U md7@sdZddlZddlmZddlmZddlmZm Z ddl m Z m Z m Z mZddd gZGd d d ZGd dde ZGd dde ZdS) aY Multivariate Conditional and Unconditional Kernel Density Estimation with Mixed Data Types References ---------- [1] Racine, J., Li, Q. Nonparametric econometrics: theory and practice. Princeton University Press. (2007) [2] Racine, Jeff. "Nonparametric Econometrics: A Primer," Foundation and Trends in Econometrics: Vol 3: No 1, pp1-88. (2008) http://dx.doi.org/10.1561/0800000009 [3] Racine, J., Li, Q. "Nonparametric Estimation of Distributions with Categorical and Continuous Data." Working Paper. (2000) [4] Racine, J. Li, Q. "Kernel Estimation of Multivariate Conditional Distributions Annals of Economics and Finance 5, 211-235 (2004) [5] Liu, R., Yang, L. "Kernel estimation of multivariate cumulative distribution function." Journal of Nonparametric Statistics (2008) [6] Li, R., Ju, G. "Nonparametric Estimation of Multivariate CDF with Categorical and Continuous Data." Working Paper [7] Li, Q., Racine, J. "Cross-validated local linear nonparametric regression" Statistica Sinica 14(2004), pp. 485-512 [8] Racine, J.: "Consistent Significance Testing for Nonparametric Regression" Journal of Business & Economics Statistics [9] Racine, J., Hart, J., Li, Q., "Testing the Significance of Categorical Predictor Variables in Nonparametric Regression Models", 2006, Econometric Reviews 25, 523-544 N)optimize) mquantiles)KDEMultivariate KernelReg)gpke LeaveOneOut _get_type_pos _adjust_shapeSingleIndexModel SemiLinear TestFFormc@s*eZdZdZd ddZddZddZd S) r a] Nonparametric test for functional form. Parameters ---------- endog : list Dependent variable (training set) exog : list of array_like objects The independent (right-hand-side) variables bw : array_like, str Bandwidths for exog or specify method for bandwidth selection fform : function The functional form ``y = g(b, x)`` to be tested. Takes as inputs the RHS variables `exog` and the coefficients ``b`` (betas) and returns a fitted ``y_hat``. var_type : str The type of the independent `exog` variables: - c: continuous - o: ordered - u: unordered estimator : function Must return the estimated coefficients b (betas). Takes as inputs ``(endog, exog)``. E.g. least square estimator:: lambda (x,y): np.dot(np.pinv(np.dot(x.T, x)), np.dot(x.T, y)) References ---------- See Racine, J.: "Consistent Significance Testing for Nonparametric Regression" Journal of Business & Economics Statistics. See chapter 12 in [1] pp. 355-357. dcCsD||_||_||_||_||_||_t|||dj|_||_ dS)N)bwvar_type) endogexogrfform estimatornbootrr _compute_sigsig)selfrrrrrrrrh/home/sam/Atlas/atlas_env/lib/python3.8/site-packages/statsmodels/sandbox/nonparametric/kernel_extras.py__init__PszTestFForm.__init__cCsd|j}|j}|||}|||}t|d}||}|t|}|||_t d}d|d}d|d} ||} | |} | |} t |j df} t |j D]j}| }tjjdd|fd}|| k}| |||<||}|||}|||}||}||| |<q| |_d}|jt| dkr4d}|jt| d krJd }|jt| d kr`d }|S) Nrg@g@sizezNot Significantg?*gffffff?z**gGz?z***)rrrrnpshapemean_compute_test_statZ test_statsqrtemptyrrangecopyrandomuniformZ boots_resultsr)rYXbmnZresidZsqrt5Zfct1Zfct2u1u2rZI_distjZu_bootZprobindZY_bootZb_hatZm_hatZ u_boot_hatrrrrrZsD         zTestFForm._compute_sigcCsFt|d}t|j}t|dddf}d}d}t|D]\}}t|} t| } t|j | |j|ddf |j dd} ||| | } | j| jkst || 7}|| d 7}t |dkst t |dks@t q@|d||d9}t|j d} |j | } |d| ||d9}||t| |}|S)NrF)data data_predictrZtosumrg?)rr rr__iter__ enumeratenextsqueezerrrAssertionErrorsumrrprodr#)rur-ZXLOOZuLOOZivalZS2iX_not_iZu_jKZf_iZix_conthpTrrrr"s0   zTestFForm._compute_test_statN)r )__name__ __module__ __qualname____doc__rrr"rrrrr ,s# &c@s:eZdZdZddZddZddZd d d Zd d ZdS)r a Single index semiparametric model ``y = g(X * b) + e``. Parameters ---------- endog : array_like The dependent variable exog : array_like The independent variable(s) var_type : str The type of variables in X: - c: continuous - o: ordered - u: unordered Attributes ---------- b : array_like The linear coefficients b (betas) bw : array_like Bandwidths