>h:vSSKJr SSKrSSKJr SSKJr SS/r"SS5r "SS5r "S S \5r g) )GLSN)LikelihoodModelResults)sparseSURSem2SLScJ\rSrSrSrS SjrSrSrSrSr S S jr S r S r g)raF Seemingly Unrelated Regression Parameters ---------- sys : list [endog1, exog1, endog2, exog2,...] It will be of length 2 x M, where M is the number of equations endog = exog. sigma : array_like M x M array where sigma[i,j] is the covariance between equation i and j dfk : None, 'dfk1', or 'dfk2' Default is None. Correction for the degrees of freedom should be specified for small samples. See the notes for more information. Attributes ---------- cholsigmainv : ndarray The transpose of the Cholesky decomposition of `pinv_wexog` df_model : ndarray Model degrees of freedom of each equation. p_{m} - 1 where p is the number of regressors for each equation m and one is subtracted for the constant. df_resid : ndarray Residual degrees of freedom of each equation. Number of observations less the number of parameters. endog : ndarray The LHS variables for each equation in the system. It is a M x nobs array where M is the number of equations. exog : ndarray The RHS variable for each equation in the system. It is a nobs x sum(p_{m}) array. Which is just each RHS array stacked next to each other in columns. history : dict Contains the history of fitting the model. Probably not of interest if the model is fit with `igls` = False. iterations : int The number of iterations until convergence if the model is fit iteratively. nobs : float The number of observations of the equations. normalized_cov_params : ndarray sum(p_{m}) x sum(p_{m}) array :math:`\left[X^{T}\left(\Sigma^{-1}\otimes\boldsymbol{I}\right)X\right]^{-1}` pinv_wexog : ndarray The pseudo-inverse of the `wexog` sigma : ndarray M x M covariance matrix of the cross-equation disturbances. See notes. sp_exog : CSR sparse matrix Contains a block diagonal sparse matrix of the design so that exog1 ... exogM are on the diagonal. wendog : ndarray M * nobs x 1 array of the endogenous variables whitened by `cholsigmainv` and stacked into a single column. wexog : ndarray M*nobs x sum(p_{m}) array of the whitened exogenous variables. Notes ----- All individual equations are assumed to be well-behaved, homoskedastic iid errors. This is basically an extension of GLS, using sparse matrices. .. math:: \Sigma=\left[\begin{array}{cccc} \sigma_{11} & \sigma_{12} & \cdots & \sigma_{1M}\\ \sigma_{21} & \sigma_{22} & \cdots & \sigma_{2M}\\ \vdots & \vdots & \ddots & \vdots\\ \sigma_{M1} & \sigma_{M2} & \cdots & \sigma_{MM}\end{array}\right] References ---------- Zellner (1962), Greene (2003) Nc .^[T5S-S:wa[S[T5-5eU(a"UR5S;a[SU-5eX0l[TSSS25nX@l[ R "U4Sj[U555nXPl[ R"TSSS25nX`l [URSRS5Ul /n/nTSSS2V s/sH=oRUR[ RR!U 5- 5PM? n TSSS2V s/sH3oR[ RR!U 5S- 5PM5 n [ R"U5Ul[ R"U5Ul[&R("[+URU-5[+[ R,"UR$S-5545n [ R."[ R0"SUR$S-455Ul[U5HPn TSSS2U XUR-U S-UR-2UR2U UR2U S-24'MR U R55Ul[ R8"U5(a[ R"U5nOUc/n [U5H^n U R[;XkUSS2UR2U UR2U S-245R=5R>5 M` [ R"U 5RAUS5n URCU 5nX l"[ RRG[ RRIURD55RJUl&URO5 gs sn fs sn f) NrQsys must be a list of pairs of endogenous and exogenous variables. Got length %s)dfk1dfk2zdfk option %s not understoodc3b># UH$n[R"TSSS2U5v M& g7f)rNr )npasarray).0isyss mC:\Users\julio\OneDrive\Documentos\Trabajo\Ideas Frescas\venv\Lib\site-packages\statsmodels/sandbox/sysreg.py SUR.__init__..js+JArzz#add)A,77s,/)(len ValueErrorlower_dfk_Mr column_stackrangeexogrendogfloatshapenobsappendlinalg matrix_rankdf_residdf_modelr lil_matrixintsumcumsumhstack_colstocsrsp_exoganyrfitresidreshape_compute_sigmasigmacholeskypinvT cholsigmainv initialize) selfrr8dfkMr!r"r)r*_r2rresidss ` r__init__ SUR.__init__\s8 s8a<1 %'*3x01 1 99;/1 !?3!GHH ADqD N JqJJ  3ss8$ $**Q---a01 HKADqD R 1RYY%:%:1%== > R@CADqD J 1..q1A5 6 J 8,  8, ##S1%5 t}}Q' (%*+YYryy!T]]1_)=>? qA58AYq\ diiK1dii/JJqM$**QqS/12 3}} 66%==JJu%E ]F1X c%(4JJqM$**QqS/112,3447CE%%AZZ'//"5F''/E II..ryy~~JJ0 !!"  