezVARMAX.__init__.._slice s9_S)F%v 67K f F& &)orderr$error_cov_typemeasurement_errorenforce_stationarityenforce_invertibility)?r>r?r@rAr=intk_ark_ma ValueErrorrrr$ trend_offsetrpolynomial_trendk_trendsize_trend_is_constrk_exogmle_regressionrr4 asanyarraymax_k_ordershape setdefaultr r super__init__ssm_time_invariantr3k_endogsumvaluesk_paramsrnobs _trend_data _final_trendzerosr" diag_indicesr5_idx_state_intercept_idx_transition_idx_state_cov tril_indices_idx_lower_state_cov _idx_obs_cov _params_trend _params_ar _params_ma_params_regression_params_state_cov_params_obs_cov _final_exog _init_keyslistkeys)r9endogr!r=r$r>r?r@rArFkwargs _min_k_arrVr#r" trend_dataidxr;r+ __class__s` r:rSzVARMAX.__init__s7 -!2$8!%:" aMM aMM  != = =/00 0 9>>di1nn?@@ @ 9q==TY]] =" $ $ $  (.@.L.L+t| $ 5 :a ?!> $ 5a 8A = +400d#kAo t,, )M%((E 1%% !DI- +a.T]* *L999 ,.?(.JKKK   x(  FL   ;??t|a//8L/',DH $#'<$,#> $ a$) ; $ a$) ;(, t{(B %  * , ,+/!>>     N 2 2(* (E(ED %  ! M ,rt|/L/L LD  ' ' ' ' ' %+VGV%<%<"F"(&v"6"6"(&v"6"6*0&v*F*F')/ V)D)D&'-vi'@'@$f  5557;FKKMM7J7JK Kr<c "|j|fd|i|S)Nr!)_clone_from_init_kwds)r9ror!rps r:clonez VARMAX.clone7s")t)%EEdEfEEEr<c"dttfiS)Nfit) VARMAXResultsVARMAXResultsWrapper)r9s r: _res_classeszVARMAX._res_classes:s ';<==r<c tj|jtj}t j|j}|}tj | d}d}|j dkr*|j dkrtj |j|jf}n%|j dkr|j}n|j dkr|j}tjtj|r;tjtj|d}||}|||}tjd}tjd}|j dks |j dkrtj||}|tj||z}|j dkr|d|j j}|jdkr||j dj}g}|jdkr|jnd} t1j|} | | dd} |jdkr0tj| jj}| j}|jdkr|jr| |j|jz|jj |j|j|jj} tCdgtE| z} | stGd |dz}g}|j$dkrt1j|}||j$dd}tj|jj}|j%r| |j|j$z|jj |j|j|j$j} tCdgtE| z}|stGd |dz}|jdkr|j dks|j&r| |j|jz|jj |j|j|jj} tj'|jtj(| dz }|j dkrtj||}|j&dkrtj||}|j dkr|||j)<|jdkr |||j*<|j$dkr |||j+<|j&r|||j,<|j-d kr"| j./||j0<nV|j-d krKtj1| j.}||j2||j0<|j3rN|j$dkr"|j./||j4<n!| j./||j4<|S) NdtypeW) requirementsrr )axisn)maxlagsicr$z\Non-stationary starting autoregressive parameters found. Using zeros as starting parameters.z\Non-stationary starting moving-average parameters found. Using zeros as starting parameters.rr)5r4r]rYfloat64pd DataFramerocopy interpolaterequirebfillrHrKc_r[r!anyisnanlinalgpinvdotTrVrCrVARryarrayparamsravelresidr@reshaperrmrrDrArLeyerWrerfrgrhr>sigma_urricholeskyrcr?rj)r9rror!mask trend_params exog_paramstrendexog_params ar_paramsrCmod_arres_arcoefficient_matricesr ma_paramsmod_mares_ma invertibletmp cov_factors r: start_paramszVARMAX.start_params>s$-rz::: TZ__..//!!## 5;;==s;;; >  RVD"233 3E|a/  >@ |a.t|}}=?   IMMtyyqu%%DT== 9q==//17799I  9q==T6=!!L49,dlgdlDL$)<g}jtjtsgjdkrvt jD]a}jdD]?}|dkr|d|zgz }|dkr|d|zgz }.|d||fzgz }@b|fdt jDz }|fdt jDz }|fdt jDz }jd kr$|fd t jDz }n.jd kr#|fd t jDz }j r#|fd t jDz }|S)Nrz intercept.