>hDSrSSKrSSKJr SSKJr SSKJr Sr Sr "SS \5r "S S \5r "S S \5r g)a Created on Mon Jul 26 08:34:59 2010 Author: josef-pktd changes: added offset and zero-inflated version of Poisson - kind of ok, need better test cases, - a nan in ZIP bse, need to check hessian calculations - found error in ZIP loglike - all tests pass with Issues ------ * If true model is not zero-inflated then numerical Hessian for ZIP has zeros for the inflation probability and is not invertible. -> hessian inverts and bse look ok if row and column are dropped, pinv also works * GenericMLE: still get somewhere (where?) "CacheWriteWarning: The attribute 'bse' cannot be overwritten" * bfgs is too fragile, does not come back * `nm` is slow but seems to work * need good start_params and their use in genericmle needs to be checked for consistency, set as attribute or method (called as attribute) * numerical hessian needs better scaling * check taking parts out of the loop, e.g. factorial(endog) could be precalculated N)stats) factorial)GenericLikelihoodModelcZ[R"[R"X- 55$)Nnpmaxabsarr1arr2s oC:\Users\julio\OneDrive\Documentos\Trabajo\Ideas Frescas\venv\Lib\site-packages\statsmodels/miscmodels/count.pymaxabsr$s 66"&&% &&c`[R"[R"X- S- 55$)Nrr s r maxabsrelr's 66"&&q) **rc$\rSrSrSrSrSrSrg) PoissonGMLE,aJMaximum Likelihood Estimation of Poisson Model This is an example for generic MLE which has the same statistical model as discretemod.Poisson. Except for defining the negative log-likelihood method, all methods and results are generic. Gradients and Hessian and all resulting statistics are based on numerical differentiation. c[R"URU5nURn[R"U5X2-- [R "[ U55-$ Loglikelihood of Poisson model Parameters ---------- params : array_like The parameters of the model. Returns ------- The log likelihood of the model evaluated at `params` Notes ----- .. math:: \ln L=\sum_{i=1}^{n}\left[-\lambda_{i}+y_{i}x_{i}^{\prime}\beta-\ln y_{i}!\right] )rdotexogendogexplogr)selfparamsXBrs r nloglikeobsPoissonGMLE.nloglikeobs:sI"VVDIIv & vvbzUX%y/?(@@@rc[US5(d[eURnURn[R "[R "X55n[R"USS9$)zOreturn frozen scipy.stats distribution with mu at estimated prediction resultr)loc) hasattr ValueErrorr%r rrrrpoisson)rrr%r mus rpredict_distribution PoissonGMLE.predict_distributionOsStX&& [[F]]Ft,-B==+ +rN)__name__ __module__ __qualname____firstlineno____doc__r"r+__static_attributes__r-rrrr,s A* ,rrc6^\rSrSrSrSU4SjjrSrSrU=r$)PoissonOffsetGMLE_aYMaximum Likelihood Estimation of Poisson Model This is an example for generic MLE which has the same statistical model as discretemod.Poisson but adds offset Except for defining the negative log-likelihood method, all methods and results are generic. Gradients and Hessian and all resulting statistics are based on numerical differentiation. c >Ub/URS:Xa USS2S4nUR5UlOSUl[TU]"X4SU0UD6 g)Nrmissing)ndimraveloffsetsuper__init__rrrr<r9kwds __class__s rr>PoissonOffsetGMLE.__init__lsT  {{a$ ,,.DKDK  g  rcUR[R"URU5-nURn[R "U5X2-- [R "[U55-nU$r)r<rrrrrrr)rr r!rnlogliks rr"PoissonOffsetGMLE.nloglikeobs|sV$[[266$))V4 4 &&*(266)E2B+CCr)r<NNnone r.r/r0r1r2r>r"r3 __classcell__rAs@rr5r5_s   rr5c6^\rSrSrSrSU4SjjrSrSrU=r$) PoissonZiGMLEaMaximum Likelihood Estimation of Poisson Model This is an example for generic MLE which has the same statistical model as discretemod.Poisson but adds offset and zero-inflation. Except for defining the negative log-likelihood method, all methods and results are generic. Gradients and Hessian and all resulting statistics are based on numerical differentiation. There are numerical problems if there is no zero-inflation. c >SUl[TU]"X4US/S.UD6 Ub/URS:Xa USS2S4nUR 5UlOSUlUc'[ R"URS45Ul URRSUl [ R"[ R"UR5S45Ul U=RS- sl S/Ulg)Nrzi)r9extra_params_namesr8r start_params)k_extrar=r>r:r;r<ronesnobsrshapenparamshstackrQ cloneattrr?s rr>PoissonZiGMLE.__init__s   3g$(6 3-1 3  {{a$ ,,.DKDK <1 .DIyyq) IIrwwt||'/BFF2J.%(:RVVIeDT=UU ffTBFFG1H4E3E,F%FGGqr)rXrrRrVr<rQrFrHrJs@rrLrLs *6rrL)r2numpyrscipyr scipy.specialrstatsmodels.base.modelrrrrr5rLr-rrrcsM:#9'+ /,(/,f2.2hD*Dr