5 SJs :AP  :PcURUR5UlURUR5Ul[ R RUR5Ul[ R"UR[ R"UR55Ul S[ R/0Ul SUlg)Nparamsr)whitenr"wendogr2wexogrr'r: pinv_wexogdot transposenormalized_cov_paramsinfhistory iterationsr>s rr=SUR.initializeskk$**- [[. ))..4%'VVDOO T__-&/" BFF8, c@URSRU5 g)NrF)rOr&)r>rFs r_update_historySUR._update_historys X%%f-rScURnURn[R"XR5nUR (dUnGO'UR R 5S:Xa|[R"US-5n[U5HBn[U5H0nURUS-URUS--S-XVU-'M2 MD URX"5 O[R"US-5n[U5HTn[U5HBnU[R"URUS-URUS-5- XVU-'MD MV URX"5 [RR[RRXE- 55RUlXE- $)zB Computes the sigma matrix and update the cholesky decomposition. r r rg?)rr%rrKr;rrzerosr r*r6maxr'r9r:r<)r>rBr@r%sigdivrjs rr7SUR._compute_sigmasc GGyyffVXX&yyC YY__ & (((1a4.C1XqA!%q!1!!3!]]1-a/!147 9C!H" KK ((1a4.C1XqA#bffT]]1-=a-? a(*', ,C!H" KK II..ryy~~cg/FGIIwrScURnXRLaZ[R"[R"UR [R "U55URSS55$XRLaF[R"UR [R "X"55U-R5$g)z SUR whiten method. Parameters ---------- X : list of arrays Data to be whitened. Returns ------- If X is the exogenous RHS of the system. ``np.dot(np.kron(cholsigmainv,np.eye(M)),np.diag(X))`` If X is the endogenous LHS of the system. rrN) r%r"rrKkronr<eyer6r2rtoarray)r>Xr%s rrG SUR.whitens yy ?66"''$"3"3BFF4LA "Q! ! ,, KK 1 1 4%''()*1') 4rSc[R"UR5(d+URURUR 5UlUR n[R"URUR5nURU5 U=RS- sl U(d[XUR5nU$URSnU(Ga[R"[R"USUS- 5U:5(GaURU:GaUR U-R#US5nURU- n URU 5UlUR%UR5Ul UR%UR 5Ul[R(R+UR&5Ul[R"UR[R,"UR55Ul [R"URUR5nURU5 U=RS- sl U(aN[R"[R"USUS- 5U:5(aURU:aGM[XUR5nU$)aB igls : bool Iterate until estimates converge if sigma is None instead of two-step GLS, which is the default is sigma is None. tol : float maxiter : int Notes ----- This ia naive implementation that does not exploit the block diagonal structure. It should work for ill-conditioned `sigma` but this is untested. rrFr)rr3r8r7r"r!rrKrJrHrUrP SysResultsrMrOabsr2r6rGrIr'r:rL) r>iglstolmaxiterr@betasur_fitconv fittedvaluesrBs rr4SUR.fits"vvdjj!!,,TZZCDJ GGvvdoot{{3 T" 1 T-G-GHGN||H%rvvd2hb&9:S@AA7* LL-66qdesigns rpredict SUR.predict rS)rr0rr<r*r)r"r!rOrPr%rMrJr8r2rHrINN)Fgh㈵>d) __name__ __module__ __qualname____firstlineno____doc__rCr=rUr7rGr4rt__static_attributes__rrrSrrrs.GR3j.640*X rSc.\rSrSrSrSSjrSrSrSrg) ria Two-Stage Least Squares for Simultaneous equations Parameters ---------- sys : list [endog1, exog1, endog2, exog2,...] It will be of length 2 x M, where M is the number of equations endog = exog. indep_endog : dict A dictionary mapping the equation to the column numbers of the the independent endogenous regressors in each equation. It is assumed that the system is entered as broken up into LHS and RHS. For now, the values of the dict have to be sequences. Note that the keys for the equations should be zero-indexed. instruments : ndarray Array of the exogenous independent variables. Notes ----- This is unfinished, and the design should be refactored. Estimation is done by brute force and there is no exploitation of the structure of the system. Nc >[U5S-S:wa[S[U5-5e[USSS25nX@lUSSS2UlUSSS2UlUSSS2Vs/sH"n[ R RU5PM$ snUlX0l 0nUR5Vs/sHoVRU/5PM nUH5nX'H*nXgRURUSS2U45 M, M7 X l [ R"[ R"SUR455n UHn[!X'5 M UR'U5Ulgs snfs snf! [#S[%X'5<SU<35e=f)Nr rr rz=The values of the indep_exog dict must be iterable. Got type z for converter )rrrr"r!rr'r(_K instrumentskeys setdefaultr& _indep_endogr.r/iter TypeErrortyperGrI) r>r indep_endogrr@rA instr_endogeq_keyvarcol_col_maps rrCSem2SLS.__init__s s8a<1 %'*3x01 1 ADqD N1X 1I 58AY?Y299((+Y?' /:/?/?/AB/A!  " %/AB!F%-#**499V+YrIrreqinstr_eqnewRHSrLHSyhats rrGSem2SLS.whitenNs '' && .BuuRHYYr]'')F&x0EAs$4$4599;HHD371[_Q//01 LL ! rSc/nURnURn[UR5H:nUR [ X4X$5R 5R5 M< U$)z )rIr"r rr&rr4rF)r>deltarIr"r\s rr4 Sem2SLS.fitdsX  twwA LLUXux0446== >  rS)rrrr"r!rrIrw) ryrzr{r|r}rCrGr4r~rrrSrrrs.1.h,rSc6^\rSrSrSrSU4SjjrSrSrU=r$)rfinz Not implemented yet. cF>[TU]XX45 UR5 grq)superrC _get_results)r>modelrFrMscale __class__s rrCSysResults.__init__rs! % . rScgrqrrrQs rrSysResults._get_resultswrvrSrr)Ng?) ryrzr{r|r}rCrr~ __classcell__)rs@rrfrfns   rSrf) #statsmodels.regression.linear_modelrnumpyrstatsmodels.base.modelrscipyr__all__rrrfrrrSrrsD39 ) k k `jjX  '  rS