%sr zdrift.%sz trend.%d.%sc g|]G}tjD]0}tjD]}d|dz||fz1HS)z L%d.%s.%sr )rangerCrV.0jik endog_namesr9s r: z&VARMAX.param_names..s   49%%  4<((    1Q3 A A? ?     r<c g|]G}tjD]0}tjD]}d|dz||fz1HS)z L%d.e(%s).%sr )rrDrVrs r:rz&VARMAX.param_names..s   49%%  4<((    ac;q>;q>B B     r<cpg|]2}tjD]}dj|d|3S)zbeta..)rrK exog_names)rrrrr9s r:rz&VARMAX.param_names..se   4;''   :DOA& 9 9Q 9 9    r<rc&g|] }d|zS)z sigma2.%srrrs r:rz&VARMAX.param_names..s2k!n,r<rcg|]<}t|dzD]'}||kr d|znd|d|(=S)r z sqrt.var.%sz sqrt.cov.r)r)rrrrs r:rz&VARMAX.param_names..s{qs 4566Q//>[^>>k!n>>r<c&g|] }d|zS)zmeasurement_variance.%srrs r:rz&VARMAX.param_names..s2*KN:r<) r isinstancermrHrrVrGnonzeror>r?)r9 param_namesrrrs` @r:rzVARMAX.param_namess & $*D11 (&-K diagrirTrcrrrfrrVrrrDrArrgrhr?rj)r9 unconstrained constrainedr)state_cov_lower coefficientsrvariances r:transform_paramszVARMAX.transform_paramss,a888 h}2-:MNNN +88J*K D&' 9q==T6="j00GM$2H$I1$LMM $66"$(48K+@+F1>1D#F#F#F"$"89  9:F?O4EFF )9AA dlTY688L2, JJ + (+?+E+E+G+GK ( (+8+IK ( 9q==T7=t|=3FGGGI(9AA dlTY688L2, JJ + (+?+E+E+G+GK ( (+8+IK ( $1 2 D+,  * , ,d45q8 . / / N 2 2d45 . /  ! 8d23Q6 , -r<ctj|d}tj|j|j}||j||j<|jdkr|jr|jdkr tj ||j }ne|jdkrZtj|j dj|j}||j ||j <tj ||j}||j|j|j|jz}t%||\}}|||j<n||j||j<|jdkr|jrtj|j|j}||j|j|j|jz}t%||\}}|||j<n||j||j<||j||j<|jdkr||j dz||j <n |jdkr||j ||j <|jr||jdz||j<|S) a Transform constrained parameters used in likelihood evaluation to unconstrained parameters used by the optimizer. Parameters ---------- constrained : array_like Array of constrained parameters used in likelihood evaluation, to be transformed. Returns ------- unconstrained : array_like Array of unconstrained parameters used by the optimizer. r rr~rrrr)g?)r4rr]rPrrerCr@r>rrirTrcrrrfrrVrrrDrArrgrhr?rj)r9rrr)rrunconstrained_matricesrs r:untransform_paramszVARMAX.untransform_paramsRs h{!444 !2+:KLLL -88J,K d() 9q==T6="j00GK0F$GHH $66"$(48K+@+F1<1B#D#D#D  67  9:F?O4EFF 't7?? dlTY688L4L)LL - "H-C-I-I-K-KM$/ * *-8-IM$/ * 9q==T7=t|;3DEEEI&t7?? dlTY688L4L)LL - "H-C-I-I-K-KM$/ * *-8-IM$/ * / 0 d-.  * , ,D23S8 $0 1 1 N 2 2D23 $0 1  ! 8D0136 $. /r<ct|tjt |jdd}dtj|j|D\}}}}}}|j ro|j dkrd|j dks |j dkrN| |}t||dk}|r|std|jrk|jdkrb|j s |jdkrT| |}t||dk}|r|stddSdSdSdSdS)Nr,c3>K|]}|VdSr2)tolist)rarrs r: z2VARMAX._validate_can_fix_params..sC,J,J CJJLL,J,J,J,J,J,Jr<rr zCannot fix individual autoregressive parameters when `enforce_stationarity=True`. In this case, must either fix all autoregressive parameters or none.zCannot fix individual moving average parameters when `enforce_invertibility=True`. In this case, must either fix all moving average parameters or none.)rR_validate_can_fix_paramsr4cumsumrmr3rX array_splitrr@rCrV issupersetlen intersectionrErArD) r9rix_ar_namesma_namesfix_allfix_anyrts r:rzVARMAX._validate_can_fix_paramss ((555 YtDO224455 6 6ss ;,J,J$&N43CR$H$H,J,J,J(Hh1a  $ "Q|a49q==%00:: 00::;;a?"7"$!"""  % "$)a--| "ty1}}%00:: 00::;;a?"7"$!""" " "-- "" -}""r<c||||}|jr||j|j|jj}tj|j dd|}|j|j |j <|j ,tj|j ||j dd|jdf<|j dkr|js)tjd|j}||j dddf<||j|j|j j}|jr|}n"tj|jdd|}|j |j xx|jz cc<|jT|j djdkr>|j dd|jddfxxtj|j|jz cc<|jr@|j 9tjtj|j} | |j dd|jdf<||j|j|j|jz} ||j|j|j|jz} tj| | f|j |j<|jdkr||j|j |j<nm|jdkrbtj|j d j |j} ||j| |j!<tj| | j|j d <|j"r||j#|j |j$<dSdS) N) transformedincludes_fixedr r-r,rr~rrr))% handle_paramsrLrhrrVrKrr4rr!rTr_rkrHrrrerJr[r\stopnanrfrCrgrDrr`r>rirar]rPrcr?rjrd) r9rrr complex_stepr interceptzerorrr&r'rs r:updatez VARMAX.updates##F 3A$CC   3 !89AA dk+++, ty}k::I2;+DHT. /+AC$kB3B3*MT\M2=>  / / /!-1"5:b@@*MT\M233>???26%|D5D5567???   A4#3#;(26666C=@DH&  r9 :DO $ , , L$,244 DO $ , , L$,244)+r2v%&  * , ,t-. HT( ) ) N 2 2 htx '<'B-3\;;;Ot-. D5 6$&F?O>8D::d2A2h??DD!:::$*(+T[N(;(;(;(;(+DJ&9:::: $D  + EEE*D   {D  * * * *s#A%%AB+6C C c  ||5tj||f||||||| | | | d | }dddn #1swxYwY|S)N) measurement_shocks state_shocks initial_stateanchor repetitionsr! extend_model extend_kwargsrr)rrRsimulate)r9r nsimulationsrrrrrr!rrrrrpoutrts r:rzVARMAX.simulate)s  ! !$ ' '  "%''" 9K);T)'   C                s'A  AA)NrrrFTTr r2)TFF) NNNNNNNNTF)__name__ __module__ __qualname____doc__rSrwpropertyr|rrrrrr contextlibcontextmanagerrrr r __classcell__rts@r:rrsbbH>ABGBFqKqKqKqKqKqKfFFFF>>X>~~X~@==X=~QQQfKKKZ"""""<?D!AGAGAGAGF&+&+&+PXh'((@D?C;?FK     )(     r<rcNeZdZdZ dfd ZddZejdZejdZ e e j j dfd Z e e j j dfd Z dd Ze e jjdfd ZxZS)rza Class to hold results from fitting an VARMAX model. Parameters ---------- model : VARMAX instance The fitted model instance Attributes ---------- specification : dictionary Dictionary including all attributes from the VARMAX model instance. coefficient_matrices_var : ndarray Array containing autoregressive lag polynomial coefficient matrices, ordered from lowest degree to highest. coefficient_matrices_vma : ndarray Array containing moving average lag polynomial coefficients, ordered from lowest degree to highest. See Also -------- statsmodels.tsa.statespace.kalman_filter.FilterResults statsmodels.tsa.statespace.mlemodel.MLEResults Nc tj|||||fi|tdi|jj|jj|jj|jj|jj|jj |jj |jj |jj |jj |jjd |_d|_d|_|jj dkr~t%j|j|jj}|jj}|jj } ||| z|j||| j|_|jj dkrt%j|j|jj} |jj}|jj } | || z|j||| j|_dSdS)N) r>r?r@rArFr=rCrDr$rHrKrr)rRrSrmodelr>r?r@rArFr=rCrDr$rHrK specificationcoefficient_matrices_varcoefficient_matrices_vmar4rrrfrVrrrg) r9rrfilter_resultscov_typecov_kwdsrprrVrCrrDrts r:rSzVARMAXResults.