Methods ------- fit(): Computes the fitted values ``E[Y|X] = g(X * b)`` and the marginal effects ``dY/dX``. References ---------- See chapter on semiparametric models in [1] Notes ----- This model resembles the binary choice models. The user knows that X and b interact linearly, but ``g(X * b)`` is unknown. In the parametric binary choice models the user usually assumes some distribution of g() such as normal or logistic. cCs~||_t||_|jd|_t|d|_t||j|_t|jd|_|j|_ d|_ d|_ d|_ |j |_|\|_|_dS)Nrrgaussian wangryzinaitchisonaitken)rlenr@r rrrr nobs data_typeckertypeokertypeukertype_est_loc_linearfunc _est_b_bwr+r)rrrrrrrrs   zSingleIndexModel.__init__cCsVtjj|jdfd}tj|j|dd}|d|j}||jd}||}||fS)NrrrZdisp)rr'r(r@rfmincv_looZ_set_bw_boundsrZparams0Zb_bwr+rrrrrRs  zSingleIndexModel._est_b_bwc Cst|}|d|j}||jd}t|j}t|j}d}t|D]r\}}t|} |j || t ||dddf t |j||dddf| dd} ||j|| d7}qJ||j S)Nrrrrr4r5) rasarrayr@rrrr6r7r8rQdotrK) rparamsr+rLOO_XLOO_YLr>r?r)GrrrrUs   "zSingleIndexModel.cv_looNc Cs|dkr|j}n t||j}t|d}t|f}t||jf}t|D]z}|j|j|j t |j|j dddft |||dddf|j d}|d||<t |d}|||ddf<qN||fS)Nrrr4) rr r@rr r$r%rQrrrYr+r9)rr4N_data_predictr!mfxr>mean_mfxmfx_crrrfits      zSingleIndexModel.fitcCsVd}|dt|jd7}|dt|jd7}|d|jd7}|d7}|d7}|S) Provide something sane to print.zSingle Index Model Number of variables: K =  zNumber of samples: nobs = Variable types: BW selection method: cv_ls Estimator type: local constant strr@rKrrreprrrr__repr__szSingleIndexModel.__repr__)N rCrDrErFrrRrUrdrorrrrr s ' c@s:eZdZdZddZddZddZd d d Zd d ZdS)r a} Semiparametric partially linear model, ``Y = Xb + g(Z) + e``. Parameters ---------- endog : array_like The dependent variable exog : array_like The linear component in the regression exog_nonparametric : array_like The nonparametric component in the regression var_type : str The type of the variables in the nonparametric component; - c: continuous - o: ordered - u: unordered k_linear : int The number of variables that comprise the linear component. Attributes ---------- bw : array_like Bandwidths for the nonparametric component exog_nonparametric b : array_like Coefficients in the linear component nobs : int The number of observations. k_linear : int The number of variables that comprise the linear component. Methods ------- fit Returns the fitted mean and marginal effects dy/dz Notes ----- This model uses only the local constant regression estimator References ---------- See chapter on Semiparametric Models in [1] cCst|d|_t|||_t||_t||j|_||_t|jd|_ ||_ |j |_ d|_ d|_ d|_|j|_|\|_|_dS)NrrrGrHrI)r rrrJr@exog_nonparametrick_linearrr rKrrLrMrNrOrPrQrRr+r)rrrrqrrrrrrr;s   zSemiLinear.__init__cCsNtjj|j|jfd}tj|j|dd}|d|j}||jd}||fS)z Computes the (beta) coefficients and the bandwidths. Minimizes ``cv_loo`` with respect to ``b`` and ``bw``. rrrSN)rr'r(rrr@rrTrUrVrrrrRKs zSemiLinear._est_b_bwc Cst|}|d|j}||jd}t|j}t|j}t|j}t|j|dddf}d}t |D]\} } t |} t |} t| |dddf} | | }|j ||| |j| ddf dd}|| ddf}||j| ||d7}qr|S)a Similar to the cross validation leave-one-out estimator. Modified to reflect the linear components. Parameters ---------- params : array_like Vector consisting of the coefficients (b) and the bandwidths (bw). The first ``k_linear`` elements are the coefficients. Returns ------- L : float The value of the objective function References ---------- See p.254 in [1] rNrWr5) rrXrrrrrr6rqrYr7r8rQ)rrZr+rr[r\ZLOO_ZZXbr]iir?r)ZZXb_jZYxr^ltrrrrUXs*   zSemiLinear.cv_looNc Cs|dkr|j}n t||j}|dkr,|j}n t||j}t|d}t|f}t||jf}|jt ||j dddf}t |D]P}|j |j ||j||ddfd}|d||<t|d} | ||ddf<q||fS)z+Computes fitted values and marginal effectsNrr_r)rr rrrqr@rr r$rrYr+r%rQrr9) rZ exog_predictZexog_nonparametric_predictr`r!rar)r>rbrcrrrrds$      zSemiLinear.fitcCsVd}|dt|jd7}|dt|jd7}|d|jd7}|d7}|d7}|S)rez'Semiparamatric Partially Linear Model rfrgzNumber of samples: N = rhrirjrkrmrrrroszSemiLinear.__repr__)NNrprrrrr s . ) )rFnumpyrZscipyrZscipy.stats.mstatsrZstatsmodels.nonparametric.apirrZ&statsmodels.nonparametric._kernel_baserrrr __all__r r r rrrrs   oq