__init__Ss 6>8X  AG   #  "j7!%!=$(J$C%)Z%E J3Z%JOJOZ%z)j'#& &   *)-%(,% :?Q  TZ-B!CDDIj(G:?D!!'D.'::<ggw--a  ) :?Q  TZ-B!CDDIj(G:?D!!'D.'::<ggw--a  ) ) )  r<c |L||j|j|dd}|j}|jd}|jd}n|jd}|jd}|d|j|jjz|jj|fd|i|}t|j d|||j _ |j ||j} n||j} | S) Nr )r!).r).r,rFr!knownconstantstationary_cov)get_predictionrZprediction_resultspredicted_statepredicted_state_covrQrrFrwrr"rTrsmoother_resultssmoothrfilter) r9ror!rpfcast fcast_resultsrinitial_state_covmodress r:extendzVARMAXResults.extends  '' 4948'LLE!4M)9&AM - A& I   09M $ 8 A .$)dj6M*MNNNdju::4:6::!/ L'M,".".".  ,**T[))CC**T[))C r<c#K|j}||5|jjdddf}||j|dd|jdf|jjd|jdf< dV||jjdddf<n#||jjdddf<wxYw ddddS#1swxYwYdS)az Set the final state intercept value using out-of-sample `exog` / trend Parameters ---------- exog : ndarray Out-of-sample `exog` values, usually produced by `_validate_out_of_sample_exog` to ensure the correct shape (this method does not do any additional validation of its own). out_of_sample : int Number of out-of-sample periods. Notes ----- This context manager calls the model-level context manager and additionally updates the last element of filter_results.state_intercept appropriately. Nr,r-)rrr r-rrrV)r9r!rrs r:rzVARMAXResults._set_final_exogsI(j   & & I I-=aaaeDK JJt{ # # #%| |R78   /  b0@ A I=H#3AAArE::[#3AAArE:HHHH: I I I I I I I I I I I I I I I I I Is*AB79B=B7B&&B77B;>B;c#DK|o|jjdk}|r>t|jjddt jd|jjfg}|jjdkr,t|jjdd|ddg}nd}|jj|j zdz }|j |||}|j j dddf}|j j dddddf} |j|| ||jddd } | j dddf|j j dddf< dV|rtj|j j dddf<dSdS#|rtj|j j dddf<wxYw) a Set the final predicted state value using out-of-sample `exog` / trend Parameters ---------- exog : ndarray Out-of-sample `exog` values, usually produced by `_validate_out_of_sample_exog` to ensure the correct shape (this method does not do any additional validation of its own). out_of_sample : int Number of out-of-sample periods. Notes ----- We need special handling for forecasting with `exog`, because if we had these then the last predicted_state has been set to NaN since we did not have the appropriate `exog` to create it. rr,Nr )r!rFrT)rr return_ssm)rrKrror4r]rVr!rFrZrwr rrrTinitialize_knownrrr) r9r! out_of_sampleflag tmp_endogtmp_exogtmp_trend_offsettmp_modrrtmp_ress r:_set_final_predicted_statez(VARMAXResults._set_final_predicted_states(6!2Q!6  0  %rxDJ4F0G'H'H JKKIz 1$$!4:?233#7bqb"BCC#z6BQF j&&yx4D'FFG*:111b5AH!0DQQQ2XNN K ( ((8F ) H H HnnT[d48T%KKG '2.   /2 6 D EEE D=?V#3AAArE::: D Dt D=?V#3AAArE:CCCCs E>>!FF predictedc |d}|j|||d\}} } } |j|| }i} |jjdkr|jj|jz| d<||5||| 5tj d||||||| d|} dddn #1swxYwYdddn #1swxYwY| S)NrT)silentrF)startenddynamicinformation_setindexr!rr) r_get_prediction_index_validate_out_of_sample_exogrHrFrZrr+rRr)r9r/r0r1r2r3r!rp_start_endr$rrrrts r:rzVARMAXResults.get_predictions =E J , ,UCt , L L 'mQz66t]KK :  ! ! '$)3 . ) ! !$ ' ' ; ;00}EE ; ;,egg,;S'$35t"/;;4:;; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ;  s6C"C; CC CC CC"%C"c ||dkrd} n,|dkr|j} n|j|\} } } | dkr |j| z} | |jkrtdt | |z|jz d} |j|| }||| 5tj|f|||||||| d| }dddn #1swxYwY|S)Nr/rr0z4Cannot anchor simulation after the estimated sample.)rrrrrr!rr) rZr_get_index_locrErNr5r+rRr)r9rrrrrrr!rrrpilocrr$rrts r:rzVARMAXResults.simulatesl >Vw..DD u__9DD226::JD!Q !889t#D $)  ()) )D</$);Q?? z66t]KK  , ,T= A A  "%''"1C);T)   C                s/$CC#&C#cd}|j|jz }|jjdkr|dkr|jj| d}t j5} | |j|| ||||j |j ||||} dddn #1swxYwY| S)Nr)r/r0revisions_details_start state_index) rZrrKr!r ExitStack enter_contextrr+rnews) r9previousr/r0periodsr<r=r!r$stackrs r:_news_previous_resultsz$VARMAXResults._news_previous_results(s1 HM1 : q ]Q%6%6:?M>??3D ! # # )u    > >t D D E E E    C Cm!%!% & & &',,)C(?'-))C  ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) sA;CCC皙?Tc . ddlm|j}|jdkr!|jdkrd}d|jd|jd}n$|jdkr d}d|jz}n d }d|jz}|jdkr|d z }||zg}|jdkr|d |jr|d t ||| }|rtj t|j}dfd } |jj} |jj} |jj} |jj} |jj}g}t#| D]}g}d}| dkr7|tj ||| | zz| || | zz }| dkrI|| z| z}|dz| z| z}||tj ||z|| | dzzz }| dkrI|| z| z}|dz| z| z}||tj ||z|| | dzzz }|dkr<||tj ||z|dz|zz|| |zz }|jjr9|tj|jj|z dz dtj|}|||jj}t-|t.s|g}d||z}| |||}|j|tj t|j|jj}| ||dd}|j|g}||gfD]P}tj|}t|dkr||Qtj|}tjt/t7|t7|}t|dkr)| ||dd}|j||S)Nr)summary_paramsVARMA(,)rz(%s)VMAXrzmeasurement error)alphar/ model_namedisplay_paramsTc||j||j||j||j|||f}g}t j|jj| D]Z}|r1d | ddd}n|}||j vrd|z}| |[ |d|d|S)Nrr,z %s (fixed)F)ynamexnamerNuse_ttitle)rbsezvaluespvaluesconf_intr4rdatarrjoinsplit fixed_paramsappend) r9rrU strip_endrrname param_namerNrGs r: make_tablez)VARMAXResults.summary..make_tablefsT[.|D)4<+=}}U++D13! HTY%:;;DAHHJJ33D *%(XXdjjoocrc.B%C%C %) t000%1J%> &&z2222%~c[,1eMMMMr<r r%rzResults for equation %szError covariance matrixF)r_zOther parameters)T)statsmodels.iolib.summaryrGrrCrDrKrHr^r?rRsummaryr4arangerrrrVrrrY concatenaterrrmtablesriflattenset difference)r9rNr/separate_paramsspecrOr=rdindicesrbrVrCrDrHrK endog_masksrmasksr+r0rrrUtablestate_cov_maskm inverse_maskrGrts ` @r:rdzVARMAXResults.summaryEs0<<<<<<! 9q==TY]] J0 00DI000EE Y]]Jdi(EEJdi(E ;?? # J 5() rs ......))))))333333//////'''''''''======66666666==========******X X X X X XX X X vHHHHHJHHHV /////,///*M:::::r<